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.

  • Resolved

DRV8837: overshoot values

Intellectual 760 points

Replies: 3

Views: 59

Part Number: DRV8837

Hi all,

Could you please help me answering following support request from our customer:

 “While making test protocol, calculating limits for measurements, and doing also some preliminary measurements, I discovered that some parameters are not completely defined into the datasheet of the DRV8837.

 Therefore, I kindly ask for your support.

 I need to know values for the overshoot and undershoot for the bridge output, as timing and amplitude.

The same information is needed for the input.

Do you have please scope waveforms for the inputs and the outputs, while this H-Bridge is driving a DC brush motor?

 

One more inquiry to add: what is the voltage on the output pins in the situation when the PWM enable signals switched to low?

There is table 1 in the datasheet, but I measured the outputs at about 1.5V in the first 2 combinations, and I have no explanation for this.

The VM=5V in our application.”

 

Many thanks in advance.
Regards,
Brian

  • Brain,

    The switching node voltage will have:

    1. one body diode forward voltage plus the high side FET parasitic inductance voltage above the input voltage on higher side.

    2. one body diode forward voltage plus the low side FET parasitic inductance voltage below the ground on lower side.

    The body diode forward could take 0.7V to 1V. I don't have the switching node waveform.

    If you saw 1.5V overshoot and undershoot, please measure the overshot and undershoot near the IC pin to avoid the trace parasitic inductance to get the real number.

    The input signal counts as a small signal. Datasheet's -0.5V spec should be good.

    Best Regards,

    Wang Li

    Motor Drive Solutions

  • In reply to Wang5577:

    Hi Wang,

    many thanks  for the answer.

    1. For the outputs, they will re-make the measurements accordingly.

            2.For the inputs, the issue they had is a minor undershoot,  with a measured value of -0,55V for about 20ns.

    "But I think this result is not a value to damage the inputs, so we can consider another timing value to measure this parameter.

    To explain: in the test protocol they specified 20ns for overshoot measurement. But I think that in the particular case of this IC where the input signal is under 100kHz, they can define the measurement time for overshoot with the same value as the rise-time, respectively fall-time, in this case 188ns in the datasheet (round-up to 200ns).

    Do you agree with this?

    One more question was addressed, maybe you could help also answering this?

    Adding to the question:
    "the issue that we have seen: while the nSLEEP pin is on LOW, as pe the functionality table in the datasheet, the DC Motor is in COAST mode.

    We measured the voltage of 1,48V at both outputs, with DC Brushed Motor connected.

    Please confirm, is this a correct value? Is there a calculation for this?"

    many thanks in advance.

    Best regards,
    Brian

  • In reply to Brian Zoro:

    Brian,

    For the inputs 20ns 0.55V undershoot issue I agree with you. And think it could be a measurement issue again. The probe's ground point difference could make different result.

    "while the nSLEEP pin is on LOW, as pe the functionality table in the datasheet, the DC Motor is in COAST mode" At nSLEEP pin low, all the FETs are turned off. But, the high side FET and the low side FET are still connected between input and ground. If the input voltage is not zero, the output pin can get some voltage even the high side FET and low side FET both are high impedance.

    Best Regards,

    Wang Li

    Motor Drive Solutions

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.