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.

Possible blown VREF input on DRV8825

Other Parts Discussed in Thread: DRV8825

I have an issue with a board that I designed using the DRV8825 (two per board).  We designed the boards, built and tested them, and everything seemed fine.  Recently, I got a report about the stepper motors not operating properly.  When I investigated, I found that there was a problem with the VREF signal.  (One divider sets AVREF and BVREF for both parts).

I am setting VREF with a divider (1.30K over 1.00K and 0.1uF cap to ground) from a 3.3V supply.  This gives me a nominal 1.43V reference.  Combined with my 0.5 ohm sense resistors, I should be looking at about 0.6A peak current.  The circuit does work as designed.  However, on this particular board, VREF is sitting at 0.069V and phase current is accordingly lower.  Unpowered, the resistance from that node to ground is about 27 ohms instead of about 770 ohms typical for a working board.  There is anecdotal evidence that at least one other board in the field has this same issue (out of about 8 currently deployed).

I am using an independent 3.3V supply to generate VREF (instead of V3P3OUT) and divider resistance is quite a bit lower than what is shown in some of the reference schematics.  I have an E-stop switch, which cuts the VM power (24V regulated) while leaving 3.3V intact.

Is there anything that I might have done to damage that input?  Is it sensitive to backpowering?  Is it particularly sensitive to ESD?  I cannot find anything in the datasheet that would explain the issue.

  • Hi Karl, the only part I don't follow is "divider resistance is quite a bit lower than what is shown in some of the reference schematics."  Are you referring to your 1.3k and 1k divider, or a different one?  Could you share part of your schematic?

    Given the high failure rate (out of 8), if there's a voltage spike occurring it might be easy to replicate and observe.  Try scoping the 3.3V supply while power cycling the system and starting/stopping the motors.  Also, try pressing e-stop repeatedly to see if that can cause it.

    Best regards,
    RE

  • Yes, the divider resistance in question is the one that sets the VREF voltage.  I am using 1.3K and 1.0K, while I have seen reference designs with 50K and 70K or a 10K potentiometer.  If there was a backpowering issue when the VM supply is turned off, it could be worse in my case because the lower resistance would feed more current into the pin.

    I will take a look at the supply for good measure, but I don't know how likely a spike is.  The supply that feeds VM has about 4000uF of bulk capacitance and the motor windings are only being fed with 600mA.

    Thanks

  • I just tested applying 3.3V directly to AVREF while VM was disconnected, and current was in the nanoamps.  I increased the voltage some and power cycled, and it stays consistent, and unpowered DC resistance stays the same (about 1.7 Mohms).

    It's really not clear what's causing your issue.  Maybe if you try enough conditions, you could reproduce the failure and give us a hint on where to look next.

    Best regards,
    RE

  • Okay, I will play around with it and see if it happens to a new board.

    I don't suppose there is any history of those pins being more susceptible to ESD than some of the other pins?  The boards are now conformally coated, but there is a chance that one was zapped before coating or that the motor winding signals were zapped through the external IO pins.

  • All pins are rated and tested for 2kV HBM ESD.

    Are all your grounds tied together?  What device is producing 3.3V?  If you have a diff probe, try probing 3.3V vs VM during powerup, operation, motor stopping, motor load changing, and e-stopping.

    Best regards,
    RE

  • I realize that this is a nearly two year old thread, but we are currently experiencing a similar issue and were wondering what the final resolution was for you? Just was curious about your findings . . . will start my own thread later today.
  • Well, we never got a final answer on the issue.  I did not replace the chip on that board because of the extra effort involved with the thermal pad.  Our design is low volume, so we are not a good sample size, but we have probably 30 of these boards of varying ages in the field and we have not seen the issue again.

    Good luck.  I would be interested to find out if there is either a manufacturing issue or something that can occur externally to cause this.