DRV8432 BURN-OUT

Hi Davorin

We need more info about the issue. Could you show the block schematic of this application? Also capture of the layout will be good as well as some spec or description of the load. We need to understand the current level and the load condition. Some test waveform of the normal running will be welcome too if you still have one.

Also could you tell whether there is a heat sink or whether the part go very hot during normal running? or any thing special happened at the failure point?

Thanks.

Hi Zuo,

During normal operation, when we have all control signals connected to DRV8432 works fine. With EVM and also with our board.

We have adequate heatsink with fan and measure temperature very close to DRV8432.

Current limit resistor is 24k. We measure current with ACS711, and limit the duty cycle. Current limit is set to +/-7A per channel (A-B, C-D).

In both cases when we burn DRV8432, PVDD (48v) and GVDD (12V) was ON.

Also load (TEC)  was connected all the time. The only think we do is connecting and disconnecting input control signals.

Best regards

Davorin

Hi Davorin

I think the attachment of your last post is only including the power supply blocks. Could you show the DRV8432 block schematic?

We don't have similar experience like this before. Also I think it will be very rough to analysis the root cause with so limited information. I suggest to do more reproducing test until we can identify the main condition which is causing the issue. 

We can make some conditions worse than normal to increase the rate of the failure. Most burn out issue is caused by high voltage over shooting or outputs short or some ESD event.  Also we need to take some measures to limit the damage of the failure like setting more limits on the test power supply to avoid PCB broken so that we can easily swap the chip to test again. (apply enough samples before the reproducing test)

Thanks.

Hi Zuo

Thank you for your help. Last time I forget to attach the DRV8432 schematic.  

Situation when we burn the DRV8432:

PWM Freq =300kHz

PWM-A  2-98%, PWM-B inverted,          RESET A-B - HIGH,         OUT A-B = Load TEC cca 6ohm

PWM-C LOW, PWM-D HIGH,                 RESET C-D HIGH,          OUT C-D = NO LOAD (OPEN OUT)

Maybe is in some connections with bootstrap capacitor charging , or PWM inputs connected to LOW / HIGH all the time, or no load connected to input. 

In data sheet I found that bootstrap capacitor might not be able to be charged if no load, or extremly light load is presented at output.

Davorin

Hi Davorin

I haven't seen the bulk capacitor (>100uf, prefer 470uF to 1000uF) on all the PVDD pins except the four 0.1uf. Am I missing it on your schematic?

It could be a big risk without the bulk capacitor placed close to the PVDD pins. Especially at the turn on moment of big current because the PVDD voltage maybe unstable at that time. That could cause a burning issue when the internal logic is affected by the unstable monent.

Thanks.

Hi Zuo,

Bulk capacitor is 560uF, look at  previous attachment with power  supply block.

Davorin