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.

DRV8301: DRV8301 Latched OC on a few boards.

Part Number: DRV8301
Other Parts Discussed in Thread: CSD19532Q5B,

Over-Current Latched Failures:

I'm seeing what appears to be random behavior on the first run command (post a reset or a pwr up) issued to a motor (most boards OK, small percentage fail). If the first run command works, then subsequent run commands work OK without failure, even on a bad board. I found setting the OC behavior of the 8301 to Over-current Latch mode produces the OC error on bad boards, however, changing the behavior to Current Limiting seems to fix the bad board with no errors on the first (or subsequent) runs. Wonder what is different about the first Run (actually a Ramp command) command that might setoff the bad boards. Also increasing the OC threshold from 0.138V to 0.155V seems to fix the problem as well. Lowering the Threshold a few notches and the bad board continues to work most of the time, indicating the rare OC is not real perhaps.

Also, monitoring the 48V current supply, I don't see an over current. My normal Ramp produces a peak current during acceleration of about 140mA and when I get the OC latch error the scope doesn't trigger. If this were a real OC I would think I would see something. I don't see any punch through current when looking across the sense resistors either so I think the dead-band is sufficient.

I had a great deal of trouble trying to pass emissions testing. Tried setting the drive current down to 0.25A but no help (little to no effect). I adjusted the gate resistors to 70 Ohms, going above 150 Ohms caused the DRV to stop working and I don't know why. Next, I had to add RC snubbers (C=3300pf, R = 1 Ohm across each MOSFET, CSD19532Q5B) which finally fixed the bad emissions. Not sure if any of this is causing the internal workings of the DRV to work inconsistently on a few boards or not. I also have dead-band dialed into the Piccolo, but don't remember how much. 

We tested many boards in OC limiting mode in development and never saw this and perhaps was masking some other underlying issue. These boards also work in OC latched mode and these older boards still work OK but it seems like the new batch are failing? Is it possible we have either bad parts (ESD damaged perhaps, given the supply chain issues) or something else.

Any suggestions appreciated.

Thanks.

  • Hi Robert, 

    Let's confirm if some things are actually happening or not before making any assumptions. 

    To confirm if OCP is occurring, can you please:
    - Share the OCP_ADJ_SET used
    - Read address 0x00 to see if any FETxx_OC bit(s) get set
    - Monitor nFAULT (falling edge trigger) with the VDS of each FET to see if the VDS of that FET exceeds the OCP_ADJ_SET value. For instance, monitor nFAULT, GH_x/GL_x, and SH_x/SL_x for each FET to confirm if an OCP event occurs at that FET

    In regards to emissions testing, lowering the gate drive current will help to ensure you aren't slamming the MOSFETs on/off and causing EMI ringing due to fast VDS slew rates (<100ns usually). However, ringing and transients may also occur from poor PCB layout which can add parasitic inductance that contributes to ringing at the switch node. Can you ensure that on the powerstage layout of the PCB:
    - The gate drive traces are at least 10mil wide 
    - There is enough copper and vias to dissipate the max motor current in the PCB (follow current path of battery, HS FET, switch node, LS FET, motor ground)
    - RC snubbers are great and glad it's working. This is a good fix ifs the layout/components add parasitic inductance; however layout is usually the primary concern in regards to ringing if lowering gate drive current does not help. 

    Could you also please share:
    - Motor voltage & max current
    - Is the motor unloaded or loaded during the test?
    - Does this issue happen across multiple devices or one device?

    Thanks,
    Aaron

  • Hi Aaron, Please refer to my private post I sent to you. Thank you.

  • Hi Robert,

    Replied over PM. I'll give you a call later today. 

    Thanks,
    Aaron

  • Hi Aaron,

    I hope I'm replying on the right location in this thread, hard to tell. Anyway, Can you describe the process for a customer return since I want to send back the DRV8301 chip?

    Thanks,

    Rob

  • Hi Robert,

    We are in the process of looping in a field application engineer to assist with the customer return process offline and I am providing a response to your feedback now. 

    Thanks,
    Aaron

  • Hi Robert,

    We have assigned an FAE to assist with the customer return process, let's continue discussions offline so we can continue to investigate the batch of DRV8301 units. We appreciate your patience. 

    Thanks,
    Aaron