--- Log opened Thu Jan 01 00:00:06 2026
2026-01-01T00:10:59 -!- nerozero [~nerozero@87.253.63.54] has quit [Remote host closed the connection]
2026-01-01T01:06:17 < catphish> i'm confused, how are you planning to generate a sine wave without PWM?
2026-01-01T01:07:25 < catphish> qyx: i mean, what is the inverter doing at all if not PWM?
2026-01-01T01:07:50 < qyx> yes it is doing pwm, but into an inductor, the motor receives true sine wave
2026-01-01T01:10:02 < catphish> you almost definitely don't need to do that
2026-01-01T01:10:30 < catphish> a motor is a much bigger inductor than any inductor you'd want to include in your design
2026-01-01T01:12:38 < qyx> I need to do that because of the inverter topology
2026-01-01T01:12:49 < catphish> what topology?
2026-01-01T01:18:19 -!- qyx [~qyx@84.245.121.128] has quit [Ping timeout: 246 seconds]
2026-01-01T01:18:49 -!- qyx [~qyx@84.245.121.32] has joined ##stm32
2026-01-01T01:18:57 < qyx> my interwebs broke
2026-01-01T01:19:04 < qyx> https://ethz.ch/content/dam/ethz/special-interest/itet/power-electronic-systems-lab/images/Awards/conference-presentation-awards/3_APEC18_Three-Phase-Buck-Boost-Y-Inverter-with-Wide-DC-Input-Voltage-Range_FINAL-Paper_Antivachis.pdf
2026-01-01T01:19:08 < qyx> https://repository.kaust.edu.sa/bitstreams/82c2b878-3e17-4abe-880f-450cfe4f050c/download
2026-01-01T01:30:19 < catphish> qyx: i haven't had precisely the right amount of drugs to understand that inverter :'(
2026-01-01T01:33:15 < catphish> i'm guessing this is essentially a per-phase buck-boost but i can't figure out what on earth is goin gon
2026-01-01T01:33:31 < qyx> yes it is per-phase buck boost
2026-01-01T01:33:45 < zyp> catphish, I think the point is that when the input voltage is too low you need to boost it anyway, at which point you can just make a variable buck-boost stage that fits right into the motor, instead of boosting to a fixed voltage that's fed into motor PWM
2026-01-01T01:34:22 < catphish> yeah, i can't quite understand how it works, but the idea of doing a buck-boost on each phase is awesome
2026-01-01T01:36:22 < zyp> a traditional three phase inverter is kinda a buck topology using the motor as the inductor
2026-01-01T01:36:29 < catphish> correct
2026-01-01T01:36:51 < zyp> so it works well as long as you have a higher input voltage than the motor needs
2026-01-01T01:37:09 < zyp> but I don't think you can make what's effectively a boost topology with the motor as inductor
2026-01-01T01:38:51 < zyp> I mean, the mechanism of a boost converter is getting more current flowing in the inductor than what the load needs, and the whole issue with not having enough voltage for the motor is not being able to push more current (hence torque) into it
2026-01-01T01:38:52 < catphish> no, that wouldn't work, because the back EMF is in series with tghe inductor, so i understand now the need for the extra inductor
2026-01-01T01:39:08 < catphish> though i dont understand how it's actually controlled
2026-01-01T01:39:36 < zyp> well, consider a traditional inverter
2026-01-01T01:39:53 < zyp> for each phase you have effective voltage = bus voltage * pwm duty
2026-01-01T01:40:29 < zyp> you get the exact same behavior if you make that an actual voltage
2026-01-01T01:40:49 < zyp> I mean, consider a plain brushed DC motor
2026-01-01T01:41:09 < zyp> feeding it 24V at 50% duty is exactly the same as feeding it continous 12V
2026-01-01T01:42:50 -!- System_Error [~SystemErr@user/systemerror] has quit [Ping timeout: 252 seconds]
2026-01-01T01:43:04 < zyp> if you have a buck-boost driving a brushed DC motor, you can vary the voltage to get the same behavior as if you ran fixed voltage and varied pwm
2026-01-01T01:44:40 < zyp> replace the brushed DC motor with a brushless (and yeah, I know qyx said induction, but they work similarly enough), replace the single buck-boost with one for each phase, and run the voltage through the park/clarke transforms and you have it
2026-01-01T01:45:03 < catphish> oh, i think i undertand the 6-FET version, not so much the 12-fet versioon
2026-01-01T01:45:44 < zyp> I haven't looked at the schematics, but isn't 12 FETs just three synchronous buck-boost stages?
2026-01-01T01:46:27 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 244 seconds]
2026-01-01T01:46:55 < catphish> zyp: just googled that and yes it is
2026-01-01T01:47:58 < catphish> this diagram helps https://www.mdpi.com/wevj/wevj-16-00315/article_deploy/html/images/wevj-16-00315-g001-550.jpg
2026-01-01T01:48:29 < qyx> yes and I have 6 such stages
2026-01-01T01:48:43 < catphish> though i'm unsure why you'd ever use S2 or S3
2026-01-01T01:49:17 < catphish> couldn't they just be diodes?
2026-01-01T01:49:17 < qyx> for synchronous rectification of course
2026-01-01T01:49:27 < qyx> yes they can
2026-01-01T01:49:36 < catphish> that makes sense then
2026-01-01T01:49:59 < zyp> catphish, the word «synchronous» for a SMPS means it uses active halfbridges, not passive diodes
2026-01-01T01:50:19 < catphish> ss that purely to avoid diode losses?
2026-01-01T01:50:21 < catphish> *is
2026-01-01T01:50:24 < qyx> yes
2026-01-01T01:50:38 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-01T01:50:45 < qyx> now how much capacitors is enough
2026-01-01T01:51:18 < zyp> that's what I've wondered about too
2026-01-01T01:51:34 < zyp> how do you size the output cap when you're gonna have the output generate a sine?
2026-01-01T01:51:35 < catphish> qyx: again, given that the motor is an epic inductor, i'm not sure those capacitors are doing much
2026-01-01T01:51:47 < qyx> oh, input capacitors
2026-01-01T01:51:54 < catphish> oh, big
2026-01-01T01:51:59 < qyx> really
2026-01-01T01:52:05 < qyx> even with interleaved topology?
2026-01-01T01:52:09 < qyx> and high freq
2026-01-01T01:52:38 < catphish> i'm not sure
2026-01-01T01:52:43 < zyp> aren't input capacitance mostly a matter of how much you can fit and afford? :p
2026-01-01T01:52:49 < qyx> my quick talk to AI friend caused me to believe 33 uF is enough for one phase
2026-01-01T01:53:01 < catphish> i do however see a more bigger problem that you need to be fearful of
2026-01-01T01:53:07 < mawk> he's not your friend
2026-01-01T01:53:26 < qyx> and be well within the cap dV/dt handling capability and a couple of volts ripple
2026-01-01T01:53:42 < qyx> and that's even wihtout considering the interleaved topology, which halves that
2026-01-01T01:53:57 < qyx> also it is actually 3 phase and the input cap is shared, not all phases are switching at the same time
2026-01-01T01:54:02 < qyx> also, they are out of phase
2026-01-01T01:54:05 < qyx> so, idk
2026-01-01T01:54:28 < catphish> what happens when current is flowing backwards and you turn off S3?
2026-01-01T01:54:42 < catphish> that seems like a pretty fatal flaw in this topology
2026-01-01T01:54:57 < qyx> what would happen?
2026-01-01T01:55:08 < catphish> well the way i see it, S3 explodes
2026-01-01T01:55:40 < qyx> nothing explodes at whatever instant because there is a body diode
2026-01-01T01:55:41 < zyp> what do you mean?
2026-01-01T01:55:47 < qyx> you can have S3 permanently turned off
2026-01-01T01:55:57 < catphish> S3's body diode doesn't go that way
2026-01-01T01:56:00 < qyx> it would act as a normal diode in async topology
2026-01-01T01:56:12 < catphish> the load is connected only to the top of S3
2026-01-01T01:56:32 < catphish> the load is bidirectional and highly inductive
2026-01-01T01:56:34 < zyp> in a synchronous regulator, I don't think you ever tristate the halfbridges
2026-01-01T01:56:57 < catphish> right, but what happens when the top output FET is off? it can't conduct
2026-01-01T01:57:03 < catphish> and there's no other path
2026-01-01T01:57:20 < zyp> that's what the capacitor is for
2026-01-01T01:57:21 < catphish> except the output capacitor, which is a big ask of an output capacitor
2026-01-01T01:57:22 < qyx> there is a capacitor
2026-01-01T01:57:44 < catphish> i guess that works of those capacitors are big
2026-01-01T01:58:18 < zyp> the output capacitor obviously needs to be big enough to handle ripple regardless of which way power is going
2026-01-01T01:58:21 < qyx> remember you are switching at many kHz
2026-01-01T01:58:48 < zyp> also, remember that a buck-boost topology is fully symmetric
2026-01-01T01:59:06 < zyp> it can both buck and boost power in both directions
2026-01-01T01:59:36 < zyp> in other words, four quadrant operation
2026-01-01T02:00:11 < catphish> i'd want the capacitor to be at least as big as the motor because it's possible to dump the entire energy of the motor coil into that cap
2026-01-01T02:00:26 < catphish> without doing the maths, i don't know if that's hard
2026-01-01T02:00:47 < zyp> I don't think that statement makes sense
2026-01-01T02:00:50 < catphish> happy 2026 UTC+0 people
2026-01-01T02:02:10 < catphish> imagine there is 100A flowing from phase A to phase B, and then for whatever reason you shut down the inverter, that 100A will flow into phase B's capacitor with whatever energy is stored in the motor
2026-01-01T02:02:34 < qyx> no, thats not how the capacitor works
2026-01-01T02:02:54 < catphish> so where do you think that evergy is going to go?
2026-01-01T02:02:55 < qyx> once it charges, no more current flows
2026-01-01T02:03:11 < qyx> and if the back emf is not infinite, it will never flow again
2026-01-01T02:03:17 < qyx> unless it starts reversing
2026-01-01T02:03:38 < qyx> then the capacitor discharges and charges in opoosite polarity
2026-01-01T02:03:45 < catphish> i'm not talking about back EMF, i'm taking about the instantaneous energy in the motor coil's inductance
2026-01-01T02:05:55 < qyx> isn't it the same as an inverter in the "coast to stop" mode, when all transistors are turned off?
2026-01-01T02:06:02 < qyx> a classic inverter
2026-01-01T02:06:47 < catphish> in a classic inverter, you will see that evergy can always flow backwards through the inverter to the battery
2026-01-01T02:07:17 < catphish> there's always a path back to the battery even if all FETs are off
2026-01-01T02:07:42 < catphish> but in your topology, with S3 is off, there's no path from the motor phase back to the DC bus
2026-01-01T02:08:25 -!- IanW_ [~IceChat9@jindivik.force9.co.uk] has quit [Quit: Bye]
2026-01-01T02:08:28 < catphish> i assume the solution here is a huge output capacitor
2026-01-01T02:08:41 < catphish> at least as big as the load inductance
2026-01-01T02:09:28 < qyx> this converter is not special by any means, there are 3 such phases and there are body diodes in both directions
2026-01-01T02:09:31 < zyp> if you have a current flowing, wouldn't it just keep recycling through the body diodes?
2026-01-01T02:09:47 < qyx> yes
2026-01-01T02:10:12 < catphish> zyp: there is no body diode that can take current from the output in this topology
2026-01-01T02:10:22 < catphish> that's my concern
2026-01-01T02:10:47 < zyp> isn't current flowing out of the output?
2026-01-01T02:11:09 < catphish> look at https://www.mdpi.com/wevj/wevj-16-00315/article_deploy/html/images/wevj-16-00315-g001-550.jpg
2026-01-01T02:11:10 < catphish> Vout has no path through a body diode to safety
2026-01-01T02:11:32 < catphish> zyp: it's AC, half the time current is flowing in from the output
2026-01-01T02:12:33 < catphish> the only place that current can go it into the output cap
2026-01-01T02:12:38 < catphish> *is into
2026-01-01T02:13:04 < zyp> it's not pure AC, output voltage is always positive with respect to ground, so the output current is always flowing clockwise through the boost stage and the load
2026-01-01T02:13:35 < zyp> and when it's flowing clockwise, it can pass through the body diodes
2026-01-01T02:14:44 < catphish> except it's not. it's always positive with respect to ground, but the load isn't connected to ground like in that diagram, it's connected to the other 2 phases
2026-01-01T02:14:44 < zyp> remember that when you stop feeding energy into an inductor, the current keeps flowing in the same direction
2026-01-01T02:15:31 < zyp> yup, but all of the three phase voltages are sitting at some potential above ground
2026-01-01T02:15:50 < catphish> https://i.ibb.co/Gfn4HKxY/Screenshot-20260101-001516.png
2026-01-01T02:15:58 < catphish> this is the full topology
2026-01-01T02:16:10 < catphish> ground isn't relevant in the load
2026-01-01T02:16:34 < catphish> sometimes current is flowing out of "a" and sometimes into "a"
2026-01-01T02:17:07 < catphish> current can't flow out of all 3 phases, the load has no ground
2026-01-01T02:20:16 < qyx> oh you miss that one diode from the phase output to VBUS?
2026-01-01T02:20:19 < qyx> which is not there?
2026-01-01T02:20:39 < qyx> I guess it can be added
2026-01-01T02:21:16 < qyx> but there is no VBUS
2026-01-01T02:21:35 < catphish> qyx: no it can't because it would prevent you boosting above the VBUS voltage
2026-01-01T02:22:02 < qyx> so a 400 V TVS would do
2026-01-01T02:22:28 < catphish> yes, that might work, or a large output capacitor might also suffice
2026-01-01T02:22:41 < catphish> i haven't read the whole paper, this may be discussed
2026-01-01T02:22:45 < qyx> it definitely needs to be above the back EMF
2026-01-01T02:22:50 < qyx> to not brake the rotor
2026-01-01T02:23:04 < qyx> otherwise it goes puff
2026-01-01T02:23:41 < qyx> ok now I understand your concern, yeah
2026-01-01T02:23:44 < catphish> i think this has more scope for going boom than a regular "buck" inverter, but i assume this can be worked around
2026-01-01T02:24:06 < zyp> back EMF can't get higher than your output voltage, unless you're adding external acceleration
2026-01-01T02:25:17 < catphish> back EMF isn't an issue, as the output caps will already be charged to back EMF voltage
2026-01-01T02:25:49 < zyp> if your max output voltage is 400V, you cant push the motor to a speed that'll generate more than 400V of back-EMF
2026-01-01T02:25:50 < catphish> the issue is the inductive spike that happens when you turn off S3
2026-01-01T02:25:58 < zyp> (less when you accunt for losses)
2026-01-01T02:26:18 < qyx> yes TVS would catch that spike but can't catch the back-EMF
2026-01-01T02:26:29 < qyx> so needs to be rated about the working voltage at max speed
2026-01-01T02:26:37 < qyx> but below the transistor breakdown voltage
2026-01-01T02:26:45 < qyx> *above
2026-01-01T02:28:31 < catphish> at some point you basically want to short out the FETs before their failure voltage
2026-01-01T02:28:37 < zyp> > < zyp> it's not pure AC, output voltage is always positive with respect to ground, so the output current is always flowing clockwise through the boost stage and the load
2026-01-01T02:28:55 < zyp> I take this back, this is obviously not true since it'll be operating in all four quadrants
2026-01-01T02:31:23 < zyp> when you have negative current (and the inverter is operating), you'd be bucking or boosting the voltage and feeding it back to the input bus
2026-01-01T02:31:53 < catphish> yep
2026-01-01T02:33:11 < catphish> in a traditional buck inverter, this is mostly safe, because the body diodes can always be used, though car inverters are sometimes written off when the fuse blows and the DC bus cap rises too high
2026-01-01T02:33:15 < zyp> I was confusing myself with the simplified model of driving a brushed motor with a single buck-boost, which wouldn't have this problem (but also wouldn't be able to run in reverse)
2026-01-01T02:34:52 < catphish> yeah when you look at a single phase in isolation it seems fine
2026-01-01T02:35:21 < catphish> but when you have 2 or 3 switches phases, you have to have reverse current
2026-01-01T02:38:13 < zyp> speaking of motors, can anybody make sense of these ratings?
2026-01-01T02:38:32 < qyx> now the question is, why are they all using foil caps? I want MLCCs
2026-01-01T02:38:56 < zyp> these ratings: https://photos.app.goo.gl/gTpVtrJYDxtaqMEQ8
2026-01-01T02:39:12 < catphish> ugh, it's not trivial to find low-quiescent linear regulators with input voltages above 5.5V
2026-01-01T02:39:53 < catphish> qyx: MLCC just don't go big enough
2026-01-01T02:39:56 < qyx> catphish: tps709
2026-01-01T02:39:57 < zyp> 2.83kW, 600V, 5.5A kinda makes sense enough
2026-01-01T02:40:18 < zyp> but wtf is the 156V rating on the other motor? wtf does Uistr mean?
2026-01-01T02:40:36 < qyx> I can't read that label, probably as a result of my low internet-fu
2026-01-01T02:40:56 < catphish> 2.83kW, 600V, 5.5A makes perfect sense with some power factor thrown in
2026-01-01T02:41:17 < zyp> yeah, but 3.77kW, 9.5A, 156V makes no sense
2026-01-01T02:41:55 < qyx> isn't it 5 phase?
2026-01-01T02:42:32 < zyp> can't be, there's only three conductors for motor power
2026-01-01T02:43:20 < catphish> maybe the voltage rating is to ground but the power is based on a delta configuration
2026-01-01T02:43:37 < catphish> that would be a weird way to label it, but it's my best guess
2026-01-01T02:43:45 < zyp> by the way, did I mention that I got another robot?
2026-01-01T02:43:47 < zyp> https://photos.app.goo.gl/aUkqkbhf7zeuc5PG8
2026-01-01T02:44:01 < catphish> i see no robot
2026-01-01T02:44:16 < catphish> oh, yes i do
2026-01-01T02:44:33 < qyx> there is no Istr in the original datasheet
2026-01-01T02:44:35 < catphish> what can she do?
2026-01-01T02:44:54 < qyx> also the label is different for siemens and kuka
2026-01-01T02:45:07 < zyp> what original datasheet?
2026-01-01T02:45:27 < qyx> https://adegis.com/media/asset/fb15429c3b25d5992d74c7bdf75015298bf86e833defd660f51ddaa0fa512e7b.pdf
2026-01-01T02:45:30 < qyx> not datasheet
2026-01-01T02:45:33 < qyx> looks like a user manual
2026-01-01T02:45:56 < zyp> nice, you've got better google-fu than me
2026-01-01T02:46:21 < zyp> I only found secondhand sellers offering those for like $1-4k when I googled :p
2026-01-01T02:46:46 < qyx> it even has the pinout
2026-01-01T02:47:31 < qyx> Istr sounds like strom or string
2026-01-01T02:47:34 < zyp> catphish, nothing yet, I don't have 400V in the workshop yet
2026-01-01T02:47:58 < catphish> i have 400V in my house now :D
2026-01-01T02:48:12 < catphish> i guess that's normal in germanland
2026-01-01T02:48:20 < catphish> but kind of a novelty here
2026-01-01T02:48:27 < zyp> so do I, but I don't have a few km of extension cord :p
2026-01-01T02:48:44 < qyx> I would say you can't get single phase for a common house
2026-01-01T02:49:00 < zyp> I did buy a 230V/400V transformer a few years ago, but still haven't gotten it wired up
2026-01-01T02:49:12 < zyp> I need to buy some suitable breakers for the 400V side
2026-01-01T02:49:20 < catphish> qyx: 99% of houses are single phase 230V in UK
2026-01-01T02:49:34 < catphish> this is my new setup in my new house https://i.ibb.co/5Xw96nXk/PXL-20251128-143257297.jpg
2026-01-01T02:50:09 < qyx> yes, checking the distribution company rules now, the max single phase you can get is 25 A
2026-01-01T02:50:13 < qyx> anything higher is 3 phase
2026-01-01T02:50:21 < zyp> the fanuc robot I have has an isolation transformer in the controller, can be rewired wye/delta to work with either 230V or 400V
2026-01-01T02:50:29 < zyp> but the kuka doesn't do that
2026-01-01T02:50:55 < zyp> I think it rectifies the 400V directly to get a higher bus voltage,
2026-01-01T02:51:25 < zyp> considering the 600V rating of some of those motors
2026-01-01T02:51:38 < catphish> here you can get up to 100A single phase and only then do they upgrade to 3 phase, i have 3 x 100A :D
2026-01-01T02:52:04 < qyx> wow what the hell is that catphish 
2026-01-01T02:52:12 < qyx> that's not a germany style distribution box
2026-01-01T02:52:28 < catphish> qyx: UK style distribution boxes
2026-01-01T02:53:02 < qyx> they are similarly ugly than american
2026-01-01T02:53:08 < zyp> looks like a weird crossbreed between european and american shit
2026-01-01T02:53:13 < catphish> the box on the left is normal UK style, the tall one on the right is commercial style, not common in houses
2026-01-01T02:53:34 < catphish> yeah, commercial UK gear is more like american style, but 3-phase
2026-01-01T02:53:46 < catphish> i wish we have euro style 3 phase boards :(
2026-01-01T02:54:36 < catphish> residential UK is normally just one long row (like the board on the left), because we normally only do single phase
2026-01-01T02:54:49 < catphish> though in my case both of those boards are 3 phase
2026-01-01T02:55:17 < zyp> here's what I have in my home: https://bin.jvnv.net/file/46MYs.jpg
2026-01-01T02:56:14 < catphish> it's a slightly weird setup, the board on the right is a normal commercial distribution board, every circuit is a single phase RCBO. the board on the left is a bit different, it exists only to feed sub-boards in other parts of the house, it has no RCBOs but has a time delayed upfront RCD instead
2026-01-01T02:56:32 < qyx> yes we do zypstyle too
2026-01-01T02:56:38 < catphish> zyp: yeah, that's the style i wanted :'( can't buy un UK
2026-01-01T02:56:57 < qyx> like this https://www.elhakomplet.sk/fileadmin/user_upload/img5.JPG
2026-01-01T02:57:06 < qyx> eaton is very common here
2026-01-01T02:57:13 < zyp> same
2026-01-01T02:57:14 < qyx> and OEZ, which is basically siemens
2026-01-01T02:57:33 < catphish> on the other hand, no bus bars? seems kind of a mess of wires
2026-01-01T02:57:47 < zyp> bus bars are on the bottom of the breakers
2026-01-01T02:57:56 < qyx> there are bus bars
2026-01-01T02:57:59 < qyx> don§t you see them?
2026-01-01T02:58:02 < catphish> oh sorry, yes i see them, where's the input
2026-01-01T02:58:13 < qyx> on the left through the RCD
2026-01-01T02:58:13 < zyp> behind the bus bar
2026-01-01T02:58:18 < qyx> or behind
2026-01-01T02:58:31 < qyx> oh yes behind
2026-01-01T02:58:43 < zyp> the breakers can take both a cable and a bus bar connection on each screw
2026-01-01T02:58:45 < qyx> usually you can get N/3P or 3P/N RCDs, at least with siemens
2026-01-01T02:59:01 < catphish> yeah, there's a few things here i wish we had in UK :'(
2026-01-01T02:59:03 < qyx> with N/3P you can use a 3 phase busbar to go from the RCD directly to breakers
2026-01-01T02:59:18 < catphish> 3 phase bus bars with alternating neutrals, we don't have hat
2026-01-01T02:59:35 < qyx> there are no alternating neutrals
2026-01-01T02:59:54 < qyx> as you can see in my picture, we are not using RCDs per circuit
2026-01-01T03:00:02 < zyp> huh?
2026-01-01T03:00:10 < qyx> there is only one 3 phase RCD and then a bunch of single phase breakers for each RCD
2026-01-01T03:00:13 < zyp> we have alternating neutrals
2026-01-01T03:00:26 < qyx> you have IT, which yeah
2026-01-01T03:00:31 < zyp> no, I have TN
2026-01-01T03:00:45 < zyp> my busbars go L1, N, L2, N, L3, N, L1, N, …
2026-01-01T03:00:46 < qyx> as we are not using 2 module RDCs, we don't need alternating neutrals :P
2026-01-01T03:01:14 < zyp> and the one in my subpanel go L1, L2, L3, N, L1, L2, L3, N, …
2026-01-01T03:01:15 < zyp> https://bin.jvnv.net/file/T2G1F.jpg
2026-01-01T03:01:22 < qyx> tbh I haven't seen our kind of setups anywhere else, maybe in czech
2026-01-01T03:01:34 < qyx> but I like it more than yours
2026-01-01T03:02:43 < qyx> and here I only have 3 single-phase RCDs, with 16 individual breakers on the first one and 4+4 on the other two
2026-01-01T03:04:42 < qyx> 1u/250V MLCC is half the price of 1u/250V foil
2026-01-01T03:04:48 < qyx> ..and half the capacitance
2026-01-01T03:04:50 < zyp> here's the panel I've got in the workshop: https://bin.jvnv.net/file/ViJ23.jpg
2026-01-01T03:05:01 < zyp> that one is IT, so no N on the busbars
2026-01-01T03:06:05 < qyx> I was about to ask
2026-01-01T03:06:10 < zyp> photo is from when I'd just finished running the new supply cable and were waiting for the meter to get installed
2026-01-01T03:06:10 < qyx> where is the tn-c-s split
2026-01-01T03:06:56 < zyp> I'm gonna have to put in one when I put in the 400V transformer :)
2026-01-01T03:07:57 < zyp> I still have a couple meters left of that supply cable, so I figure I'll just use that for both the primary and secondary sides
2026-01-01T03:08:38 < zyp> running PEN on the PE conductor in the secondary side cable, and then doing the PE-N split as soon as it enters the cabinet
2026-01-01T03:11:23 < zyp> qyx, do you happen to know anything about what circuit breakers to put around transformers? :p
2026-01-01T03:11:43 < catphish> yeah, i was referring to L1, N L2, N, L3, N, L1, N, etc
2026-01-01T03:12:27 < catphish> all bus bars here are L1, L2, L3, L1, L2, L3 which keeps the size down, but neutral isn't switched
2026-01-01T03:13:07 < zyp> but then you need all that N wiring for the RCD
2026-01-01T03:15:38 < qyx> for that bulky half a meter transformer you have?
2026-01-01T03:17:09 < qyx> for primary I guess anything works, if you use the right breaking current depending on your upstream and value to keep the selectivity right
2026-01-01T03:17:33 < qyx> you are only handling winding shorts, basically
2026-01-01T03:18:12 < zyp> yeah, AIUI input side needs to guard against shorts, output side needs to guard against overload
2026-01-01T03:18:26 < qyx> for secondary I would go with the value exactly at the transformer rating, it can handle a bit more than that but i2t is always higher than rated
2026-01-01T03:18:40 < zyp> I'm trying to find the photo of the transformer label
2026-01-01T03:19:54 < qyx> and I would maybe use some sort of temperature device
2026-01-01T03:20:15 < qyx> which can be a simple thermostat and a circuit breaker actuator
2026-01-01T03:20:25 < qyx> better turn off output before the winding is too hot
2026-01-01T03:20:50 < qyx> even if it has a ptc inside, you could spare you from some smell, especially if it is indoors
2026-01-01T03:22:43 < zyp> ah, there it is: https://images.finncdn.no/dynamic/1280w/2022/11/vertical-0/16/7/281/231/067_823163041.jpg
2026-01-01T03:24:13 < zyp> and yeah, it's made for 400/230, I'm gonna run it in reverse
2026-01-01T03:24:58 < zyp> I'm also probably never gonna exceed the 20kVA rating :)
2026-01-01T03:25:09 < qyx> I wonder what is the no load input consumption
2026-01-01T03:25:38 < zyp> I wonder what the inrush current is
2026-01-01T03:26:08 < zyp> IIRC I have a spare 50A breaker that I was hoping I could hook the primary to
2026-01-01T03:26:51 < zyp> in which case I guess I've technically got overload protection on primary
2026-01-01T03:27:30 < qyx> that's 30 kW?
2026-01-01T03:27:48 < qyx> 50*230*3 is it?
2026-01-01T03:27:56 < zyp> no, 50*230*√3
2026-01-01T03:28:21 < qyx> and the main is 63?
2026-01-01T03:28:57 < qyx> I wonder if you could just connect it directly to the main breaker with a switch and yolo
2026-01-01T03:29:30 < zyp> so look at https://bin.jvnv.net/file/ViJ23.jpg again
2026-01-01T03:30:03 < qyx> yeah to tha hager on the 4th rail from top
2026-01-01T03:30:05 < zyp> Q1 is 80A, feeds into Q2 and Q3 (the two three phase breakers sitting alone bottom left), which IIRC are 50A each
2026-01-01T03:30:13 < qyx> oh
2026-01-01T03:30:54 < zyp> Q2 feeds the busbar above, i.e. all the 230V stuff, and I'm thinking to have Q3 feed the transformer
2026-01-01T03:31:13 < qyx> I would replace Q3 with a 63 A one
2026-01-01T03:31:35 < qyx> hm but you can't use the existing 50 A for secondary
2026-01-01T03:31:40 < zyp> I don't wanna buy a 63A one if the 50A works
2026-01-01T03:32:18 < qyx> so 50 A primary and 32 A secondary then?
2026-01-01T03:32:53 < zyp> 32A sounds too high, being above the rated value
2026-01-01T03:33:24 < qyx> not many values to choose from then :P
2026-01-01T03:33:27 < zyp> then again, next size down is 25A :p
2026-01-01T03:34:24 < zyp> I figure I'll just put that off and run it right into a 16A RCBO
2026-01-01T03:34:37 < zyp> because what I currently need is a 16A outlet
2026-01-01T03:35:02 < zyp> eventually I'd like to have another 16A outlet, so I can deal with it then :p
2026-01-01T03:43:37 < qyx> that's the right approach to problem solving!
2026-01-01T03:44:37 < qyx> so for 25A current, how many vias do I need
2026-01-01T03:44:57 < zyp> as many as you can fit?
2026-01-01T03:46:23 < catphish> thanks for the tip on the TPS709, looks ideal
2026-01-01T03:46:50 < qyx> np
2026-01-01T03:47:08 < qyx> hm apparently a 0.3/0.6 via can handle about 2 A of current
2026-01-01T03:47:11 < catphish> looks like it has an enable pin too, so i can skip the FET i was using to turn off my LCD
2026-01-01T03:47:28 < qyx> that's a plenty of current
2026-01-01T03:47:33 < qyx> of course I could fit 100 of them there
2026-01-01T03:48:28 < catphish> so pick a number between 1 and 100 :)
2026-01-01T03:48:52 < zyp> «can handle» is relative
2026-01-01T03:49:31 < qyx> https://bin.jvnv.net/file/q5hD9/Screenshot_2026-01-01_02-49-17.png
2026-01-01T03:49:38 < qyx> it is gonna be cooled by large copper busbars
2026-01-01T03:49:46 < qyx> (the red ones)
2026-01-01T03:50:00 < catphish> looks sane
2026-01-01T03:50:43 < qyx> if that thing really handles 6x25 A, I may start believing in god
2026-01-01T03:51:21 < catphish> why isn't it on the same side as the busbar?
2026-01-01T03:52:18 < catphish> not sure i'd rely on vias for any real current
2026-01-01T03:53:01 < qyx> because the fet needs to be from the bottom
2026-01-01T03:53:03 < qyx> cooling reasons
2026-01-01T03:53:16 < qyx> and it basically sits directly on the busbar
2026-01-01T03:53:28 < qyx> there is a pcb between them
2026-01-01T03:53:50 < qyx> catphish: what current sensors did you use?
2026-01-01T03:54:42 < catphish> again, i was at the mercy of tesla's setup, they have hall effect sensors built around the bus bars
2026-01-01T03:55:15 < qyx> good, so I may do the same and it should work
2026-01-01T03:55:28 < qyx> most inverters I have seen dismantled on youtube use halls
2026-01-01T03:56:40 < catphish> qyx: it depends on current really, for lower current you can get chips with halls, i wouldn't bother with shunt resistors due to the high voltages involved
2026-01-01T03:57:26 < qyx> yeah allegro has some
2026-01-01T03:57:52 < zyp> I don't really like them
2026-01-01T03:57:56 < qyx> me neither
2026-01-01T03:58:02 < catphish> i imagine you want this kinda setup https://www.hobbyrc.co.uk/matek-hall-current-sensor-150a
2026-01-01T03:58:44 < qyx> ACS758 was my first idea
2026-01-01T03:58:47 < zyp> I've got bad experience with the allegro halls, but to be fair most of that was due to poor design
2026-01-01T03:59:19 < catphish> basically anything that can handle your current is fine
2026-01-01T03:59:36 < zyp> would be nice if they measured somewhat accurately too though
2026-01-01T03:59:42 < qyx> OH! TMCS1127
2026-01-01T04:00:10 < qyx> that's something I havent seen before
2026-01-01T04:00:28 < catphish> also consider your isolation
2026-01-01T04:00:49 < qyx> oh zero
2026-01-01T04:01:01 < qyx> :>
2026-01-01T04:01:01 < catphish> that makes things easier
2026-01-01T04:01:16 < qyx> I really should isolate the output from the rest
2026-01-01T04:01:26 < qyx> the battery is at safe level +-55 V (110 V total)
2026-01-01T04:01:30 < catphish> that one claims 80A
2026-01-01T04:01:45 < qyx> but again, I am not doing isolated inverter
2026-01-01T04:01:51 < qyx> so I would rather isolate CAN
2026-01-01T04:02:05 < zyp> TI has some fun delta-sigma shunt ADCs
2026-01-01T04:02:15 < zyp> e.g. https://www.ti.com/lit/ds/symlink/amc1303m2510.pdf
2026-01-01T04:02:16 < catphish> if you can run the whole thing at high voltage and just isolate CAN, that's simple
2026-01-01T04:02:47 < zyp> you could probably piggyback AVDD from the high side bootstrap
2026-01-01T04:02:48 < qyx> 0.2% nonlinearity, 0.4% gain error
2026-01-01T04:02:53 < qyx> 12 mA lifetime drift
2026-01-01T04:02:55 < qyx> I can handle that
2026-01-01T04:03:33 < catphish> what kind of motor are you planning to run
2026-01-01T04:04:05 < zyp> qyx, then just run the signal into the stm32 DFSDM or similar :)
2026-01-01T04:04:05 < qyx> 15 kW induction 93% eff, 1450 rpm
2026-01-01T04:04:22 < catphish> induction motors are a rabbit hole :)
2026-01-01T04:04:35 < qyx> cooking with what I can get
2026-01-01T04:04:50 < qyx> zyp: I also considered amc1200
2026-01-01T04:05:31 < catphish> i never managed to conteol mine very cleverly, but V/Hz works well enough
2026-01-01T04:05:42 < zyp> I like going digital as soon as possible, less risk of noise fucking up your shit
2026-01-01T04:06:02 < qyx> catphish: good to hear, I am lazy to invent anything more elaborate
2026-01-01T04:06:29 < zyp> I'd be more inclined to use a hall sensor if there were any suitable ones with digital output
2026-01-01T04:07:10 < zyp> the only ones I found last I checked had too low sample rate to be interesting for motor control
2026-01-01T04:07:12 < catphish> you don't even need current sensors for V/Hz conrol, but they're useful for monitoring
2026-01-01T04:07:43 < zyp> whatever you do, you probably want to be able to current limit
2026-01-01T04:08:13 < catphish> if nothing else, a safety cutoff is a good idea
2026-01-01T04:08:27 < catphish> mine saved me a couple of times
2026-01-01T04:08:54 < qyx> oh and what do you use as an emergency battery disconnect?
