• State
  • Locked Locked
  • Replies 9 replies
  • Subscribers 29 subscribers
  • Views 1340 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

DRV8842: DRV8842 Creating Scratchy Motor Noise at low Duty Cycles

Jamey Hurd
Prodigy 70 points
Part Number: DRV8842

Operating with 20kHz PWM Frequency, and at low duty cycles, a constant but inconsistent scratching noise (around the 500-1khz freq) occurs. The lower the duty cycle, they louder it gets. Varying the PWM frequency/period does not affect the scratching frequency. Stalling (holding the motor shaft) does not affect the frequency (It isn't mechanical, bearings, etc.). We've taken a different motor controller and ran the same code and the noise is no longer there. Is this some sort of decay problem or slew rate issue with the DRV8842?

Thank you for the help.

  • 0
    TI__Guru** 114905 points
    Hi Jamey,

    What is considered a low duty cycle?
    Which decay mode being used?
    Are you using any current regulation features?
    When running the different motor controller have you observed the outputs of both controllers to determine what is different?
  • 0 in reply to Rick Duncan
    Prodigy 70 points
    What is considered a low duty cycle? A: Less than 20 - 30%
    Which decay mode being used? A: We've tried all three with the same results.
    Are you using any current regulation features? A: We've the resistor options of the controller set to 100% (We're are trying different settings now)
    When running the different...? A: The DRV8842 has sharp pulse edges. The NXP part has rounded pulse edges.
  • 0 in reply to Jamey Hurd
    Prodigy 70 points
    Tested with different current regulations, same results.
  • 0 in reply to Jamey Hurd
    Prodigy 70 points

    I have another observation. With a steady 19.29kHz pulse into the DRV8842, the output to the motor shows a pwm signal that varies from that of the input (See Screenshot). The frequency shifts around inconsistently and also changes in width. I've also been able to correlate the scratching noise with the DRV8842 enabling and disabling inconsistently around the threshold of where the DRV8842 starts to register input pulses. This is around 2% duty cycle. The noise disappears as the RPM is increased. The shuttering of the output pwm does not get better with higher duty cycles.

    So based on these results, it seems we are running into two problems:

    Problem 1:

    The DRV8842 does not want to operate below 2% duty cycles. Output from the DRV8842 only enables after it receives duty cycle above 1.52%. The DRV8842 output enables/disables inconsistently until it reaches a duty above 2%.

     (Screenshot 1, Blue highlight shows the inconsistent PWM, Orange highlight shows inconsistent enabling/disabling)

    Problem 2: 

    The DRV8842 exhibiting erratic duty cycles and frequency/periods at any duty sent to the input.

    (Screenshot 1 & 2, Blue shows the inconsistent pwm, Orange shows some odd flutter that occurs. As the RPM is increased, this area rises to eventually become the same level as the output (18VDC))

  • 0 in reply to Jamey Hurd
    Prodigy 70 points
    Still looking for help on this. Observing this problem more today, this shaking PWM output is causing acoustic noise at certain frequency windows. For Example 2-5% duty, 15-20% duty, 30-35% duty exhibits loud acoustic noise, even with the motor stalled.
  • 0 in reply to Jamey Hurd
    TI__Expert 4650 points

    Hi Jamey,

    I apologize for the delay in response.  We're working on debug with our designer, I should have a response to you in the next 24 hours.

  • 0 in reply to Phil Beard
    Prodigy 70 points

    Screenshot of one of the acoustic noise spots. (motor running at ~40% duty)

    Same duty cycle but with the motor stalled.

     

    Very odd that the output pulse is being clobbered.

  • 0 in reply to Jamey Hurd
    TI__Expert 4650 points
    Hi Jamey,

    From your post above it looks like the device is hitting the current limit and entering into an interval of Mixed Decay. If you use Slow Decay is the acoustic performance any better?

    The shuttering displayed in your outputs is explained by the input deglitch circuit inside of the DRV8842. Once an input signal is present on the INx pins, it must first pass through a deglitcher before the DRV8842 will change the state of the outputs. This delgitcher will take between 400- and 800-ns randomly, so when triggering on the INX signal that is consistently timed, the corresponding output will shutter within this deglitch spec.

    With the motor output condition shown above I have a couple questions:

    1.) What is the temperature of the device during this testing?
    2.) How much current is being driven through the motor?
    3.) What does the VM Voltage look like during this PWM frame?
    4.) What is the state of the nFAULT Pin?
  • 0 in reply to Phil Beard
    Prodigy 70 points

    1. Room temperature.

    2. 2A at stall, 50mA with no load

    3. ?

    4. Normal

    I'm concerned that this deglitcher is a show stopper. It's taking a nice pulse and outputting a pulse with a randomly delayed 400 to 800ns pulse that even changes in width randomly. This is causing acoustic noise that we can't work with. And this is apparently by design within the deglitch spec as you stated. I don't see the word deglitch in any of the DRV8842 datasheet. Unfortunately we have a full production run of these assembled parts in house.