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.

DRV10983 destroyed when power interrupted

Other Parts Discussed in Thread: DRV10983

Hi,

I have been using the drv10983 chip to power a small 3 phase water pump. I have used both the EVM as well as a board I designed. I am finding that in either case, it is very common for the drv10983 chip to burn out and smoke when the DC power to the chip is briefly interrupted and then restored. It happens even if the motor is idle when the power disconnect happens.

The chip is being controlled under i2c from an attiny85 microcontroller that is being powered by the VREG (pin 6) output of the drv10983. The output is set to 5v, but the microcontroller runs at 3.3 v (It has it's own regulator). I have now seen this happen on 7 different chips, including on the EVM. It the case of the EVM there was no I2C control at all, the motor was connected but was not turning.

When the chip blows, it turns white hot in the area of pin 23. 

What can I do to protect the chip from this failure. I can reproduce it within seconds on any chip.

The phase resistance of my motor is 2.78 ohms and the Phase to Phase Kt is 205 mV/Hz. It is a 24 pole motor. 

The schematic is literally a copy of the EVM schematic, with the addition of the microcontroller.

Thanks,

Don

  • Hi Don,

    Sorry to hear of your issue. We will contact our experts on this device and have them reply.

    Is it possible that when the power is restored the voltage is exceeding the maximum voltage of the device? What is the duration of the typical power interruption? Also how is power interrupted? Is a power switch used or is this some type of hot swap scenario?
  • Hi Don,

    I have contacted the expert for this device and they should respond shortly.

    In the meantime, can you take an o-scope and examine what happens to the voltage when you interrupt the power supply? Are there any voltage spikes on the PCB that might be damaging the IC? 

  • I have checked it with a storage scope and I did not see any significant spikes. It is kind of hard to capture it because it doesn't fail every time and I am running out of chips. I don't think it is voltage spikes though. It happens both when the chip is running on my highly regulated bench supply and when it is running off of the actual system supply (which is a brick style cord mounted supply, kind of like a laptop supply). The nominal supply voltage is 24 volts. It acts more like a lock up condition in the chip I think. I have seen the failure occur when I was running from my bench supply set at as low as 12 volts with a current limit of .8 amps. The chip doesn't smoke under those circumstances, it just turns into a dead short.

    I am wondering about a bulk capacitor near the chip. Perhaps I should add a 100uF or so capacitor near the chip??

    Don
  • Hi,

    As I mentioned to Nicholas, I have been unable to see any significant spikes with a storage scope. The power is generally interrupted because someone disconnects the unit from it's power supply, which is a cord mounted brick type power supply. There is a typical DC power connector between them. I think the failure happens when the DC connector is plugged in hot and it does several breaks / makes before it connects good. So the duration could be very short (milliseconds). But it fails even if the motor is not running at all, in fact it happens when the chip should be in standby.

    I am wondering about the value of a bulk capacitor near the chip. Maybe 100 uF or something like that. Any recommendations?

    Thanks,

    Don
  • Hi Don,

    What is the current bulk capacitance for the PCB power rail? I would definitely recommend adding a 100uF and see if the issue improves.

     

  • It is 10 uF (same as the evaluation module). Is there a particular value you would recommend?

    DOn
  • I'll let the engineer responsible for the device answer for that. Generally bulk capacitance sizing is heavily application dependent.

    You can refer to section 10 (page 29) in the below datasheet for some general information regarding bulk capacitance.

    http://www.ti.com/lit/ds/symlink/drv8301.pdf

  • Hi Don,

    From your description of fault scenario, most probable reason of device failure appears to be over-voltage stress during hot plug condition with frequent make/brake cycle. As Nick suggested, try adding 100uF decoupling cap this will help to prevent over-voltage surge during such scenario and would solve the issue.

    Best Regards

    Milan-Motor Application Team

  • That did it. Thanks.