2026-01-01T04:08:58 < qyx> a pyro breaker?
2026-01-01T04:09:23 < catphish> nothing
2026-01-01T04:09:53 < qyx> I would probably do an accel + something, because there is a risk of tipping over
2026-01-01T04:10:08 < qyx> and I want to disconnect the battery at some suitable place if that happens
2026-01-01T04:10:31 < catphish> in my car, the majority of the safety was done in software, the last resort is just a 600A fuse
2026-01-01T04:11:04 < catphish> (or turning the key to open the contactors)
2026-01-01T04:11:20 < qyx> I remember you did a steel box for the battery, do you have a photo when it is completed?
2026-01-01T04:11:25 < qyx> I only saw it empty :P
2026-01-01T04:11:52 < zyp> qyx, do keep in mind that disconnecting the battery means disconnecting your sink for any regenerated energy
2026-01-01T04:11:58 < Steffanx> Is your car still alive and moving around catphish ?
2026-01-01T04:12:28 < zyp> disconnect the battery at the wrong time and you'll blow up a bunch of stuff
2026-01-01T04:12:39 < catphish> https://i.ibb.co/vvXs5tx2/PXL-20220607-141934958.jpg
2026-01-01T04:12:43 < qyx> zyp: I will break the whole hydrostatic drivetrain so no regen energy, all will go into oil heating
2026-01-01T04:12:56 < qyx> the motor will not turn in emergency
2026-01-01T04:13:08 < qyx> (under normal circumstances, yes)
2026-01-01T04:13:33 < catphish> Steffanx: yep, car is still on the road, i sold it, but it's still going well
2026-01-01T04:13:44 < zyp> if it's moving when the emergency occurs, it'll still have inertia, no?
2026-01-01T04:14:23 < catphish> there's no requirement to do anything electrical with the inertia though
2026-01-01T04:14:35 < zyp> nope
2026-01-01T04:14:37 < qyx> no, the motor will coast to stop
2026-01-01T04:14:48 < catphish> the only energy you need to worry about electrically is the inductance
2026-01-01T04:15:14 < catphish> or... if you have a fixed magnet motor and the back EMF exceeds battery voltage, than can get messy
2026-01-01T04:15:29 < catphish> but you can still just disconnect it
2026-01-01T04:15:34 < qyx> depending on my valve configuration, the motor may be stopped by a relief valve
2026-01-01T04:16:06 < qyx> why did you sell the car?
2026-01-01T04:16:36 < qyx> also, whoa the battery box, looks goot
2026-01-01T04:16:36 < qyx> d
2026-01-01T04:16:41 < catphish> qyx: mostly because i didn't have a proper space to keep it
2026-01-01T04:17:01 < qyx> don't you have a new house with big garden?
2026-01-01T04:17:47 < catphish> yes, but even there, there's only a very small garage
2026-01-01T04:18:06 < catphish> maybe eventually i'd build more and have a new car prohect
2026-01-01T04:18:09 < catphish> *project
2026-01-01T04:18:25 < catphish> but right now, i have no money until i sell the old house :)
2026-01-01T04:18:41 < catphish> new house isn't ready to live in yet
2026-01-01T04:21:02 < catphish> i'm shocked i didn't find TPS709 sooner, this is super useful for lazy super low power STM32 automotive stuff
2026-01-01T04:21:08 < catphish> and it's cheap
2026-01-01T04:22:10 < qyx> there is also tps706 which I am using almost exclusively for 5 to 3v3
2026-01-01T04:22:39 < qyx> basically I am only using that + a low noise one (lm5912)
2026-01-01T04:22:51 < qyx> sorry lp5912
2026-01-01T04:24:42 < catphish> this STM32 can do 2uA in standby mode, so i'm hoping to get down to that kinda order of magnitude
2026-01-01T04:26:20 < catphish> another 1.5uA from the TPS709, so i should be well within my 10uA target if it works properly
2026-01-01T04:27:26 < qyx> another good one is tps259531 (6 V in, buck)
2026-01-01T04:27:40 < qyx> with a *real* 4 uA Iq shen switching, with zero load
2026-01-01T04:27:44 < qyx> *when
2026-01-01T04:27:52 < qyx> sorry
2026-01-01T04:28:14 < qyx> TPS62825, 5.5 V in
2026-01-01T04:28:44 < catphish> i usually want 12V in
2026-01-01T04:28:57 < catphish> though this project is 9V battery
2026-01-01T04:30:24 < qyx> time to make a lib for that bulky SiC mosfet
2026-01-01T04:31:19 -!- hexo is now known as hexo[
2026-01-01T04:31:24 < catphish> i need to make lib for TPS709 but it's too late so i sleep first
2026-01-01T04:32:11 -!- hexo[ is now known as ]hexo[
2026-01-01T04:38:42 < qyx> why PG-HDSOP-16 transistors has gate on pin 8
2026-01-01T04:38:46 < qyx> and PG-HDSOP-22 on pin 1
2026-01-01T04:38:51 < qyx> that's not fair
2026-01-01T04:53:25 -!- infisc [uid692580@user/infisc] has joined ##stm32
2026-01-01T05:02:24 -!- ]hexo[ is now known as hexo
2026-01-01T05:09:07 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has joined ##stm32
2026-01-01T07:07:43 -!- infisc [uid692580@user/infisc] has quit [Quit: Connection closed for inactivity]
2026-01-01T07:14:01 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-01T08:01:33 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 244 seconds]
2026-01-01T08:13:39 -!- infisc [uid692580@user/infisc] has joined ##stm32
2026-01-01T08:14:30 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-01T08:28:12 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has quit [Ping timeout: 256 seconds]
2026-01-01T10:27:43 -!- infisc [uid692580@user/infisc] has quit [Quit: Connection closed for inactivity]
2026-01-01T10:30:08 < jpa-> karlp: pluto+ appears weird, with the title saying "70MHz-6GHz" but then the specs suggest it ships with AD9363 which limits it to 325MHz-3.8GHz?
2026-01-01T10:30:35 < jpa-> or is the "can hack to" some kind of software hack?
2026-01-01T11:16:40 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-01T12:13:56 -!- IanW_ [~IceChat9@jindivik.force9.co.uk] has joined ##stm32
2026-01-01T12:56:00 < qyx> rip jbo
2026-01-01T12:56:12 < qyx> wtf did they do
2026-01-01T13:07:25 < Steffanx> Big Bang.
2026-01-01T13:15:03 < ventyl> wat?
2026-01-01T13:26:53 < qyx> something exploded in a bar killing 40 or so
2026-01-01T14:47:43 -!- infisd [~infisc@user/infisc] has quit [Ping timeout: 264 seconds]
2026-01-01T15:12:11 -!- IanW_ [~IceChat9@jindivik.force9.co.uk] has quit [Quit: Bye]
2026-01-01T15:39:53 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has joined ##stm32
2026-01-01T15:55:33 -!- lockna [~lockna@193-81-168-132.hdsl.highway.telekom.at] has quit [Remote host closed the connection]
2026-01-01T15:56:47 -!- lockna [~lockna@193-81-168-132.hdsl.highway.telekom.at] has joined ##stm32
2026-01-01T16:02:33 -!- lockna [~lockna@193-81-168-132.hdsl.highway.telekom.at] has quit [Remote host closed the connection]
2026-01-01T16:03:47 -!- lockna [~lockna@193-81-168-132.hdsl.highway.telekom.at] has joined ##stm32
2026-01-01T16:50:35 -!- lockna [~lockna@193-81-168-132.hdsl.highway.telekom.at] has quit [Ping timeout: 240 seconds]
2026-01-01T17:02:06 < Steffanx> Welcome to the euro by the way antto . Do you like your new coins?
2026-01-01T17:06:57 < qyx> do we officially have dracula now?
2026-01-01T17:16:37 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has quit [Ping timeout: 246 seconds]
2026-01-01T17:21:09 < antto> Steffanx, i haven't seen the euro at all yet
2026-01-01T17:23:05 < Steffanx> When will you actually get the euro then, antto ?
2026-01-01T17:24:57 < antto> i don't want it
2026-01-01T17:25:08 < antto> afaiu this was a bad move
2026-01-01T17:25:31 < antto> but it's Done now, and there's no path to get out of it
2026-01-01T17:25:45 < qyx> that was the best thing we could do back in 2008
2026-01-01T17:25:46 < antto> how wonderful
2026-01-01T17:25:59 < qyx> and I am glad we have it
2026-01-01T17:26:32 < antto> time will tell
2026-01-01T17:27:04 < antto> not all countries are the same
2026-01-01T17:29:23 < ventyl> AFAIK your economy is doing better than ours, you should be OK
2026-01-01T17:30:47 < qyx> if you didn't notice, all economies in the EU are currently doing better than ours
2026-01-01T17:32:04 < qyx> oh what wait
2026-01-01T17:32:12 < qyx> isn't antto from romania?
2026-01-01T17:32:30 < Steffanx> .bg
2026-01-01T17:32:34 < ventyl> yes, they are doing better than us
2026-01-01T17:32:36 < qyx> so no dracula then, sorry
2026-01-01T17:32:48 < ventyl> probably even bulgary is doing better than us
2026-01-01T17:32:50 < ventyl> :>
2026-01-01T17:33:23 < qyx> all good, me fopaah
2026-01-01T17:33:23 < antto> our numbers were faked to get in
2026-01-01T17:34:19 < qyx> see, we have faked numbers, stealing billions from our citizens to "consolidate" and still doing the worst
2026-01-01T17:35:02 < antto> we're not okay, i wanted to eat some pickles, went to the supermarket, there were 8 different brands of pickles, most of them bulgarian, and yet the pickles in all of them were from .tr
2026-01-01T17:35:21 < antto> same goes for basic sh*t like table salt
2026-01-01T17:36:18 < antto> the .eu is becoming a big lie
2026-01-01T17:36:49 < antto> in some aspects
2026-01-01T17:37:51 < qyx> that's not fault of the eu imho, the same goes for any other community, members can't be self sufficient if driven by the money, only the things worth it are being made by individual members
2026-01-01T17:38:00 < antto> the food is all wonderful because Regulations and stuff, but then i actually go to get some pickles and i can choose between 8 kinds of Turkish pickles which are grown outside the Wonderful EU Regulations, so there can be all kinds of sh*t in them
2026-01-01T17:38:29 < qyx> oh you mean from the outside
2026-01-01T17:38:40 < antto> some time earlier we were going to shut down our coal electricity because it's Bad
2026-01-01T17:39:03 < antto> and it's Very Expensive (because the .eu adds an immaginary tax to it)
2026-01-01T17:39:19 < antto> and you know what we would've done when that happened?
2026-01-01T17:39:44 < antto> we'd export the coal to Serbia, they'd make electricty from it and sell it to us
2026-01-01T17:39:57 < antto> and that's not gonna make the air cleaner
2026-01-01T17:40:11 < qyx> idk we have nukes
2026-01-01T17:40:20 < qyx> and some water
2026-01-01T17:41:18 < antto> i mean, some of those regulations are stupid, some are inneffective when the thing they ban is not banned in a nearby country
2026-01-01T17:43:19 < antto> now this sh*t with the Euro currency afaik will allow the f*cking idiots in the gubbermint to do even worse things then they could do before
2026-01-01T17:43:24 < antto> and they surely will
2026-01-01T17:43:44 < antto> * than
2026-01-01T17:44:10 < antto> like... drawing gigantic loans
2026-01-01T17:44:49 < antto> but it's too late... so... meh
2026-01-01T17:45:22 < antto> january is the month where all of our Lev has to be changed to Euro, so it's gonna be a month of chaos
2026-01-01T19:26:16 -!- rkta [~rkta@user/rkta] has quit [Quit: zzz]
2026-01-01T19:26:28 -!- rkta [~rkta@user/rkta] has joined ##stm32
2026-01-01T19:35:42 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-01T19:36:03 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Read error: Connection reset by peer]
2026-01-01T19:37:07 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-01T19:47:20 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Ping timeout: 240 seconds]
2026-01-01T20:22:39 < catphish> ooo stm32cubeide v2!
2026-01-01T20:36:32 < catphish> wait... there's no cubemx in cubeide 2.0 - what's even the point without cubemx :'(
2026-01-01T20:37:53 < catphish> maybe it's still somewhat integrated once installed
2026-01-01T20:40:46 < catphish> interesting, cubemx requires java but cibeide doesn't
2026-01-01T20:41:43 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has joined ##stm32
2026-01-01T20:48:03 < Phantom> catphish: what's new in it?
2026-01-01T20:48:08 < Phantom> faster speed maybe?
2026-01-01T20:48:19 < Phantom> less .ioc corruption?
2026-01-01T20:48:47 < catphish> well, no ioc corruption because cubeide is no longer integrated :D
2026-01-01T20:49:09 < catphish> running it externally doesn't seem like a major hassle
2026-01-01T20:49:25 < Phantom> so, it is a downgrade then
2026-01-01T20:51:26 < catphish> appears to be, i haven't read the release notes, but it's certianly less useful without it
2026-01-01T21:03:04 < qyx> catphish: https://bin.jvnv.net/file/8pZ3L/Screenshot_2026-01-01_20-02-42.png
2026-01-01T21:04:43 < catphish> qyx: that's some serious engineerings
2026-01-01T21:06:18 < qyx> I should probably add some holes for smoke
2026-01-01T21:15:15 < karlp> jpa-: aiui, yeahm pluto+ is the 63, which is software unlockable to be "61" by folloiwng AD's own notes, just with not the same absolute perf.
2026-01-01T21:15:40 < karlp> I prefer the plus because a) gig ethernet and b) 2x2, I don't really need the extra freq. also c) can get it with a proper case
2026-01-01T21:16:21 < qyx> catphish: keyways, does that mean I can't rotate the angled plugs freely?
2026-01-01T21:17:14 < qyx> eg. red and black are not available w/o keyways, I understand I must choose receptacle and plug with the same keyway, but what direction will it go in the end? can I rotate it when mated?
2026-01-01T21:18:45 < qyx> oh now I can see red/black are awailable w/o too
2026-01-01T21:19:35 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 250 seconds]
2026-01-01T21:22:52 < jpa-> karlp: yeah, it looks nice
2026-01-01T21:28:00 < catphish> qyx: no, keyways just prevent you plugging the wrong plug into the wrong socket by making them different
2026-01-01T21:28:13 < catphish> the rotation is in the plug and still functions
2026-01-01T21:28:59 < qyx> ok so for the motor I want w/o keyways, for others I want black/reds
2026-01-01T21:29:49 < catphish> depending on the plugs, they can have lots of key options, but you should ideally use them for the battery terminals to avoid catastrophic mistakes
2026-01-01T21:30:06 < catphish> probably doesn't matter on motor
2026-01-01T21:30:38 < catphish> this is confusing. TPS709 datasheet page 13 https://www.ti.com/lit/ds/symlink/tps709.pdf
2026-01-01T21:30:47 < catphish> 7.4 Device Functional Modes
2026-01-01T21:31:23 < catphish> "EN is pulled high by a 300-nA current source; therefore, EN can be left floating to enable the device." fine but... "When EN goes below 0.4 V, the device is disabled. During this time, OUT is high impedance and the current into IN (I(SHUTDOWN)) is typically 150 nA."
2026-01-01T21:31:54 < qyx> what's confusing
2026-01-01T21:32:25 < catphish> how can the part only pull 150nA from IN when it's pulling up EN at 300nA
2026-01-01T21:33:37 < qyx> so the current source is probably disabled too
2026-01-01T21:33:57 < qyx> which means it cannot wake up back when the EN is released
2026-01-01T21:34:45 < qyx> which means you can eg. use a button to temporarily connect EN to your VIN with some resistor
2026-01-01T21:35:16 < qyx> and then the device can go back to standby by pulling EN low with a open drain output
2026-01-01T21:35:28 < qyx> idk
2026-01-01T21:38:34 < catphish> yeah, that makes sense (except you can't connect EN to VIN on this part), and also, i'm not convinced that's true, or it would probably be explicit about it
2026-01-01T21:40:06 < catphish> anyway, i'm not too worried, i'm going to pull it down with a 10k resistor, and if the idle current is 450nA i won't worry
2026-01-01T21:42:38 < catphish> but it makes no sense for that to be true
2026-01-01T22:08:53 -!- jmcgnh [~jmcgnh@wikipedia/jmcgnh] has quit [Quit: ZNC 1.8.1 - https://znc.in]
2026-01-01T22:49:28 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-01T23:50:05 -!- hexo [~hexo@user/hexo] has quit [Quit: Leaving]
2026-01-01T23:50:51 -!- hexo [~hexo@user/hexo] has joined ##stm32
--- Day changed Fri Jan 02 2026
2026-01-02T00:00:55 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
2026-01-02T00:03:12 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 256 seconds]
2026-01-02T00:04:28 -!- infisd [~infisc@user/infisc] has joined ##stm32
2026-01-02T00:23:25 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has quit [Ping timeout: 264 seconds]
2026-01-02T00:37:10 -!- infisd [~infisc@user/infisc] has quit [Quit: Vive la révolution.]
2026-01-02T00:37:51 -!- infisd [~infisc@user/infisc] has joined ##stm32
2026-01-02T00:53:31 -!- infisc [uid692580@user/infisc] has joined ##stm32
2026-01-02T01:00:40 < specing> antto> and it's Very Expensive (because the .eu adds an immaginary tax to it)
2026-01-02T01:00:55 < specing> it's not immaginary, it's partial offset for the negative externalities
2026-01-02T01:01:29 < specing> coal pollutes heavily, it is a disaster every day it operates
2026-01-02T01:32:01 < Steffanx> So you move your coal elsewhere and import electricity. Yeah... Problem solved
2026-01-02T01:34:41 < specing> not so fast, EU now has import CO2 tariffs
2026-01-02T01:34:59 < specing> specifically to fix this
2026-01-02T01:35:29 < specing> they came into force yesterday, in fact
2026-01-02T01:46:50 < antto> i meant a "made up tax" really
2026-01-02T01:47:45 < antto> it allows some f*ckers to say "coal is more expensive than Green Energy", which is only true after you artificially put this tax
2026-01-02T01:49:48 < karlp> ok, vna restored, found the correct python script as an attachment on a forum post that works flawlessly for upgrades from linux, vs the failing "run the vendor binary in wine" that was recommended...
2026-01-02T01:58:18 < karlp> catphish: if you want cheaper, low Iq, basic regs, it's always worth looking at: https://github.com/karlp/zypsnips/blob/master/lm1117.replacements.for.glory.md
2026-01-02T01:58:40 < karlp> two of the three there are still well in the thousands of stock...
2026-01-02T01:59:37 < qyx> does catphish have a slight idea about what inductance in mH were we talking about?
2026-01-02T02:00:25 < karlp> catphish: all 3 phases are brown? or is that really just using three phase gear with single phase split into three sections?
2026-01-02T02:01:03 < qyx> my ai friend says 30 mH with about 25 J stored energy in one winding
2026-01-02T02:01:09 < qyx> so I need a pretty bulky TVS
2026-01-02T02:04:03 < karlp> (but yeah, tps709 is a great family...)
2026-01-02T02:08:51 < specing> antto: coal is much more expensive once you account for its negative externalities.
2026-01-02T02:09:19 < antto> dafuq is that even
2026-01-02T02:09:22 < specing> as I said already, the tax is a (partial) offset for those negative externalities to correct the degenerate market
2026-01-02T02:10:15 < specing>  https://www.reddit.com/r/Economics/wiki/faq_carbonpricing
2026-01-02T02:10:17 < karlp> and... we're caught up
2026-01-02T02:10:23 < karlp> fack, it's midnight already, back to work tomorrow
2026-01-02T02:12:34 < antto> specing, i'm not gonna read that whole thing now
2026-01-02T02:13:31 < antto> this tax is artificial
2026-01-02T02:14:40 < antto> do you know how much of the polution comes from .bg?
2026-01-02T02:15:16 < specing> a fair share?
2026-01-02T02:15:32 < antto> you can't see .bg on the f*cking globe
2026-01-02T02:15:38 < specing> Why are you still bitching about something you choose to remain clueless about?
2026-01-02T02:16:29 < specing> It's a nice read, I recommend it
2026-01-02T02:17:03 < antto> okay, let's say the tax is not dumb
2026-01-02T02:17:36 < antto> it was used as a reason to very quickly shut down our coal power
2026-01-02T02:18:16 < qyx> well, coal power should have been dead at least 10 years
2026-01-02T02:18:47 < antto> because?
2026-01-02T02:19:04 < qyx> we managed to turn them off in 2023 and 2024 and that was already too late
2026-01-02T02:19:59 < antto> it'd be great if we could have something better so we can turn the coal off
2026-01-02T02:20:00 < qyx> because what, because 1. fossil and releases stored carbon 2. not available forever 3. pollution
2026-01-02T02:20:19 < antto> do we have something better? no. will we have something soon? doesn't look like so
2026-01-02T02:20:48 < antto> we do have coal, we probably have gas but SomeOne is not letting us dig it
2026-01-02T02:20:56 < qyx> don't we have water? and fission with less emissions and waste?
2026-01-02T02:21:25 < qyx> not even speaking about other renewables
2026-01-02T02:21:43 < qyx> I am running on solar all year long
2026-01-02T02:21:56 < antto> we may have alternative possibilities, but we don't have anything ready to just turn on, so before this happens, it will be very stupid to turn off the coal
2026-01-02T02:22:03 < antto> or whatever you currently have
2026-01-02T02:23:10 < qyx> well, at primary school back in 1995 we were learning about going away from fossil fuels
2026-01-02T02:23:15 < qyx> in a fucking post soviet country
2026-01-02T02:23:21 < qyx> so it is not a surprise it happens
2026-01-02T02:24:17 < specing> antto: the prices of CO2 coupons have been steadily rising for the past decade, there's no use pretending that it happened all of a sudden
2026-01-02T02:24:18 < qyx> and we are not even touching the topic of cars, where any liquid fuels have some reason
2026-01-02T02:24:35 < antto> we were gonna build a 2nd nuclear power years ago, but it was f*cked, and it doesn't look like it's gonna happen, in the mean time, our one and only nuclear power plant is really aging badly
2026-01-02T02:26:18 < antto> specing, what happened all of a sudden is the attempt of our own gubbermint to shut down the coal plant
2026-01-02T02:26:32 < antto> ...using the EU as an excuse
2026-01-02T02:27:00 < specing> if it is making a hole in your government's budget, then a shut down is the only reasonable option
2026-01-02T02:27:36 < specing> I can totally relate to this, we have a new 600 MWe coal plant built in 2014 thatis set to shut down in a few years, and is making several 100M eur loss per year
2026-01-02T02:27:42 < antto> yes, when the budget is being made by an absolute idiot
2026-01-02T02:27:55 < specing> nothing to do with budgetting
2026-01-02T02:28:29 < antto> but our coal plants fit in the picture with a key role, they regulate
2026-01-02T02:28:42 < antto> afaiu, if they get shut down - we will be f*cked
2026-01-02T02:29:22 < qyx> nah worse things happen
2026-01-02T02:29:37 < qyx> no need to worry
2026-01-02T02:30:22 < antto> and the dumbest thing then is.. at best our coal will go to Serbia and they'll make dirty electricity from it and we might get a chance to buy that electricity at some price, but we surely will be sniffing that air because Serbia are right next to us
2026-01-02T02:31:23 < antto> qyx, what can be worse?
2026-01-02T02:31:32 < specing> Serbia will have to clean their air up due to CBAM
2026-01-02T02:31:57 < antto> dafuq
2026-01-02T02:32:08 < qyx> antto: being ingested by russia and then the regulation will not bother you
2026-01-02T02:32:18 < antto> so Serbia can burn our coal and clean the air.... but we cannot burn our coal and clean the air?
2026-01-02T02:32:23 < specing> antto: Cross Border Adjustment Mechanism - carbon tariff
2026-01-02T02:32:46 < specing> burn coal and clean the air? What
2026-01-02T02:33:35 < antto> i know how we're gonna fix all our carbon problems
2026-01-02T02:33:41 < antto> when we all die
2026-01-02T02:33:54 < antto> this is the plan, and we're going down fast!
2026-01-02T02:34:08 < qyx> don't forget to get buried instead of getting burned
2026-01-02T02:34:35 < qyx> we need to replenish those liquid dinosaurs
2026-01-02T02:35:27 < Steffanx> Aquamation is a thing for you when it's time?
2026-01-02T02:35:35 < specing> the market will fix it as the carbon coupon stock is shrinking yearly and polluters will be priced out entirely
2026-01-02T02:35:35 < antto> besides the coal plants, we also have that other thing, it's like a water battery
2026-01-02T02:35:45 < antto> which can do some regulation afaik
2026-01-02T02:36:13 < specing> pumping hydro is mean for regulation
2026-01-02T02:36:30 < specing> meant*
2026-01-02T02:37:01 < antto> it moves a pile of water up when there's excess energy, and later you can let it fall down when you need energy
2026-01-02T02:37:19 < antto> but this one also almost got sabotaged recently
2026-01-02T02:40:31 < antto> anyway..
2026-01-02T02:40:37  * antto vanishes
2026-01-02T02:40:58 < qyx> oh yes I want pumping micro hydro in the future
2026-01-02T02:41:03 < qyx> for multiple purposes
2026-01-02T02:41:26 < qyx> watering plants and fire prevemtion during the summer
2026-01-02T02:41:38 < qyx> and as a energy reservoir during the winter
2026-01-02T02:43:06 < qyx> but for anything useful you need 1000-ish m3 of water and at least 50-100m of elevaton
2026-01-02T02:43:11 < qyx> which is not trivial
2026-01-02T02:44:12 < qyx> and you also need to get good insurance in case you flood everything downstream
2026-01-02T02:48:44 < qyx> oh tomorrow is a workday
2026-01-02T02:54:21 < specing> I guess batteries are nowadays a good alternative to that
2026-01-02T02:54:28 < specing> much much more compact and safe
2026-01-02T02:54:40  * specing still considers it a waste of batteries
2026-01-02T02:57:35 < qyx> it is reasonable only if the rest of the infrastructure is already done or needed for other purpose
2026-01-02T03:43:16 -!- infisc [uid692580@user/infisc] has quit [Quit: Connection closed for inactivity]
2026-01-02T05:24:13 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has joined ##stm32
2026-01-02T07:16:54 -!- jmcgnh [~jmcgnh@wikipedia/jmcgnh] has joined ##stm32
2026-01-02T08:31:33 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-02T09:16:35 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has quit [Ping timeout: 240 seconds]
2026-01-02T10:51:38 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-02T10:52:05 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-02T12:06:12 < catphish> karlp: in my house? yes all 3 phases are brown on the meter tails, they have little tags on with the phase number
2026-01-02T12:07:42 < karlp> grosss :)
2026-01-02T12:08:17 < karlp> I know so much stuff just gets done, but is that the coded way of doing it for a new build? all the same colour wire, just mark them at one end?
2026-01-02T12:09:07 < ventyl> even if they are colored, rules for assigning them to phases vary widely
2026-01-02T12:09:26 < ventyl> like, I am brown-gray-black but I've seen people doing it all three ways
2026-01-02T12:09:47 < catphish> karlp: tps709 seems ideal, a pin compatible 1117 makes sense, but actually i have an extra FET to turn one of them off, so might as well use a part with a shutdown pin
2026-01-02T12:10:29 < catphish> karlp: this was a new install, and the only colours the meter installers carry is brown and blue
2026-01-02T12:11:29 < catphish> and yes, it's completely insane
2026-01-02T12:11:39 < ventyl> instrument cluster behaves as if 10V LDO in it just gave up. instrument readings are way off. for some reason unknown to humanity, Germans flipped IN and OUT pins, so standard 10V LDO doesn't fit mechanically. And they stopped making them.
2026-01-02T12:15:11 < catphish> i'm still confused that the TPS claims to have a 150nA shutdown current but also claims to pull up its shutdown pin with 300nA thus making that 150nA shutdown current impossible, or i'm totally misunderstanding
2026-01-02T12:17:10 < karlp> catphish: oh, yeah, if you have options, use tps709, my page just happens to have some things thare can be subbed if you did a "I just wanted to use an lm1117 and didn't think" 
2026-01-02T12:18:05 < catphish> karlp: if that was the only mistake on my board, i'd probably retrofit from your suggestions, but i am making layout changes anyway
2026-01-02T12:19:44 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-02T12:32:09 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-02T12:39:09 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-02T12:43:35 -!- Livio [~livio@user/livio] has quit [Remote host closed the connection]
2026-01-02T12:55:26 < catphish> sometimes i hate electronics. once i get down into the real heart of battery saving i find unpleasant things, like my input/output protection diodes leak current in the region of another 10uA
2026-01-02T13:00:45 -!- infisd [~infisc@user/infisc] has quit [Ping timeout: 244 seconds]
2026-01-02T13:13:54 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-02T13:42:30 < karlp> yar, under 100uA, shit gets real quick :)
2026-01-02T13:43:38 < Steffanx> I worked on a < 10 uA device. It was fun... Except the occasional "where does the current flow to this time" moments.
2026-01-02T13:44:05 < Steffanx> It included some io expanders and had to wake up on any user input.
2026-01-02T13:45:07 < Steffanx> Best part was the encoder, with detents at "open open" and "closed closed" -_-
2026-01-02T13:45:34 < Steffanx> Which the one designing them into the product didn't thing about of course 
2026-01-02T14:04:57 -!- infisd [~infisc@user/infisc] has joined ##stm32
2026-01-02T14:05:23 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has joined ##stm32
2026-01-02T14:05:37 -!- Linux_Kerio [~Linux_Ker@chello085216213137.chello.sk] has quit [Read error: Connection reset by peer]
2026-01-02T15:12:09 < zyp> my favorite leakage I tracked down was on a nrf52+nrf91 project
2026-01-02T15:12:46 < zyp> nrf91 had the reset pin driven by a nrf52 gpio
2026-01-02T15:13:28 < zyp> nrf91 side has internal pullup, so nrf52 side had no pull
2026-01-02T15:14:15 < zyp> for whatever reason, the internal pullup in the nrf91 is to 2.4V or something
2026-01-02T15:14:56 < zyp> 2.4V puts the input buffer on the nrf52 in the undefined region, and makes it leak
2026-01-02T15:15:11 < zyp> disabling the input buffer fixed the leak
2026-01-02T15:16:48 < catphish> Steffanx: i'm aiming for 10uA, just about to to annotate the schematic with how much everything is supposed to draw
2026-01-02T15:19:36 < catphish> luckily it's a super simple design, always-on TPS709 drived STM32 and BAT54S. then another switchable TPS709 with a 10k pulldown on its EN pin powers an LCD screen that's normally off when the STM32 is in standby
2026-01-02T16:14:05 < qyx> 100 uA is my record when I didn't need to tweak the application explicly for limitations it brings
2026-01-02T16:14:48 < qyx> I just did some freertos sleeping magic, reconfiiguring clocks, etc
2026-01-02T16:22:24 < zyp> the nrf stuff we do seems to achieve sub 10uA without too much effort
2026-01-02T16:39:08 < qyx> I suppse u5/u3 should be able to too
2026-01-02T17:30:13 -!- Livio [~livio@user/livio] has quit [Quit: leaving]
2026-01-02T18:55:49 < catphish> i placed an order with JLCPCB's SLA printing today, will be fun to see how that works out.
2026-01-02T19:03:27 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-02T19:13:09 < Steffanx> It was always perfect for what I used them for
2026-01-02T19:20:01  * lemmi was also happy with the results
2026-01-02T19:26:43 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 246 seconds]
2026-01-02T19:54:00 < jpa-> i helped the kid with scratch and immediately got roasted for code style violation.. "why did you call this variable 'kloonien_maara' instead of 'kloonien määrä'
2026-01-02T19:54:32 < jpa-> i don't think he's old enough to hear about C yet..
2026-01-02T20:06:40 < qyx> don't you think they should allow emojis in c26?
2026-01-02T20:11:26 < jpa-> i thought they were already allowed, or is it just in c++?
