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.

DRV8323: Overheating

Part Number: DRV8323
Other Parts Discussed in Thread: DRV8301

Hi, my DRV8323S is overheating when operating at a VM of 48V and applying PWM to 3 phases.  This seems to be a problem at 36V+.  When the board sitting on the bench in open air, the DRV8323 case surface temperature reaches 150C in under 2 minutes.  Eventually, it will trigger thermal fault/warning/shutdown.  If the board is enclosed, the driver will fail and need to be replaced.  I've isolated the current draw of the chip and I only read 62mA peak, but I will test this again when the package gets hot, to be sure.  Is this heat to be expected at 48V, even though the device is rated to 60V?  How can I keep it from getting above 125C?

Attached are the driver and half-bridge schematic.  Thank you.

drv8323S.pdfhalf_bridge.pdf

  • Hello Jed,

    An expert has been notified and should be able to answer or ask some more questions on Monday or Tuesday.

    For now, if I understand you correctly, it sounds like you do not hit thermal shut down VM < 36V? Do you expect higher current through the FETs when greater than 36V based on your application? 

    I'm going to assume the expert is going to want to look at the DRV8323 and FET (power stage) layout, if you can provide it. This is because the situation sounds like it might be intended operation but the layout or performance has not been optimized for your application. I'd recommend you check out the layout app note for motor drives if you are unsure of typical thermal layout techniques: http://www.ti.com/lit/an/slva959a/slva959a.pdf . Furthermore, using the EVM in the same condition might be able to help determine this if the EVM does not experience thermal shutdown. 

    There is the possibility of abnormal behavior in regards to the DRV8323, so providing more details about the use case (VM, Output Current, Output Power, application, features of the DRV8323 that are being used) will helpful.

    Thank you,

    -Cole

  • Jed,

    Could you share your SPI settings as well?

    Regards,

    -Adam

  • Hi Adam,

    Some background: We have similar firmware running on a C2000 DSP to a DRV8301 with seemingly no issue (obvious differences between the 8301 and 8323 aside). We switched to DRV8323 mostly to save board space. PWM is at 100kHz in 6x Mode. With the PWMs in

    There doesn’t seem to be much of a thermal difference when changing the value of i_drive_p/n (hs & ls) through its range. Nominally we run: i_drive_p = 10 mA, i_drive_n = 20 mA
    t_drive = 500 ns
    t_dead = 50 ns

    With the PWMs off, we see package temp of 55C, but with PWMs enabled but commanding 0 current we see > 130C

  • Jed,

    Are you using 394 ohms on each gate?

    Can you share a waveform of the SHx node, High Side Vgs, and Low side VGS?

    Regards,

    -Adam

  • 392 ohms, yes.  We were using 10 but had a pretty severe transient/ringing so I increased the value until the switching waveform was dampened.  I scoped either side of the gate resistor with 10 ohms populated in all 6 places and then 392 to make sure there wasn't an increase in gate power draw.  There was actually a decrease.  The driver overheats in any case.  Here's 392ohms then 10:

  • Jed,

    Can you share your layout?

    What FET are you using?

    Regards,

    -Adam

  • All,

    To summarize the private discussions, the layout showed that there was minimal thermal vias and powerpad did not have great thermal escape. 

    We are trying a heatsink to see if this helps the issue.

    Regards,

    -Adam