• State Resolved
  • Locked Locked
  • Replies 21 replies
  • Subscribers 30 subscribers
  • Views 4497 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.

DRV8871 startup

Hiroshi Sato
Expert 2590 points
Other Parts Discussed in Thread: DRV8871, INA225

Hello,

Our customer now trying to use DRV8871, but their startup current is too big(about 5A/1us) and the device could not operate correctly.

It seems that it due to current regulation (not OCP), so they had changed Rilim  from 39k to 15k ohm.

The result was sometimes it worked correctly but sometimes not.

To work it correctly, 

What value of Rilim should be used (what is the smallest)?

Please let us know the settings to extend tBLANK or tDRIVE time, if there is.

 

Thanks and best regards,

Hiro

  • 0
    TI__Guru** 114905 points
    Hi Sato-san,

    Can you provide a scope capture of the current? 5A for 1us should not create current regulation.

    I would recommend trying ~23kOhm. This will increase the current to 2.5A to 3A.

    Using a 15kOhm resistor can create OCP conditions (OCP minimum is 3.7A)


    What value of Rilim should be used (what is the smallest)?

    We recommend setting the current to <3.5A. This is ~20k across the 59 to 69 kV constant.

    Please let us know the settings to extend tBLANK or tDRIVE time, if there is.

    These are fixed by the device.
  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    Thank you for your  comment.

    Pease see attached file, there are two circuits for driving two different motors.

    In their measurement, the motors had not driven with Rilim=39k ohm, but they had driven with Rilim=15k ohm.

    The customer want to know how to use DRV8871 with 5A/1us startup current.

    I think if there is any way to disable current regulation function, it will be possible.

    Or, if there is any way to reduce the startup spike current, it will be ok too.

    Thanks and best regards,

    Hiro

    schematics_and_waveforms.xlsx

  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points

    Hi Sato-san,

    Thank you for the scope captures.

    Is there a capacitor across the motor terminal? Is this causing the current spike?

    Hiroshi Sato said:
    The customer want to know how to use DRV8871 with 5A/1us startup current.

    The device is designed to operate with a 5A/1us startup current. The 5A/1us startup is ignored during the 2us blanking time.

    In the scope captures labeled abnormal operation, the 1.64A current limit is causing the device to enter current regulation. The DRV8871 is operated as designed.

    In the scope captures labeled forward1/2 and reverse1/2 with a 15kOhm resistor, the current does not stop because the current is less than the chopping current.

    After the 2us blanking time, the current will stop only if:

    1) There is an overcurrent event (current > 3.6A) or

    2) There is an overtemperature event

    3) The current is being regulated by the device.

    Has the customer tried the recommended 23kOhm resistor?

  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    Thank you for your detailed comment.

    The customer have tired 23k Ohm resistor and any abnormal waveform was not observed.

    And they have tried with 15k Ohm again, there was no abnormal waveform too. 

    The datasheet describes the minimum resistance is 15k Ohm, but do the current regulation operate unstable with 15k Ohm in their application ?

    If so, they want to know the mechanism of the current regulation function/circuit (the reason for 23k Ohm or gureater, circuit detail ).

    And if the circuit is a kind of current shunt monitor, they think that the startup current is reduced or omitted by adding input RC filter, is it correct ?

    And, they think the timing parameters are very important now.

    Because the reason  that their startup current has no problem  is  the time is shorter than tBLANK(2us) and tocp(1.5us).

    Culd you provide us the min/max raw data or designed value for tBLANK and tocp ? If possible, they want to know tOFF and tRETRY min/max too.

     

    Thanks and best regards,

    Hiro

  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points
    Hi Sato-san,

    The datasheet describes the minimum resistance is 15k Ohm, but do the current regulation operate unstable with 15k Ohm in their application ?

    >> At 15k, the current regulation is typically 4.2A. This is in the OCP range. When current regulation is set in the OCP range, device to device variation can cause some devices to regulate properly and others to trip overcurrent. It is recommended to regulate current below the OCP range to avoid device to device variations.

    If so, they want to know the mechanism of the current regulation function/circuit (the reason for 23k Ohm or gureater, circuit detail ).

    >> I suggested 23 kOhm to set the maximum current across the Vilim constant (59 to 69 kV) to 3A. The resistor could be set to 20k Ohm, which is 3.45A for the max Vilim constant of 69.

    And if the circuit is a kind of current shunt monitor, they think that the startup current is reduced or omitted by adding input RC filter, is it correct ?

    >> I don't understand this. Where would the input RC filter be added?

    And, they think the timing parameters are very important now.

    Because the reason that their startup current has no problem is the time is shorter than tBLANK(2us) and tocp(1.5us).

    Culd you provide us the min/max raw data or designed value for tBLANK and tocp ? If possible, they want to know tOFF and tRETRY min/max too

    >> We will research this information. It may take a few days to gather the data.
  • 0 in reply to Rick Duncan
    TI__Guru** 114905 points
    Hi Sato-san,

    tBLANK can vary +/- 50% from the typical value
    tOCP can vary +/- 50% from the typical value
    tOFF can vary +/- 30% from the typical value
    tRETRY can vary +/- 25% from the typical value
  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    Excuse me for my poor explanation,

    I attached the file for the detail of the input RC filter.

    Please see it and give us advice.

    Thanks and best regards,

    Hiro

    current_regulation-filter_plan.pptx

  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    Thank you for your sending the data.

    Now, the customer understand that the way to operate their motor is

    The peak current duration makes shorter than tBLANK 2.0us and tOCP 1.5us.