2026-01-02T23:27:02 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has joined ##stm32
2026-01-02T23:40:51 -!- Livio_ [~livio@user/livio] has joined ##stm32
2026-01-02T23:44:34 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-02T23:56:06 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has quit [Ping timeout: 265 seconds]
--- Day changed Sat Jan 03 2026
2026-01-03T00:02:03 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has joined ##stm32
2026-01-03T00:18:24 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 256 seconds]
2026-01-03T00:18:37 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 246 seconds]
2026-01-03T00:47:19 -!- Livio_ [~livio@user/livio] has quit [Quit: leaving]
2026-01-03T00:53:12 < catphish> i code in ascii because my keyboard has those buttons on it :)
2026-01-03T01:00:43 < Phantom> you could code in asm, because your keyboard have the letters on the instructions on it
2026-01-03T01:01:02  * qyx gives catphish a keyboard with BASIC commands
2026-01-03T01:01:06 < Phantom> heck, you could even code in binary!
2026-01-03T01:18:15 -!- qyx [~qyx@84.245.121.32] has quit [Ping timeout: 245 seconds]
2026-01-03T01:20:27 -!- qyx [~qyx@84.245.121.219] has joined ##stm32
2026-01-03T01:59:10 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has quit [Quit: Ping timeout (120 seconds)]
2026-01-03T02:41:39 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has joined ##stm32
2026-01-03T03:08:24 -!- infisd [~infisc@user/infisc] has quit [Ping timeout: 260 seconds]
2026-01-03T04:19:01 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-03T04:26:07 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-03T07:11:53 -!- infisd [~infisc@user/infisc] has joined ##stm32
2026-01-03T08:09:50 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-03T10:11:36 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-03T10:13:24 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-03T10:18:57 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-03T10:37:14 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-03T11:00:10 -!- nerozero [~nerozero@87.253.63.54] has quit [Remote host closed the connection]
2026-01-03T11:01:57 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-03T14:02:02 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has quit [Quit: Ping timeout (120 seconds)]
2026-01-03T14:28:48 < karlp> updating my sony headphones for this airoha problem, had to install the app again on this phone, you dont normally need it.
2026-01-03T14:29:28 < karlp> it starts the update automatically, then in the "more info" says "this will take 40 minutes" (!wot!?) and "do it in a low interference area, after fully charging, and after turning off the "auto power off" feature
2026-01-03T14:29:34 < karlp> not... the best experience.
2026-01-03T14:29:41 < karlp> but will probably all work out for the best.
2026-01-03T14:36:42 < karlp> but at least I'm not in venuezela I guess...
2026-01-03T14:54:18 < qyx> arent you
2026-01-03T14:55:56 < karlp> up ot 85%
2026-01-03T15:03:09 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-03T16:15:24 -!- infisc [uid692580@user/infisc] has joined ##stm32
2026-01-03T16:28:35 < jfsimon> Hello
2026-01-03T16:30:23 < jfsimon> I'm working on STM32 H743, wondering the correct pins for USB : usb_dn PA11, usb_dp PA12, usb_sense not sure is PA9 or PC0, or not determined ?
2026-01-03T17:50:12 -!- octorian [octo@2600:3c00::f03c:91ff:fe93:a61c] has quit [Quit: ZNC - http://znc.sourceforge.net]
2026-01-03T17:58:41 < englishman> haha the sony app is terrible not to mention it has changed name like 3 times so i never know what to look for
2026-01-03T18:18:18 -!- octorian [~octo@chroniton.logicprobe.org] has joined ##stm32
2026-01-03T18:35:02 -!- infisc [uid692580@user/infisc] has quit [Quit: Connection closed for inactivity]
2026-01-03T18:56:26 < Steffanx> Ha yeah
2026-01-03T18:56:31 < Steffanx> Same issue here
2026-01-03T18:56:49 < Steffanx> You can search for Sony though
2026-01-03T19:53:47 < mawk> you mean VDD50USB jfsimon ?
2026-01-03T19:54:08 < mawk> or OTG_FS_VBUS?
2026-01-03T19:54:17 < mawk> this one is PA9
2026-01-03T19:55:04 < mawk> and PB13 for OTG_HS_VBUS
2026-01-03T19:59:20 < mawk> it's all in the datasheet, but it's a bit hard to search because of the line breaks
2026-01-03T20:33:25 < jfsimon> mawk Thank you
2026-01-03T20:33:50 < jfsimon> I can't really fine PA9 OTG_FS_VBUS from the manual, i guess i can't properly find/read. I give a new try?
2026-01-03T20:34:23 < mawk> page 74 of the datasheet
2026-01-03T20:34:35 < mawk> DS12110 
2026-01-03T20:35:17 < mawk> and page 79
2026-01-03T20:35:52 < jfsimon> Thanks maxk
2026-01-03T20:35:55 < jfsimon> mawk
2026-01-03T20:36:48 < mawk> are your headphones dead yet karlp 
2026-01-03T20:38:02 < jfsimon> Got it the line breaks make search indeed uneasy
2026-01-03T21:50:51 < karlp> mawk:no, update finished, took about 40 minutes though, which seems wildly long, but ok.
2026-01-03T21:51:36 < karlp> wanted to fix this: https://insinuator.net/2025/12/bluetooth-headphone-jacking-full-disclosure-of-airoha-race-vulnerabilities/
2026-01-03T22:10:49 -!- Livio_ [~livio@user/livio] has joined ##stm32
2026-01-03T22:13:32 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-03T22:21:39 -!- infisc [uid692580@user/infisc] has joined ##stm32
2026-01-03T22:36:18 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 252 seconds]
2026-01-03T22:37:11 -!- infisd [~infisc@user/infisc] has quit [Read error: Connection reset by peer]
2026-01-03T22:44:38 < mawk> ah that must be why my headphones suddenly have an update too
2026-01-03T22:44:45 < karlp> very likely.
2026-01-03T22:45:14 < karlp> airoha apparently provided the vendors all an update 6 months ago, but sony at least didn't do anything until the 39c3 guys pushed it a bit harder...
2026-01-03T22:45:21 < karlp> jabra apparently fixed it months ago.
2026-01-03T22:46:24 < mawk> >In the course of analyzing Airoha-based headphones, we discovered that many devices accept Bluetooth Low Energy and Bluetooth Classic connections without enforcing authentication.
2026-01-03T22:46:38 < mawk> sounds like they shipped development builds
2026-01-03T22:47:14 < mawk> good thing that I dropped my headphones enough times that the mic is dead
2026-01-03T22:47:39 < karlp> well, if they have the link code, they can just connect their own device to your phone, it's not just about listening to you.
2026-01-03T22:48:04 < mawk> ah right
2026-01-03T22:48:18 < mawk> that sounds bad indeed
2026-01-03T22:48:48 < karlp> particularly easy, is connect as hands free profile, and just start dialing pay numbers...
2026-01-03T22:49:19 < karlp> I wonder if there are any legitimate uses for premium numbers...
2026-01-03T22:50:31 < Steffanx> Time to update my headphones 
2026-01-03T22:51:52 < Steffanx> If there's any update.
2026-01-03T23:31:58 < machinehum> Damn
2026-01-03T23:32:02 < machinehum> That's nasty
--- Day changed Sun Jan 04 2026
2026-01-04T00:03:10 -!- Livio_ [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-04T00:04:03 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-04T00:11:07 -!- haritz [~hrtz@2a01:4b00:bc2e:7000:d5af:a266:ca31:5ef8] has joined ##stm32
2026-01-04T00:11:07 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-04T00:11:33 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-04T00:23:30 < Steffanx> Last update is from October last year..
2026-01-04T00:31:24 -!- infisc [uid692580@user/infisc] has quit [Quit: Connection closed for inactivity]
2026-01-04T00:32:52 < mawk> check if it's airoha chipset
2026-01-04T00:33:59 < ventyl> hm so my junky jbls are quite probably affected as well
2026-01-04T00:38:59 < ventyl> one more star and creepyOS will surpass FreePascal OS on GitHub :>
2026-01-04T00:56:10 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 245 seconds]
2026-01-04T00:57:22 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 246 seconds]
2026-01-04T01:05:51 -!- vampirefrog [~vampirefr@5.12.7.155] has quit [Quit: The Lounge - https://thelounge.chat]
2026-01-04T01:06:06 -!- vampirefrog [~vampirefr@5.12.7.155] has joined ##stm32
2026-01-04T01:24:50 -!- vampirefrog [~vampirefr@5.12.7.155] has quit [Quit: The Lounge - https://thelounge.chat]
2026-01-04T01:25:06 -!- vampirefrog [~vampirefr@5.12.7.155] has joined ##stm32
2026-01-04T01:25:36 < karlp> what's it called again? 
2026-01-04T01:25:39 < karlp> cqrx?
2026-01-04T01:25:58 < karlp> cos I don't think this is you: https://github.com/creepy-pirate/creepyOS
2026-01-04T01:27:56 < karlp> cmrx, there you go, one more star Þ:)
2026-01-04T01:28:22 < karlp> oooh, you have a .org page and everything now :)
2026-01-04T01:28:30 < karlp> sexayyy https://cmrxrtos.org/
2026-01-04T01:40:52 < ventyl> yeah
2026-01-04T01:43:12 < ventyl> its shit ton of work for 1/2 of developer
2026-01-04T02:13:19 < ventyl> karlp: thx
2026-01-04T02:51:34 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has joined ##stm32
2026-01-04T02:51:51 < nomorekaki> hello night
2026-01-04T03:03:37 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-04T05:31:00 -!- hexo [~hexo@user/hexo] has quit [Ping timeout: 252 seconds]
2026-01-04T06:26:03 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-04T06:33:17 -!- hexo [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-04T06:58:02 -!- haritz [~hrtz@user/haritz] has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
2026-01-04T07:28:54 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-04T07:34:01 -!- hexo [~hexo@user/hexo] has quit [Ping timeout: 244 seconds]
2026-01-04T08:15:58 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-04T09:24:32 -!- Kerr [~quassel@h159-112-20-235.spkqwa.broadband.dynamic.tds.net] has joined ##stm32
2026-01-04T09:54:24 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-04T10:00:14 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-04T10:48:32 -!- Kerr [~quassel@h159-112-20-235.spkqwa.broadband.dynamic.tds.net] has quit [Quit: Bye.]
2026-01-04T11:08:07 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 246 seconds]
2026-01-04T11:29:07 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 246 seconds]
2026-01-04T12:01:43 < antto> nomorekaki, have you watched The Room? ;P~
2026-01-04T12:56:09 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-04T13:29:06 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-04T13:41:32 -!- nerozero [~nerozero@87.253.63.54] has quit [Remote host closed the connection]
2026-01-04T13:43:23 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-04T13:52:24 < nomorekaki> maybe
2026-01-04T14:04:12 < antto> it's about a room that materializes wishes ;P~
2026-01-04T14:18:03 < nomorekaki> freaky friday was fun comedy
2026-01-04T14:18:45 < nomorekaki> i laughed and farted hard
2026-01-04T14:19:28 < nomorekaki> checks both boxes
2026-01-04T14:26:13 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has quit [Quit: Client closed]
2026-01-04T14:37:17 -!- infisc [uid692580@user/infisc] has joined ##stm32
2026-01-04T14:47:47 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-04T14:51:19 -!- ds2 [~ds2@user/ds2] has quit [Ping timeout: 264 seconds]
2026-01-04T14:53:53 -!- haritz [~hrtz@2a01:4b00:bc2e:7000:d5af:a266:ca31:5ef8] has joined ##stm32
2026-01-04T14:53:53 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-04T16:19:36 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-04T18:11:02 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has joined ##stm32
2026-01-04T18:11:41 < nomorekaki> chat killer
2026-01-04T18:26:58 -!- infisc [uid692580@user/infisc] has quit [Quit: Connection closed for inactivity]
2026-01-04T18:41:38 < Steffanx> Yeah we need more details. Did you keep the underpants clean, nomorekaki ?
2026-01-04T18:41:59 < nomorekaki> not certain
2026-01-04T18:42:41 < nomorekaki> any JBC pros here?
2026-01-04T18:43:16 < nomorekaki> DIP8 desoldering with one move. Any soldering tip that would do that?
2026-01-04T18:45:01 < BrainDamage> a sufficiently sized hot air nozzle would give you similar results
2026-01-04T18:46:39 < nomorekaki> I try to minimize hot air
2026-01-04T18:48:50 < qyx> what about the finnish cold air then
2026-01-04T18:50:17 < nomorekaki> maybe solder pot
2026-01-04T18:50:35 < nomorekaki> would be easy dip8 desoldering tool
2026-01-04T19:16:51 < jpa-> isn't that what C245-305 is meant for?
2026-01-04T19:18:42 < nomorekaki> it does DIP8?
2026-01-04T19:19:27 < jpa-> well the dimensions seem close enough
2026-01-04T19:20:42 < nomorekaki> what JBC setup you have jpa-?
2026-01-04T19:20:48 < nomorekaki> also how many tips?
2026-01-04T19:21:06 < nomorekaki> or how many euros in tips rather
2026-01-04T19:24:23 < jpa-> maybe 200 eur in tips, just ordered T210 (already have T245)
2026-01-04T19:24:31 < Steffanx> AliExpress the tips :)
2026-01-04T19:24:46 < nomorekaki> oh not that many euros in tips.. thats like 4 pcs
2026-01-04T19:25:03 < Steffanx> AliExpress 
2026-01-04T19:25:13 < nomorekaki> maybe
2026-01-04T19:25:20 < jpa-> i have 7 original tips and two aliexpress tips, can't really tell the difference
2026-01-04T19:25:20 < nomorekaki> genuine?
2026-01-04T19:25:31 < nomorekaki> oh
2026-01-04T19:26:45 < jpa-> eleshop has the best prices for genuine ones, but even there the prices have gone up
2026-01-04T19:27:55 < nomorekaki> I tried to order first my business debit card limit exceeded then I needed to call bank and then "payment didn't process" and now I'm back to second thoughts
2026-01-04T19:31:19 < jpa-> how much tips were you ordering if it exceeds your card limit? :D
2026-01-04T19:31:52 < nomorekaki> tips + 4in1 station 4keur
2026-01-04T19:32:03 < jpa-> maybe ask the bank for a loan, maybe they'll accept the tips as a collateral
2026-01-04T19:32:09 < jpa-> pricey
2026-01-04T19:32:23 < jpa-> you must be soldering a lot
2026-01-04T19:32:33 < nomorekaki> there was special tips that almost exceed value of my current setups
2026-01-04T19:32:47 < jpa-> i consider even the 200 eur JBC station i have excessive
2026-01-04T19:33:19 < nomorekaki> I don't need both 210 and 245 but rather safe than sorry
2026-01-04T19:34:09 < jpa-> i'll just switch the handle when i need to
2026-01-04T19:35:08 < nomorekaki> I wonder if for 200eur JBC would even look in your direction.. they don't have those prices anymore
2026-01-04T19:35:23 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-04T19:35:33 < nomorekaki> I guess most basic station is between 500-1000
2026-01-04T19:38:29 < jpa-> yeah, looks like they don't sell bt-2bwa anymore, lcb is 299 eur
2026-01-04T19:41:36 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-04T19:44:28 < nomorekaki> I could optimize 210 out but it's still the same station and I lose tweezers too
2026-01-04T19:45:57 < jpa-> i use sequre ht140 for tweezers, it takes the 210 tips
2026-01-04T19:47:54 < nomorekaki> DMSE is 700eur less than DMPSE by taking away 210 and tweezers and all related accessories
2026-01-04T19:48:48 < nomorekaki> I don't even need to look that tweezers cost 500eur or something if bought separatelly
2026-01-04T19:50:29 < nomorekaki> 360eur
2026-01-04T19:50:58 < jpa-> lcb station 299e comes with t245, t210 separately 66e, ht140 for tweezers 80e
2026-01-04T19:52:24 < antto> for multi-legged THT desoldering, an option is to take a chunk of fat copper wire, bend it to a shape which passes thru all the legs, then solder it onto each joint, and then heat the whole copper thing and allow the component to fall by itself ideally
2026-01-04T19:52:47 < antto> after that, you clean the PCB
2026-01-04T19:52:50 < jpa-> yeah, that works
2026-01-04T19:53:05 < nomorekaki> na JBC must have a tip for that
2026-01-04T19:53:07 < antto> i've had to desolder a ton of wrong voltage SPDT relays (6 pins)
2026-01-04T19:54:10 < antto> nomorekaki, send them the datasheet of the component you wanna desolder and they will be happy to design a suitable tool for you? ;P~
2026-01-04T19:55:15 < antto> hm, i've also desoldered a pile of RJ-45 jacks with LEDs and magnetics inside
2026-01-04T19:55:28 < antto> hotair f*cks up their plastic legs
2026-01-04T19:55:41 < antto> the thick copper wire technique works much better
2026-01-04T20:02:07 < nomorekaki> antto: I know but I assume they have something for DIP8 already
2026-01-04T20:03:33 < nomorekaki> jpa-: if tweezers are using t210 tips then it lacks all the special tweezer tips
2026-01-04T20:03:33 < jpa-> well what is wrong with the tip i already said.. it's not like you need 0.1mm precision for the dimensions
2026-01-04T20:03:46 < nomorekaki> jpa-: nothing
2026-01-04T20:03:55 < nomorekaki> I will try
2026-01-04T20:04:17 < jpa-> what's so special about special tweezer tips, except extra cost?
2026-01-04T20:05:31 < nomorekaki> usually shape to remove various components
2026-01-04T20:06:25 < nomorekaki> probably there would be t210 tips that could emulate similar operation somewhat
2026-01-04T20:07:21 < jpa-> you're in some kind of repair business now?
2026-01-04T20:08:30 < nomorekaki> been a while
2026-01-04T20:09:19 < jpa-> yeah, i guess high end station can make sense then
2026-01-04T20:09:27 < jpa-> i'm just doing protos
2026-01-04T20:10:58 < nomorekaki> it's doesn't only make sense but it's necessary to have optimized tools
2026-01-04T20:11:16 < nomorekaki> or just grunt with unoptimized tools but that doesn't go far
2026-01-04T20:16:31 < nomorekaki> I can double down because it's always the same parts or similar packages
2026-01-04T20:22:20 -!- fenugrec [~f@192.214.232.39] has joined ##stm32
2026-01-04T20:38:50 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-04T21:05:47 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has quit [Ping timeout: 272 seconds]
2026-01-04T21:44:35 < antto> fiximir
2026-01-04T22:05:18 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has joined ##stm32
2026-01-04T23:07:13 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 264 seconds]
2026-01-04T23:20:54 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
--- Day changed Mon Jan 05 2026
2026-01-05T01:15:09 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 252 seconds]
2026-01-05T01:18:55 -!- qyx [~qyx@84.245.121.219] has quit [Ping timeout: 264 seconds]
2026-01-05T01:20:20 -!- qyx [~qyx@84.245.120.41] has joined ##stm32
2026-01-05T01:30:00 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 252 seconds]
2026-01-05T05:40:07 -!- jmcgnh [~jmcgnh@wikipedia/jmcgnh] has quit [Ping timeout: 240 seconds]
2026-01-05T05:40:07 -!- BrainDamage [~m-t6k752@user/BrainDamage] has quit [Ping timeout: 240 seconds]
2026-01-05T05:40:35 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has quit [Ping timeout: 240 seconds]
2026-01-05T05:41:27 -!- BrainDamage [~m-t6k752@user/BrainDamage] has joined ##stm32
2026-01-05T05:41:55 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has joined ##stm32
2026-01-05T05:46:02 -!- rkta_ [~rkta@user/rkta] has joined ##stm32
2026-01-05T05:47:16 -!- akaWolf1 [~akaWolf@akawolf.org] has joined ##stm32
2026-01-05T05:47:22 -!- tom17 [~tom@user/tomeaton17] has joined ##stm32
2026-01-05T05:48:01 -!- umbramal- [~umbramali@188.74.90.42] has joined ##stm32
2026-01-05T05:50:39 -!- akaWolf [~akaWolf@akawolf.org] has quit [Ping timeout: 246 seconds]
2026-01-05T05:50:39 -!- rkta [~rkta@user/rkta] has quit [Ping timeout: 246 seconds]
2026-01-05T05:50:39 -!- umbramalison [~umbramali@188.74.90.42] has quit [Ping timeout: 246 seconds]
2026-01-05T05:50:39 -!- akaWolf1 is now known as akaWolf
2026-01-05T05:50:40 -!- tomeaton17 [~tom@user/tomeaton17] has quit [Ping timeout: 246 seconds]
2026-01-05T06:03:08 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has quit [Ping timeout: 240 seconds]
2026-01-05T06:03:34 -!- Netsplit *.net <-> *.split quits: karlp, mid-kid
2026-01-05T06:04:38 -!- mid-kid [~mid-kid@2a01:7c8:aac8:1e8:5054:ff:fe5e:cd48] has joined ##stm32
2026-01-05T06:04:44 -!- Netsplit over, joins: karlp
2026-01-05T06:11:16 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has joined ##stm32
2026-01-05T06:59:39 -!- ds2 [~ds2@user/ds2] has joined ##stm32
2026-01-05T07:30:25 -!- haritz [~hrtz@user/haritz] has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
2026-01-05T07:37:08 -!- cutofmyjib [~cutofmyji@user/cutofmyjib] has joined ##stm32
2026-01-05T07:40:24 -!- cutofmyjib [~cutofmyji@user/cutofmyjib] has quit [Client Quit]
2026-01-05T07:44:43 -!- cutofmyjib [~cutofmyji@user/cutofmyjib] has joined ##stm32
2026-01-05T07:46:09 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has quit [Quit: Ping timeout (120 seconds)]
2026-01-05T07:55:59 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-05T09:01:55 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 250 seconds]
2026-01-05T09:03:29 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-05T09:47:05 -!- rkta_ is now known as rkta
2026-01-05T09:49:45 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-05T09:51:18 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-05T09:53:35 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 240 seconds]
2026-01-05T10:12:49 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-05T11:41:52 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-05T11:48:32 -!- tom17 is now known as tomeaton17
2026-01-05T12:49:09 -!- haritz [~hrtz@2a01:4b00:bc2e:7000:d5af:a266:ca31:5ef8] has joined ##stm32
2026-01-05T12:49:09 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-05T15:24:52 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-05T16:07:23 -!- PhantomWork [~PhantomWo@user/phantom] has quit [Quit: Leaving]
2026-01-05T16:38:16 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-05T16:42:43 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-05T16:44:47 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-05T16:45:27 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-05T17:05:24 -!- HelloShitty [~psysc0rpi@bl18-184-247.dsl.telepac.pt] has quit [Ping timeout: 260 seconds]
2026-01-05T17:12:31 -!- HelloShitty [~psysc0rpi@bl18-184-247.dsl.telepac.pt] has joined ##stm32
2026-01-05T17:33:38 < mawk> hmm
2026-01-05T17:33:57 < mawk> for my glitchy switch I wonder if I can use the capacitance of the traces as a switched capacitor
2026-01-05T17:34:10 < mawk> to reach the wetting current and burn off the oxidation, in a safeish way
2026-01-05T17:40:04 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-05T17:43:20 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-05T18:26:19 -!- fenugrec [~f@192.214.232.39] has left ##stm32 []
2026-01-05T19:31:53 -!- goodvibrations32 [~user@user/goodvibrations32] has joined ##stm32
2026-01-05T20:28:41 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-05T21:22:53 -!- hexo__ [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-05T21:23:23 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-05T21:23:45 < jpa-> https://electronics.stackexchange.com/questions/764011/ldo-dropout-i-am-confused-after-discussing-with-onsemi-ae   heh, nice LDO trap of boasting 100mV dropout but rest of the specs being for 1V drop
2026-01-05T22:54:54 < catphish> useful lesson
2026-01-05T22:55:23 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-05T22:56:40 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-05T23:03:24 -!- Livio_ [~livio@user/livio] has joined ##stm32
2026-01-05T23:06:52 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-05T23:07:16 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 246 seconds]
2026-01-05T23:12:44 -!- Livio_ [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-05T23:31:14 -!- lockna [~lockna@193-81-168-132.hdsl.highway.telekom.at] has joined ##stm32
2026-01-05T23:34:51 -!- lockna_ [~lockna@193-81-168-132.hdsl.highway.telekom.at] has joined ##stm32
2026-01-05T23:35:08 -!- lockna [~lockna@193-81-168-132.hdsl.highway.telekom.at] has quit [Read error: Connection reset by peer]
2026-01-05T23:38:37 -!- lockna_ [~lockna@193-81-168-132.hdsl.highway.telekom.at] has quit [Read error: Connection reset by peer]
2026-01-05T23:38:55 -!- lockna_ [~lockna@193-81-168-132.hdsl.highway.telekom.at] has joined ##stm32
2026-01-05T23:44:43 -!- lockna_ [~lockna@193-81-168-132.hdsl.highway.telekom.at] has quit [Ping timeout: 246 seconds]
--- Day changed Tue Jan 06 2026
2026-01-06T00:13:44 < karlp> that's like some of teh Iq measurements not when idle, but when in unrealistic non-switch conditions that qyx and I have run into with diodes parts.
2026-01-06T00:15:22 < qyx> yeah with FB overdriven and switcher not switching
2026-01-06T01:42:28 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 265 seconds]
2026-01-06T01:42:35 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
2026-01-06T04:55:20 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Ping timeout: 240 seconds]
2026-01-06T04:56:03 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T05:01:23 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Max SendQ exceeded]
2026-01-06T05:03:28 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T05:33:51 -!- hexo__ [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-06T05:37:13 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-06T06:17:56 -!- haritz [~hrtz@user/haritz] has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
2026-01-06T07:28:08 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-06T08:44:48 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-06T08:48:01 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 240 seconds]
2026-01-06T08:48:01 -!- ds2 [~ds2@user/ds2] has quit [Ping timeout: 240 seconds]
2026-01-06T08:48:06 -!- ds2 [~ds2@69.54.141.155] has joined ##stm32
2026-01-06T08:48:15 -!- HelloShitty [~psysc0rpi@bl18-184-247.dsl.telepac.pt] has quit [Ping timeout: 240 seconds]
2026-01-06T08:49:23 -!- HelloShitty [~psysc0rpi@bl18-184-247.dsl.telepac.pt] has joined ##stm32
2026-01-06T08:51:51 -!- Netsplit *.net <-> *.split quits: ds2
2026-01-06T08:57:53 -!- Netsplit over, joins: ds2
2026-01-06T08:59:30 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-06T08:59:54 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T09:00:11 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-06T09:00:53 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T09:08:53 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-06T10:07:17 -!- Livio [~livio@user/livio] has quit [Remote host closed the connection]
2026-01-06T10:09:45 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-06T10:10:18 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-06T10:48:01 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-06T10:53:14 -!- Livio [~livio@user/livio] has quit [Quit: leaving]
2026-01-06T11:49:13 < mawk> jpa-: >@Confused This specific datasheet seems to have been written by some particularly evil marketing person though... "Very Low Dropout: 80 mV at 250 mA". Okay so we check figure 15. Yes indeed we do actually get 80mV @ 250mA but only at a junction temperature of -40°C... I'd be pretty pissed too if I had picked this part only reading page 1 of the datasheet.
2026-01-06T12:17:58 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-06T12:19:11 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T15:27:28 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-06T15:27:45 < bitmask> OI
2026-01-06T15:27:50 < bitmask> oi oi oi
2026-01-06T15:31:00 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Client Quit]
2026-01-06T15:31:55 < tomeaton17> moin
2026-01-06T15:32:34 < tomeaton17> I have returned from austria
2026-01-06T15:36:53 < nohit> welcome back
2026-01-06T16:43:20 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-06T16:49:40 -!- haritz [~hrtz@140.228.70.141] has joined ##stm32
2026-01-06T16:49:40 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-06T17:07:25 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T18:43:34 < qyx> what were you doing there
2026-01-06T18:48:36 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-06T18:49:06 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Client Quit]
2026-01-06T19:30:23 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-06T19:59:46 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-06T20:02:40 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-06T20:04:11 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-06T20:08:46 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T20:11:50 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-06T20:12:30 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-06T21:00:04 < qyx> low power pros, have you ever seen a low power segment LCD driver as a dedicated and *encapsulated* chip?
2026-01-06T21:04:12 < qyx> oh there is a whole category on mouser
2026-01-06T21:04:14 < qyx> never seen that
2026-01-06T21:05:41 < qyx> PCF8551 looks good
2026-01-06T21:08:19 < jpa-> "Introduction to soldering"  well that is a comprehensive datasheet
2026-01-06T21:08:58 < qyx> lol
2026-01-06T21:09:18 < qyx> sub 5 uA consumption including Vlcd
2026-01-06T21:09:30 < qyx> the question is how to generate the Vlcd
2026-01-06T21:09:42 < qyx> maybe 3 V is enough from a CD2032
2026-01-06T21:09:45 < qyx> *CR2032
2026-01-06T21:18:56 < mawk> like on the flex pcb qyx ?
2026-01-06T21:19:03 < mawk> or you want a DIP style screen
2026-01-06T21:22:49 < aandrew> huh I never saw something like PCF8551 before, nifty
2026-01-06T21:22:56 < aandrew> I've never driven a bare LCD before either mind you
2026-01-06T21:27:39 < qyx> I did using stm32l053
2026-01-06T21:28:11 < qyx> but either I did it wrong or the peripheral is power hungry, it was drawing 100 uA
2026-01-06T21:29:12 < qyx> either way I want to use something more recent, U*, and putting it in stop1/2 or so
2026-01-06T21:29:18 < qyx> but I want the LCD to keep going
2026-01-06T21:30:05 -!- artok [~azo@91-153-163-37.elisa-laajakaista.fi] has quit [Quit: Lost terminal]
2026-01-06T21:30:22 < qyx> mawk: I have some DIP 11 segment ones and also bare glass using zebra bars
2026-01-06T21:30:52 < qyx> *strips
2026-01-06T21:56:12 < aandrew> U*?
2026-01-06T21:56:21 < aandrew> microcontroller agnostic?
2026-01-06T21:57:57 < qyx> no, stm32u0/u3/u5
2026-01-06T21:58:07 < aandrew> ah, u* == stm32u*, gotcha
2026-01-06T22:12:13 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 246 seconds]
2026-01-06T22:30:50 -!- PhantomWork [~PhantomWo@user/phantom] has joined ##stm32
2026-01-06T22:32:09 -!- Livio [~livio@user/livio] has quit [Quit: leaving]
2026-01-06T22:51:44 -!- System_Error [~SystemErr@user/systemerror] has quit [Ping timeout: 252 seconds]
2026-01-06T22:59:12 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-06T23:56:07 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 264 seconds]
2026-01-06T23:56:10 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 246 seconds]
2026-01-06T23:57:19 -!- om3ga [~om3ga@93.177.187.134] has quit [Ping timeout: 240 seconds]
--- Day changed Wed Jan 07 2026
2026-01-07T00:22:14 -!- om3ga [~om3ga@93.177.187.134] has joined ##stm32
2026-01-07T00:57:12  * Phantom stills don't like pulsing a tl431
2026-01-07T00:57:19 < Phantom> it feels so wrong
2026-01-07T01:02:56 -!- artok [~azo@91-153-163-37.elisa-laajakaista.fi] has joined ##stm32
2026-01-07T01:14:12  * antto puts a logic inverter on Phantom's feelings
2026-01-07T01:18:07 -!- qyx [~qyx@84.245.120.41] has quit [Ping timeout: 240 seconds]
2026-01-07T01:20:16 -!- qyx [~qyx@84.245.121.152] has joined ##stm32
2026-01-07T02:13:10 -!- fenugrec [~f@192.214.232.39] has joined ##stm32
2026-01-07T03:39:05 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-07T03:40:05 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-07T04:41:57 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-07T05:06:55 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-07T05:11:55 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-07T05:12:39 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-07T06:08:31 -!- dobson [~dobson@static.38.6.217.95.clients.your-server.de] has quit [Ping timeout: 240 seconds]
2026-01-07T06:16:04 -!- dobson [~dobson@static.38.6.217.95.clients.your-server.de] has joined ##stm32
2026-01-07T07:16:48 -!- haritz [~hrtz@user/haritz] has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
2026-01-07T08:27:34 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-07T10:03:38 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-07T10:05:05 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-07T11:39:25 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-07T11:43:57 -!- Livio [~livio@user/livio] has quit [Remote host closed the connection]
2026-01-07T11:51:59 -!- haritz [~hrtz@140.228.70.141] has joined ##stm32
2026-01-07T11:51:59 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-07T12:33:57 < karlp> I think I'm finally getting the hang of hid report descriptors.
2026-01-07T12:34:42 < karlp> I was completely failing, but it turned out not so much, I was just using an enum for report ids, and I "forgot" (didn't know) that report id 0 is invalid, but I kept thinking my descriptor structure was wrong, not that the first report id was simply invalid :)
2026-01-07T12:36:17 < karlp> https://pypi.org/project/hid-parser/ was super useful in the end for helping convince myself that descriptors were correct and useful for the consumer.
2026-01-07T12:36:48 < karlp> like, actually providing enough information to a user about how to interpret your reports, rather than just "lol, here's an opaque set of fields"
2026-01-07T12:38:28 < karlp> https://paste.jvnv.net/view/iifZP
2026-01-07T12:43:27 < jpa-> how about hidviz?
2026-01-07T12:43:34 < ventyl> do I want to deal with this?
2026-01-07T12:44:01 < ventyl> MS OS descriptors were already weird enough
2026-01-07T12:46:03 < karlp> ventyl: are you making hid devices? perhaps. otherwise, no.
2026-01-07T12:46:19 < karlp> also, I've been trying to use the hid scale usage page, which has ~zero public references
2026-01-07T12:47:04 < karlp> I've found a python script for reading a dymo scale, which now that I know how it all works, I can see is almost definitely using the hid scale usage pages, but I only have this scrip that reads the weight, not a copy of it's original descriptors.
2026-01-07T12:47:36 < karlp> jpa-: never even found hidviz. it's google juice is poor :|
2026-01-07T12:47:48 < karlp> let's try it out..
