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.

DRV8305-Q1: current amplifier output distorts at positive peak when max duty is 91%

Part Number: DRV8305-Q1
Other Parts Discussed in Thread: DRV8302

Hi,

The current shunt is 1mohm and the peak current is 125A. I found, when the max duty is 91% set by MCU, the current amplifier output (SO1) distorts at positive peak (as Channel 3 shows). All the tree phases have this issue.Meanwhile, the negative peak is always fine. The actual Phase A current is measured by current probe in Channel 4, no obvious distortion happens.

When the max duty is set as 80%, the SO1 distortion disappears (as Channel 1 shows). 

At postive peak, the high side FET is at its max duty, so I checked the gate drive. Channel 2 is the high-side drive, Channel 3 is the low-side drive, Channel 1 is SO1. The gate drive resistors are 0 ohm,  sink and source current (Idrive) are set as max level. No abnormal gate drive is found.

I also check the current amplifer. The CL is 270pF. I think it is not the slew rate issue.

The positive peak of SO1 is below 2.5V (offset is 1.65V for 0A), so it is not the amplifier output maximum limit issue.

I would like to know why positive peak of SO1 distorts when max duty is set as 91%. Any other measurement I could do for this issue?

Thanks.

  

  • Check the processor side PWM signal for crosstalk - switching when not expected to or logic levels between the two digital values. I've seen at higher current operation (higher Duty cycle) board noise can affect these signals and cause poor motor commutation suddenly. Secondly, make sure the shunt resistors have plenty of pad contact and solder or else you will get voltage drop there.
  • Another possibility, check your traces for current feedback from the shunt. Make sure these are a differential pair from across the shunt back to the driver chip. Make sure there is no ground anywhere on that differential pair between the shunt and driver chip. If there is then large current will flow down that trace as a shared path to ground (sinking motor current) and causing voltage drop across that trace. The issue would appear at higher current levels.
  • Hi Jerry,

    Which version of the device is being used (DRV83053Q, 55Q, NE, or NQ)?
    Would you please re-run the scope captures with Channel 1 and Channel 2 off?

    Also, please zoom in on the distorted area. This will help confirm the behavior.
  • Hi, Rick

    DRV8305NQPHPQ1

  • Hi Scott,

    We measured the waveform on the current shunt. Even if it was noisy, we could also see the distortion at the positive peak, so it is not caused by the amplifier. The customer said on anther board, which has MCU, driver and MOSFET from ST on it, it also has this distortion. We think it might be a common issue for low side current sensing in large duty operation.

    Could you help confirm the 91% max duty operation on our EVM? I need the screenshot in this condition.

    I will also ask customer to use the firmware of our EVM (on C2000) to check again. Meanwhile, BEMF of the motor will be checked as well.

    Thanks.

  • Jerry, 

    I cannot speak to the specifics of your IC or application. I used the drv8302 in the past. 

    Another thing that comes to mind that we had to work out is sampling the current value, it must be sampled at the center of the low FET on cycle if using a low side shunt sense. At high duty cycle this window gets very small and you need to trigger and get the sample fast. Otherwise your sample is not quite correct. This is all in the MCU analog/PWM section. To check the sampling is correct at high duty cycle, maybe save the values to a buffer and print them out and plot on a graph. It's worth the effort to verify this.