    Rilim should be chosen greater 20K or 23K ohm.

    But in their test with Rilim=39K ohm with above two condition is cleared, the current regulation had activated and the output current was stopped.

    They have another question, please see attached file and give your comment

    Thanks and best regards,

    Hiro

    DRV8871_ITRIP_waveform2.xlsx

  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points

    Hi Sata-san,

    Can you ask for scope captures of the current plus the two outputs?

    If the time between outputs off and on is ~25us, this is normal current regulation.

    If the time between outputs off and on is ~3ms, this is an overcurrent event.

    If the time varies, this could be an overtemperature event.

    Based on the ppt file, this is not an overcurrent event. An OCP event measures current in one direction only.

  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points
    Hi Sato-san,

    Sorry for the delayed response. Adding a sense resistor in series with the ground pin is not recommended.

    Adding a sense resistor in series with the ground pin can create unexpected results. The sense resistor is raising the ground reference of the pin. This can cause inputs to be interpreted as a logic 0 instead of a logic 1.

    If using an INA225 with the DRV8871, it is recommended to place the resistor in series with the VM pin. The INA225 group can then provide recommendations on RC values and placement.
  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    Thank you for your comment.

    And I'm sorry for I could not inform you correctry what the customer want to do.

     Please see attached ppt and give us any advise.

    Thanks and best regards,

    Hiro

    current_regulation-filter_plan-2.pptx

  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    Thank you for your comment. We understand that an current regulation or an overtemperature event happened.

    Now, the cusotmer  have two questions.

    1. Why dose the current regulation activate otherwise a peak current duration is shorter than tBLANK ?

      (They know that the peak current value itself was exceeded the Itrip they have set, but the positive peak period did not exceed tBLANK.)

    2. When the current regulation funciton is not used, what is the recommended value for RILIM ?

      (They know that the minimum value you recommended is 23k ohm, but they assume it is the min. value with using current regulation function)

    Thanks and best regards,

    Hiro

  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points
    Hi Sato-san,

    The designer says if GND and PGND are separated, Rilim should be connected to GND. He also said there should be no additional capacitance on the pin.
  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points

    Hi Sato-san,

    1. Why dose the current regulation activate otherwise a peak current duration is shorter than tBLANK ?

     (They know that the peak current value itself was exceeded the Itrip they have set, but the positive peak period did not exceed tBLANK.)

    Current regulation should not activate if the peak current is shorter than tBLANK.

    Can you please capture the output voltages and current during this event so we can determine if it is current regulation or overtemperature? Please show images at 100us/div, 2ms/div, and 100ms/div. If the output re-enables in approximately 25us, it is current regulation.

    2. When the current regulation funciton is not used, what is the recommended value for RILIM ?

     (They know that the minimum value you recommended is 23k ohm, but they assume it is the min. value with using current regulation function)

    If not using current regulation, it is recommended to use a resistor that will act as a safeguard. The value should be set such that current regulation activates in abnormal operation. For example, if the motor current should not exceed 1A, the RILIM resistor can be set for a 2A current.

    Although the minimum allow resistor in the datasheet is 15kOhm, I suggest a minimum of 20kOhm. A 20kOhm resistor sets the current regulation less than the OCP value ( 3.2 +/- 0.25A).

    Using a 15kOhm resistor can cause overcurrent events because the current regulation is 4.25A +/- 0.35A.

  • 0 in reply to Rick Duncan
    Expert 2590 points
    Hello Rick-san,
    Thank you for your confirmation and sending comment, I understand.
    Then, the customer's idea(for cutting peak current, adding an external filter in front of the "internal current sense") may correct technically,
    but it could not apply to DRV8871 due to its structure, isn't it ?

    Thanks and best regards,
    Hiro
  • 0 in reply to Rick Duncan
    Expert 2590 points
    Hello Rick-san,
    Thank you for your comment.
    About 1. , the customer will send us output voltage and current waveform in next week, please wait.
    About 2., they understand now.

    Thanks and best regards,
    Hiro
  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points

    Hi Sato-san,

    You are correct. Adding an external filter is technically correct, but does not apply to the DRV8871 due to its structure.

  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    Thank you for your sending comment, I understand.

    I'll inform our customer of it.

    Thanks and best regards,

    Hiro

  • 0 in reply to Rick Duncan
    Expert 2590 points

    Hello Rick-san,

    I'm sorry for late replying, the customer took long time for capturing the waveforms.

    Please see attached file and let us know about below.

    * In "before" waveform, why TRIP had occured though tBLANK was shorter than 1us.

    * In "after" waveform, please let us know whether the waveform has any problem or not.

    Thanks and best regards,

    Hiro

    DRV8871_ITRIP_waveform20160924.xlsx

  • 0 in reply to Hiroshi Sato
    TI__Guru** 114905 points
    Hi Sato-san,

    In "before" the output does not trip until OUT1 drops low. This is the current regulation which has a 25us off time.

    The second positive current spike is above 1.67A, which is probably causing the outputs to be disabled.

    Using the previously suggested 20kOhm (current regulation at 3.2 +/- 0.25A) should prevent the chopping. This should allow the device to operate like the "after" while avoiding potential OCP events.
  • 0 in reply to Rick Duncan
    Expert 2590 points
    Hello Rick-san,
    Thank you for checking and comment.
    The customer understand that and I asked them to change RILIM from 15k ohm to 20k ohm.

    Thanks and best regards,
    Hiro