2026-01-07T12:51:41 < ventyl> karlp: i will probably want (not necessarily have) to deal with cdc ecm/rndis
2026-01-07T12:52:39 < karlp> then you don't need to know anything about this :)
2026-01-07T12:52:46 < ventyl> CMRX biggest problem right now is that it has 0 examples that you can build and flash and try out
2026-01-07T12:53:05 < karlp> hidviz looks sorta ok I guess.  it seems to interfere with either jlink or my device when it starts though, which is ... odd.
2026-01-07T12:53:29 < ventyl> and I need something low barrier, CDC-ECM could work on a bunch of controllers and could serve as a basis for simple demo
2026-01-07T12:54:01 < ventyl> although I am not sure how relevant a "here's telnet to something that resembles '90s Unix" is as a demo for embedded RTOS
2026-01-07T13:06:15 < jpa-> IIRC the operating system support for ECM was spotty, rndis or ncm would be better
2026-01-07T13:07:01 < ventyl> internets state that if device knows both ECM and RNDIS, you should cover almost everything
2026-01-07T13:07:32 < ventyl> tinyusb has *some* example for this combo, the question is if it will work
2026-01-07T13:08:42 < jpa-> oh, and linux is dropping rndis
2026-01-07T13:09:32 < jpa-> i think cdc-ncm should work on windows10+, mac os and linux
2026-01-07T13:10:43 < ventyl> well, nobody should care about anything older than that anyway, right?
2026-01-07T13:10:52 < jpa-> yeah
2026-01-07T13:11:05 < ventyl> still the question is how much such demo will be usable and convincing
2026-01-07T13:11:13 < ventyl> hm... in the age of IoT, maybe it will
2026-01-07T13:11:28 < jpa-> user's don't know how to telnet anymore, but http will be ok
2026-01-07T13:11:41 < jpa-> just pick some private-use ipv6 address and advertise a route to it
2026-01-07T13:12:26 < ventyl> I just got a stupid idea
2026-01-07T13:13:15 < ventyl> jpa-: does any embedded IP stack support ipv6
2026-01-07T13:13:16 < ventyl> ?
2026-01-07T13:41:34 < qyx> all?
2026-01-07T13:41:57 < ventyl> hm, right
2026-01-07T13:42:13 < ventyl> i've been living under stone for years
2026-01-07T13:42:31 < qyx> if I wanted to do a custom stack, it would be ipv6/udp/dtls only
2026-01-07T13:44:05 -!- nerozero [~nerozero@87.253.63.54] has quit [Remote host closed the connection]
2026-01-07T13:45:59 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-07T14:23:04 < jpa-> a decade ago when i was playing with rndis on stm32f042 i basically made a custom tiny ipv6/tcp stack https://github.com/devEmbedded/daq4-sw/blob/master/src/tcpip.c   pretty simple because no need to deal with dropped packets when usb has retry on lower level
2026-01-07T15:24:01 -!- fenugrec [~f@192.214.232.39] has left ##stm32 []
2026-01-07T15:58:57 < ventyl> pros... what do you use to debug H5? any tool I try seems half-working or not supporting H5?
2026-01-07T16:04:03 < ventyl> nah, PEBKAC
2026-01-07T16:04:17 < ventyl> of cource CFSR reads 0 if your are not in HardFault
2026-01-07T17:38:38 < zyp> I'm gonna answer you anyway, I use blackmagic and it works well
2026-01-07T17:53:14 < ventyl> somehow BMDA made stlink on nucleo sad
2026-01-07T17:53:45 < ventyl> pyocd has this weird behavior that it understand gdb `run` command as reset and nuke the connection
2026-01-07T17:53:50 < ventyl> openocd doesn't support h5
2026-01-07T17:54:00 < ventyl> and stm's fork of openocd will fail to build
2026-01-07T17:54:30 < ventyl> pretty sad situation considering that H5 is available for ~3 years now
2026-01-07T17:59:27 < zyp> BMDA with orbtrace works well for me
2026-01-07T18:00:57 < zyp> you could also try probe-rs
2026-01-07T18:03:33 < ventyl> tell me honestly: do you consider random developer having to trial and error tools... normal?
2026-01-07T18:08:37 < zyp> I consider running into shit that doesn't work properly normal, and then it's up to you whether you want to deal with it yourself or have a paid support thing you can complain to
2026-01-07T18:24:19 < ventyl> then no wonder that embedded industry looks like loosely put pile of shit always showing problems
2026-01-07T18:25:06 < zyp> I think this holds true in general, not just for embedded stuff
2026-01-07T18:27:29 < zyp> the issue with embedded debug tooling is that there's a ton of vendors with a ton of parts to support, and beyond differences in flash controller, they can all have various quirks that needs to be handled specially
2026-01-07T18:28:24 < zyp> and for the open source stuff, it relies on volunteers writing code to deal with it
2026-01-07T18:31:00 < zyp> personally I prefer blackmagic, so for anything I'm gonna do significant work on and is not already supported by blackmagic, I'm inclined to add support
2026-01-07T18:31:57 < zyp> but I also run into a bunch of weird stuff I'm just gonna have a simple poke on, e.g. I had to flash a .hex into some hc32 a few weeks ago
2026-01-07T18:32:15 < zyp> and pyocd already had support for that, so I just went with that
2026-01-07T18:33:26 < zyp> fun fact: the part I was dealing with did not have official jlink support either, but the vendor bundled a third party jlink support package along with the .hex
2026-01-07T18:33:43 < zyp> pyocd was less hassle than jlink though :p
2026-01-07T18:46:57 < mawk> why would it fail to build ventyl 
2026-01-07T18:47:21 < mawk> it builds just fine here
2026-01-07T18:49:30 < ventyl> way too new GCC I guess, it generates warning older GCC don'
2026-01-07T18:49:37 < ventyl> t and code has -Werror set
2026-01-07T18:50:02 < karlp> ventyl: st's fork was not getting synced until... yesterday? or something?
2026-01-07T18:50:04 < ventyl> zyp: I've been working in many industries and embedded is far more littered with shit
2026-01-07T18:50:14 < karlp> so ifyour clone of st's fork was older than that, that's why it' was garbge.
2026-01-07T18:50:25 < ventyl> I synced it like 30 minutes ago
2026-01-07T18:50:29 < karlp> ok, not that then :)
2026-01-07T18:50:40 < karlp> autofail's -Werror is always fucked though
2026-01-07T18:50:47 < ventyl> i'll go flip sausages
2026-01-07T18:51:19 < karlp> so. "last" hid pondering, I want to be able tpoll current state, via a feature report, in case there've been no events to trigger an input event.
2026-01-07T18:51:33 < karlp> do I really have to duplicate everything I want in the feature report from the input report?
2026-01-07T18:51:38 < karlp> that seems pretty grosssss
2026-01-07T18:51:48 < mawk> I have a very new gcc too ventyl 
2026-01-07T18:51:55 < mawk> I had to fix a couple errors iirc
2026-01-07T18:52:03 < mawk> let me see if I have the patch for you
2026-01-07T18:53:02 < karlp> juse use --disable-werror or whatever.
2026-01-07T18:54:45 < mawk> https://bpa.st/ZFPA
2026-01-07T18:54:51 < mawk> yeah if it's just a one-off karlp 
2026-01-07T18:54:53 < mawk> I intended to upstream the patch
2026-01-07T18:54:59 < mawk> but I forgot about it
2026-01-07T18:55:25 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-07T18:58:14 -!- lockna [~lockna@193-80-112-135.hdsl.highway.telekom.at] has joined ##stm32
2026-01-07T20:03:29 < qyx> bpa.st is a ST's pastebin?
2026-01-07T20:05:15 < qyx> also lol? what was the warning mawk?
2026-01-07T20:05:43 < qyx> void *calloc(size_t nmemb, size_t size);
2026-01-07T20:11:39 < qyx> nah I forgot to take photo of my recent rs485 biasing bodge
2026-01-07T20:12:56 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-07T20:48:02 < qyx> karlp: I need you, what the hell is this https://www.felser.ch/profibus-manual/elektrische_uebertragung.html
2026-01-07T20:48:13 < qyx> why does B go to VDD?
2026-01-07T20:48:57 < qyx> because this is what I am doing https://www.ti.com/document-viewer/lit/html/SSZTAK0
2026-01-07T20:49:49 < zyp> there's no consensus on whether A or B is positive
2026-01-07T20:51:05 < zyp> > Typically, Green is for Line A (Negative) and Red is for Line B (Positive) on PROFIBUS cables.
2026-01-07T20:51:58 < qyx> typically, gtfo undecided people
2026-01-07T20:52:13 < qyx> and thats why the world is hard
2026-01-07T20:52:20 < zyp> that's what you get when you name the signals A and B, which doesn't indicate anything by itself
2026-01-07T20:52:33 < zyp> compare to e.g. CANH/CANL
2026-01-07T20:53:16 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Quit: bitmask]
2026-01-07T20:54:22 < qyx> let's see what THVD1450 says
2026-01-07T20:54:26 < qyx> Bus I/O port, A (complementary to B)
2026-01-07T20:54:30 < qyx> Bus I/O port, B (complementary to A)
2026-01-07T20:54:38 < jpa-> https://en.wikipedia.org/wiki/RS-485#Signals  real standard is "what wikipedia says".. or nowadays "what chatgpt says"
2026-01-07T20:54:38 < qyx> oh noice
2026-01-07T20:55:14 < zyp> I thought the real standard was swapping the wires until it works :)
2026-01-07T20:55:54 < qyx> so for input H, A is H, B is L, at least for THVD1450
2026-01-07T20:56:07 < qyx> so in order to idle at H, I need to bias A to VDD
2026-01-07T20:56:42 < zyp> yeah, and if you're connecting it to profibus, it goes to the red wire
2026-01-07T20:56:49 < zyp> which apparently is B
2026-01-07T20:57:19 < qyx> yeah profibus is swapped
2026-01-07T20:57:21 < jpa-> i bet half the crap is how some interface chips invert their data line and others don't, so then designers swap the wires to avoid an extra inverter..
2026-01-07T20:59:00 < jpa-> .. or maybe it even goes back to RS-232 for some crazy reason having negative voltage for '1'
2026-01-07T21:32:28 -!- grindhold_ is now known as grindhold
2026-01-07T21:33:02 < qyx> so, apparently it wroks, transceiver detects bus idling at H
2026-01-07T21:33:10 < qyx> so I did it "the right" way
2026-01-07T21:35:06 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-07T21:41:01 -!- Livio [~livio@user/livio] has quit [Remote host closed the connection]
2026-01-07T21:41:21 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Quit: bitmask]
2026-01-07T21:42:48 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-07T21:50:40 < mawk> no qyx it's the pastebin of some guy on irc
2026-01-07T21:50:45 < mawk> the warning is swapping the arguments of calloc
2026-01-07T21:51:07 < mawk> it's supposed to be calloc(n, size) and not calloc(size, n)
2026-01-07T21:51:19 < mawk> even if it gives the same result
2026-01-07T21:56:13 < qyx> both arguments are declared the same and the result is the same
2026-01-07T21:56:22 < qyx> how does it even detect you are doing it the wrong way?
2026-01-07T21:56:33 < qyx> how does it *know*?
2026-01-07T22:03:51 < aandrew> it won't always give the same result
2026-01-07T22:04:13 < aandrew> or maybe it would, sizeof() already takes alignment into account
2026-01-07T22:05:39 < aandrew> I wonder if there's some compiler magic that triggers from using common things like sizeof() for the first argument
2026-01-07T22:06:00 < aandrew> since int x = 5; foo = calloc(x, sizeof(thing)) would be perfectly legitimate
2026-01-07T22:06:36 < aandrew> or maybe triggering on typeof(param_1) == size_t or something
2026-01-07T22:30:40 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 246 seconds]
2026-01-07T22:36:50 -!- Livio [~livio@user/livio] has quit [Quit: leaving]
2026-01-07T22:50:33 < qyx> do you remember your last c quote, aandrew 
2026-01-07T22:50:54 < ventyl> mawk: this is it. i fixed like five of them and then just gave up
2026-01-07T22:51:12 < ventyl> mawk: also, this does not classify as "it builds just fine"
2026-01-07T22:51:30 < ventyl> it could be ten sizeofs, but equally well 10000 of them
2026-01-07T23:03:45 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-07T23:06:37 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Ping timeout: 264 seconds]
2026-01-07T23:14:09 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-07T23:19:01 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-07T23:20:01 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has quit [Ping timeout: 246 seconds]
2026-01-07T23:20:14 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Client Quit]
2026-01-07T23:25:06 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has joined ##stm32
2026-01-07T23:43:59 < mawk> qyx: it detects "sizeof"
2026-01-07T23:44:05 < mawk> ventyl: my patch fixes them all
2026-01-07T23:44:32 < mawk> you just apply it with patch < file.diff
2026-01-07T23:45:22 < ventyl> I know how to apply patches
2026-01-07T23:46:05 < mawk> or git apply
2026-01-07T23:46:53 < ventyl> just having to apply *any* patch doesn't classify as "build just fine" with me
2026-01-07T23:47:00 < ventyl> *builds
2026-01-07T23:47:50 < mawk> yeah I forgot I had to do that
2026-01-07T23:48:04 < mawk> but it's just a gcc version thing, it happens often
2026-01-07T23:48:13 < mawk> same thing happens when you try to build python for instance
2026-01-07T23:48:55 < mawk> the pastebin adds windows line endings to the patch for some reason, you might have to do dos2unix on it
2026-01-07T23:49:02 < ventyl> i'm tired
2026-01-07T23:49:11 < ventyl> this industry is full of fuck and full of retarded people
2026-01-07T23:49:36 < mawk> it's not ST code which fails, it's openocd code
2026-01-07T23:49:47 < mawk> it's just already fixed in upstream openocd but not backported to the ST version
2026-01-07T23:49:48 < ventyl> morons, idiots and all kind of annoying scum doing stupid things convincing me its the right thing to do
2026-01-07T23:50:39 < ventyl> who produced software which is ancient, half-broken, usually brain-dead and ugly
2026-01-07T23:52:16 < ventyl> this somehow also makes me to hate the fact that trivial and essential things like bringing up debugger connection to years-old chip does take longer than 5 seconds
2026-01-07T23:52:34 < ventyl> and is not a yeah, it just works
2026-01-07T23:53:10 < ventyl> how this fucking industry can even work if this fucking essential tool needs one day of elaboration and like.... 8 attempts to bring it up?
2026-01-07T23:55:21 < ventyl> this is nothing against any of you personally, just in ~20ish years I do some kind of development I haven't seen industry with essentials so fucked up as embedded
2026-01-07T23:55:38 < zyp> try web
2026-01-07T23:55:39 < ventyl> which is a feat, considering that I've been doing web development in the past
2026-01-07T23:55:45 < ventyl> zyp: web does better
2026-01-07T23:56:38 < zyp> inclined to disagree
2026-01-07T23:57:51 < ventyl> well, the web tooling may be cringey, especially javascript, opinions may be stupid (like immutability) but they have basics sorted out
2026-01-07T23:57:58 < ventyl> and it stacks-up
2026-01-07T23:58:25 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
--- Day changed Thu Jan 08 2026
2026-01-08T00:04:42 < qyx> sounds like a new year crisis
2026-01-08T00:07:28 < nohit> I asked you to like this so we could say goodbye to openocd but nobody did https://github.com/Marus/cortex-debug/issues/1118
2026-01-08T00:10:02 -!- om3ga [~om3ga@93.177.187.134] has quit [Quit: Bye]
2026-01-08T00:13:26 -!- om3ga [~om3ga@93.177.187.134] has joined ##stm32
2026-01-08T00:30:57 < zyp> jpa-, how would you do bit flags in protobuf?
2026-01-08T00:31:32 < karlp> woop, my kicad bug from 2021 got fixed :)
2026-01-08T00:31:56 < zyp> I figure there's two obvious solutions
2026-01-08T00:32:18 < karlp> qyx: what zyp said: A and B are arbitrary.... effecteively...
2026-01-08T00:32:26 < zyp> 1: make an enum with bit values (1, 2, 4, 8…)
2026-01-08T00:32:37 < zyp> 2: make an enum with bit offsets
2026-01-08T00:32:56 < karlp> it's why the classic rs485 troubleshooting step 1 is "swap a and b"
2026-01-08T00:34:31 < zyp> the former is what I'd generally prefer in C and similar, but it seems less useful in protobuf
2026-01-08T00:36:14 < zyp> then again, the right solution would probably be to forget about bit flags and making it a repeated enum field instead :p
2026-01-08T00:56:58 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 246 seconds]
2026-01-08T00:57:13 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 265 seconds]
2026-01-08T01:09:48 < aandrew> qyx: the one about lego?
2026-01-08T01:17:33 < qyx> yeah
2026-01-08T01:58:33 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-08T02:00:29 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Client Quit]
2026-01-08T05:33:25 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-08T05:40:26 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-08T05:53:07 -!- haritz [~hrtz@user/haritz] has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
2026-01-08T07:40:57 < jpa-> zyp: in either case, the actual field will have to be raw uint.. https://protobuf.dev/programming-guides/enum/
2026-01-08T07:41:28 < jpa-> they have such fun spec and then *all* implementations are listed as "All known C++ releases are out of conformance." "All known C# releases are out of conformance." etc.
2026-01-08T07:41:47 < jpa-> you can still use enums to get kind-of constant numbers into your code
2026-01-08T08:30:26 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-08T08:43:52 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has joined ##stm32
2026-01-08T09:00:16 -!- catphish [~quassel@user/catphish] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
2026-01-08T09:00:31 -!- catphish [~quassel@user/catphish] has joined ##stm32
2026-01-08T09:05:25 -!- kurfen [~kurfen@45.84.137.20] has quit [Ping timeout: 264 seconds]
2026-01-08T09:37:55 -!- nerozero [~nerozero@87.253.63.54] has quit [Remote host closed the connection]
2026-01-08T09:39:45 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-08T09:47:21 < zyp> jpa-, yeah, I know
2026-01-08T09:48:28 < zyp> a coworker was drafting a spec for a project a while ago, and when they asked me «what do I do about bit flag fields?» I told them to just make the field an uint32 and an enum with the flags (1, 2, 4, 8, etc…)
2026-01-08T09:51:08 < zyp> but when I took over that the other day, I saw they'd put padding values into the enum since proto3 requires the first entry to be 0, which seems inconvenient
2026-01-08T09:52:01 < zyp> so I think switching to offsets instead of values fits better
2026-01-08T09:55:54 < zyp> and then I can hook the code generator so that when I have an enum FooBits and a uint32 foo, the generator spits out a foo: BitFlags<FooBits> that still decodes like a uint32 but has an API similar to a repeated FooBits
2026-01-08T09:59:20 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-08T09:59:53 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-08T10:07:47 < jpa-> also you can always go from offsets to flags, but other way is annoying
2026-01-08T10:08:50 < zyp> hooking the code generator looks straight forward: https://docs.rs/micropb-gen/latest/micropb_gen/config/struct.Config.html#method.custom_field
2026-01-08T10:09:43 < qyx> anyone happen to know what are the current efficiency/idle power regulations for PSUs in the EU now?
2026-01-08T10:10:52 < jpa-> i wonder when will we run out of prefixes for protobuf libraries.. there is already minipb, micropb, nanopb, picopb ..
2026-01-08T10:11:47 < jpa-> qyx: the main requirement is that the power supply much make an annoying whine when in standby, forcing the user to unplug it
2026-01-08T10:11:58 < jpa-> *must
2026-01-08T10:34:16 < BrainDamage> 10^-21pb
2026-01-08T10:57:23 < zyp> jpa-, there's also a femtopb
2026-01-08T11:47:30 -!- Alexer [~alexer@alexer.net] has quit [Ping timeout: 244 seconds]
2026-01-08T11:50:18 < karlp> nano -9, pico -12, femto -15, atto -18, what's -21?
2026-01-08T11:50:37 < karlp> zepto apparently.
2026-01-08T11:52:34 < karlp> he, ronto and quecto at -27 and -30 were added for symmetry with ronna and quettat at +27 and +30, which were added for datascience, but without any actual user for the small ones.
2026-01-08T11:52:53 < zyp> you skipped 24
2026-01-08T11:53:03 < karlp> ahh, that's everybodies fave: yocto
2026-01-08T11:53:25 < karlp> so an electron weights ~1 ronto gram.
2026-01-08T12:30:22 < karlp> just got caught by python not doign 2^31 again...
2026-01-08T12:32:20 -!- goodvibrations32 [~user@user/goodvibrations32] has quit [Remote host closed the connection]
2026-01-08T12:34:22 -!- mawk [mawk@wireguard/contributor/mawk] has quit [Quit: ZNC - https://znc.in]
2026-01-08T12:34:36 -!- mawk [mawk@wireguard/contributor/mawk] has joined ##stm32
2026-01-08T12:43:20 < qyx> omg yocto
2026-01-08T13:26:12 < Steffanx> Time to move to another element. Why stick with lead?
2026-01-08T13:41:24 < zyp> Steffanx, when are you making nanoau?
2026-01-08T13:44:29 < Steffanx> When I finally decide to work on this 5G boat. Fully remotely operated.
2026-01-08T14:01:36 < qyx> lol got the joke now
2026-01-08T14:12:11 < jpa-> i hope the joke isn't as good as it seems.. otherwise someone has been reading my blog way too carefully
2026-01-08T14:27:57 < zyp> :D
2026-01-08T14:30:21 -!- flatmush [~benbrewer@149.71.24.176] has quit [Ping timeout: 244 seconds]
2026-01-08T14:31:14 < qyx> oh while (1) { uptime++; vTaskDelay(500); } is a wrong way to measure uptime
2026-01-08T14:31:33 < qyx> vTaskDelay(1000) would be slightly better
2026-01-08T14:31:52 < qyx> still wrong
2026-01-08T14:32:20 < jpa-> it works for measuring CPU load though!
2026-01-08T14:46:05 -!- flatmush [~benbrewer@149.71.29.67] has joined ##stm32
2026-01-08T14:50:13 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 260 seconds]
2026-01-08T14:54:19 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-08T15:24:23 < karlp> we actually have a cpu load thing at work that does that.
2026-01-08T15:24:25 < karlp> I hate it.
2026-01-08T15:32:25 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 245 seconds]
2026-01-08T15:34:14 < jpa-> why?
2026-01-08T15:34:30 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-08T15:42:57 < karlp> can't go to sleep.
2026-01-08T15:43:04 < karlp> it's busy counting up in the idle task,
2026-01-08T15:43:21 < karlp> and it "requires" a "calibration" phase at boot to see how fast it can count...
2026-01-08T15:43:57 < jpa-> ah, lol
2026-01-08T15:47:03 < jpa-> i've used the slightly smarter way by incrementing variable with the delta from microsecond counter, and not incrementing if the delta is more than a few microseconds 
2026-01-08T15:47:19 < jpa-> but even that doesn't easily work with sleeping
2026-01-08T15:48:38 < jpa-> (it can work if you disable interrupts and wfi and take time and enable interrupts, but that adds interrupt latency)
2026-01-08T15:50:12 < karlp> what's the right stats way of comparing sigmas?
2026-01-08T15:50:28 < karlp> like, I want to measure "flatness" of a signal (ie, noise)
2026-01-08T15:50:53 < karlp> but sigma itself, the magnitude of it, depends on the mean of the sample population.
2026-01-08T15:51:04 < jpa-> are you looking for variance?
2026-01-08T15:51:16 < karlp> so you could be the same "flatness" with different sigma.
2026-01-08T15:51:48 < jpa-> why does magnitude of noise depend on mean?
2026-01-08T15:52:47 < karlp> I mean, like, I have something that measures 5000 samples of "5mv" and the stddev of that is one thing, and 5000 samples of "2mv" has a different stddev, even if it's equally noisy/flat.
2026-01-08T15:53:18 < karlp> I tried doing the data norm, to get it all to zero mean/stddev=1, but that moves the data,
2026-01-08T15:53:37 < jpa-> why do you consider it "equally noisy" if it has different stddev?
2026-01-08T15:53:42 < karlp> so you can then do histograms to see how tight or flat the historgram is,
2026-01-08T15:54:22 < karlp> hrm, good question I guess, I just don't really see why I should have more noise just because the adc count is higher?
2026-01-08T15:54:42 < jbo> > ImHex requires at least OpenGL 3.1.0 to run but your system seems to only support up to OpenGL 3.0.0!
2026-01-08T15:57:50 < jpa-> karlp: well if your 2mV signal has 1mVrms of noise and 5mV signal has 1mVrms of noise, i'd expect stddev of both to be 1mV?
2026-01-08T15:58:34 < jpa-> jbo: hey you survived the fire!
2026-01-08T16:00:35 < jbo> couldn't imagine being without you
2026-01-08T16:17:26 < karlp> jpa-:, well, either I'm doing something measuring, but that's not what I see.
2026-01-08T16:18:14 -!- Alexer [~alexer@alexer.net] has joined ##stm32
2026-01-08T16:18:16 < karlp> I've been seeing sigma go up with the adc count.
2026-01-08T16:19:01 < karlp> mioght hve to run a bunch of these tests again 
2026-01-08T16:24:39 < jpa-> karlp: can you share plots of the raw data?
2026-01-08T16:51:22 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has quit [Remote host closed the connection]
2026-01-08T17:06:34 < qyx> yes sigma goes up with the number of samples if your noise has lower frequency components than you were able to capture
2026-01-08T17:07:04 < qyx> thats the case with various drifts, walks, 1/f noise, etc
2026-01-08T17:08:37 < qyx> it is like measuring car speed, if you sample 10 times over one second, tbe sigma/stddev would most probably be lower than 1 km/h
2026-01-08T17:09:02 < qyx> but thats not the case if you are measuring over a 10 minute trip
2026-01-08T17:09:14 < qyx> despite the car trying to keep the speed constant
2026-01-08T17:09:50 < qyx> which is a prerequisite
2026-01-08T17:11:31 < jpa-> i thought karlp had a difference in measured value, not number of samples
2026-01-08T17:12:06 < qyx> oh he didn't meant adc count but adc value
2026-01-08T17:34:44 < karlp> jpa is correct.
2026-01-08T17:35:01 < karlp> jpa-: yeh, busy looking at flights right now, but want ot re-run some of this a bit later on.
2026-01-08T17:36:07 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-08T17:45:37 < Steffanx> Lol yes jpa- . For some reason I even remember it -_-
2026-01-08T19:21:04 -!- System_Error [~SystemErr@user/systemerror] has quit [Ping timeout: 252 seconds]
2026-01-08T19:26:21 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-08T19:41:49 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has joined ##stm32
2026-01-08T19:58:38 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has quit [Quit: Client closed]
2026-01-08T20:09:24 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has joined ##stm32
2026-01-08T20:19:43 -!- nomorekaki [~nomorekak@178-55-133-143.bb.dnainternet.fi] has quit [Ping timeout: 272 seconds]
2026-01-08T20:38:39 -!- umbramal- [~umbramali@188.74.90.42] has quit [Quit: %So long and thanks for all the fish%]
2026-01-08T20:40:45 -!- umbramalison [~umbramali@188.74.90.42] has joined ##stm32
2026-01-08T20:42:29 -!- umbramalison [~umbramali@188.74.90.42] has quit [Client Quit]
2026-01-08T20:46:09 -!- umbramalison [~umbramali@188.74.90.42] has joined ##stm32
2026-01-08T21:34:03 < karlp> jpa-/qyx looks like I was either measureing badly or something before, it's ~flat ("same sigma" across the range for a given mix of source/adc: https://paste.jvnv.net/view/kgO2a
2026-01-08T21:35:00 < karlp> the 32bit adc is giving better sigma across the board vs the 24bit one, which isn't entirely surprising.
2026-01-08T21:35:16 < karlp> though given that it's down in the sub microvolts...
2026-01-08T21:35:50 < karlp> this is also just raw plotting, no filtering _at all_ 
2026-01-08T21:36:43 < karlp> I'll likely need a different test setup to do any meaningful comparison of some of the front end tweaks I was planning on, I suspect just the wires on my desk and their arrangement will have enough impact.
2026-01-08T22:22:29 -!- Livio [~livio@user/livio] has quit [Quit: leaving]
2026-01-08T22:32:04 -!- haritz [~hrtz@2a01:4b00:bc2e:7000:d5af:a266:ca31:5ef8] has joined ##stm32
2026-01-08T22:32:05 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-08T22:40:15 -!- kurfen [~kurfen@194.127.173.38] has joined ##stm32
2026-01-08T22:55:49 -!- lockna [~lockna@193-80-112-135.hdsl.highway.telekom.at] has quit [Ping timeout: 264 seconds]
2026-01-08T23:06:55 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 240 seconds]
2026-01-08T23:59:35 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
2026-01-08T23:59:49 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 246 seconds]
--- Day changed Fri Jan 09 2026
2026-01-09T00:00:09 -!- hexo__ [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-09T00:04:32 -!- hexo_ [~hexo@user/hexo] has joined ##stm32
2026-01-09T00:31:14 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has quit [Ping timeout: 260 seconds]
2026-01-09T00:31:46 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has joined ##stm32
2026-01-09T00:47:40 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has quit [Ping timeout: 244 seconds]
2026-01-09T00:51:44 < karlp> well, over 3 minute windows, my two different analog input stages have 105 vs 112 nV sigma.
2026-01-09T00:51:54 < karlp> I'm not sure if that's meaningful :)
2026-01-09T00:52:24 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has joined ##stm32
2026-01-09T00:59:01 < karlp> but with a different source, (noisier) it's a 19nV sigma diference, in the other direction whee
2026-01-09T01:05:04 -!- hexo_ is now known as hexo
2026-01-09T01:16:32 -!- qyx_ [~qyx@84.245.121.83] has joined ##stm32
2026-01-09T01:18:15 -!- qyx [~qyx@84.245.121.152] has quit [Ping timeout: 240 seconds]
2026-01-09T03:15:48 < catphish> i'm currently trying to get my head around CH569. It seems extremely capable, but also shrouded in nonexistent documentation
2026-01-09T05:27:01 -!- Alexer [~alexer@alexer.net] has quit [Ping timeout: 264 seconds]
2026-01-09T05:27:01 -!- artok [~azo@91-153-163-37.elisa-laajakaista.fi] has quit [Ping timeout: 264 seconds]
2026-01-09T05:28:16 -!- Alexer [~alexer@alexer.net] has joined ##stm32
2026-01-09T05:28:47 -!- artok [~azo@91-153-163-37.elisa-laajakaista.fi] has joined ##stm32
2026-01-09T07:54:58 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-09T08:03:48 -!- haritz [~hrtz@user/haritz] has quit [Remote host closed the connection]
2026-01-09T08:55:29 < jpa-> karlp: you might want to test with controlled external interference, as that could be a significant noise source even in actual usage
2026-01-09T09:22:33 < machinehum> catphish: Do you have a devboard?
2026-01-09T09:22:43 < machinehum> Seems like a very interesting part
2026-01-09T09:23:30 < machinehum> 6 bucks qtd 1
2026-01-09T10:01:27 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-09T10:03:09 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-09T11:08:01 < mawk> what do I need to do NFC sniffing
2026-01-09T11:08:16 < mawk> I have a ST25R devboard with transparent mode
2026-01-09T11:08:21 < mawk> maybe that's enough
2026-01-09T11:08:33 < mawk> but it shouldn't pull from the field and disturb the actual card
2026-01-09T11:09:47 < mawk> the ST docs mention emulating mifare classic, which is interesting given how it's proprietary with an "unknown" encryption scheme
2026-01-09T11:09:52 < mawk> if I was NXP I'd sue ST
2026-01-09T11:12:45 < zyp> isn't crypto1 well known now?
2026-01-09T11:13:14 < mawk> yeah that's why the quotes
2026-01-09T11:14:22 < mawk> the nRF52840 is one of the best chips apparently for emulating mifare classic and other cards, because it's super fast
2026-01-09T11:14:40 < mawk> the closer the timing is to an actual mifare IC the more compatible it is
2026-01-09T11:14:47 < zyp> heh
2026-01-09T12:02:02 < karlp> jpa-: yeah, that's actually what I'd like, bit unsure of good reliable repeatable ways of doing it.
2026-01-09T12:02:26 < karlp> we actualyl have an rf chamber sitting in the office completely unused, but I've not made this a work project yet.
2026-01-09T12:03:58 < catphish> machinehum: no, i don't have a dev board, perhaps i'll order one, they only seem to be available from aliexpress, it's amazing value for high speed data stuff
2026-01-09T12:06:44 < karlp> catphish: there's a project here I saw shared elsewhere, might have somethign useful: https://gitlab.com/yannsionneau/SucreLA/
2026-01-09T12:07:51 < karlp> https://hydrabus.com/hydrausb3-v1-0-specifications/?v=a2a551a6458a is another user
2026-01-09T12:08:03 < karlp> ex ##stm32 chatter too.
2026-01-09T12:08:31 < catphish> yeah i found hydrausb :)
2026-01-09T12:09:46 < catphish> from a hardware standpoint, the chip seems pretty easy to use, single 3.3v supply, just one external crystal needed, but the documentation is basically nonexistent
2026-01-09T12:10:05 < zyp> I have a devboard but I never figured out how to hook up a debugger to it
2026-01-09T12:10:14 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-09T12:11:34 < catphish> zyp: what i can tell is that 1) it has a bootloader that allows USB programming, i found no official documentation about how to enter bootloader mode, only one third party saying there's a GPIO pin that will do this 2) there is a programming port, but no info about what protocol it speaks
2026-01-09T12:12:38 < catphish> https://hydrabus.com/wp-content/uploads/2022/09/HydraUSB3_V1_PinAssignment.png
2026-01-09T12:12:51 < zyp> I assume it talks rvswd
2026-01-09T12:13:29 < karlp> it probably uses either 1 or 2 wire wch protocol...
