• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 531 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

DRV8353R: Getting DRV FAULT when using Matlab/Simulink ePWM module for 3-phase PMSM

Steve Abel
Prodigy 60 points
Part Number: DRV8353R
Other Parts Discussed in Thread: TMS320F28379D, DRV8353

We have a custom board utilizing a TMS320F28379D processor, a DRV8353SRTAT and 3-half-bridges using Infineon IPP051N15N5 MOSFETs (6). When using the ePWM modules out of Matlab/ Simulink running at 20 kHz we immediately get a DRV FAULT once setting ENABLE TRUE. Sub-faults seen on SDO include GDUV and/or GDF. If I switch PWM control to slower running GPIO outputs (3 hi and 3 lo) I can avoid this fault if I set DIS_GDF in the SPI SDI DRV Control Word (reg2 bits 8 and/or 9). I do have SPI SDI register writing ability and have experimented with IDRIVE, TDRIVE and DEADTIME. My concerns are (1) that I am starting PWM switching too soon, (2) my gate drive voltages may not be the correct levels or (3) I have a wrong Control setting. I am including scope shots of various data. For these cases bus voltage is 24vdc. Our motor is not yet connected to the 3 phase outputs. Please ask questions.

Photo5811 is 2 hi-side gates switching (100 Hz and 50% duty) with an Off level of 24 volts (VM) and an On level of 35 volts (VCP). YEL is gate A hi, BLU is gate B hi, VIO and GRN are gate B lo and gate A lo, which are not command On

Photo5809 is a case where only gate A lo is commanded (100 Hz, 50% duty). VIO shows the gate B lo response (GND and 11 volts) and BLU shows the gate B hi response (odd looking)

Steve Abel

  • 0
    Prodigy 60 points

    To clarify some things.

    My VBUS is the same as hi-side VDRAIN

    VM is designed to be 1 or 2 volts below VBUS

    I observe that VCP is equal to VM when ENABLE false and equal VM+11 volts when ENABLE true

    External (measurable) VGLS is 15 volts I think with ENABLE true. The internal VGLS is 11 volts as implied and seen at a lo-side gate when commanded ON

    By "odd behavior" of the hi-side gate from the photo I mean that when commanded OFF (1) its level is GND when the lo-side is commanded OFF (both OFF), (2) it slowly ramps up when the lo-side is commanded ON and (3) eventually is clamped at the VBUS level. What should it be doing?

    We can supply circuit drawings as you request them.

    Steve

  • 0 in reply to Steve Abel
    TI__Genius 13690 points

    Hi Steve,

    Thank you for posting to the Motor Drivers forum!

    Steve Abel said:
    My concerns are (1) that I am starting PWM switching too soon, (2) my gate drive voltages may not be the correct levels or (3) I have a wrong Control setting

    From what it sounds like, I do believe that you are switching your gate driver outputs too soon as it sounds like the charge pump is not fully charged when you start commutating the motor which is why you are seeing a GDUV. I would recommend waiting a bit for VCP to charge up before PWMing the gate drive outputs to give the charge pump time to charge up.

    Steve Abel said:
    Photo5809 is a case where only gate A lo is commanded (100 Hz, 50% duty). VIO shows the gate B lo response (GND and 11 volts) and BLU shows the gate B hi response (odd looking)

    The behavior that you are seeing for GHx is correct as when the LS MOSFET is turned of GHx floats at the drain voltage which in your case is VBUS.

    Best,

    ~Alicia

  • 0 in reply to Alicia Rosenberger
    Prodigy 175 points

    I found that the issue is that the DRV8353 does not pull the low side FET below 1V.  If I lower the BUS voltage from 30V to 11V the fault goes away.

    I tried this across all drive strengths and the fault remained.

  • 0 in reply to Michael Margolese
    TI__Genius 13690 points

    Hi Michael,

    Would you mind creating a new thread by "Asking a related question" as it sounds like the issue you are describing will involve a different debug?

    Best,

    ~Alicia

  • +1 in reply to Alicia Rosenberger
    TI__Genius 13690 points

    Hi Steve,

    I am going to close this thread for now, if you have any further questions, please feel free to re-open this thread. If your questions have been answered, please mark this thread as resolved.

    Best,

    ~Alicia

  • 0 in reply to Alicia Rosenberger
    Prodigy 175 points

    Hi the above reply is the same project and question steve is working on.