DRV8323R: Driver failure

Hi Alex,

1. All of our testing should be done at 20kHz PWM.

2. I reviewed the old forum post, and it seems like your drive current is incredibly low, leading to very long turn on and turn off times for your MOSFETs. I'm wondering if these long turn on and turn off times for the FETs is actually degrading them over time which is causing your driver failures.

Has any investigation been done on the MOSFETs themselves in these failed cases? I would be curious to see what the results of those investigations would be. 

My first guess is that the MOSFETs eventually fail due to thermal fatigue/gate degradation over time. If you think about it from a system level, the DRV device is really not doing anything volatile. Just supplying gate current in response to input. The MOSFETs however are subject to a slow switching cycle which will definitely increase the thermal stress they are subject to. I would try experiments regarding FET lifetime and drive current.

Thanks,

Joseph

Hi Joseph,

Thank you for your answer. So with a 30kHz PWM frequency, service life will be reduced.

Yes drive currents are very low with IdriveP 10mA and IdriveN 20mA. But before that we used IdriveP 60mA and idriveN 240mA, and failure seemed to occur even more than now!

For each failure we test the Mosfets, and when we change the driver, all works fine, so we didn't analyze the Mosfets so far.

We keep on going with testing at 10kHz PWM frequency, we believe this could reduce the failure rate.

Another question please: what' the humidity level for the service life specification?

Thanks,

Best regards

Alex

Hi Alex,

Thank you for the extra information. Yes, you can test the 10kHz PWM frequency and see if you observe a reduction in failures.

We don't have a specification for the humidity level, however I'm wondering if this may be the source of the issue?

Can potential humidity build up cause a short between some pins on DRV device? Do the DRV device have a resin coating on them to protect the pins from moisture level? If you tested this, would a protective coating remove the issue?

Thanks,

Joseph

Hi Joseph,

Thank you for your answer.

We've seen in another application with high moisture and high ambiant temperature that failure rate was even higher.

Another critical point: we use the integrated LMR16006. So one more question please about service life specification: does this data include the simultaneous operation of the DC/DC converter?

Many thanks,

Alex

Hello,

In this case, I would recommend using some sort of resin or potting to protect the system from moisture. It sounds like this is a direct cause for failures.

Regarding the LMR16006, I believe you would need to ask the team responsible for that device for lifetime operation. Did you see any failures with this DC/DC converter specifically?

Thanks,

Joseph

Hello Joseph,

thank you for the recommendation, that's what we did.

I come back to the LMR16006, my question is: when you give a service life of 100 000 hours at 105°C for the DRV8323, is it with LMR16006 active?

No we didn't see any failure with the converter, it still works when DRV8323 gate outputs fail.

Thanks,

Alex

Hello,

Yes, the buck is active during service life test.

Thanks,

Joseph