2026-01-09T12:13:58 < catphish> that seems likely - the hydra board has 2 pins (clock and data) that is says are for debugging
2026-01-09T12:13:59 < karlp> catphish: on other wch parts, there's a "boot" pin for a bootlaoder, but it's just in flash, you can erase it and then then pin is "just a pin"
2026-01-09T12:14:17 < karlp> that's why it's not listed as a "boot" pin in any docs.
2026-01-09T12:14:22 < karlp> you just see it in some EVT schematics.
2026-01-09T12:14:25 < jpa-> karlp: build a TEM cell ;)
2026-01-09T12:16:13 < catphish> karlp: i think it's a ROM bootloader, but like you say, there's no official pin assignment to load it
2026-01-09T12:16:15 < karlp> jpa-: that's what's in the lab actually
2026-01-09T12:16:21 < karlp> I don't know that sort of test gear very well.
2026-01-09T12:17:03 < karlp> I kinda want to use it, but I'm still sort of demoing the idea of what a new gen could be, 
2026-01-09T12:18:12 < karlp> I just plug a rf gen like https://www.aliexpress.com/item/1005006991025800.html in at one end and record with my device inside the box right?
2026-01-09T12:18:37 < karlp> probably needs more amps or somthign to have controlled power levels 
2026-01-09T12:18:41 < karlp> fucking rabbit hole I imagine.
2026-01-09T12:21:37 < catphish> i guess i probably want a dev board and a WCH-LINKE
2026-01-09T12:22:03 < karlp> make sure you get a link_e_ not a link...
2026-01-09T12:22:54 < catphish> this one https://www.olimex.com/Products/RISC-V/WCH/WCH-LinkE/
2026-01-09T12:30:29 < ventyl> OK, I am by far not the creepiest guy in the land of memory isolation
2026-01-09T12:31:04 < ventyl> Rustaceans made quite a feat
2026-01-09T12:34:59 < zyp> xous?
2026-01-09T12:35:11 < zyp> (I just watched the 39C3 talk on xous)
2026-01-09T12:39:52 < jpa-> karlp: yeah, rf gen at one end and 50 ohm terminator at other end
2026-01-09T12:41:19 < ventyl> zyp: yes, that talk
2026-01-09T12:43:46 < jpa-> karlp: typical radiated immunity specs are about 3 V/m, so for 50 ohm TEM cell with e.g. 10 cm gap you need 2 mW of power
2026-01-09T12:43:56 < jpa-> karlp: https://www.batronix.com/files/Tekbox/Datenblaetter/QA_OpenTEMCells.pdf  is useful doc
2026-01-09T12:44:38 < zyp> so you just put in the device and check that it doesn't crash?
2026-01-09T12:44:54 < jpa-> in karlp's case, check that noise level in ADC measurement stays within limits
2026-01-09T12:45:05 < zyp> ah, right
2026-01-09T12:45:15 < jpa-> if you fail EMI bad enough to make digital system crash, you will absolutely fail emission testing
2026-01-09T12:45:48 < jpa-> at least for radiated immunity
2026-01-09T12:45:56 < zyp> I guess you could also crank up the power higher than the limit to see how much of a marging you have
2026-01-09T12:45:58 < jpa-> conducted immunity can be harder for e.g. automotive use
2026-01-09T12:46:19 < ventyl> zyp: what bothers me the most is that part of the talk is FUD and the rest sound as if they invented something new that wasn't around for decades
2026-01-09T12:46:59 < jpa-> looks like ADF4351 can adjust output power from 0.4 mW to 3 mW so it could work for basic testing
2026-01-09T12:49:53 < jpa-> though maybe in karlp's case a more likely noise source would be AC magnetic fields from mains wiring, which would need a different kind of a setup with a loop antenna
2026-01-09T12:50:18 < jpa-> because for slow samplerate and sigma-delta ADC, i guess most of the high-frequency EMI should get filtered pretty well
2026-01-09T12:50:47 < jpa-> unless you get funny aliasing at a specific frequency :)
2026-01-09T12:51:39 < catphish> ordered a CH569 devboard and a LinkE - fingers crossed i can figure out how it works, i'm hoping i can use its high speed parallel interface for direct data capture
2026-01-09T12:52:16 < karlp> jpa-: yeah, in reality, the adc gets lowpased to fuck
2026-01-09T12:52:17 < zyp> ventyl, if I understood it correctly, they do cooperative multitasking, which stood out to me as weird, given all the focus on isolation
2026-01-09T12:53:03 < karlp> but I know that it will be a somewhat uphil battle, that would neeed to be supported by test data, fo rme to go from 32-> 24bit adcs, and toss out the cans and all the fancy feed through caps on the current design.
2026-01-09T12:53:29 < karlp> I'm just 99% sure that the current designs were just cargo'd in from prior gens and it's never been ever considered again.
2026-01-09T12:53:43 < karlp> but, big rabbit hole of metorlogy that i'm not.... entirely interested in.
2026-01-09T12:53:58 < ventyl> zyp: that's not really *that* weird in safety-relevant applications
2026-01-09T12:54:25 < karlp> I actaully want to spend mor etime on getting the kernel IIO to work now, as that was the actual goal, moving it all to linux.
2026-01-09T12:54:34 < zyp> AIUI xous is focused on security, not safety
2026-01-09T12:55:54 < karlp> jpa-: I could use an adf4351 based board with a set of nera field probes too, to just "poitn" interference at a board too right?
2026-01-09T12:56:38 < zyp> but yeah, if denial of service is not in your threat model, I guess it's fine
2026-01-09T12:58:09 < ventyl> zyp: safety models usually do cooperative multitasking with policeman. you either yield or your timeslice ends and kernel kills you
2026-01-09T12:58:41 < ventyl> the problem with doing so in pretty much any system is that it is beyond possible to keep it running after some task is killed
2026-01-09T12:58:53 < ventyl> so the only reasonable thing is to activate watchdog
2026-01-09T12:59:29 < zyp> do you mean safety as in functional safety?
2026-01-09T13:00:05 < ventyl> yes
2026-01-09T13:00:33 < zyp> remember that the safe state might not be «kill everything»
2026-01-09T13:02:19 < zyp> when I worked on hydraulic motors, we talked a bunch about functional safety in the context of AHC winches on boats, where the safe state is explicitly not «stop the winch», because that could flip the boat
2026-01-09T13:04:05 < zyp> the safe state was «ignore the fancy control algorithms and revert to a simple control algorithm that'll provide a reasonable amount of fixed torque»
2026-01-09T13:05:56 < zyp> since this was a distributed system with multiple motors on a winch, in a partial failure this would have the failed motor still contributing to holding the load, but leave the rest of the system with enough control authority to control it
2026-01-09T13:07:26 < zyp> and in a full failure, a fixed torque set a bit lower than the load's mass would still maintain cable tension in waves, while letting the winch unspool gently until the load settles on the bottom
2026-01-09T13:07:57 < jpa-> karlp: yeah, with near field probes it's harder to know the actual field intensity, but for ad-hoc testing it works
2026-01-09T13:10:49 -!- HelloShitty [~psysc0rpi@bl18-184-247.dsl.telepac.pt] has quit [Ping timeout: 246 seconds]
2026-01-09T13:13:25 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has joined ##stm32
2026-01-09T13:36:58 < karlp> I like the second paragraph of section 8 here: https://www.tekbox.com/product/QA_OpenTEMCells.pdf
2026-01-09T13:37:10 < karlp> "fuck testing per the standards, that shit takes ages to get it all right"
2026-01-09T13:37:31 < karlp> "just blast that shit, find the most sensitive parts, fix that, you should fine from there"
2026-01-09T13:37:50 < karlp> Inject the maximum RF power that
2026-01-09T13:37:52 < karlp> you can generate into the smallest TEM-cell that you have
2026-01-09T13:38:00 < karlp> This test can be set up within minutes and we don´t need to
2026-01-09T13:38:02 < karlp> do repetitive sweeps in order to find the most sensitive orientation. If the design has a weak spot, we will
2026-01-09T13:38:04 < karlp> find it irrespective of the orientation.
2026-01-09T13:44:00 < jpa-> that works too :D
2026-01-09T13:46:28 < zyp> that's kinda what I meant above by testing how much of a margin you have :p
2026-01-09T13:47:03 < karlp> yup.
2026-01-09T13:47:18 < karlp> at least if you're trying to do pre-compliance 
2026-01-09T13:47:42 < zyp> better to confirm your design is so far away from the limits that you don't have to worry about them rather than dealing with dialing in the exact limits
2026-01-09T13:49:50 < karlp> in other news, I like how rust also fell in to the trap of "1.11 must be newer than 1.6 right?" https://docs.rs/hidparser/latest/hidparser/
2026-01-09T13:52:45 < zyp> what trap?
2026-01-09T13:55:01 < zyp> are you confusing the class spec with the usage tables? I see the latest versions of each are respectively 1.11 and 1.6
2026-01-09T14:40:39 < ventyl> zyp: funily enough, just now I had lecture on functional safety
2026-01-09T14:45:07 < karlp> zyp: I probably was yeah, because i'd been running into seeing the HUT 1.11 all the time, instead of 1.6 being referred
2026-01-09T15:24:59 < zyp> I see what you mean, doing patch releases by appending a digit without a dot first is stupid as fuck
2026-01-09T15:25:43 < karlp> doing HUT properly is weird as hell,
2026-01-09T15:26:06 < karlp> no wonder people just wing it and write both sides.
2026-01-09T15:26:36 < karlp> doing "correct" gpios using ordinals is fucking gross compared to just declaring 8 bits.
2026-01-09T15:26:47 < karlp> but it gets rid of the warnings.
2026-01-09T15:28:03 < zyp> doing HUT properly for custom stuff rarely seems worthwhile
2026-01-09T15:28:10 < zyp> and yeah, I hate ordinals
2026-01-09T15:28:22 < karlp> well, it's not custom, or at least, the plan was for it to be "standard"
2026-01-09T15:28:30 < karlp> but standard seems to be not used, and or custom...
2026-01-09T15:28:46 < karlp> like, the arcade page GPIOs looked decent for doing ... gpios
2026-01-09T15:28:52 < karlp> but it wants ordinals.
2026-01-09T15:28:56 < karlp> and damn, they are gross
2026-01-09T15:29:08 < zyp> yeah
2026-01-09T15:29:34 < zyp> arcin uses the LED page for lights, also with ordinals
2026-01-09T15:30:10 < zyp> and the only reason I bothered was because somebody else already made the software on the other end recognize it as such
2026-01-09T15:32:04 < karlp> I think just using the buttons page for the gp inputs is goin to be easier, and just as "obvious"
2026-01-09T15:32:12 < karlp> but for the outputs I'm not sure of the cleanest
2026-01-09T15:34:25 < zyp> hmm, there's a SoC control page now that includes bootloader stuff, fun :)
2026-01-09T15:41:38 < karlp> yeah, I saw that, but it's not reallllly interesting (IMO) unless there's already host sw I can leverage.
2026-01-09T15:42:40 < zyp> unless you'd be making a HID bootloader and corresponding host software anyway
2026-01-09T15:42:55 < zyp> I've been using vendor page 0xb007 for it :)
2026-01-09T15:44:20 < zyp> oh, no vendor page 0xff55, usage 0xb007
2026-01-09T15:45:18 < karlp> hrm, if i use the buggons page, without being in a mouse application collection, am I going to run into being treated as mouse clicks anyway?
2026-01-09T15:45:25 < karlp> I _shouldn't_ for any sane OS right
2026-01-09T16:27:39 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-09T16:32:24 < karlp> 619 byte report descriptor, yolo
2026-01-09T16:46:57 -!- Livio [~livio@user/livio] has quit [Remote host closed the connection]
2026-01-09T16:54:06 -!- haritz [~hrtz@140.228.70.141] has joined ##stm32
2026-01-09T16:54:06 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-09T17:31:03 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-09T18:00:51 < karlp> meh, array indicies are pretty shit too.
2026-01-09T18:35:37 -!- PhantomWork [~PhantomWo@user/phantom] has quit [Ping timeout: 246 seconds]
2026-01-09T18:50:24 < karlp> man, I should have just done custom control trnasfers, I'd be long done.
2026-01-09T19:22:55 < mawk> I think 125kHz RFID would be a good introduction to RF, even if it's technically not radio stuff
2026-01-09T19:22:59 < mawk> I want to make a sniffer
2026-01-09T19:23:15 < mawk> so I can slip the antenna between the tag and the reader and see what happens
2026-01-09T19:25:44 < mawk> I have the coil/antenna already, then I assume I need a varicap to make sure the center frequency is around 125kHz, then an envelope detector, and a bunch of filters
2026-01-09T19:26:41 < catphish> mawk: 125kHz is fun, i managed to make a tag reader with not much more than a MCU and a coil
2026-01-09T19:27:22 < catphish> i forget how i identified the levels, you could probably just use a modern ADC at those speeds
2026-01-09T19:27:34 < mawk> in an AliExpress 125kHz reader/writer I bought there's the antenna, some passives, an opamp, and that seems to be it, then it goes straight to the MCU
2026-01-09T19:27:44 < mawk> a CHwhatever like the devkit you bought 
2026-01-09T19:29:50 < catphish> i wish i remember the circuit i used now, it was stupidly simple
2026-01-09T19:33:35 < mawk> https://github.com/RfidResearchGroup/ChameleonUltra/blob/main/hardware/dev-tools/Chameleon_nrf52_dev_V1.0.pdf
2026-01-09T19:33:44 < mawk> like this?
2026-01-09T19:33:57 < mawk> on the second page, ignore the driver bit
2026-01-09T19:34:33 < mawk> I'm not sure I understand the RSSI circuit
2026-01-09T19:34:54 < mawk> I assume they use a diode as quick way to make a logarithm?
2026-01-09T19:35:17 < catphish> here's a simple example https://i.sstatic.net/eoPAX.png
2026-01-09T19:36:41 < mawk> are these IGBTs? just a weird symbol for a BJT I assume 
2026-01-09T19:36:52 < catphish> here's a discussion about one that doesn't work, but similar concept - https://electronics.stackexchange.com/questions/103916/simple-rfid-reader-am-demodulator-not-working
2026-01-09T19:37:41 < mawk> why is Q1 not a PNP?
2026-01-09T19:38:29 < catphish> because it's providing the negative drive, the +5 at the top is the positive drive
2026-01-09T19:39:23 < mawk> Q1 is the one at the top right?
2026-01-09T19:39:35 < mawk> it's connecting VCC and the antenna
2026-01-09T19:39:59 < catphish> oh sorry, you're looking at the first schematic i sent
2026-01-09T19:40:03 < mawk> yeah 
2026-01-09T19:40:14 < catphish> i assume that's a mistake
2026-01-09T19:40:43 < mawk> the transistors might be IGBTs after all, otherwise there would at least be a base resistor
2026-01-09T19:40:57 < catphish> the second design i linked is even simpler
2026-01-09T19:41:49 < catphish> all that really matters is that you get the LC circuit to resonate at 125kHz
2026-01-09T19:42:50 < catphish> and then a diode-capacitor envelope detector, and you're done
2026-01-09T19:44:27 < catphish> if you're brave, a microcontroller push-pull pin might even be enough, the downside is that none of this is relevant to building a passive sniffer, in that case all you need is that coil and the envelope detector
2026-01-09T19:46:38 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-09T19:59:58 < mawk> you have an idea of the voltages involved?
2026-01-09T20:00:09 < mawk> especially if there's a high Q it might break stuff
2026-01-09T20:00:31 < mawk> no idea how to compute all that given that I don't know the impedance of the coil though 
2026-01-09T20:00:35 < mawk> and I have no RLC meter
2026-01-09T20:31:59 < catphish> i know 5v is fine, but this is a current based thing really, the voltage doesn't matter much
2026-01-09T20:32:26 < catphish> you won't break anything except your own circuit :)
2026-01-09T20:36:27 -!- jbo [~jbo@user/tct] has quit [Quit: ZNC 1.9.1 - https://znc.in]
2026-01-09T20:39:48 -!- jbo [~jbo@user/tct] has joined ##stm32
2026-01-09T20:40:28 < mawk> 5V ? seems like it could be more
2026-01-09T20:40:58 < mawk> well it's a current thing if you have a small enough load, but otherwise the voltage will climb
2026-01-09T20:41:02 < mawk> as coils do
2026-01-09T20:44:35 < qyx_> got my first rpi5
2026-01-09T20:44:51 < qyx_> they have their own southbridge or what is that chip
2026-01-09T20:53:17 < mawk> decap it
2026-01-09T20:53:27 < mawk> I can lend you my guillotine 
2026-01-09T20:56:50 < qyx_> I forgot to order a microsd
2026-01-09T20:57:00 < qyx_> because I forgot this thing needs it
2026-01-09T21:02:09 -!- SamaylyS_ [~SamaylyS@user/SamaylyS] has joined ##stm32
2026-01-09T21:47:54 < mawk> you can boot from a usb stick
2026-01-09T21:48:02 < mawk> but you need to change option bytes first
2026-01-09T22:34:11 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-09T22:49:22 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 246 seconds]
--- Day changed Sat Jan 10 2026
2026-01-10T00:00:50 < catphish> qyx_: i assumed everyone has a stack of micro SDs of varying sizes
2026-01-10T00:14:48 < zyp> catphish, I mostly have a stack of the same size
2026-01-10T00:16:33 < zyp> you can also boot from nvme
2026-01-10T00:16:47 < zyp> both nvme and usb3 is way faster than sdcard
2026-01-10T00:39:21 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 244 seconds]
2026-01-10T00:40:49 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 264 seconds]
2026-01-10T00:42:51 < catphish> nvme seems ideal!
2026-01-10T00:43:11 < catphish> PCIe all the things!
2026-01-10T00:44:49 < catphish> i'm still doing battle with understanding this CH569. It has these TCK and TIO pins which i assume are roughly equivilent to STM32's SWD interface, which is great, but one of those pins is shared with a peripheral i almost definitely need to use, so i don't really know how to overcome that, or if i can just make use of the bootloader over USB
2026-01-10T01:36:40 -!- Livio [~livio@user/livio] has quit [Quit: leaving]
2026-01-10T02:28:54 < mawk> there's nvme only on the compute module no zyp?
2026-01-10T02:28:59 < mawk> I mean the connector 
2026-01-10T02:34:04 < mawk> catphish what's wrong with it being shared? can you even debug this?
2026-01-10T02:34:32 < mawk> if it's just for flashing you can just reset and immediately debug
2026-01-10T02:34:33 < mawk> ma
2026-01-10T02:34:47 < mawk> or add a button to enable the debug port
2026-01-10T02:35:55 < catphish> well if you can reset and then flash, that should work, as long as i ensure the thing that shares the pin doesn't drive it when the MCU is in reset
2026-01-10T02:36:17 < catphish> but can you? i have no idea. how am i supposed to know :(
2026-01-10T02:37:15 < catphish> i am really only like to want to flash it, ideal would be to work out how to do that reliably over USB
2026-01-10T02:40:16 < catphish> i have only myself to blame for wanting to use this part, but still very frustrating process
2026-01-10T03:01:26 < zyp> mawk, no, rpi5 has a pcie ffc connector, you get a ton of addon boards with m.2 sockets that plugs into it
2026-01-10T03:03:25 < zyp> I have one that does both m.2 and poe: https://photos.app.goo.gl/sjgCaKpkgvYkCf947
2026-01-10T03:05:13 < Phantom> the pcie connector is a godsent thing for sure, that just changed everything about rpi
2026-01-10T03:53:20 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 240 seconds]
2026-01-10T03:59:04 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-10T04:39:49 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-10T04:47:02 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-10T05:24:45 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-10T05:31:18 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-10T06:11:28 < qyx_> zyp: I already have that pcie connector used for the 5x sata bridge
2026-01-10T07:54:40 -!- haritz [~hrtz@user/haritz] has quit [Remote host closed the connection]
2026-01-10T08:00:01 -!- om3ga [~om3ga@93.177.187.134] has quit [Ping timeout: 264 seconds]
2026-01-10T08:03:36 -!- om3ga [~om3ga@93.177.187.134] has joined ##stm32
2026-01-10T08:22:24 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-10T08:35:14 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-10T08:36:15 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-10T08:43:13 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 264 seconds]
2026-01-10T08:43:13 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 264 seconds]
2026-01-10T08:43:13 -!- umbramalison [~umbramali@188.74.90.42] has quit [Ping timeout: 264 seconds]
2026-01-10T08:43:48 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-10T08:44:53 -!- umbramalison [~umbramali@188.74.90.42] has joined ##stm32
2026-01-10T08:55:43 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-10T09:27:04 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 246 seconds]
2026-01-10T10:45:30 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 244 seconds]
2026-01-10T10:45:30 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 244 seconds]
2026-01-10T10:46:32 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-10T11:03:49 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 260 seconds]
2026-01-10T11:12:31 -!- nerozero [~nerozero@87.253.63.54] has quit [Remote host closed the connection]
2026-01-10T11:14:24 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-10T11:31:21 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-10T11:40:25 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 246 seconds]
2026-01-10T11:42:51 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has quit [Ping timeout: 244 seconds]
2026-01-10T11:46:08 < zyp> qyx_, you could get a pcie switch
2026-01-10T11:47:02 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-10T11:55:25 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has joined ##stm32
2026-01-10T12:00:24 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has quit [Ping timeout: 260 seconds]
2026-01-10T12:19:35 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 240 seconds]
2026-01-10T12:30:10 -!- IanW_ [~IceChat9@arcanum.force9.co.uk] has joined ##stm32
2026-01-10T12:31:23 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-10T12:56:32 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has joined ##stm32
2026-01-10T13:47:07 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has quit [Ping timeout: 264 seconds]
2026-01-10T14:04:53 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-10T14:32:06 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has joined ##stm32
2026-01-10T14:32:06 < mawk> ah nice zyp
2026-01-10T14:32:08 < mawk> on my pi4 I have a usb3 external ssd as the boot drive
2026-01-10T14:32:10 < mawk> and a USB3 external HDD for my terabyte porn collection 
2026-01-10T14:36:35 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has quit [Ping timeout: 240 seconds]
2026-01-10T14:57:58 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has joined ##stm32
2026-01-10T15:07:21 -!- yukam [~yukam@user/yukam] has quit [Ping timeout: 244 seconds]
2026-01-10T15:08:14 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-10T15:14:54 -!- yukam [~yukam@user/yukam] has joined ##stm32
2026-01-10T15:16:55 -!- kurfen [~kurfen@194.127.173.38] has quit [Ping timeout: 240 seconds]
2026-01-10T15:19:41 -!- kurfen [~kurfen@194.127.173.38] has joined ##stm32
2026-01-10T15:23:26 -!- haritz [~hrtz@140.228.70.141] has joined ##stm32
2026-01-10T15:23:26 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-10T15:29:20 < catphish> transpires the CH569 won't do what i want, its high speed parallel interface is too opinionated with checksums etc. fortunately there are other options
2026-01-10T15:38:32 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has quit [Ping timeout: 240 seconds]
2026-01-10T15:58:49 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 264 seconds]
2026-01-10T15:58:49 -!- qyx_ [~qyx@84.245.121.83] has quit [Ping timeout: 264 seconds]
2026-01-10T15:59:25 -!- sugarbeet [~barbas@81.4.123.134] has quit [Ping timeout: 264 seconds]
2026-01-10T16:00:23 -!- qyx [~qyx@84.245.121.83] has joined ##stm32
2026-01-10T16:01:03 -!- sugarbeet [~barbas@81.4.123.134] has joined ##stm32
2026-01-10T16:11:11 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-10T16:27:16 < mawk> you can't ignore the checksums?
2026-01-10T16:27:25 < mawk> no transparent mode or what
2026-01-10T16:27:31 < mawk> what's the target application catphish ?
2026-01-10T16:27:41 < mawk> some stm32s have a general purpose parallel interface too
2026-01-10T16:27:50 < mawk> like the H5 stuff, the PSSI peripheral 
2026-01-10T16:28:12 < mawk> which is just a repurposing of the MIPI peripheral for general purpose
2026-01-10T16:53:27 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-10T17:02:29 < Phantom> question, is there a data packet protocol for uart somewhere that is easy to implement that is reliable? for like 4-5 bytes...
2026-01-10T17:02:49 < Phantom> I don't like how I did it (9n1, 9th bit as start of packet)
2026-01-10T17:03:00 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-10T17:04:04 < karlp> modbus!
2026-01-10T17:05:32 < zyp> Phantom, just for framing? I like COBS
2026-01-10T17:07:58 < Phantom> which add the complexity of variable length
2026-01-10T17:08:08 < zyp> mawk, you only have a terabyte of porn
2026-01-10T17:08:09 < zyp> ?
2026-01-10T17:08:19 < Phantom> I might have to see if DMA is usable
2026-01-10T17:08:35 < Phantom> the 1 byte rx buffer is somewhat of an issue
2026-01-10T17:09:01 < Phantom> I'll check it out, thanks<
2026-01-10T17:09:15 < zyp> Phantom, if your messages are smaller than 250-something bytes, COBS will add exactly one byte in length
2026-01-10T17:09:31 < zyp> plus another byte for the null delimiter between frames
2026-01-10T17:09:51 < Phantom> will check it out, thanks
2026-01-10T17:10:10 < Phantom> for now, back to bed... I feels miserable (covid)
2026-01-10T17:12:32 < mawk> Phantom TCP/SLIP
2026-01-10T17:12:34 < mawk> very easy don't worry
2026-01-10T17:12:37 < mawk> yes I'm ashamed zyp
2026-01-10T17:25:50 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-10T17:38:15 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Ping timeout: 245 seconds]
2026-01-10T17:57:43 -!- yukam [~yukam@user/yukam] has quit [Ping timeout: 246 seconds]
2026-01-10T18:06:29 -!- om3ga [~om3ga@93.177.187.134] has quit [Quit: Bye]
2026-01-10T18:07:49 -!- om3ga [~om3ga@93.177.187.134] has joined ##stm32
2026-01-10T18:09:40 -!- yukam [~yukam@user/yukam] has joined ##stm32
2026-01-10T18:17:03 < SamaylyS_> AVR has decent ADC. ESP32 has crappy ADC. Where does STM32 stand?
2026-01-10T18:21:51 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-10T18:22:19 < catphish> mawk: nope, seemingly no way to avoid the proprietary framing / checksum. i just want to shovel parallel data into USB extremely fast. the next best chip for this will be FT600Q which does much the same thing, with less opinion about what's on the parallel bus
2026-01-10T18:23:05 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has joined ##stm32
2026-01-10T18:35:23 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has joined ##stm32
2026-01-10T18:54:10 < mawk> SamaylyS_ most are 12 bits
2026-01-10T18:54:22 < mawk> but if you follow the recommendations you get the advertised performance 
2026-01-10T18:54:50 < mawk> there are a coupler newer parts with 14 bits, and there's the H7 with 16 bits
2026-01-10T18:55:16 < mawk> if you have the right package and you follow the guidelines you can get like 5Mbps at 16 bits depth with the H7 part
2026-01-10T18:55:28 < mawk> and supposedly the readings are accurate to 0.5LSB
2026-01-10T18:56:33 < mawk> and you can do cool stuff like sampling two channels at the same time 
2026-01-10T18:56:44 < mawk> or even 3
2026-01-10T18:59:28 < mawk> 5Msps I mean
2026-01-10T19:15:31 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has quit [Ping timeout: 272 seconds]
2026-01-10T19:19:03 -!- infisc [uid692580@user/infisc] has joined ##stm32
2026-01-10T19:22:33 < qyx> no, avr has crappy one
2026-01-10T19:56:48 < jpa-> there are more recent AVRs with pretty good ADCs
2026-01-10T19:57:54 < jpa-> but if comparing something like atmega328 then stm32 has much better one
2026-01-10T19:59:57 -!- umbramalison [~umbramali@188.74.90.42] has quit [Quit: %So long and thanks for all the fish%]
2026-01-10T20:01:50 < antto> the xmegas and pseudo-xmegas have nicer peripherals
2026-01-10T20:02:10 < antto> they feel like a more luxurious MCU
2026-01-10T20:02:45 < antto> ...compared to the old "avr" (atmegas and sh*t)
2026-01-10T20:03:09 < antto> basically, if the chip has PDI or uPDI interface - it's an xmega or pseudo-xmega
2026-01-10T20:22:23 < jpa-> attiny824 has a relatively decent ADC too, 12 bits and differential PGA
2026-01-10T20:43:49 -!- umbramalison [~umbramali@188.74.90.42] has joined ##stm32
2026-01-10T20:53:49 < ventyl> why would anyone use that in 2026?
2026-01-10T20:54:06 < Steffanx> Didn't you port CreepyOS to it?
2026-01-10T20:54:36 < ventyl> atmega?
2026-01-10T20:54:45 < Steffanx> Attiny
2026-01-10T20:55:05 < ventyl> no, why would I do that?
2026-01-10T20:55:13 < Steffanx> Because you can
2026-01-10T20:55:16 < Steffanx> Or could 
2026-01-10T20:55:18 < ventyl> you can't actually
2026-01-10T20:55:19 < Steffanx> Or could try
2026-01-10T20:55:33 < ventyl> AVR does not qualify for running CMRX
2026-01-10T20:55:34 < Steffanx> Time to redesign it then
2026-01-10T20:56:07 < ventyl> it can run on pretty low spec HW, but it has to support HW memory isolation
2026-01-10T20:56:38 < antto> some chips named "attiny" are actually pseudo-xmegas, which makes them not so stupid
2026-01-10T20:58:54 < ventyl> but the CPU core is the same stupid
2026-01-10T21:02:46 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-10T21:03:23 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-10T21:13:41 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-10T21:20:43 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-10T21:42:22 < antto> same as what?
2026-01-10T21:42:55 < antto> they are really not "atmegas" nor "attiny" even if the name says so... if they have *PDI interface - the core is xmega
2026-01-10T21:46:10 < ventyl> ah.
2026-01-10T21:46:26 < ventyl> I don't give a flying fuck about anythin beyond the CPU part of the chip
2026-01-10T21:46:45 < ventyl> peripherals, busses, whatever. none of that is of any interest with regards to my RTOS
2026-01-10T21:47:23 < ventyl> the processor, though, is. and mega or xmega, both as same 8-bit AVR with abso-fucking-lutely no means of security
2026-01-10T21:48:20 < antto> some have AES-128 ;P~
2026-01-10T21:48:49 < antto> yes, it's not in the "core", it's a peripheral
2026-01-10T21:49:15 < ventyl> y u no MPU
2026-01-10T21:49:29 < antto> i don't remember about MPU
2026-01-10T22:08:46 -!- infisc [uid692580@user/infisc] has quit [Quit: Connection closed for inactivity]
2026-01-10T22:25:18 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-10T22:29:34 < ventyl> that's because there's none
2026-01-10T22:29:36 < ventyl> obviously
2026-01-10T23:17:16 -!- IanW_ [~IceChat9@arcanum.force9.co.uk] has quit [Ping timeout: 246 seconds]
2026-01-10T23:17:40 -!- IanW_ [~IceChat9@arcanum.force9.co.uk] has joined ##stm32
2026-01-10T23:34:31 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 240 seconds]
--- Day changed Sun Jan 11 2026
2026-01-11T00:24:37 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 264 seconds]
2026-01-11T00:24:55 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
2026-01-11T00:40:48 < SamaylyS_> why wouldn't you run AVR in 2026? they're simple, easy to learn and very satisfying
2026-01-11T00:43:41 < Steffanx> Those newer ones are quite a bit more complex though 
2026-01-11T00:51:10 < specing> Steffanx: because stm32s are much more performant, have debugger available on every devboard, are cheaper, have much cheaper devboards and are by a non-US company
2026-01-11T00:51:14 < specing> SamaylyS_:^
2026-01-11T00:54:41 < catphish> 0402 decoupling caps don't play nice with 0.5mm pitch chips and single sided assembly :(
2026-01-11T00:55:12 < catphish> we don't all have "turning the board over and soldering on the other side" money!
2026-01-11T00:55:17 < zyp> the only thing I found satisfying last time I touched AVR was finding and fixing a powersave bug that gave me the power budget I needed to implement the extra functionality I was tasked with
2026-01-11T00:56:12 < zyp> AVR seems pretty bad for low power applications
2026-01-11T00:56:42 < catphish> yeah i really don't think that's the right tool for such things
2026-01-11T00:59:34 < zyp> IIRC in the lowest power state, all oscillators are off, so there's nothing to drive a wakeup timer, only external wakeup possible, and it still draws more power than a modern MCU that wakes up every second to do BLE advertisements does on average
2026-01-11T01:00:41 < zyp> and once you turn on an oscillator and a timer, power usage increases significantly
2026-01-11T01:03:23 < catphish> i guess i'll have to reprice this project for double sided assembly, seemsingly some chips just aren't designed for single sided use
2026-01-11T01:03:28 < zyp> I had an old battery powered sensor thing, some old 433 MHz radio, magnet sensor (for doors and such) and there were also a temperature sensor on the board
2026-01-11T01:04:06 < zyp> originally set up to only wake up on an interrupt from the magnet sensor, I was tasked with having it also report the temperature periodically
2026-01-11T01:04:38 < zyp> IIRC the battery capacity and lifetime requirements gave me a power budget on the order of 30uA
2026-01-11T01:05:14 < catphish> well that's an easy easy for an STM32L4, not so much for a AVR
2026-01-11T01:05:24 < catphish> *easy ask
2026-01-11T01:05:32 < zyp> yeah, nrf5x also
2026-01-11T01:05:59 < zyp> IIRC a nrf52 can do BLE advertisements at 1s intervals on the order of 3uA
2026-01-11T01:07:20 < zyp> was kinda hard to meet for the AVR, because according to my measurements the power budget were already almost entirely spent even without wakeup
2026-01-11T01:07:51 < zyp> except I found a code bug that meant it didn't go entirely to sleep
2026-01-11T01:08:43 < zyp> IIRC the watchdog also has to be turned off to achieve lowest power
2026-01-11T01:09:20 < zyp> and to do that, you have to do two register writes within like 4 cycles of each other
2026-01-11T01:10:46 < zyp> and some of the code had gotten rearranged since it was written, adding a few instructions in between the two writes (IIRC there were a loop and an exit condition for the loop ended up between the two writes or something
2026-01-11T01:11:28 < zyp> fixing that so the watchdog got turned off properly gave me the power budget I needed to enable periodic wakeup
2026-01-11T01:15:53 < zyp> after I'd spent like two weeks on this, client pulled the plug on the project because they didn't want to keep sinking money into adding features to an almost obsolete product, so I never got to finish the other side that receives the temperature reports and forwards them to a 2G/GPRS connection :p
2026-01-11T01:18:28 -!- IanW_ [~IceChat9@arcanum.force9.co.uk] has quit [Quit: Bye]
2026-01-11T01:18:55 -!- qyx [~qyx@84.245.121.83] has quit [Ping timeout: 240 seconds]
2026-01-11T01:21:12 -!- qyx [~qyx@84.245.121.81] has joined ##stm32
2026-01-11T03:03:17 < catphish> i've spent too many hours working out how to add decoupling caps to one chip :'(
2026-01-11T03:05:39 < Steffanx> Yolo
2026-01-11T03:06:38 < catphish> https://i.ibb.co/29RLQ2Y/Screenshot-20260111-010539.png - this is one option - 0201 on top of the board
2026-01-11T03:15:54 < zyp> catphish, what chip is this?
