This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Blown Capacitor on TPA3121D2 power amp. Sparks, Fire, & Bad smells etc.

Hi, 

We are currently in production of a radio product that goes into Fire/Police/Ambulances and have seen a problem on a couple of our units where the PVCCL bypass capacitor (connected to pin 3) has blown spectacularly in the field (it was a tantalum cap->lots of fire/molten metal).

They have strayed from the EVM and recommendations by using a 220 16v  tant. cap instead of the 470 35v one used in TI's EVM board.  We are also only using the left channel for our product so the other channel and components related to it are not there.  So that capacitor is connected to everything AVCC is on including the Right channel PVCC pins.

We have protection: 20v MOVS, etc which seem to work ok for spikes and things.  We are not sure if we are using the wrong voltage/type of caps or if the TPA3121 or something else has caused the capacitor to blow?  We are not sure if it was during power up or not.  It had been in the field for a few days working.

I guess at this point we just want to find out if we are using the wrong type/voltage capacitor on that pin?  or if it shouldn't be connected to the right channel VCC as well.  There are a lot of other possibilities and maybe the capacitor is just a symptom of it, but just want to rule out the capacitor first.

Attached is a .jpg of a snippet of our schematic with the TPA3121.  C38 is the one that blew.

Thanks,

-Vic

Raven Electronics 

  • Hi, Vic,

    what is the power supply voltage?

    I also noticed that you have the unused input dc coupled to ground. I am not sure if this is causing the problem, but this is not a good idea. You should AC couple it to ground. 

    D2

  • Tantalum capacitors are know for their spectacular failures under conditions of high inrush current. Do a web search for 'tantalum capacitor inrush current' and you'll find several relevant hits. Several manufactures make specialty tantalum capacitors that are 'surge current' rated, but 'general purpose' tantalum capacitors are typically very poor under surge conditions.

  • Thanks Steve, I think this is the problem.

    I've looked it up.  Since we are already in production with it,  the easiest way to help with this problem is to double the working voltage of the cap and to do what you suggested, look for a specialized surge current cap that fits.

    It's only happened twice out of maybe 200 units, once a few months ago when i immediately turned on a bench supply and then once in a fire truck in the field, I don't know the conditions/if that happened at power on.

  • Hi Don,

    The power supply voltage is the vehicle battery voltage 12-14 volts, we have it clamped to 24 volts or so max with mov's  and they have tested it on specialized automotive surge testing/spiking equipment and I think it passed ok.

    We will have to put a capacitor to ground on the unused audio input in the next iteration.

    I'm thinking its the surge problem to the cap suggested by Steve above.  We will look at doubling the cap rated voltage and maybe look for specialized tantalum cap with surge rating that still fits on our board.  

    I'm happy to get all these good suggestions and help from everyone.

    Thanks

  • At least a fire truck was nearby :)

  • Hi, Vic,

    I hope that the cap is the issue.

    Steve, thanks for chiming in with your experience! Great work!

    -d2