• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 83 subscribers
  • Views 753 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.

DRV-69M-KIT based driver OEM

Mikael Sundberg
Intellectual 740 points


Hi,

We have developed our own inverter based on the design in the DRV-69M-KIT. Now we have an issue with the driver.

When running the ID-process with lab2c we get good values and everything goes well. Motor spins up and valueas are repeatable and very similar to that from the KIT. However, if we now reenable the RUN command we just get an erratic jumping around motor with our driver(not the kit) and sometimes a complete driver failure. (We have 10 pcbs that all reproduce these results.)

We are really trying to identify the problem here. We have made huge efforts on the pcbs, with solid 5oz copper and kelvin connections to the three current sensing resistors connected(0.002Ohm) The trace lenghts are also very short, no single one more than 2 inches.

Anyone have an idea what could be the issue? We think since the identification is all well, this should work. Any ideas also what we could measure, change or anything. We are pretty desperate at this moment.

Would it be beneficial to double the current sensing resistor to enhance sensitivity? (And of course adjust current scaling accordingly.)

I am certain there is an issue with either the voltage feedback cirquit or current feedback. Which one is used during the ID process?

Could it be noise from oversaturated gates on the mosfets? We shall try with higher resistance on the driver gates.

Any input is highly appreciated! 

Br,

Mikael

  • 0
    TI__Guru** 116140 points
    Both 3 phases current and voltage sampling circuit are used for motor parameters identification and running. These two sampling circuits are very sensitive for identification and running.

    Did the motor not run well on low speed or high speed, with high load or without load?

    It's better to follow below rules for PCB design if the board is for high current motor.
    1. Low-side current sensing route differentially, and add two small resistors and capacitor near the shunt resistor if the route is long to DRV or amplifier.
    2. Ensure pwm signal without crossover with SOA/B/C of DRV
    3. Ensure minimal loop VM to DGND and DVDD to PGND of DRV device.
    4. Place DRV or amplifier near the MCU, and add a RC filter near ADC of MCU for current and voltage sampling 56(OHM)/2.2(nF).
  • 0 in reply to Yanming Luo
    Intellectual 740 points

    Hi Yanming,

    Thank you very much for the proposals.

    1. Are you talking about adding a low value resistor to the current resistor diff pair? We allready use differential routing btw.

    2. You are talking about the PWM driving the mos gates here crossing the voltage feedback routes?

    Your other suggestions are allready implemented.

    The VFB is looking ok on the scope and very similar to the ref-kit during identification. I am thinking it coul get corrupted on the way somewhere or that the current sensong is off.

    The motor ONLY runs during the full ID process. After that the motor is completely unusable and we get an overcurrent situation. This is what is annoying. If I could understand what the difference is from when identifying the motor to when running it maybe I could get a better clue of the problem source.

    Br,

    Mikael

  • 0 in reply to Mikael Sundberg
    TI__Guru** 116140 points
    Did you get the similar identified parameters of motor on OEM board and TI EVM Kit? If no, it means there is some issue on current or voltage sensing circuit. If yes, it means the low current is ok for motor running, but the high current will cause noise on current/voltage ADC channel. You can use a DAC or Datalog (as lab01b or lab01c) to check the current/voltage feedback signal.
  • 0 in reply to Yanming Luo
    TI__Guru** 116140 points
    Hi Mikael,

    Any update from your side and have any further questions on this? We will close this thread if no.
  • 0 in reply to Yanming Luo
    Intellectual 740 points

    Hi,

    The update is that we remade the design slightly with more layers, different ground plane routing and so on. We have no conclusions yet if this solved the issues. Further testing this week will reveal if the issues persists.

    /Mikael

  • 0 in reply to Mikael Sundberg
    Intellectual 740 points
    Finally, this issue was completely resolved. An improved layout with more solid ground planes 6 layers, a via grid and good signal/power separation seemes to have made the disturbances vanish. Now all is just excellent!
  • 0 in reply to Mikael Sundberg
    TI__Guru** 116140 points
    Congrats. Great, thanks for sharing.