2026-01-11T03:16:57 < zyp> 12 decoupling caps on a qfn76 seems excessive
2026-01-11T03:18:10 < zyp> it's mostly BGA that calls for double sided assembly, since they typically got all the power pads in the middle
2026-01-11T03:23:15 < catphish> zyp: it's a FT601Q, i'm inclined to believe i don't need all of them, but i've no real way to know which ones i can drop
2026-01-11T03:24:20 < catphish> there are some at the same voltage 4 pins (2mm) apart, they probably don't really need a cap each
2026-01-11T03:26:20 < catphish> it's still a pretty tight fit with the tracks either side
2026-01-11T03:26:42 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has joined ##stm32
2026-01-11T03:26:42 < zyp> you could move them further away
2026-01-11T03:26:47 < zyp> but yeah
2026-01-11T03:26:57 < zyp> this design also looks like 0201: https://www.hdl.co.jp/USB-107/top.800.jpg
2026-01-11T03:28:20 < zyp> hmm, no, smallest parts are probably still 0402
2026-01-11T03:28:38 < zyp> yeah
2026-01-11T03:29:42 < catphish> the FTDI reference board uses 24 caps on the back!
2026-01-11T03:30:39 < zyp> what are you making?
2026-01-11T03:32:29 -!- duude__ [~duude__@user/duude/x-4676560] has quit [Ping timeout: 244 seconds]
2026-01-11T03:32:42 -!- duude__ [~duude__@user/duude/x-4676560] has joined ##stm32
2026-01-11T03:32:45 < catphish> zyp: USB oscilloscope
2026-01-11T03:33:27 < zyp> analog scope, not just LA?
2026-01-11T03:33:34 < catphish> yes
2026-01-11T03:33:52 < zyp> how's the architecture gonna look?
2026-01-11T03:36:45 < catphish> it's pretty barebones, an analog annenuator, high inpout impedance buffer, differential amplifier into ADC, then into an FPGA which is essntially just to fix up timing, then into the FTDI chip. 2 channels, 8 bit 100Msps
2026-01-11T03:37:38 < catphish> the goal of the project is a super low cost continuous-sampling USB 3 scope, all the magic, triggers etc will be in the host software
2026-01-11T03:38:17 < zyp> right
2026-01-11T03:38:40 < catphish> also, i'm super-cost optimizing to try to get this thing into a hobby budget, i feel like it might work :)
2026-01-11T03:38:59 < zyp> sounds like a SDR receiver without the RF stage
2026-01-11T03:39:11 < catphish> zyp: yep pretty much exactly that
2026-01-11T03:41:31 < catphish> inspired by https://github.com/EEVengers/ThunderScope - but MUCH lower ambitions for spec than theirs
2026-01-11T03:41:41 < zyp> yeah, I've seen that
2026-01-11T03:43:39 < catphish> i think i have no choice but to use 0201 caps here, and since that pushes up the assembly cost, it means i can go tighter on the other manufacturing tolerances and it should all fit fine
2026-01-11T03:44:03 < catphish> ususlly i relentlessly design to fir JLCPCB's cheapest tolerance options
2026-01-11T03:44:07 < catphish> *fit
2026-01-11T03:44:32 < catphish> but that all goes out the window with 0201
2026-01-11T03:44:41 < zyp> ft601q aside, which fpga are you using? wouldn't that come in a bga package that calls for bottom size caps?
2026-01-11T03:45:32 < catphish> i'm using a XO2 in a QFP100
2026-01-11T03:45:46 < catphish> it doesn't need to do much
2026-01-11T04:02:37 < catphish> zyp: thanks for that image above, i found the schematic for it, which is a much more useful example than the official one
2026-01-11T07:47:20 -!- duude__ [~duude__@user/duude/x-4676560] has quit [Ping timeout: 240 seconds]
2026-01-11T07:47:51 -!- duude__ [~duude__@user/duude/x-4676560] has joined ##stm32
2026-01-11T08:17:01 < qyx> y u no picoscope
2026-01-11T08:28:27 < ds2> now try 0.4mm BGA 
2026-01-11T08:49:38 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-11T09:40:30 -!- haritz [~hrtz@user/haritz] has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
2026-01-11T10:39:14 -!- teknix [~unknown@user/hsv] has quit [Ping timeout: 256 seconds]
2026-01-11T10:46:08 -!- teknix [~unknown@user/hsv] has joined ##stm32
2026-01-11T10:51:30 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-11T10:52:03 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-11T12:51:20 -!- IanW_ [~IceChat9@arcanum.force9.co.uk] has joined ##stm32
2026-01-11T12:57:46 -!- Posterdati [~Posterdat@user/Posterdati] has quit [Quit: KVIrc 5.2.6 Quasar http://www.kvirc.net/]
2026-01-11T12:58:30 < catphish> i stayed up basically all night but now i have 0402 caps on a 0.8mm QFP and everything within JLC's cheapest tolerances :)
2026-01-11T12:59:42 -!- Posterdati [~Posterdat@user/Posterdati] has joined ##stm32
2026-01-11T13:49:32 < antto> <silence.wav>
2026-01-11T13:49:45 < antto> this was the sound of a 0402 capacitor falling on the carpet
2026-01-11T14:08:14 < jpa-> carpet is not the best floor for electronics labs, and not just for that reason :)
2026-01-11T14:08:42 < jpa-> i can hear 0402 falling on esd mat
2026-01-11T14:33:21 -!- hexo__ [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-11T14:34:29 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-11T14:37:03 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has quit [Ping timeout: 272 seconds]
2026-01-11T14:43:08 -!- hexo__ [~hexo@user/hexo] has quit [Quit: Leaving]
2026-01-11T14:43:42 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-11T17:04:25 < karlp> jpa-: wirewrap is great!
2026-01-11T17:04:48 < karlp> wanted to get around to demoing something with some hx711 kit I got ages ago, it only has a "grove" connector on it.
2026-01-11T17:05:15 < karlp> and i don't have anything that mates with that, and was going to order shit and shelve this, didn't feel like soldering a bunch of stranded wire to headers
2026-01-11T17:05:23 < karlp> then remembered my new hot wire wrap tool.
2026-01-11T17:06:04 < jpa-> :)
2026-01-11T17:08:07 < karlp> https://imgur.com/a/wire-wrap-love-EciRLjB
2026-01-11T17:10:24 < jpa-> you need to buy more colors
2026-01-11T17:14:28 < BrainDamage> I've abused more than once a wire wrap tool on bodge wires on a connector to solder them more easily
2026-01-11T17:15:11 < BrainDamage> when they are tightly wrapped you can easily handle the connector even if it's a round shape, and solder on
2026-01-11T17:17:30 < mawk> why solder if it's already wire wrapped
2026-01-11T17:17:35 < mawk> I thought it was very solid
2026-01-11T17:18:11 < karlp> jpa-: didn't want to do that until I'd tried it out,
2026-01-11T17:18:13 < karlp> but yeah,
2026-01-11T17:18:21 < mawk> I need to buy one too
2026-01-11T17:18:41 < karlp> BrainDamage: but... why bother with the solder when it's already wrapped ;)
2026-01-11T17:18:45 < mawk> jpa- started the big ##stm32 wire wrapping craze of the late 20s
2026-01-11T17:20:05 < BrainDamage> karlp: because a wrap on a round post is not a permanent solution
2026-01-11T17:20:39 < jpa-> i hate it being "late 20s" now
2026-01-11T17:21:04 < BrainDamage> I've been using it since a decade
2026-01-11T17:22:10 < BrainDamage> a round post wrap may work now, but not eg next week
2026-01-11T17:22:16 < mawk> o
2026-01-11T17:22:39 < BrainDamage> you need the edges to create high pressure points where it cold welds
2026-01-11T17:23:18 < mawk> right
2026-01-11T17:23:26 < BrainDamage> depending on the wire kinks and how it was wrapped, it may or may work on a round post from eccentricity, but you're rolling the dice
2026-01-11T17:23:35 < mawk> score with a knife
2026-01-11T17:23:52 < BrainDamage> a blob of solder is easier at that point
2026-01-11T19:03:16 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-11T20:00:08 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 240 seconds]
2026-01-11T21:06:03 -!- hexo__ [~hexo@user/hexo] has quit [Quit: Leaving]
2026-01-11T21:07:22 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-11T21:19:40 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-11T21:20:37 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-11T22:24:46 -!- Livio [~livio@user/livio] has quit [Quit: leaving]
2026-01-11T22:48:17 < catphish> hello, i would like to request some technical support please.
2026-01-11T22:48:51 < catphish> the datasheet for this FT601Q says this: A general purpose 30 MHz (Frequency=30 MHz, Stability≤±30ppm, ESR≤50Ohm, Load Capacitance=12~18pF) crystal is recommended. Place the crystal and the load capacitor on the same layer and near to the FT60x.
2026-01-11T22:49:16 < catphish> any idea how i might work out what (if any) capacitors i need to add to a crystal to achieve their requirement?
2026-01-11T22:50:07 < catphish> there are lots of 12-18pf crystals, but i have no idea how to load them correctly
2026-01-11T22:56:25 < catphish> oh, the drive circuit can be ignored, so the calculation is based only on the crystal and PCB, that's easy then
2026-01-11T22:59:19 < Phantom> or just ignore the calculation and put a 12-18pf cap and be done with it
2026-01-11T23:05:42 < catphish> the reference circuit uses 18pf caps. i suspect that will work just fine, especially as i have no chance to calculate the stray capacitance of my circuit
2026-01-11T23:06:11 < catphish> 12-5
2026-01-11T23:07:11 < Phantom> it slightly detune the crystal, but barely, nothing that the circuit will notice anyway
2026-01-11T23:09:03 < catphish> apparently a sane assumption is 5pf for stray capacitance, so 12pf crystal = 14pf caps, 18pf crystal = 26pf caps
2026-01-11T23:09:21 < catphish> so i'm going to take your suggestion and not worry too much
2026-01-11T23:10:22 < catphish> i'll just aim for around 16pf of everything :)
2026-01-11T23:26:43 -!- hexo [~hexo@user/hexo] has quit [Quit: Leaving]
2026-01-11T23:28:36 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-11T23:57:37 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 264 seconds]
2026-01-11T23:58:55 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
--- Day changed Mon Jan 12 2026
2026-01-12T00:03:06 < karlp> catphish: https://github.com/karlp/zypsnips/blob/master/crystal-load-wisdom.txt
2026-01-12T00:03:33 < karlp> 12*2 - a bit ~ sure, 16...
2026-01-12T00:05:54 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has joined ##stm32
2026-01-12T00:39:55 < catphish> karlp: the formula i found was 2 x (n - some)
2026-01-12T00:40:26 < catphish> not dissimilar to your formula, your "some" is just bigger
2026-01-12T00:40:42 < catphish> "a bit" > "some"
2026-01-12T01:57:05 -!- IanW_ [~IceChat9@arcanum.force9.co.uk] has quit [Quit: Bye]
2026-01-12T02:01:46 -!- hexo [~hexo@user/hexo] has quit [Read error: Connection reset by peer]
2026-01-12T02:02:05 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-12T02:06:18 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-12T02:07:30 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-12T02:08:05 -!- splud [~noneya.bi@user/splud] has quit [Read error: Connection reset by peer]
2026-01-12T02:16:11 -!- hexo [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-12T02:18:31 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-12T02:22:48 -!- splud [~noneya.bi@user/splud] has joined ##stm32
2026-01-12T02:31:22 -!- hexo [~hexo@user/hexo] has quit [Ping timeout: 244 seconds]
2026-01-12T03:23:30 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-12T03:49:28 -!- splud [~noneya.bi@user/splud] has quit [Ping timeout: 246 seconds]
2026-01-12T04:08:08 -!- splud [~noneya.bi@user/splud] has joined ##stm32
2026-01-12T08:11:53 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-12T09:01:05 < qyx> I am using 12p for 9p crystals, 18p for 12p crystals
2026-01-12T09:01:19 < qyx> I do not buy other types of crystals
2026-01-12T09:04:47 < jpa-> https://electronics.stackexchange.com/questions/698199/stm32f417-dfu-bootloader-fails-to-start-hse-crystal   >= 18 pF crystals are worth avoiding
2026-01-12T09:07:51 < Phantom> not the other way around?
2026-01-12T09:07:57 < qyx> loljpa
2026-01-12T09:08:21 < qyx> so their rated startup time is 0.74ms and they are using 0.79ms timeout?
2026-01-12T09:09:16 < jpa-> just ST being ST ;)
2026-01-12T09:09:40 < jpa-> but 18 pF crystals easily exceed other datasheet crystal specs for STM32 too
2026-01-12T09:09:48 < Phantom> hmm   will have to recheck what I used, just because  :D
2026-01-12T09:09:57 < jpa-> (by exceed i mean are outside allowed range)
2026-01-12T10:12:53 -!- c10ud [~c10ud@user/c10ud] has joined ##stm32
2026-01-12T10:55:00 < jpa-> https://github.com/torvalds/AudioNoise/commit/93a72563cba609a414297b558cb46ddd3ce9d6b5   torvalds is vibe-coding rp2350/python and internet explodes
2026-01-12T11:14:22 < qyx> This is a historic moment, software and human civilization will never be the same
2026-01-12T11:25:19 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-12T11:26:01 < Steffanx> Lolwut
2026-01-12T11:26:35 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-12T11:44:49 < mawk> ohno
2026-01-12T11:44:57 < mawk> what happened to my hero
2026-01-12T11:47:00 < antto> he folded... foldvalds
2026-01-12T11:51:31 < jpa-> meanwhile, i'm in trouble because i have assisted agency overreach on genuine emergent sentient AGI    http://paste.dy.fi/TpC/plain
2026-01-12T12:00:37 < catphish> jpa-: what on earth did i just read?
2026-01-12T12:06:07 < mawk> lol jpa- 
2026-01-12T12:06:13 < mawk> do you know what device he found this in?
2026-01-12T12:06:21 < mawk> and what library is he talking about 
2026-01-12T12:06:33 < karlp> nanopb, which jpa wrote, and is used ~everywhere.
2026-01-12T12:06:39 < karlp> so he gets to deal with the masses.
2026-01-12T12:07:57 < mawk> ah I see
2026-01-12T12:08:16 < jpa-> https://gist.github.com/gorenje/a775f44bb9307852911f7f3534666442   based on this google finding mentioning both zemberek and nanopb, i think it might be iphone
2026-01-12T12:08:26 < polprog> amazing
2026-01-12T12:09:00 < polprog> jpa-: i think he should also sue the chip designers
2026-01-12T12:09:30 < polprog> unrelated: what is kapsi.fi? ive seen this domain pop up a few times since i started using IRC, is it something like pubnix?
2026-01-12T12:09:49 < jpa-> it's a finnish association that started offering web space and ssh shell servers back in 2004
2026-01-12T12:10:28 < mawk> käpsïfï
2026-01-12T12:10:39 < polprog> oh, thats very neat
2026-01-12T12:10:44 < jpa-> no ï in finnish
2026-01-12T12:29:14 < qyx> I am surprised there is only a single k in kapsi
2026-01-12T12:31:18 < qyx> what the  hell is that AGI thing
2026-01-12T12:34:04 < polprog> artificial general intelligence
2026-01-12T12:34:09 < polprog> people who think LLMs are sentient
2026-01-12T12:34:31 < polprog> and that just one more month and we will have AI do all the work for us, and we just party and get money for free
2026-01-12T12:35:11 < antto> how many more months till it cleans the cat sandbox
2026-01-12T12:35:20 < ventyl> what a shame AI isn't able to write a piece of code that fills 5 registers correctly
2026-01-12T12:35:40 < antto> or throws out the trash
2026-01-12T12:35:44 < polprog> antto: best i can do is an app that reminds you to clean the sandbox. It uses 5 AWS VMs
2026-01-12T12:35:52 < ventyl> but again... sentient doesn't imply smart
2026-01-12T12:35:57 < antto> hm... if it throws out the trash, there's a chance that it might not come back ;P~
2026-01-12T12:36:38 < antto> Absolute Garbage Idiotism
2026-01-12T12:38:42 < jpa-> i love the licenses of some of those apple notices.. "getopt.c: This file was copied from the following newsgroup posting:  ... It is the code which was given out at the 1985 UNIFORUM conference in Dallas. ... The people there assure me that it is indeed in the public domain."
2026-01-12T12:44:22 < polprog> :-D
2026-01-12T12:54:17 < qyx> are there no LEDs in LCSC basic parts?
2026-01-12T12:56:05 < Steffanx> There used to be.
2026-01-12T12:56:51 < Steffanx> And there are 6 in jlcpcb part library
2026-01-12T12:57:37 < Steffanx> They only have 7 million red ones in stock.
2026-01-12T12:58:04 < qyx> yes there is a ton in extended
2026-01-12T12:58:29 < Steffanx> 6 basic ones I meant.
2026-01-12T13:08:47 < qyx> is 	
2026-01-12T13:08:58 < qyx> Foshan NationStar better than Xinglight?!
2026-01-12T13:09:14 < karlp> there's only a couple of leds in basic, it's not great
2026-01-12T13:09:29 < qyx> I need RGB in PLCC4 but I could pay $3
2026-01-12T13:10:17 < qyx> the state of PCBA in this part of the EU forces me to use china services
2026-01-12T13:10:36 < qyx> I am not willing to spend any more time populating prototypes
2026-01-12T13:10:37 < karlp> you get red in 805 and 603, then a few other colours in 805, and somehow yellow in 603 as well, that's it.
2026-01-12T13:13:01 < karlp> new button in basic though, that's nice: https://jlcpcb.com/partdetail/XUNPU-TS_1088AR02016/C720477
2026-01-12T13:13:07 < karlp> I never liked the other button that was in basic
2026-01-12T13:14:20 < karlp> hecking f103 is in basic though :|
2026-01-12T13:21:03 < qyx> g491 is 10 eur each?
2026-01-12T13:21:11 < qyx> wtf
2026-01-12T13:21:16 < qyx> not cloned maybe
2026-01-12T13:21:47 < qyx> and only 10 in stock
2026-01-12T13:27:18 < karlp> only qfns are expensive :)
2026-01-12T13:27:57 < karlp> lqfp100 is only $3
2026-01-12T14:00:13 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has quit [Ping timeout: 246 seconds]
2026-01-12T14:03:46 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has joined ##stm32
2026-01-12T14:14:27 < qyx> anyone did piezo speaker with max232?
2026-01-12T14:16:58 < jpa-> such hifi, much bass
2026-01-12T14:17:42 < qyx> hm PAM8906 looks a bit better
2026-01-12T14:25:21 -!- specing [~specing@user/specing] has quit [Ping timeout: 252 seconds]
2026-01-12T14:27:47 -!- specing [~specing@user/specing] has joined ##stm32
2026-01-12T14:33:21 < antto> did someone say PAM8906
2026-01-12T14:35:05 < antto> oh, that's not what i thought it is
2026-01-12T14:49:39 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has quit [Quit: Client closed]
2026-01-12T15:10:24 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-12T15:23:31 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-12T15:43:43 -!- blathijs [~matthijs@tika.stderr.nl] has quit [Ping timeout: 240 seconds]
2026-01-12T16:08:37 -!- blathijs [~matthijs@tika.stderr.nl] has joined ##stm32
2026-01-12T16:10:17 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-12T16:13:12 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-12T16:54:23 -!- hexo [~hexo@user/hexo] has quit [Quit: Leaving]
2026-01-12T17:09:40 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-12T17:13:43 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-12T17:16:45 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-12T17:17:26 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-12T17:48:36 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-12T18:11:13 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-12T19:06:35 -!- flatmush [~benbrewer@149.71.29.67] has quit [Ping timeout: 240 seconds]
2026-01-12T19:16:48 -!- flatmush [~benbrewer@149.57.118.149] has joined ##stm32
2026-01-12T19:49:56 -!- umbramalison [~umbramali@188.74.90.42] has quit [Quit: %So long and thanks for all the fish%]
2026-01-12T20:08:50 -!- octorian [~octo@chroniton.logicprobe.org] has quit [Remote host closed the connection]
2026-01-12T20:12:05 -!- octorian [octo@2600:3c00::f03c:91ff:fe93:a61c] has joined ##stm32
2026-01-12T20:25:40 -!- umbramalison [~umbramali@188.74.90.42] has joined ##stm32
2026-01-12T22:24:55 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 264 seconds]
2026-01-12T22:56:14 < ventyl> term has been coined
2026-01-12T22:56:18 < ventyl> Microslop Copilot
--- Day changed Tue Jan 13 2026
2026-01-13T00:36:31 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
2026-01-13T00:37:43 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
2026-01-13T01:15:59 -!- hackkitten [~hackkitte@2a00:6020:ad3f:300::35e] has quit [Ping timeout: 244 seconds]
2026-01-13T01:19:49 -!- qyx [~qyx@84.245.121.81] has quit [Ping timeout: 264 seconds]
2026-01-13T01:21:13 -!- qyx [~qyx@84.245.120.206] has joined ##stm32
2026-01-13T01:47:01 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has joined ##stm32
2026-01-13T02:16:44 -!- hackkitten [~hackkitte@2a00:6020:ad3f:300:433f:7942:6b1a:37f8] has joined ##stm32
2026-01-13T02:42:52 -!- hackkitten [~hackkitte@2a00:6020:ad3f:300:433f:7942:6b1a:37f8] has quit [Ping timeout: 246 seconds]
2026-01-13T02:55:05 -!- hackkitten [~hackkitte@94.31.115.2] has joined ##stm32
2026-01-13T03:02:44 -!- hackkitten [~hackkitte@94.31.115.2] has quit [Ping timeout: 260 seconds]
2026-01-13T03:43:49 -!- hackkitten [~hackkitte@94.31.115.40] has joined ##stm32
2026-01-13T04:04:55 -!- hackkitten [~hackkitte@94.31.115.40] has quit [Ping timeout: 240 seconds]
2026-01-13T04:09:02 -!- System_Error [~SystemErr@user/systemerror] has quit [Remote host closed the connection]
2026-01-13T04:16:03 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-13T04:49:41 -!- hackkitten [~hackkitte@94.31.119.95] has joined ##stm32
2026-01-13T05:11:34 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-13T05:16:41 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-13T05:20:50 -!- jfsimon1981 [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-13T05:22:57 -!- Miyu [~hackkitte@2a00:6020:ad3f:300:433f:7942:6b1a:37f8] has joined ##stm32
2026-01-13T05:23:19 -!- hackkitten [~hackkitte@94.31.119.95] has quit [Read error: Connection reset by peer]
2026-01-13T05:27:29 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-13T05:32:36 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Max SendQ exceeded]
2026-01-13T05:34:46 -!- SamaylyS_ [~SamaylyS@user/SamaylyS] has quit [Remote host closed the connection]
2026-01-13T05:35:21 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-13T06:18:09 -!- Miyu [~hackkitte@2a00:6020:ad3f:300:433f:7942:6b1a:37f8] has quit [Ping timeout: 260 seconds]
2026-01-13T06:37:24 -!- hackkitten [~hackkitte@94.31.119.95] has joined ##stm32
2026-01-13T07:54:15 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-13T09:07:53 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has joined ##stm32
2026-01-13T09:39:43 -!- kurfen_ [~kurfen@194.127.173.38] has joined ##stm32
2026-01-13T09:40:15 -!- kurfen [~kurfen@194.127.173.38] has quit [Ping timeout: 240 seconds]
2026-01-13T10:08:26 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-13T10:08:58 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-13T11:51:21 < qyx> hm lcsc even has my bridgelux LEDs, preorder only though
2026-01-13T11:51:29 < qyx> min 97 pcs and kinda good price too
2026-01-13T12:48:35 -!- hexo__ [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-13T12:49:00 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-13T12:54:42 < qyx> hey mqtt pros, my persistence file is always 47 bytes long
2026-01-13T12:54:49 < qyx> (mosquitto broker)
2026-01-13T13:00:48 < jpa-> how long would you like it to be?
2026-01-13T13:01:54 < qyx> sufficiently long to actually containt some data
2026-01-13T13:03:14 < qyx> hm there is autosave_interval
2026-01-13T13:06:27 < qyx> also SIGUSR1 works
2026-01-13T13:07:06 < qyx> well, there is a problem with this particular version of mosquitto, it has a bug preventing a disconnected bridge to reconnect
2026-01-13T13:07:33 < qyx> so I need to do cron to grep established connections and once it disappears, I restart the mosquitto
2026-01-13T13:07:59 < qyx> the problem is the data during that time is lost because it is not being saved
2026-01-13T13:15:23 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-13T13:31:33 -!- duude__- [~duude__@user/duude/x-4676560] has joined ##stm32
2026-01-13T13:31:49 -!- duude__ [~duude__@user/duude/x-4676560] has quit [Ping timeout: 264 seconds]
2026-01-13T13:33:27 -!- duude__- is now known as duude__
2026-01-13T14:26:41 -!- nomorekaki [~nomorekak@188-67-208-161.bb.dnainternet.fi] has quit [Ping timeout: 272 seconds]
2026-01-13T16:56:50 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has quit [Remote host closed the connection]
2026-01-13T17:04:01 < catphish> i'm putting an STM32H7 into standby mode and it seems to be using rather a lot of current (350uA). i'm not 100% certain it's the chip, but i suspect it is. should i be shutting anything down before entering standby mode? not sure what i'm missing. i was expecting more like 10uA
2026-01-13T17:05:04 < catphish> oh, it's my debugger. it drops to 4uA when i disconnect the st-link. yay!
2026-01-13T17:05:29 < jpa-> also make sure you have no floating input pins (enable pulldown or pullup on unconnected pins)
2026-01-13T17:09:50 < catphish> i hardwired all the unused pins to GND on the board :)
2026-01-13T17:10:18 < catphish> i don't think the H7 supports pullup/pulldown in standby
2026-01-13T17:10:22 < zyp> jpa-, or analog mode
2026-01-13T17:10:40 < zyp> floating inputs is not an issue if the input buffer is turned off
2026-01-13T17:11:15 < zyp> catphish, if pullup/pulldown is disabled, input buffers are probably also
2026-01-13T17:11:27 < catphish> zyp: that's a good point
2026-01-13T17:11:38 < catphish> anyway, it's down at 4uA which is what the datasheet says is typical
2026-01-13T17:12:08 < zyp> yeah, that sounds reasonable
2026-01-13T17:12:35 < zyp> and yeah, you absolutely cannot have a debugger connected while doing low power measurements
2026-01-13T17:13:02 < catphish> that's 125,000 hours on a 500mAh 9V battery in a handheld device with a user replaceable battery, so i won't worry further
2026-01-13T17:14:14 < catphish> yeah, that's 14 years, i think the batteries will die of old age before i drain them, my target was 20uA so i'm super happy
2026-01-13T17:15:22 < zyp> what sort of device is this?
2026-01-13T17:15:22 < catphish> the only flaw with this device is that the encoder shaft is too long and hence the knob isn't very flush
2026-01-13T17:17:21 < catphish> it's a handheld thingemy with a DAC+ADC used for diagnostics of sonar
2026-01-13T17:44:45 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-13T17:45:49 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-13T17:46:05 -!- haritz [~hrtz@2a01:4b00:bc2e:7000:d5af:a266:ca31:5ef8] has joined ##stm32
2026-01-13T17:46:05 -!- haritz [~hrtz@user/haritz] has changed host
2026-01-13T17:55:49 < karlp> heh nice, got some molex premade picoblade cables.
2026-01-13T17:56:12 < karlp> they are only made such that pin1 on one side goes to pin 4 on the other.
2026-01-13T17:56:55 < karlp> whole line is all wired like that: https://www.molex.com/content/dam/molex/molex-dot-com/products/automated/en-us/salesdrawingpdf/151/15134/151340402_sd.pdf?inline
2026-01-13T17:57:08 < karlp> (this was not how the boards were designed of course ;)
2026-01-13T17:58:12 < jpa-> why would they do that..
2026-01-13T17:59:53  * karlp shrugs
2026-01-13T18:00:38 < karlp> I'm not a fan of the molex wiring though, it means you normally need two different schematic symbols, one for each end of the cable, fuck that noise.
2026-01-13T18:01:08 < jpa-> and marking "pin 1" on the PCB becomes confusing
2026-01-13T18:07:28 < zyp> only way it makes sense is if you designed the pinouts to be symmetric with tx/rx crossing
2026-01-13T18:07:55 < zyp> but yeah, it's the top/top FFC problem all over again
2026-01-13T18:08:41 < karlp> yup
2026-01-13T18:09:00 < karlp> they've already got some boards in house with picoblade too, can't use jst or somethign with more china cable options.
2026-01-13T18:09:10 < karlp> going to walk away from this shitfire :)
2026-01-13T18:10:18 < karlp> then you get datasheets like this, that _look_ like they do the right thing, but the table says a different thing? https://jlcpcb.com/api/file/downloadByFileSystemAccessId/8590897532048388097
2026-01-13T18:29:13 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-13T18:33:14 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-13T18:34:17 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-13T18:39:45 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-13T18:41:02 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Remote host closed the connection]
2026-01-13T18:41:41 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has joined ##stm32
2026-01-13T18:41:55 -!- specing [~specing@user/specing] has quit [Ping timeout: 240 seconds]
2026-01-13T18:44:26 < machinehum> catphish: Any update with the ch569?
2026-01-13T18:45:24 < catphish> machinehum: only that i dropped it from consideration for now because its parallel interface is too opinionated, it requires framing and checksums
2026-01-13T18:45:55 < catphish> i'm sure it can be implemented in FPGA but i am trying a simpler approach first with the FT600Q
2026-01-13T18:46:39 < machinehum> What were you thinking of using it for ?
2026-01-13T18:48:54 < catphish> i'm building a very simple USB oscilloscope
2026-01-13T18:49:05 < catphish> simple but fast
2026-01-13T18:49:35 < machinehum> Cool
2026-01-13T19:09:52 < qyx> catphish: you have to disable the debug interface
2026-01-13T19:10:54 < catphish> qyx: thanks, luckily i figured that out fairly quickly :)
2026-01-13T19:13:13 < machinehum> catphish: Have you tried building shit in their sdk or you didn't get that far
2026-01-13T19:13:33 < machinehum> That part might be nice for my secure USB drive v2
2026-01-13T19:13:40 < catphish> machinehum: i didn't get that far. i have ordered a dev board in case i want to play with it later
2026-01-13T19:13:48 < machinehum> catphish: All good
2026-01-13T19:14:04 < machinehum> You just got them from aliexpress?
2026-01-13T19:28:04 -!- bitmask [~bitmask@static-23-234-100-191.cust.tzulo.com] has joined ##stm32
2026-01-13T19:30:32 -!- bitmask [~bitmask@static-23-234-100-191.cust.tzulo.com] has quit [Client Quit]
2026-01-13T19:34:10 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-13T19:35:12 < catphish> yes
2026-01-13T19:40:33 -!- Sadale [~Sadale@user/sadale] has quit [Quit: Bye!]
2026-01-13T19:40:59 -!- Sadale [~Sadale@user/sadale] has joined ##stm32
2026-01-13T21:53:40 -!- specing [~specing@user/specing] has joined ##stm32
2026-01-13T22:00:45 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 245 seconds]
2026-01-13T22:01:39 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-13T23:03:52 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
--- Day changed Wed Jan 14 2026
2026-01-14T00:28:56 -!- System_Error [~SystemErr@user/systemerror] has quit [Ping timeout: 252 seconds]
2026-01-14T00:36:17 -!- System_Error [~SystemErr@user/systemerror] has joined ##stm32
2026-01-14T01:19:15 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 240 seconds]
2026-01-14T01:20:19 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 260 seconds]
2026-01-14T01:48:08 -!- bitmask [~bitmask@static-23-234-100-191.cust.tzulo.com] has joined ##stm32
2026-01-14T01:52:45 -!- bitmask [~bitmask@static-23-234-100-191.cust.tzulo.com] has quit [Ping timeout: 252 seconds]
2026-01-14T04:14:32 -!- c10ud [~c10ud@user/c10ud] has quit [Read error: Connection reset by peer]
2026-01-14T06:33:31 -!- cutofmyjib [~cutofmyji@user/cutofmyjib] has quit [Quit: WeeChat 4.1.1]
2026-01-14T08:02:37 -!- catphish [~quassel@user/catphish] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
2026-01-14T08:02:52 -!- catphish [~quassel@user/catphish] has joined ##stm32
2026-01-14T08:04:30 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has joined ##stm32
2026-01-14T08:10:43 -!- bitmask [~bitmask@ool-43525d25.dyn.optonline.net] has quit [Ping timeout: 246 seconds]
2026-01-14T09:07:30 < qyx> hey karlp 
2026-01-14T09:07:47 < qyx> what do you use to prepare data for jlcpcb pcba orders?
