BPODs suddenly frying

vvega

My colleague and I were setting up 4 BPODs (2.5) and Analog Output Modules, and after connecting them to the computer and flashing them, they seemed to work just fine. However, when we connected them to a second computer and plugged them into power (DC), they made a high-pitched noise and then turned off--LED powered off, no longer connected. now on my bench-testing computer, they don't turn on at all and I can't detect the board.
The power supplies we have are 12V, 2.0A, 24W. Did those fry the boards? Nothing appears burnt on the boards, and there are no visible shorts anywhere. 1 board made the noise and turned off, but it powers on just fine when connected via USB to my bench-testing computer. The other 3 don't. The ones that don't turn on seem to get very hot (qualitative), close to the usb port, when plugged in, so I suspect something to do with the power?
I tried testing 5V to gnd on the teensy with a multimeter and read something like 4.8V-4.9V, so hopefully it's not completely toast...
Does anyone have tips for how to test the boards to see if they can be salvaged? Or what could be causing this?

Josh

Hello vvega,

We ship power supplies with each Bpod state machine.
If you used your own, please share the make and model.
Also, please check the polarity of the power supply. Most are positive-center but others are negative-center. Bpod requires a positive-center power supply. You'll want to look for these symbols on the adapter to tell them apart. The positive-center symbol is also drawn next to the jack on the state machine enclosure.

If any part of a state machine gets uncomfortably hot to the touch when plugged in to the USB jack only, that's bad news. A component may have burned out, creating an electrical short.

Please let me know what you find.

Thanks,
Josh

vvega

Hi Josh,

We used Triad Magnetics's WSX120-2000 (DigiKey). Thankfully the polarity is correct, positive-center. After further testing, now 2 of the 4 boards will connect to my computer--I reflashed the teensy by holding down the button on the teensy until its small red light came on, and then I re-uploaded the firmware. Now for those two boards, they connect properly, but the RGB LED doesn't turn on and I'm not sure why.
for the other 2 boards, nothing has worked--trying to reset the teensy does nothing, and they both do have a component on the teensy that gets hot to touch when plugged into USB only. Maybe they are dead dead? I haven't tried poking every component on those boards, but I have pinpointed the heat to just 1 component on the teensy itself. Do you think it's reasonable to desolder the teensys and replace them if they're dead?
I also haven't tried plugging any of the 4 boards into DC power again since the first incident.

Josh

Hi vvega,

Looking at the specs of the adapter, I don't see anything that should be of concern.
If Bpod software connects but the light doesn't turn on, it's possible that a different firmware version was loaded. Using LoadBpodFirmware, the one you want is StateMachine_Bpod_2_5 from the dropdown menu, or if you're using Arduino you want the version here.

It sounds like the other two boards are dead dead. Desoldering a Teensy is technically possible, but it would take a lot of time to get right. Are these units you built?

The DC power jack only connects to the valve driver chip. It is galvanically isolated from everything else on the board, so it should have no effect on the Teensy (unless there is a short circuit somewhere).

Josh

vvega

Hi Josh,

Okay, I'm glad to hear that the adapters should be compatible! I'm not sure what would have caused the systemic problem then.
I did use that specific version of the firmware linked through Arduino on all of the boards, both before the incident and after. Now, though, the two detectable boards don't light up the RGB LED at all. The LED doesn't look obviously burnt out, and measuring across the cathode/anode with a multimeter in diode mode gives a nonzero voltage (~0.5-2V)

I did build the units: reflow oven for the SMD parts -> manual cleanup of any bridges + short-circuit testing w a multimeter -> soldering iron for the through-hole parts -> quick short-circuit testing of the teensy.
I'll do another round of short-circuit testing, but they all passed the first two rounds.

Josh

I'd check to make sure the resistance between barrel jack ground and USB ground is infinite. I'd also check the barrel jack pin vs. USB ground, which should be infinite. Please let me know what you find.

vvega

Josh Apologies for the delay in response -- I get infinite for both conditions.

Josh

If that's the case, there is no path for current between the barrel jack and the microcontroller, so the issue is probably unrelated to the DC adapter.

Where did you source the components from? In my experience it's best to buy Teensy boards directly from PJRC, or from a major industrial supplier (e.g. Digikey, Mouser, Adafruit). If you send photos of the top and bottom of the PCB to support@sanworks.io I'll look them over.

Thanks,

-Josh

vvega

Good to hear! I bought them from DigiKey, so they should be real/reputable. I will send pictures over. Thank you for your help!