DRV8302: nFAULT permantly active, Chip gets hot

Rüdiger,

Thanks for posting on the MD forum!

Please add me as a friend and we can share the schematics in private message.

Have you checked the device output pins using a scope while the application is running and loaded?

If you remove a suspected problem unit from the PCB and compare resistance measurements to GND with a known good unit, do you notice any pin differences?

Regards,

-Adam

Hi Adam,

thanks for your reply.

>>Have you checked the device output pins using a scope while the application is running and loaded?

If you are talking about the Gatevoltage for the high and lowside, yes I checked this, there is nothing but I would wonder if there would be something because of the pending error and because of GVDD is ~0V.

>> If you remove a suspected problem unit from the PCB and compare resistance measurements to GND with a known good unit, do you notice any pin differences?

Only for that I understood the right way. I remove the Gatedriver from a "bad" and from a "good" PCB and measure the resistance of the 6 gates against GND?

-> I will do this tomorrow!

Best regards,

Rüdiger

Rüdiger,

Please also check the GVDD, 

Regards,

-Adam

Huhu,

today I unsoldered the Gatedriver from a "bad" and a "good" PCB.

The measurement was not  really exiting, the impedance against GND for the low side corresponds to the pull down for the gate, for the high side it's about 5MOhm.

More interesting was the measurement of "bad" PCB's with placed Gatedriver and unsoldered FET. For some Outputs, Highside or Lowside, the impedance I measured for about 2 Ohm! It seems like a shortcut inside the chip which would explain that it gets warm/hot.

What could be the reason for this?

Rüdiger,

Unfortunately it sounds like you may have damaged those outputs due to transients on the output or similar. Please check the output pins on a working unit with a scope under load and see if you see any violations of the ABS max table.

Regards,

-Adam

Hi Adam,

Checking the pins during work is not so easy, the motor drives a bike.

I was thinking about HOW the transients damage the Gatedriver. Does it "work" over the the FET and Source-Gate line which is really high-impedance or do the transients shoot into the Sense inputs? For the sensing input I only found a max. current of 1mA, but what voltage does this correspond to? Does TI recommend an explicit transient protection via for example ESD diode?

Regards,

Rüdiger

Rüdiger,

Damage can occur on any of the output pins or sense lines, this is why I have asked to check the voltage on those nodes during operation. 

Some customers use schottky diodes on the SHx nodes and from gate to source on all FETs to help protect against transients. You could try this.

Regards,

-Adam

Hi Adam,

I wish you a happy new year.

Unfortunately I had no opportunity to do the measurements yet. 

Regarding your suggestion to place Shottkys:

- SHx nodes against what? GND or VSS? What is the goal? Support for the body diode? Is the assumption that the body diode is not fast enough to lead the current away?

Regards,

Rüdiger

sorry, I meant GND or VDD...

Rüdiger,

Once you take your measurements we can decide if and where to add diodes. If the transients are on the gates then we can add a diode between gate and source, if the transients are on the switch node then we can add diode from SHx to GND.

Regards,

-Adam