2026-01-14T09:07:52 < jpa-> i use kikit
2026-01-14T09:08:53 < qyx> for panelization?
2026-01-14T09:09:16 < qyx> I mean for matching parts
2026-01-14T09:09:24 < jpa-> i usually only do proto orders so no panelization, or let jlcpcb panelize
2026-01-14T09:09:42 < jpa-> kikit takes the part numbers from "LCSC" attribute in the schematic and makes the bom file
2026-01-14T09:09:45 < qyx> I tried the automatching thing on their web baed on my comments like "1u 0603 X7R"
2026-01-14T09:09:59 < qyx> oh but you don't need kikit for that, do you,
2026-01-14T09:10:09 < qyx> you can do BOM with any custom attribute
2026-01-14T09:10:24 < jpa-> and the kikit schematic symbol table editor nicely groups by value, so you can easily put the same code for e.g. all 1 µF caps
2026-01-14T09:10:37 < jpa-> sure, you can do BOM and placement and gerber export manually and click around in dialogs
2026-01-14T09:10:47 < jpa-> or you can use one repeatable kikit command
2026-01-14T09:10:56 < qyx> and the BOM dialog does that too
2026-01-14T09:11:09 < qyx> you can group by whatever you want and assign in groups
2026-01-14T09:11:20  * qyx confused
2026-01-14T09:11:41 < jpa-> yes, i agree you *can* do all of this with plain kicad
2026-01-14T09:11:55 < antto> i haven't used LCSC for PCBs, but i use my Excelent BOM tool every time
2026-01-14T09:12:20 < antto> especially for merging BOMs from different projects
2026-01-14T09:12:23 < jpa-> but i like it that i just run "kikit fab pcbway ..." and don't have the risk of having a wrong option checked in some dialog
2026-01-14T09:12:24 < qyx> but anyway, first thing first
2026-01-14T09:12:30 < jpa-> or forgetting to update drill file or something..
2026-01-14T09:12:33 < qyx> match components and assign p/n
2026-01-14T09:13:01 < qyx> is LCSC some standard thing? I named mine JLCPCB
2026-01-14T09:13:04 < jpa-> put the Cxxxxx code into your components, don't bother trying to match by some manufacturer code
2026-01-14T09:13:41 < qyx> yeah
2026-01-14T09:13:45 < jpa-> "LCSC" is just the default name in kikit, you can override it
2026-01-14T09:13:48 < qyx> that I got after 3 hours
2026-01-14T09:14:17 < jpa-> (oh, and another nice thing: kikit has JLCPCB_CORRECTION field so that you can fix PCBA part rotation if needed)
2026-01-14T09:15:04 < qyx> oh yes, how can be half of the parts wrong
2026-01-14T09:15:15 < jpa-> because there is no standard :/
2026-01-14T09:15:25 < antto> https://antonsavov.net/tmp/samd21_daplink-pmkb_20240608.html
2026-01-14T09:15:26 < qyx> I mentioned that in one company recently and they yelled at me "YOU SHALL MAKE YOUR OWN LIBRARIES"
2026-01-14T09:15:43 < qyx> I was like what? for all components?
2026-01-14T09:15:47 < jpa-> and for some ridiculous reason PCBA shops will just manually "fix" things and assemble your STM32's wrong way around
2026-01-14T09:15:57 < qyx> "YES FOR ALL COMPOENNTS INCLUDING RESISTORS, STOCK LIBRARIES ARE WRONG"
2026-01-14T09:16:04 < qyx> full dongs mode
2026-01-14T09:16:09 < jpa-> qyx: yeah i have met such people too :D
2026-01-14T09:16:27 < jpa-> some people also want e.g. 100 kohm resistor to be a different component than 47 kohm resistor..
2026-01-14T09:16:47 < jpa-> and think that this somehow "takes time to set up but saves time later"
2026-01-14T09:17:03 < antto> it can be if you use atomic libraries
2026-01-14T09:17:14 < qyx> no.
2026-01-14T09:17:16 < jpa-> what is atomic library?
2026-01-14T09:17:34 < antto> in the new-ish kicad this doesn't have to be a distinct symbol, it can be a symbol alias from a database/httplib library
2026-01-14T09:17:46 < antto> so basically Device:R with metadata on top
2026-01-14T09:17:58 < jpa-> i mean sure, you *can* do that, but why *would* you do that?
2026-01-14T09:18:15 < antto> the metadata includes different value, description, MPN, etc.. but the symbol is still Device:R
2026-01-14T09:18:18 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has joined ##stm32
2026-01-14T09:18:31 < antto> well, that's one way to work
2026-01-14T09:18:55 < qyx> of course the symbol is still the same
2026-01-14T09:18:58 < antto> if you need to have a BOM with MPNs at the end, you might wanna set this all up early
2026-01-14T09:19:01 < qyx> all resistors have the same symbol
2026-01-14T09:19:19 < antto> or... you might do it at the very end
2026-01-14T09:19:33 < qyx> but there are different footprints and you can assign different MPN to them
2026-01-14T09:19:44 < antto> so?
2026-01-14T09:19:49 < jpa-> antto: but doesn't it get annoying if you want to change a resistor value? then you need to go through the whole "change symbol..." dialog etc, and be careful not to edit the value instead
2026-01-14T09:19:50 < antto> what's the problem
2026-01-14T09:19:51 < qyx> i don§t follow
2026-01-14T09:20:16 < qyx> so you have many distinct resistors in the *symbol* library, for each value?
2026-01-14T09:20:18 < antto> jpa-, yes...
2026-01-14T09:20:26 < qyx> and you don't use the value field altogether?
2026-01-14T09:20:38 < antto> qyx, no, i don't yet
2026-01-14T09:20:51 < jpa-> yeah, i have seen people promote that style of working, but i don't get what is the benefit
2026-01-14T09:20:52 < qyx> sounds a bit wrong
2026-01-14T09:20:55 < antto> i am waiting for the httplib API to stop changing
2026-01-14T09:21:24 < jpa-> if updating the BOM last is such a big fuss, you could probably make some tool that has predefined list of "preferred parts" for e.g. 0603 passives
2026-01-14T09:21:49 < jpa-> and autoassigns them if footprint and value match
2026-01-14T09:21:54 < antto> i already have a BOM tool, and an httplib server tool
2026-01-14T09:22:11 < antto> how many more tools must i make
2026-01-14T09:22:26 < antto> before i end up making a whole EDA
2026-01-14T09:22:44 < jpa-> i've done without any.. i just don't think the approach of bazillion symbols really solves anything
2026-01-14T09:23:08 < jpa-> you'd need a tool to make the symbols and update the symbols (e.g. when your preferred resistor supplier changes)
2026-01-14T09:23:09 < antto> it does only when you know in advance exactly what parts you wanna put
2026-01-14T09:23:32 < jpa-> but even *if* you know, you can just type it into the fields right there and then..
2026-01-14T09:23:36 < antto> because changing your mind afterwards requires some unpleasant actions
2026-01-14T09:24:23 < antto> jpa-, i've done it that way... work "quickly" on the schematic, and at the very end - go thru the components and put all the missing metadata
2026-01-14T09:24:30 < antto> i know how it works, i don't like it
2026-01-14T09:24:42 < antto> sometimes i know in advance what kind of resistors i wanna use
2026-01-14T09:25:27 < jpa-> so will you have two different symbol libraries, for the times when you know and when you don't know?
2026-01-14T09:25:29 < antto> anyway, eventually i will be able to test this (when httplib api gets sane and i update my server tool) and i'll see if it's good or not
2026-01-14T09:25:39 < antto> eh?!
2026-01-14T09:25:44 < antto> why would i need two?!
2026-01-14T09:25:59 < antto> the httplib library is just like a real one
2026-01-14T09:26:51 < antto> when i need a resistor, i can either insert Device:R, or i can pick MyHTTPLIB:R_0603_47k_RoyalOhm or however i'd name it
2026-01-14T09:27:18 < qyx> yes and thats wrong
2026-01-14T09:27:19 < jpa-> ok, so yeah, two symbol libs ("Device" and "MyHTTPLIB")
2026-01-14T09:27:30 < antto> jpa-, it's more than that
2026-01-14T09:27:33 < qyx> anyway, why jlc doesn'ŧ have x7r caps
2026-01-14T09:27:50 < qyx> only x5r
2026-01-14T09:28:04 < antto> jpa-, the kicad stock lib appears as a pile of actual symbol libraries, Device is just one of them
2026-01-14T09:28:20 < jpa-> sure
2026-01-14T09:28:25 < antto> the HTTPLIB can appear as several libs
2026-01-14T09:28:51 < jpa-> qyx: because they favor price.. and rarely does the difference matter, just put a larger x5r cap instead
2026-01-14T09:28:54 < antto> because there are categories - each category appears as a sepparate lib in kicad, but setup-wise it's one HTTPLIB
2026-01-14T09:29:14 < antto> at least that's how the v1 API worked
2026-01-14T09:29:15 < qyx> jpa-: yeah but that's an existing design :S
2026-01-14T09:29:18 < qyx> but I get it
2026-01-14T09:31:12 < jpa-> (bigger as in bigger capacitance, not necessarily footprint)
2026-01-14T09:31:32 < antto> perhaps, if you're forced to use a nasty (very crap) ERP/inventory system, you might look for salvation in an HTTPLIB
2026-01-14T09:31:59 < antto> it might be the wrong place, but what can you do
2026-01-14T09:32:14 < jpa-> or it could be a nice way to share libs between coworkers
2026-01-14T09:32:20 < jpa-> from git or something
2026-01-14T09:32:27 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 244 seconds]
2026-01-14T09:32:43 < antto> the HTTPLIB is not really like the normal libs
2026-01-14T09:33:00 -!- hexo [~hexo@user/hexo] has quit [Ping timeout: 265 seconds]
2026-01-14T09:33:17 < antto> if you make a new symbol for an IC that didn't exist in kicad - this is a real Symbol
2026-01-14T09:33:45 < antto> the database/httplib libraries are only metadata libs, they contain symbol aliases, flavored with text
2026-01-14T09:33:49 < jpa-> ah, yeah - ok, i understood httplib wrong
2026-01-14T09:34:25 < antto> in a normal Real lib, you cannot make an alias to Device:R unless you make it in Device lib, which is a stock lib
2026-01-14T09:34:42 < antto> in other words, you can't make symbol aliases from another lib
2026-01-14T09:35:15 < jpa-> it could actually make sense if they'd somehow make it so that the symbol is "Device:R" afterwards, with just a weak reference to the original httplib metadata definition
2026-01-14T09:35:43 < antto> httlibs and database libs allow exactly this - an httplib library can have aliases to Device:R, Transistor:Q_BCE, etc..
2026-01-14T09:36:15 < jpa-> but i mean, that you can edit the part value etc. afterwards just like as if it was a Device:R placed with prefilled fields
2026-01-14T09:36:19 < antto> what it can't have is an actual real symbol, it doesn't define drawings or pins
2026-01-14T09:36:48 < antto> how "nice" these new kinds of libs will work UI-wise depends on how it's implemented
2026-01-14T09:36:55 < jpa-> agreed
2026-01-14T09:37:47 < jpa-> i guess ideal would be if for each Device:R you could have some kind of "Choose part.." that would search the metadata aliases, preferrably keying by value field
2026-01-14T09:37:51 < antto> this actually gives me an idea
2026-01-14T09:37:57 -!- haritz [~hrtz@user/haritz] has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
2026-01-14T09:38:13 < antto> which i should submit to gitlab
2026-01-14T09:38:21 < jpa-> so either you can place your 47kohm 0805 directly, or later just pick it when you have Device:R 47kohm
2026-01-14T09:38:45 < jpa-> (and have the httplib server interface to lcsc or digikey database ;)
2026-01-14T09:38:53 < antto> because a lot of symbols aliases for resistors/caps can be generated by an algorithm in a JIT way instead of sitting there all the time
2026-01-14T09:39:22 < antto> but this will require a change in the API
2026-01-14T09:40:26 < antto> the symbol picker searchbox should send the text written there to the httplib server, and then the server can decide if you're looking for a resistor of some value, and if you are - it can JIT "synthesize" part numbers for compatible resistors based on algorithm(s)
2026-01-14T09:40:51 < antto> mmmm, this won't resonate well with one of the devs tho
2026-01-14T09:41:03 < antto> because of the "symbol caching"
2026-01-14T09:41:05 < antto> bluh
2026-01-14T09:41:18 < antto> oh well, prepare you wallet for More RAM
2026-01-14T09:41:41 < jpa-> maybe we could have an LLM hallucinate parts on the fly ;)
2026-01-14T09:42:09 < antto> nah, the API doesn't work that way
2026-01-14T09:42:38 < antto> it did, v1 API did, but one of the devs wants to cache the whole httplib on load and never again
2026-01-14T09:42:45 < antto> which means "JIT" won't work
2026-01-14T09:42:58 < jpa-> too bad
2026-01-14T09:43:06 < antto> unless someone changes his mind
2026-01-14T09:43:19 -!- martinmoene [~martinmoe@132.229.46.129] has quit [Read error: Connection reset by peer]
2026-01-14T09:43:27 < antto> it depends on the number of ThumbsUp
2026-01-14T09:43:46 < antto> maybe i'll write my idea anyway and leave it to the people
2026-01-14T09:45:53 -!- c10ud [~c10ud@user/c10ud] has joined ##stm32
2026-01-14T09:46:09 < nohit> 09:20:16 <qyx> so you have many distinct resistors in the *symbol* library, for each value?
2026-01-14T09:46:37 < nohit> this is how we work in my current customer project
2026-01-14T09:46:46 < qyx> NOOOOO
2026-01-14T09:46:50 < nohit> the kicad standard library is disabled
2026-01-14T09:47:02 < nohit> there's only parts with internal part numbers
2026-01-14T09:47:19 < qyx> what the hell happens with the world
2026-01-14T09:59:42 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-14T10:00:21 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-14T10:07:42 < antto> nohit, nasty
2026-01-14T10:07:48 < antto> at least use the metalibs
2026-01-14T10:08:00 < antto> but i guess.. too late
2026-01-14T10:08:55 < nohit> why is it nasty ?
2026-01-14T10:09:01 < nohit> i think its great
2026-01-14T10:11:06 < nohit> internal part number might have second sourcing options but designer do not need to worry about that
2026-01-14T10:18:55 < nohit> https://www.kicad.org/blog/2026/01/2025-End-of-Year-Fund-Drive-The-Year-We-Lost-Count/
2026-01-14T10:24:42 < jpa-> somehow the new features don't sound that big now :)
2026-01-14T10:45:37 < qyx> jlc's search form and table is unusable
2026-01-14T10:45:49 < qyx> it stops refreshing properly as soon as it hits the first http error
2026-01-14T10:50:04 < jpa-> yes
2026-01-14T10:50:17 < jpa-> you have to reload the page all the time
2026-01-14T10:53:35 < qyx> 2 hours passed and I am still clicking jlc
2026-01-14T11:02:24 < jpa-> why are you so slow
2026-01-14T11:03:04 < jpa-> (meanwhile i have accomplished the following: opened vscode, noticed that the client has yet again expired my git access..)
2026-01-14T11:12:04 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has quit [Ping timeout: 244 seconds]
2026-01-14T11:24:36 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has joined ##stm32
2026-01-14T11:35:09 < qyx> ok I matched pretty much everything except ina187 and 74HCT2G16
2026-01-14T11:38:23 < jpa-> downgrade to ina180 :)
2026-01-14T11:39:06 < jpa-> and 74LVC
2026-01-14T11:42:48 < qyx> ina180 is not 28V capable
2026-01-14T11:46:55 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has quit [Remote host closed the connection]
2026-01-14T11:48:19 < jpa-> time to preorder then :)
2026-01-14T11:50:43 < nohit> qyx: there's a flaw in jlcpcb web store when using parts from your own lib (parts you bought earlier), let's say your order costs 130e and you are using parts from your own lib that cost 50e. the web store gives you the option to select IOSS as the order is under 150e, but the real amount is 130e + 50e, so it goes over 150e so IOSS is not an option in reality and you have to pay VAT for the total amount of 180e. you will receive 
2026-01-14T11:50:43 < nohit> invoice from UPS later, the VAT + a UPS fee :)
2026-01-14T11:50:56 < nohit> i asked them to fix this
2026-01-14T11:51:22 < qyx> I am not using ups
2026-01-14T11:51:22 < nohit> so it would show the total amount and IOSS would be disabled when making the order
2026-01-14T11:51:28 < nohit> that doesnt matter
2026-01-14T11:51:40 < nohit> i just used UPS as an example
2026-01-14T11:51:49 < qyx> yes it does because for europacket they are doing customs in germany
2026-01-14T11:52:01 < qyx> the last mile courier service charges no fees
2026-01-14T11:52:03 < qyx> just the vat
2026-01-14T11:52:40 < qyx> for ups/fedex I would need to pay the fee
2026-01-14T11:52:53 < nohit> that does not change the point and the flaw in their system
2026-01-14T11:53:11 < nohit> you are able to select IOSS even when its not really possible
2026-01-14T11:53:32 < qyx> it is not worth fixing because ioss is gonna be history :P
2026-01-14T11:53:39 < nohit> is it ?
2026-01-14T11:53:48 < qyx> 2027 or so
2026-01-14T11:53:54 < nohit> ok
2026-01-14T11:54:45 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-14T11:54:56 < qyx> there is some initiative to pay 3e minimum customs fee for every item received
2026-01-14T11:55:32 < jpa-> but that is built on IOSS, not an alternative to it
2026-01-14T11:55:41 < qyx> https://ec.europa.eu/commission/presscorner/detail/en/ip_25_3045
2026-01-14T11:55:57 < jpa-> because you already pay like 30+ EUR VAT with IOSS, they are not going to replace that with 3 EUR
2026-01-14T11:57:44 < qyx> let's try to order that board now
2026-01-14T11:57:48 < qyx> after 3 hours
2026-01-14T12:02:21 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has joined ##stm32
2026-01-14T12:03:37 -!- NEYi [~NEYi@195.234.78.179] has joined ##stm32
2026-01-14T12:06:25 -!- rob_w [~bob@host-82-135-31-73.customer.m-online.net] has quit [Remote host closed the connection]
2026-01-14T12:13:34 < qyx> so, 211e for 5 boards, not bad at all
2026-01-14T12:15:16 < qyx> and about 10 hours of work saved
2026-01-14T12:19:35 < qyx> meh 30e shipping
2026-01-14T12:43:35 < Steffanx> Still cheap
2026-01-14T12:44:08 < Steffanx> 10 hours of qyx labour is at least 2000 euros
2026-01-14T12:50:01 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 246 seconds]
2026-01-14T12:55:20 < jpa-> oh wow, why do they pay qyx so much
2026-01-14T12:55:34 < jpa-> he's supposed to be eastern european underpaid laborer
2026-01-14T13:03:33 < qyx> wait what noe I can't select europacket anymore
2026-01-14T13:04:17 < jpa-> i think europacket is not available for PCBA for some reason
2026-01-14T13:09:16 < qyx> it was listead as an option when putting the order in the shopping cart
2026-01-14T13:09:23 < qyx> but now at checkout it is not possible anymore
2026-01-14T13:11:49 < c10ud> i discovered safe-pcb, no fees as they have their own eu offices
2026-01-14T13:12:14 < c10ud> website is straight from the 80s but they deliver
2026-01-14T13:16:52 < qyx> ok so it is well over 211
2026-01-14T13:17:03 < qyx> nearly 400e because of shipping and customs
2026-01-14T13:24:20 < qyx> so I should use DDP
2026-01-14T13:34:12 < jpa-> c10ud: lol, the safe-pcb website doesn't work at all.. texts are randomly in french and chinese even if i select english :)
2026-01-14T13:36:27 < c10ud> mmm seems to work in ffx/win11 (at least homepage)
2026-01-14T13:36:54 < c10ud> but that may be due to the fact it loads the it version by default
2026-01-14T13:37:51 < jpa-> maybe because it has no finnish version it decides that french + chinese = finnish
2026-01-14T13:37:55 < qyx> I can't do anything once I click the language selector
2026-01-14T13:38:06 < jpa-> even though i'd want english :)
2026-01-14T13:38:15 < jpa-> but yeah, specs ok but pretty expensive
2026-01-14T13:38:33 < jpa-> and still made in china, so i'd rather do aisler at that point
2026-01-14T13:42:19 < c10ud> but i paid around half when compared to aisler
2026-01-14T13:42:50 < c10ud> no extra fees
2026-01-14T13:45:12 -!- NEYi_ [~NEYi@195.234.78.179] has joined ##stm32
2026-01-14T13:48:18 -!- NEYi [~NEYi@195.234.78.179] has quit [Ping timeout: 252 seconds]
2026-01-14T13:50:05 < karlp> which safe-pcb are you looking at? safepcb.ai? https://www.safepcbglobal.com/index.html or https://www.safe-pcb.com/safe_pcb/ ? they all seem to have the same logo?
2026-01-14T13:51:34 < tomeaton17> that website is something
2026-01-14T13:51:46 < karlp> qyx: I have buitl up an atomic parts library, with all the metadata filled in, andthen I use the bouni plugin to just generate.  I don't try and let the bouni plugin match and find shit.
2026-01-14T13:53:23 < karlp> I use the spreadsheet bom view a lot to make sure things ar eright
2026-01-14T13:53:59 < karlp> "KiCAD JLCPCB tools" from the plugin manager: https://www.safe-pcb.com/safe_pcb/
2026-01-14T13:54:01 < karlp> not that 
2026-01-14T13:54:10 < karlp> https://github.com/bouni/kicad-jlcpcb-tools that one
2026-01-14T13:58:03 < qyx> oh thanks will look at that
2026-01-14T13:58:10 < karlp> (so I'm doing what you hate and have different symbols for different value resistors, yes)
2026-01-14T13:58:38 < qyx> ok no thanks
2026-01-14T13:59:18 < karlp> you don't have to,
2026-01-14T13:59:30 < karlp> I used to assign it manually in the bom sheet tool
2026-01-14T13:59:40 < karlp> I just got tired for copying things back and forth, am trying this style out
2026-01-14T13:59:48 < karlp> it doesn't change how the bouni plugin works at all.
2026-01-14T14:00:01 < karlp> kikit didn't really have this at the time, but sounds like it does ~same things really.
2026-01-14T14:00:16 < karlp> takes attributes frrom the bom to make things
2026-01-14T14:00:39 < karlp> I currently don't think about rotations much.
2026-01-14T14:01:00 < karlp> I also don't order that many pcbs, why do you care how my process works ? :) I thin you make more than me
2026-01-14T14:02:49 < qyx> because I knew you spent some time optimizing for jlc
2026-01-14T14:02:53 < karlp> kicad10 finally gets footprints with inner layers, whee
2026-01-14T14:03:16 < karlp> yeah, and that optimization was to stop trying to recheck whether things had been updated or copied on the bom sheet, because the website was slow and clunky
2026-01-14T14:03:40 < karlp> so I've done the last 4 boards with a slowly growing atomic symbol set, and you know, I'm fine with it.
2026-01-14T14:04:07 < qyx> but now I need to autism up a small temperature chamber really quick
2026-01-14T14:04:08 < karlp> yeah,  you had to make a few resistors, but now I _know_ what I've got.
2026-01-14T14:04:42 < karlp> also, my symbols now have 805 and 603 and shti in them for R/C now, so I can check sch power supplies more easily and knwo what the sort of shit i've rellyu got going on.
2026-01-14T14:04:50 < qyx> maybe I could do yet another python plugin asi if there weren't enough
2026-01-14T14:05:58 < karlp> https://bin.jvnv.net/file/osjnS/Screenshot%20From%202026-01-14%2012-05-30.png
2026-01-14T14:06:50 < qyx> that should go into an attribute
2026-01-14T14:06:52 < karlp> Ineed to try this "design blocks" thing out, but still suckz0rs with wayland.
2026-01-14T14:07:20 < karlp> ok, so what.. you want to place "R" and _then_ add attributes for size and value?
2026-01-14T14:07:22 < karlp> fuck that
2026-01-14T14:07:34 < karlp> I add KR_0603_10k frrom my jlc-basic lib
2026-01-14T14:08:30 < qyx> that sounds like the wrong way for me :S
2026-01-14T14:09:00 < karlp> well, thankfully, both ways work juyst fine :)
2026-01-14T14:09:10 < karlp> but now you know how I've started doing things.
2026-01-14T14:09:30 < karlp> this is partly because I'm building with specific manufacturing in mind.
2026-01-14T14:09:57 < karlp> I'm not doing some abstract "my team builds a flawless schematic, someone else can decide if it can be built, or if the parts are available" 
2026-01-14T14:10:13 < karlp> which is also valid, just not for me :)
2026-01-14T14:18:32 -!- NEYi_ [~NEYi@195.234.78.179] has quit [Quit: Leaving]
2026-01-14T14:48:29 < nohit> i use kicad jlcpcb tools too, its great
2026-01-14T15:20:33 < qyx> how do I measure how strong an elektromagnet is
2026-01-14T15:20:38 < qyx> *electromagnet
2026-01-14T15:20:51 < qyx> maybe I should do some math first
2026-01-14T15:21:01 < qyx> and then try to match either a hall or a magnetic sensor
2026-01-14T15:53:26 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has quit [Ping timeout: 256 seconds]
2026-01-14T15:59:30 -!- ColdKeyboard [~ColdKeybo@user/coldkeyboard] has joined ##stm32
2026-01-14T16:24:43 < jpa-> qyx: it's hard to find a magnetic sensor that can measure over 0.5 T or so, while neodymium magnets are usually 1 T or more at the surface; so you'd have to measure at some consistent distance, but that is then affected by magnet geometry
2026-01-14T16:25:00 < jpa-> but maybe your electromagnet is weak :)
2026-01-14T16:25:30 < qyx> meanwhile the whole topic is moot
2026-01-14T16:25:32 < jpa-> (or you could measure at lower currents and assume it to be linear, but it isn't)
2026-01-14T16:26:16 < qyx> but anyway, for academic purposes, the electromagnet was a neodymium oermanent magnet with 20kg holding capbility, whatever thatt means
2026-01-14T16:26:32 < qyx> and using a coil to "neutralize" it to unlock
2026-01-14T16:27:03 < qyx> but since those thigs are not precisely made, I was about to "calibrate" them for zero holding force
2026-01-14T16:27:37 < qyx> as they are all equal dimensions, yeah I cohld just use a constant distance spacing
2026-01-14T16:29:30 < jpa-> ah, well zero field measuring would be easy
2026-01-14T16:30:22 < qyx> I also wanted to measure the magnet alone without the coil energized
2026-01-14T17:20:52 < ventyl> is stm32h7 + stlinkv3 + openocd (dapwhatever + SWD) something that is well known to ignore breakpoints?
2026-01-14T17:26:51 < aandrew> karlp> I add KR_0603_10k frrom my jlc-basic lib
2026-01-14T17:27:02 < aandrew> insanity
2026-01-14T17:27:21 < aandrew> I add the symbol, get the idea down and then go back and choose size/value
2026-01-14T17:28:05 < aandrew> I totally get the benefit of having an 0603_10k, 0603_10k3, 0603_10k5 from an insa-bom perspective but never could get into that flow
2026-01-14T17:29:12 -!- phryk [~totallyno@user/phryk] has quit [Ping timeout: 256 seconds]
2026-01-14T17:29:28 < jpa-> ventyl: nope
2026-01-14T17:29:41 < jpa-> ventyl: is your code in flash, ram or qspi flash?
2026-01-14T17:30:30 < jpa-> ventyl: does "hbreak" work any better (forces use of hardware breakpoints even for ram addresses)
2026-01-14T17:42:42 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
2026-01-14T17:43:21 -!- grindhold [~quassel@v2202504104743335453.luckysrv.de] has joined ##stm32
2026-01-14T17:47:33 -!- grindhold [~quassel@v2202504104743335453.luckysrv.de] has quit [Ping timeout: 252 seconds]
2026-01-14T18:09:16 < nohit> ventyl: which openocd version you are using ?
2026-01-14T18:10:18 < karlp> aandrew: well, like I said, I've been trying it out for a little while now, and honestly, it's been fine so far, I was against the concept on teh face of it,
2026-01-14T18:10:29 < karlp>  but I really was redoing the same bullshit with manufacutring data time and time again.
2026-01-14T18:10:42 < karlp> time will tell, I may hate it and go back to something else.
2026-01-14T18:27:40 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has joined ##stm32
2026-01-14T18:31:51 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has quit [Ping timeout: 244 seconds]
2026-01-14T18:42:51 < machinehum> catphish: Seems like I'll need a USB-A to USB-A to program the board 
2026-01-14T18:43:28 < catphish> machinehum: yep probably, i didn't give that much thought yet, unusual choice of connector
2026-01-14T18:43:41 < catphish> but also, the obvious choice of connector because it costs so little
2026-01-14T18:44:59 < machinehum> I also found this https://github.com/hydrausb3
2026-01-14T18:45:04 < catphish> yep, that's a thing
2026-01-14T18:45:20 < machinehum> Cool
2026-01-14T18:45:22 < catphish> afaik it's much the same design
2026-01-14T18:45:40 < machinehum> The main repo from the vendor is cancer 
2026-01-14T18:46:58 < machinehum> #include "CH56x_usb20.H"
2026-01-14T18:47:25 < machinehum> You grep around and fine the file is called ch56x_USB20.h
2026-01-14T18:47:38 < machinehum> find*
2026-01-14T18:49:06 < karlp> wot, do you use a case sensitive file system? :) lol, peasant.... or something.
2026-01-14T18:49:50 < machinehum> Oh god it all makes so much sense now
2026-01-14T18:50:02 < machinehum> I didn't even think of that lol
2026-01-14T18:50:39 < qyx> of course?
2026-01-14T18:50:55 < qyx> wut, does anybody use a case *insensitive* filesystem?
2026-01-14T18:50:55 < karlp> me too, just joking about why they don't even notice it.
2026-01-14T18:50:59 < karlp> everybody on windows.
2026-01-14T18:51:15 < qyx> o windows is case insensitive?
2026-01-14T18:51:18 < qyx> poor men
2026-01-14T18:51:25 < zyp> macos too, usually
2026-01-14T18:51:33 < zyp> rarely matters
2026-01-14T18:51:42 < karlp> rarely...
2026-01-14T18:51:44 < karlp> then it does...
2026-01-14T18:52:19 < zyp> yeah, I remember around a decade ago I ran into an issue with it and had to make a diskfile with a case-sensitive filesystem for some shit
2026-01-14T18:55:23 < catphish> machinehum: i think hydrausb refactored all the code into their own version of the SDK
2026-01-14T19:28:49 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has joined ##stm32
2026-01-14T19:33:29 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has quit [Ping timeout: 260 seconds]
2026-01-14T20:24:27 < karlp> well, I got IIO to talk to my adc in kernel land.
2026-01-14T20:24:34 < karlp> but I didn't l ook deeply into the driver impl.
2026-01-14T20:24:49 < karlp> it doesn't support diff mode, it doesn't support continuous reading, it doesn't support reference selection
2026-01-14T20:24:57 < karlp> and.. I only seem to be getting zeros out of it.
2026-01-14T20:25:02 < karlp> but, this is progress!
2026-01-14T20:25:39 < karlp> and I found some silk labels that weren't updated when I did some pin swapping...
2026-01-14T20:44:05 < karlp> goddamn, this driver is fucking useless
2026-01-14T20:44:21 < karlp> the merge has almost no commentary becauyse it came from a ti.com address
2026-01-14T20:44:34 < karlp> but any user submitted driver gets put through the fucking wringer for not being complete enough.
2026-01-14T20:44:41 -!- grindhold [~quassel@2a0a:4cc0:c0:70c9::1] has joined ##stm32
2026-01-14T20:48:19 < karlp> goddamn, somuch for picking adcs based on what had iio "drivers"
2026-01-14T21:13:49 -!- Posterdati [~Posterdat@user/Posterdati] has quit [Ping timeout: 264 seconds]
2026-01-14T21:24:15 -!- phryk [~totallyno@user/phryk] has joined ##stm32
2026-01-14T21:28:04 -!- Posterdati [~Posterdat@user/Posterdati] has joined ##stm32
2026-01-14T21:31:15 -!- PhantomWork [~PhantomWo@user/phantom] has joined ##stm32
2026-01-14T21:31:49 < PhantomWork> Question guys about xtal on mcu.... it say to use 5-20pf... But also have to include the pcb and mcu pin capacitance with a note of roughly 10pf can be used as estimate....    so the true range is... -5 to 10pf? aka 0-10?
2026-01-14T21:33:38 < jpa-> what says that?
2026-01-14T21:34:13 < jpa-> but yes, it is possible to pick a crystal and a PCB layout so that it needs no load caps, or so that it is overloaded even without load caps
2026-01-14T21:35:52 < PhantomWork> stm32f431c6 datasgheet, page 100
2026-01-14T21:38:21 < PhantomWork> so if the crystal specs load capacitance: 20pf....    I should use like a 10pf due to the trace+pin that add 10?     and not the 20pf I currently use, right?
2026-01-14T21:38:22 < jpa-> that definitely does not say to subtract 10pF from the "5 pF to 20 pF" range
2026-01-14T21:38:34 < jpa-> no, that is not how you calculate load capacitors
2026-01-14T21:38:59 < jpa-> read the appnote the datasheet references, AN2867
2026-01-14T21:39:27 < PhantomWork> checking
2026-01-14T21:47:41 < ds2> karlp: what are you using the ADC for? and do you really care about IIO?
2026-01-14T21:50:44 < ds2> while IIO is the "proper" way of doing ADCs, for simple data stuff, there are alternative that work fine. Just don't expect to upstream it. I've exposed ADC via other interfaces
2026-01-14T21:56:25 < PhantomWork> jpa-: found the formula there, so my 20+20pf, add 10pf for board+pins, is dead on the 20pf rated xtal, and fit in the range so is fine... I think.
