Post Go back to editing

H-bridge application of ADuM4223 for stepper motor

Hello Team,

I am using ADuM4223 in H-bridge for driving stepper motor ( model no.: 34HS46-5004D-E1000)

Below is schematic of stepper motor driver circuit. Here, 12V is supplied to ADuM4223 and  24V supplied to H-Bridge. Host micro controller is used for controlling motor. 

I am facing problem of MOSFET (csd18537nkcs) damage. Vds of MOSFET is 60V and Id is 50A.

The frequency of failure of MOSFET is not fixed.  Some times driver runs for 20 day continuously. System involves frequent ON and OFF operations in different modes.    

Please guide me on this. Let me know any other information needed.

 

Parents
  • Hello Kiran,

    The ADuM4223 has a long history of successful applications, and many hours logged of operation. I don't see anything immediately obvious on what could be causing the failures, but there are some debugging tests we can look at to investigate:

    1) The ADuM4223 has a very powerful gate drive. The turn off path has a diode with no external series gate resistor. Is it possible to get a waveform of the lowside VOB to GNDB pin? We can look for undershoot. Too much undershoot could cause an issue. This is a difficult measurement to make cleanly, and we might get what looks to be undershoot, when it is just measurement noise. A physically small measurement loop is suggested.

    2) Is there any chance of the controller accidently causing a shoot through command (for example both inputs high at the same time). How much deadtime is being added by the controller to distance the turn off edges from the turn on edges? Is it possible to get a waveform of the VIA and VIB signals?

    3) What dv/dt is being seen at the MW_A to GND and MW_B to GND nodes? Is this exceeding 25 kV/µs? I don't think it will with a bus voltage of 24 V.

    Thanks,

    RSchnell

  • Hi ,

    Thank you for reply.

    Please find below waveform details.

    1. VOB signal at 30 RPM.

    2. VOB signal at 120 RPM

    3. VIA and VIB signal at 120 RPM


    4. MWA and MWB at 20 RPM

    5. MWA and MWB at 120 RPM

  • Hi ,

    I had shared captured waveforms, as per your request. Any clue regarding this.

    Regards,

    Kiran

  • Hi ,

    Please find below the captured waveforms at the input terminals (PWM2LPWM2HPWM1L and PWM1H) of ADUM4223.  

    Also, please find below the waveforms at the gate terminal of VOB2, VOA1, motor terminal A and A'.

    I observed false signal at the gate terminals and at motor terminals. Same obeseved on other H bridge. I am not able to analyse this. Please guide me.

    Regards,

    Kiran

Reply
  • Hi ,

    Please find below the captured waveforms at the input terminals (PWM2LPWM2HPWM1L and PWM1H) of ADUM4223.  

    Also, please find below the waveforms at the gate terminal of VOB2, VOA1, motor terminal A and A'.

    I observed false signal at the gate terminals and at motor terminals. Same obeseved on other H bridge. I am not able to analyse this. Please guide me.

    Regards,

    Kiran

Children
  • Hello Kiran,

    Sorry for the late reply, I missed the E-zone message alert. I apologize for that.

    It looks like your measurements of VOA1 are taken between VOA1 and (CURR_SENSE_BR1 or GND). Do you have a differential probe that could allow you to measure between VOA1 and MW_A? It could make troubleshooting a little easier. If you don't have a differential probe, you could do oscilloscope MATH to have (VOA1 to GND) - (MW_A to GND).

    For a bootstrap setup, the highside gets charged when the lowside is on. From the capture of the PWM signals, there's a long time between current directions. During this time when all switches are off, there will be self commutation in the motor windings. Also during this time, the bootstrap capacitor for the highside is slowly discharging. It looks like it is not a problem in the waveforms you have shown, as I don't see too much voltage droop on the first cycle of VOA1. I can't post pictures right now for some reason, but I hope you understand what I mean.

    Which part of the last waveform do you think are the false signals? I think if we get a signal captured with differential probes, it should look more like we expect.

    RSchnell

  • Hello Kiran,

    Thanks for reply.

    We do not have differential probe. We did measurement between VOA1 to MW_A terminals after disconnecting all other probes . Please find below waveforms captured between

    VOA1 to MW_A terminals (Q3_VGS) at  30 RPM.

    and VOA1 to MW_A terminals (Q3_VGS) at  10 RPM.

    Secondly, in the last waveform, It was not expected highlighted part of the waveform as below.

    Reagrds,

    Kiran

  • Hi Kiran,

    I noticed one thing in the schematic. The resistors and extra capacitor on the gates of the high sides should terminate to the half-bridge midpoint:

    RSchnell

  • Hi RSchnell,

    Thanks for reply. We will implement your suggestion.

    Also, we could measure VOA1 w.r.t. MW_A and  VOA2 w.r.t. MW_A' unsing our DSO. These waveforms are same as VOB1 and VOA1.

    Regards,

    Kiran