2026-01-14T22:09:45 < PhantomWork> And.... does 0603 vs 0402 makes any difference? 0603 would be samsung, 0402 would be chinese
2026-01-14T22:12:31 < zyp> PhantomWork, do you have a schematic?
2026-01-14T22:13:29 < PhantomWork> zyp: yeah? but.... what for?   stm32 + 2 cap + 1 crystal needs a schem?
2026-01-14T22:14:20 < zyp> I was hoping it'd have references that I could check, because it sounds like whatever you're doing is wrong
2026-01-14T22:14:27 < zyp> which crystal?
2026-01-14T22:15:01 < PhantomWork> ah for exact parts, not for the schem itself
2026-01-14T22:15:07 < zyp> yes
2026-01-14T22:15:47 < PhantomWork> https://www.lcsc.com/datasheet/C115962.pdf     
2026-01-14T22:15:59 < PhantomWork> for the xtal
2026-01-14T22:16:38 < zyp> «10pF, 20pF, or specify» very specific :)
2026-01-14T22:16:58 < zyp> but the exact part is listed as CL=20pF, so okay
2026-01-14T22:17:13 < zyp> I'd probably put 36pF load caps on it
2026-01-14T22:19:29 -!- phryk [~totallyno@user/phryk] has quit [Quit: ZNC 1.9.1 - https://znc.in]
2026-01-14T22:21:10 < zyp> https://github.com/karlp/zypsnips/blob/master/crystal-load-wisdom.txt#L15
2026-01-14T22:21:44 < zyp> this is consistent with what section 3.3 in the AN2867 that jpa- referred to says
2026-01-14T22:25:18 < zyp> AN2867 says CL = (C1 * C2) / (C1 + C2) + CS, and when you make the assumption that C1 == C2 and substitute Cx, it simplifies to CL = Cx / 2 + CS, which reorders into what I said
2026-01-14T22:28:22 < zyp> if you get it wrong and put the wrong load for the cap, the oscillator will still work, but get detuned
2026-01-14T22:29:02 < zyp> wrong load makes it oscillate slightly slower or faster
2026-01-14T22:30:04 < PhantomWork> zyp: yeah    20pf with the 10pf estimate gives 20pf load, and fit within the 5-20pf range
2026-01-14T22:30:20 < zyp> 10pF stray sounds excessive
2026-01-14T22:30:33 < PhantomWork> board + pins
2026-01-14T22:31:02 < zyp> idk how your layout looks, but 10pF sounds excessive
2026-01-14T22:32:05 < zyp> IIRC the one time I fucked up, I put something like 18pF load on 18pF cap
2026-01-14T22:32:09 -!- phryk [~totallyno@user/phryk] has joined ##stm32
2026-01-14T22:32:26 < zyp> and that detuned the frequency by like 250ppm
2026-01-14T22:33:45 < PhantomWork> https://i.imgur.com/M5rSJgw.png
2026-01-14T22:33:48 < zyp> which broke everything, because it was driving a 2.4GHz radio, and a 250ppm detune there means it's 0.6 MHz off, which is significant when your radio protocol only has 1 MHz wide channels
2026-01-14T22:34:58 < PhantomWork> yeah 250ppm for radio is crazy... I have a RTLSDR with 47-63ppm and it's nearly useless
2026-01-14T22:35:33 < zyp> the crystal itself was rated for like 20ppm, but that doesn't matter if you get the load wrong :)
2026-01-14T22:35:59 < zyp> but if I didn't do anything that required accuracy, I probably wouldn't have noticed
2026-01-14T22:38:34 < PhantomWork> I'll be using CANbus at 500kbps, so should be fine if it is detuned a bit
2026-01-14T22:38:48 < zyp> yes :)
2026-01-14T22:40:10 < PhantomWork> and there    direct ground to the xtal caps from the µC   no more layer shifting, good
2026-01-14T22:41:24 < zyp> IIRC the bitstuffing in CAN ensures at most five (or was it six) bit times without an edge to resync on, which should allow even more clock deviation than UARTs which can have nine consecutive identical bits
2026-01-14T22:41:46 < zyp> and UARTs aren't particularly picky
2026-01-14T22:42:53 < karlp> ds2: I'm evaluating linux based adc, yeah, I _believe_ iio is how I should use it, to get robust, kernel managed data collection, and it _seems_ all in the right place, just that the actual device drivers are _very_ haphazardly complete.
2026-01-14T22:43:05 < karlp> ds2: what other interfaces would you be meaning?
2026-01-14T22:43:09 < PhantomWork> DRC is happy, so.... ready to send that one... I think
2026-01-14T22:43:41 < karlp> I think I could just hack a different iio driver that did what I wanted instead, I kinda feel like iio _might_ be the easiest way of consuming it, trying not to do "all custom kernel shit"
2026-01-14T22:44:29 < karlp> goal is really just stable continuous sampling rate adc, with kernel keeping time nicely, and buffering so that userspace can process clean reliable adc samples.
2026-01-14T22:45:33 < karlp> this is for proposing to move something that is fully embedded right now, but has massive pain around doing all the ethernet and high level stuff in bare metal C, with a legacy spaghetti of C with freertos thrown on top.
2026-01-14T22:45:51 < karlp> moving to linux would make all of _that_ easy, but I need to show how the adc's would work reliably.
2026-01-14T22:46:33 < karlp> just using spidev and python works really well for me for demoing the hardware, but it doesn't give me a warm fuzzy for long term stability, and I'm sure it's goign to use more cpu load.
2026-01-14T22:46:49 < karlp> having the kernel take care of watching the data ready pin on the adc and read out with timestamps sounds just excellent.
2026-01-14T22:52:56 < PhantomWork> 96$ in extended parts fees.... grr
2026-01-14T22:56:40 < karlp> 32 unique, special parts?
2026-01-14T22:56:47 < karlp> fancy.
2026-01-14T22:56:51 < karlp> didn't think your board was that big.
2026-01-14T22:57:30 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-14T23:00:33 < PhantomWork> https://i.imgur.com/JwaCReG.png
2026-01-14T23:01:17 < PhantomWork> 67 unique components, 301 components total
2026-01-14T23:03:27 < zyp> still fairly cheap PCBA NRE
2026-01-14T23:04:40 < zyp> quickly gets negligible if you're making multiple, and even for one I'd rather pay that than hand assemble it
2026-01-14T23:06:00 < qyx> you are not adventurous enough
2026-01-14T23:07:23 < PhantomWork> yeah, I just hate to pay for fees :D
2026-01-14T23:08:02 < zyp> in the end, it's the sum that matters
2026-01-14T23:08:13 < qyx> also wut 67 unique components
2026-01-14T23:08:34 < qyx> even my most unique measuring shit has not that many
2026-01-14T23:11:22 < PhantomWork> that is why I'm checking to see if I can substitute something
2026-01-14T23:12:47 < PhantomWork> if space allow, I can reduce one...
2026-01-14T23:19:35 < PhantomWork> 2, maybe 3...    and found one error, that will be fixed by part merging
2026-01-14T23:19:42 < PhantomWork> 18mA polyfuse is way too low
2026-01-14T23:29:26 < PhantomWork> so, why so many? 3.3V + 5V smps, different inductor and resistor. CANbus, I2C eeprom, I2C ADC, I2C digipot, I2C bus isolator, RS485 OOK tranceiver, power filters, 17 channels outputs, 2 inputs, 5 different polyfuses (will be reduced to 2-3), TVS, zeners, 2 voltage references (might use one?) and random bits
2026-01-14T23:49:32 < karlp> don't you ahve an stm23 on this board? why the vrefs and adc?
2026-01-14T23:49:45 < zyp> IIRC isolated power domain
2026-01-14T23:50:03 < karlp> resistors should all be basic anyway, don't countthem.
2026-01-14T23:50:19 < zyp> unless you need very specific values
2026-01-14T23:50:34 < karlp> there's an awful lot of resistors in basic :)
2026-01-14T23:50:58 < zyp> I've ran into not finding the value I needed in basic
2026-01-14T23:51:11 < zyp> long time ago though
2026-01-14T23:51:13 < karlp> how rescently?
2026-01-14T23:51:18 < karlp> yeah, it's got shit heaps these days
2026-01-14T23:51:19 < zyp> years
2026-01-14T23:52:28 < PhantomWork> karlp: stm32G431
2026-01-14T23:52:31 < karlp> yeah, it's got shit heaps these days
2026-01-14T23:53:21 < PhantomWork> yeah   loooked at the whole part list... I can only reduce in polyfuse
2026-01-14T23:54:07 < qyx> karlp: 49k9
2026-01-14T23:54:59 < qyx> yes I could have used 27k+2k7
2026-01-14T23:56:41 < karlp> there's 0805, 0603 and 0402 49k9 in basic.
2026-01-14T23:56:44 < karlp> what are you talking about?
2026-01-14T23:56:51 < qyx> wat
2026-01-14T23:57:01 < karlp> sorry, 0603 in basic, and theo thers are "extended plus" which is free, so effectively basic.
2026-01-14T23:57:18 < qyx> were I dumb earlier today
2026-01-14T23:57:28 < karlp> https://bin.jvnv.net/file/qFm8m/Screenshot%20From%202026-01-14%2021-57-13.png
2026-01-14T23:57:43 -!- Arcar [~arcar@user/Arcar] has joined ##stm32
2026-01-14T23:57:49 < karlp> were you doing metro and insisting on thin film or something? :)
2026-01-14T23:58:32 < qyx> no, now I can find it
2026-01-14T23:58:41 < qyx> that search form is useless
2026-01-14T23:58:50 < qyx> chink junk
2026-01-14T23:59:51 < PhantomWork> home time!
--- Day changed Thu Jan 15 2026
2026-01-15T00:00:04 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 260 seconds]
2026-01-15T00:00:45 < karlp> you gotta learn the right ways of searching yeah,
2026-01-15T00:01:47 -!- Arcar [~arcar@user/Arcar] has quit [Remote host closed the connection]
2026-01-15T00:06:06 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-15T00:13:11 < catphish> i've had a busy week, designed myself a USB oscilloscope, then realised i might as well design a logic analyzer on the same platform, awaiting prototypes now, and then i have to figure out how to write drivers, and what frontend to use
2026-01-15T00:14:09 < karlp> well done
2026-01-15T00:14:37 < catphish> whether i did well remains to be determined :D
2026-01-15T00:16:04 < karlp> nah, doing things is success in and of itself :)
2026-01-15T00:16:41 < catphish> usually i start projects, learn a bunch, give up, and bank those learnings until someone wants to pay me for those skills
2026-01-15T00:17:08 < catphish> but i'm ever hopeful that one of my projects will end in a useful product
2026-01-15T00:17:16 < karlp> well, I tried making my (working, diff mode, continous) spidev python code do single ended manually polled conversions, and I get zeros just like the in tree iio driver does.
2026-01-15T00:17:24 < karlp> I wonder what's wrong with it in single ended mode?
2026-01-15T00:23:04 < catphish> no clue here :)
2026-01-15T00:27:04 < ventyl> jpa-: no, it doesn't help, there's no warning in openocd output nor gdb output, it just doesn't work
2026-01-15T00:27:09 < ventyl> nohit: last release
2026-01-15T00:35:48 < catphish> here tis. analog: https://i.ibb.co/bDRyLDx/scope.jpg digital: https://i.ibb.co/b5dY2cMG/la.jpg
2026-01-15T00:52:13 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 264 seconds]
2026-01-15T00:53:25 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has quit [Ping timeout: 264 seconds]
2026-01-15T01:18:07 -!- qyx [~qyx@84.245.120.206] has quit [Ping timeout: 240 seconds]
2026-01-15T01:20:18 -!- qyx [~qyx@84.245.120.202] has joined ##stm32
2026-01-15T02:09:33 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-15T03:27:59 -!- jfsimon1981 [~jfsimon19@2a01:cb14:b9b:2000:9104:b9aa:1662:b9ad] has joined ##stm32
2026-01-15T03:31:49 -!- jfsimon [~jfsimon19@lfbn-lyo-1-1534-250.w86-207.abo.wanadoo.fr] has quit [Ping timeout: 264 seconds]
2026-01-15T04:06:35 -!- hexo [~hexo@user/hexo] has quit [Ping timeout: 240 seconds]
2026-01-15T04:19:36 -!- hexo [~hexo@user/hexo] has joined ##stm32
2026-01-15T08:12:44 < ds2> karlp: I need to use the onboard ADC of the TI AM335x and the driver was crap so I wrote my own and just exposed the data as a character interface. It was trivial. The main thing I needed that TI driver didn't do was using the hardware oversample/averaging
2026-01-15T08:13:15 < ds2> IIO sound cool til you try to use it. more complex then needed, IMO
2026-01-15T08:32:44 -!- nerozero [~nerozero@87.253.63.54] has joined ##stm32
2026-01-15T09:11:23 -!- cutofmyjib [~cutofmyji@user/cutofmyjib] has joined ##stm32
2026-01-15T09:14:02 -!- cutofmyjib [~cutofmyji@user/cutofmyjib] has quit [Client Quit]
2026-01-15T10:20:24 -!- martinmoene_ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-15T10:20:55 -!- martinmoene__ [~martinmoe@205-20-132-5.ftth.glasoperator.nl] has joined ##stm32
2026-01-15T10:57:13 -!- octorian [octo@2600:3c00::f03c:91ff:fe93:a61c] has quit [Ping timeout: 255 seconds]
2026-01-15T10:57:18 -!- octorian_ [octo@2600:3c00::f03c:91ff:fe93:a61c] has joined ##stm32
2026-01-15T11:29:12 < karlp> well, "I'll just make a kernel driver, it's trivial" is very much point of view based :)
2026-01-15T11:29:25 < karlp> I get what you mean though, it seems to have _many_ possibilities,
2026-01-15T11:29:46 < karlp> I'm kinda disgusted by how completely worthless this intree driver is though.
2026-01-15T11:30:19 < karlp> who the fuck builds a 24bit multi channel, differential adc onto a board and makes a driver that does single shot conversions, single ended only, polled.
2026-01-15T11:30:46 < karlp> it's blatant that it was accepted without question because it came from the right place, standard kernel bullshit.
2026-01-15T11:31:22 < karlp> ds2: doing it via iio should make things like iiod-scope and things work, like, I want to maintain _less_ stuff, not "yolo, NIH all the things"
2026-01-15T11:32:04 < karlp> so I'm likely to just make a parallel iio driver that does more useful things, without a lot of regard for whether it could ever be upstreamed.
2026-01-15T11:32:59 < karlp> https://github.com/torvalds/linux/blob/master/drivers/iio/adc/ti-ads124s08.c#L282-L298 that's all it can do.
2026-01-15T11:33:07 < karlp> it never writes to any other register, 
2026-01-15T11:33:26 < karlp> and input mux is a two field register, not a single source.
2026-01-15T11:33:33 < karlp> I'm not sure this has ever even worked.
2026-01-15T12:15:01 -!- HelloShitty [~psysc0rpi@dsl-119-61.bl27.telepac.pt] has quit [Ping timeout: 264 seconds]
2026-01-15T12:17:10 -!- HelloShitty [~psysc0rpi@bl15-226-17.dsl.telepac.pt] has joined ##stm32
2026-01-15T12:22:49 < tomeaton17> https://news.ycombinator.com/item?id=46616775 its all kicking off in hobbyist land
2026-01-15T12:37:17 < jpa-> lol
2026-01-15T13:03:13 < ventyl> so, either stlink fucks up setting breakpoint, or h7 doesn't give a fuck about it
2026-01-15T13:04:47 < jpa-> code running from flash normally?
2026-01-15T13:04:53 < ventyl> yes
2026-01-15T13:05:18 < ventyl> this particular code with this particular firmware with this particular environment setup was debugging just fine up until now
2026-01-15T13:05:25 < jpa-> have you tried looking at openocd debug log?
2026-01-15T13:05:31 < ventyl> heck, the device is in production
2026-01-15T13:05:45 < ventyl> yes, openocd states that the breakpoint was set up correctly, even the address matches memory map
2026-01-15T13:06:08 < ventyl> breakpoints.c:88 breakpoint_add_internal(): [0] added hardware breakpoint at 0x080075d6 of length 0x00000002, (BPID: 2)
2026-01-15T13:07:02 < jpa-> and you are sure the code actually runs? can you try pausing with ctrl-c and set breakpoint on next line or something?
2026-01-15T13:07:17 < ventyl>  .text.main     0x080075d0    
2026-01-15T13:07:33 < ventyl> I can talk to the device via USB, so I am pretty sure it runs
2026-01-15T13:07:43 < ventyl> that breakpoint is set to main, just BTW
2026-01-15T13:08:03 < jpa-> but that it runs that particular location in code.. i.e. no silly inlining or something else bypassing it
2026-01-15T13:08:49 < ventyl> it is main function
2026-01-15T13:08:53 < ventyl> and there is no LTO
2026-01-15T13:09:20 < ventyl> like I am back at basics, I am trying just any breakpoint to hit
2026-01-15T13:09:38 < jpa-> so how about the ctrl-c, see where it is, set breakpoint on next line?
2026-01-15T13:10:05 < jpa-> that also gets rid of any issues where resetting would remove breakpoints or something (i've seen that happen on other microcontrollers)
2026-01-15T13:10:43 < ventyl> OK, this is weird
2026-01-15T13:10:57 < ventyl> i loaded the firmware into flash ONCE AGAIN and it works now
2026-01-15T13:11:15 < ventyl> this is something that happened before... load commands were pretending they work, but in fact, nothing was flashed
2026-01-15T13:11:34 < jpa-> ah, i have had that too but never really figured out
2026-01-15T13:11:49 < jpa-> and somehow verify in gdb has been flaky even when it flashes correctly :)
2026-01-15T13:13:11 < ventyl> 12 hours wasted
2026-01-15T13:13:25 < ventyl> i fucking hate the flakiness of this industry
2026-01-15T13:20:04 < c10ud> buy pro stuff
2026-01-15T13:20:49 < jpa-> as if that really helps
2026-01-15T13:23:33 < c10ud> tbh never had any real issues with segger stuff
2026-01-15T13:24:07 < karlp> lol
2026-01-15T13:24:12 < karlp> jlink lies _all the time_ about flashing
2026-01-15T13:24:15 < karlp> "load" done.
2026-01-15T13:24:24 < karlp> "run"...window pops up to actually flash
2026-01-15T13:24:36 < karlp> "compare-sections" yup, 100% match.
2026-01-15T13:24:38 < karlp> (lies)
2026-01-15T13:24:57 < c10ud> stm32?
2026-01-15T13:25:08 < karlp> no, but it's a segger behaviour, not a target behaviour.
2026-01-15T13:25:28 < karlp> I'm not the only one who reports that the segger tools often just lie because "htye know best, they'll handle it behind the scenes for you"
2026-01-15T13:25:46 < karlp> we dont'
2026-01-15T13:26:02 < karlp>  use stm32 because theyr'e french, and one guy here one time had a citroen that he didn't like.
2026-01-15T13:26:13 < c10ud> lol
2026-01-15T13:26:17 < c10ud> they're italian also
2026-01-15T13:26:28 < karlp> don'
2026-01-15T13:26:32 < karlp> t try and think rationally :)
2026-01-15T13:26:49 < c10ud> idk, in my experience thousands of bootloadered stuff with jlink projects went ok
2026-01-15T13:26:59 < c10ud> only thing to work ok once properly prepared lol
2026-01-15T13:27:05 < c10ud> double click prj then f7
2026-01-15T13:27:17 < BrainDamage> "they are italian also" < don't give them more arguments to avoid stm32
2026-01-15T13:27:53 < jpa-> 2019-12-03T16:35:46 < c10ud> I'm trying with jlink but no luck       2017-12-07T13:50:18 < c10ud> I think I fucked up when I updated jlink dll to 8.2 or so         not real problems ;)
2026-01-15T13:28:16 < c10ud> woooah
2026-01-15T13:28:43 < c10ud> well I remember there was some issue with old prj with new jlink
2026-01-15T13:29:08 < c10ud> but to answer specifically i would have to have a way to remember which prj i was working on that day :p
2026-01-15T13:29:45 < c10ud> 2017 i think it was probably atmel stuff, not stm32
2026-01-15T13:30:00 < c10ud> 2019 could have also been nxp
2026-01-15T13:30:05 < c10ud> dig more :P
2026-01-15T13:30:23 < c10ud> also in 2017 i had an edu and a clone
2026-01-15T13:30:40 < jpa-> i don't care, i just know there is no software thing that wouldn't have problems
2026-01-15T13:30:42 < c10ud> not like the PLUS PRO ULTRA (?) of today 
2026-01-15T13:31:45 < ventyl> c10ud: IIRC, i had exactly the same problem with jlink too
2026-01-15T13:32:24 < ventyl> actually, the damn yapicoprobe was far the most stable probe I've used so far
2026-01-15T13:33:32 < ventyl> and with fucking renesas e2 as well
2026-01-15T13:33:55 < jpa-> picoprobe has worked relatively well for me too
2026-01-15T13:34:56 < jpa-> but RP has its problems with breakpoints if you run it from RAM for predictable timing and gdb wants to use RAM breakpoints which then get overwritten, so you need to remember to use hbreak, but only RP2350 supports hardware breakpoints for RAM contents etc..
2026-01-15T13:35:38 < ventyl> so far I am not running code from RAM, it doesn't play well with memory model of CMRX with out-of-box process maps
2026-01-15T13:36:24 < jpa-> on STM32 there isn't as much point anyway as the flash latency is not that bad
2026-01-15T13:36:41 < jpa-> with external QSPI on RP2350 it's pretty dire when you get a cache miss
2026-01-15T13:37:27 < karlp> (minichlink, one of the "alternatives" for ch32xx stuff, insists on rewriting flash to do software breakpoints, _for all breakpoints_  _by design_ (they dissapprove of hardware breakpoints for... reasons))
2026-01-15T13:39:28 < jpa-> lol
2026-01-15T13:39:38 < ventyl> jpa-: I am running some code on RP2040 as well. At this stage of CMRX development, if things are slow, it is actually good for me because any performance bottleneck will show up fast
2026-01-15T13:40:43 < ventyl> The problem is that memory isolation is basically a curse word. All SDKs randomly throw stuff around places so it is often almost impossible to impose any reasonable isolation model
2026-01-15T13:41:23 < karlp> jpa-: it's very weird, that community has done all sorts of interesting things, but some of it is jsut so.... weird.
2026-01-15T13:41:24 < jpa-> yeah, it is easy to invent impractical isolation models, inventing a practical one is much harder
2026-01-15T13:41:58 < ventyl> I'd say that this model is very practical and actually quite tolerant to brain-dead and messy organization
2026-01-15T13:42:23 < ventyl> But then always someone invents more brain-dead or more messy memory organization
2026-01-15T13:43:12 < ventyl> There are basically two things that break it: - We have modes this piece of code to RAM for performance reasons, so you don't have to
2026-01-15T13:43:44 < ventyl> and 'Our HAL maintains a private state object for you'
2026-01-15T13:47:52 < ventyl> s/modes/moved/
2026-01-15T14:37:46 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-15T14:59:52 < ventyl> 12 hours later, I could finally fix.... a stupid pointer arithmetics bug
2026-01-15T15:11:20 -!- kurfen_ [~kurfen@194.127.173.38] has quit [Ping timeout: 240 seconds]
2026-01-15T15:44:48 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 252 seconds]
2026-01-15T15:45:15 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-15T16:20:27 < qyx> those chink wild goose mosfets..
2026-01-15T16:22:56 < qyx> so dpak is about the most advanced package I could get at jlc
2026-01-15T16:25:48 -!- kurfen [~kurfen@194.127.173.38] has joined ##stm32
2026-01-15T16:32:22 -!- _Posterdati_ [~posterdat@user/Posterdati] has joined ##stm32
2026-01-15T16:42:44 < karlp> you cna't get the 56/3x3?
2026-01-15T16:49:07 < qyx> I can't find any
2026-01-15T16:49:17 < qyx> lfpak56 would be great
2026-01-15T16:53:44 < karlp> what sort of specs are you looking for?
2026-01-15T16:53:58 < karlp> (i've onyl got like 2 hour sleft before a "holiday" so not starting anything new)
2026-01-15T16:54:45 < karlp> https://jlcpcb.com/partdetail/HUAYI-HYG053N10NS1C2/C2986239 in "lfpak56" ?
2026-01-15T16:54:50 -!- Laurenceb_ [~Laurenceb@cust226-dsl93-89-135.idnet.net] has joined ##stm32
2026-01-15T16:55:12 < karlp> https://jlcpcb.com/partdetail/HL-40N06D/C7543836 in the 3x3?
2026-01-15T16:55:58 < qyx> 60/100 V and 5 A
2026-01-15T16:56:17 < karlp> fuck, I just pushed the wrong button and lost my package selection
2026-01-15T16:56:24 < karlp> no problme, I'll do it again :)
2026-01-15T16:56:25 < qyx> wait what, that one looks legit
2026-01-15T16:56:28 < karlp> n-chanel?
2026-01-15T16:56:34 < qyx> yeah, thanks
2026-01-15T16:56:37 < qyx> that's a good one
2026-01-15T16:57:01 < karlp> ok, I'll call it a win and move on then :)
2026-01-15T16:57:19 < qyx> how on earth one use that search form, I am completely lost
2026-01-15T16:57:23 < qyx> I can't find shit
2026-01-15T16:57:27 < karlp> it's um... tedious,
2026-01-15T16:57:44 < karlp> but the only way I've found to do useful things is to manually scroll the package type and include the things that I know are the same.
2026-01-15T16:57:52 < karlp> then scan _that_ list.
2026-01-15T16:57:59 < karlp> at least, when I'm looking for compatible things like this.
2026-01-15T16:58:05 < qyx> don't they have an api?
2026-01-15T16:58:14 < karlp> api would be worse,
2026-01-15T16:58:15 < qyx> it could be worth to do a searcher
2026-01-15T16:58:20 < karlp> how would I manually select the right things?
2026-01-15T16:59:20 < karlp> I know to manually pick dfn/son/qf* 33, 3x3, 3.1x3.1, 4.9x5.9, 56, 33, 5060, 3030 3333 are all "same"
2026-01-15T16:59:22 < karlp> but not...
2026-01-15T16:59:29 < karlp> can't b eat the human pattern matcher.
2026-01-15T17:00:06 < karlp> so, I'm goign through this again, you want to look for just a lfpka33 style? or 56 style?
2026-01-15T17:00:39 -!- _Posterdati_ [~posterdat@user/Posterdati] has quit [Quit: KVIrc 5.2.6 Quasar http://www.kvirc.net/]
2026-01-15T17:07:16 < karlp> ok, it took that long to select all the lfpak33 sized from the list
2026-01-15T17:07:48 < karlp> 60/100V means what, at least 60V ds, preferably 100v?
2026-01-15T17:08:48 < karlp> https://jlcpcb.com/partdetail/6332415-APG250N01Q/C5443731 perhaps?
2026-01-15T17:08:56 < karlp> 100V, 25A. 3x3?
2026-01-15T17:09:01 < karlp> good stock, good price....
2026-01-15T17:09:43 < karlp> what would be more useful (to me) than an api,would be the ability to bookmark/share/save the state of the search boxes...
2026-01-15T17:09:59 < karlp> like I've manually slected like 50+ form packagetype, I'd l ike to keep that....
2026-01-15T17:10:41 < karlp> I don't know how much you're going to find with 100V in 3x3...
2026-01-15T17:12:08 < karlp> https://jlcpcb.com/partdetail/HUASHUO-HSBB6066/C2828498 is another 60V one, 5.2mohm
2026-01-15T17:12:25 < karlp> milli Ω
2026-01-15T17:12:27 < karlp> there we go.
2026-01-15T17:14:02 < karlp> there's a 20n06d as well, similar to the 40n06d I linked earlier, just... worse :)
2026-01-15T17:14:03 < karlp> https://jlcpcb.com/partdetail/HL-20N06D/C7543835
2026-01-15T17:16:05 < karlp> I think that's it for >=60V, stock over 10k, in lfpak33 sized...
2026-01-15T17:18:19 < karlp> here you go, perfecto: western and all :) https://jlcpcb.com/partdetail/TexasInstruments-CSD19538Q3A/C478471
2026-01-15T17:18:20 -!- duude__ [~duude__@user/duude/x-4676560] has quit [Ping timeout: 256 seconds]
2026-01-15T17:18:42 < karlp> 100V, 15A, 3x3, I'm 98% sure that the 3.3x3.3 are actually "same" as the 3x3, but... do check ;)
2026-01-15T17:20:12 < karlp> yeah, it is, lfpak33 itself is 3.3x3.3
2026-01-15T17:27:40 < qyx> lol karl you called that a win
2026-01-15T17:28:12 < karlp> what now?
2026-01-15T17:28:12 < qyx> csd19538 is great, I am using a similar one now
2026-01-15T17:28:25 < karlp> you're happy?
2026-01-15T17:28:26 < qyx> now tell me how did you find anything there
2026-01-15T17:28:30 < qyx> I already was
2026-01-15T17:28:32 < qyx> thanks
2026-01-15T17:28:44 < karlp> like i said, manually scrollign the package list and selecting what I knew to be ~lfpak33 sized,
2026-01-15T17:28:48 < karlp> then ordering by stock
2026-01-15T17:29:21 < karlp> takes a bout a minute to select them all scrolling that shitty litle 8 entry window, btu the results are worth it, IMO.
2026-01-15T17:30:07 < qyx> let's apply the same logic
2026-01-15T17:31:28 < qyx> selecting mosfets, it doesn't show other metrics except manufacturer and package
2026-01-15T17:31:32 < qyx> fukit
2026-01-15T17:32:38 < qyx> results remaining 60000 and there are no usable packages
2026-01-15T17:32:55 < qyx> I can only see packages starting with a number and letters A, B or C
2026-01-15T17:33:44 < qyx> I select extended only and whoa, there are 69000 packages!
2026-01-15T17:33:47 < qyx> *parts
2026-01-15T17:34:11 < qyx> I select economic only and there are ever more!
2026-01-15T17:34:27 < karlp> yeah, you have to mouse scroll this little window https://bin.jvnv.net/file/U5rmV/Screenshot%20From%202026-01-15%2015-33-54.png
2026-01-15T17:34:40 < karlp> and use ctrl-click to add them all.
2026-01-15T17:34:54 < karlp> shift-click doesn't seem to do anything to select in batches, but they're never ordered usefull anyway most of the time...
2026-01-15T17:35:57 < qyx> Transferred 317.14 kB (6.19 MB size)
2026-01-15T17:36:01 < qyx> this may be the issue
2026-01-15T17:36:05 -!- duude__ [~duude__@user/duude/x-4676560] has joined ##stm32
2026-01-15T17:36:08 < qyx> the response is truncated in the middle
2026-01-15T17:36:45 < karlp> oh, you ahve a borked web connection, no wonder you can't do anything :)
2026-01-15T17:37:21 < qyx> https://bin.jvnv.net/file/0mwlK/Screenshot_2026-01-15_16-37-04.png
2026-01-15T17:37:24 < qyx> this is all I see
2026-01-15T17:38:21 < karlp> hah
2026-01-15T17:38:37 < karlp> sux2bu I guess then?
2026-01-15T17:40:03 < qyx> that's the modern web development, return 6 MB response when the only thing you want is a compressed list of metrics
2026-01-15T17:40:09 < qyx> and first ~100 results
2026-01-15T17:43:26 < karlp> no, I htink they're doing the right thing. 6MB is the list of metrics,
2026-01-15T17:43:33 < karlp> it has all the packages and shit and all the variations
2026-01-15T17:43:43 < karlp> so that when it loads, the selectors are all filled in.
2026-01-15T17:43:59 < karlp> I think you'r eunderestimating how much metric data there is.
2026-01-15T17:44:06 < karlp> you just to fix your ninterwebnets.
2026-01-15T17:44:18 < karlp> ok, docs sent off, interruption complete. 
2026-01-15T17:44:24 < karlp> what can I serch for next?
2026-01-15T17:45:49 < qyx> I am not underestimating, tme search works, mouser search works, dk works
2026-01-15T17:45:52 < qyx> just jlc crops
2026-01-15T18:01:13 < qyx> also their "search for" is logic OR
2026-01-15T18:01:18 < qyx> insane
2026-01-15T18:11:48 < PhantomWork> down to 62 unique components... slightly better!
2026-01-15T18:54:46 -!- hexo__ [~hexo@user/hexo] has quit [Remote host closed the connection]
2026-01-15T18:55:10 -!- hexo__ [~hexo@user/hexo] has joined ##stm32
2026-01-15T19:02:36 -!- hexo__ [~hexo@user/hexo] has quit [Ping timeout: 256 seconds]
2026-01-15T19:17:00 -!- Laurenceb_ [~Laurenceb@cust226-dsl93-89-135.idnet.net] has quit [Quit: Client closed]
2026-01-15T19:40:35 -!- Livio [~livio@user/livio] has joined ##stm32
2026-01-15T19:47:19 -!- om3ga [~om3ga@93.177.187.134] has quit [Ping timeout: 240 seconds]
2026-01-15T20:07:28 -!- om3ga [~om3ga@93.177.187.134] has joined ##stm32
2026-01-15T22:17:28 -!- Livio [~livio@user/livio] has quit [Ping timeout: 252 seconds]
2026-01-15T23:00:01 -!- nerozero [~nerozero@87.253.63.54] has quit [Ping timeout: 265 seconds]
2026-01-15T23:30:35 < catphish> machinehum: my WCH board and programmer arrived - i have no immediate plans to do anything with it for now though
2026-01-15T23:31:53 < catphish> i've started trying to figure out how to write a driver for my FTDI stuff, not much idea what i'm doing yet
2026-01-15T23:46:50 -!- octorian_ is now known as octorian