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.

DRV10987: Abnormal Kt problem

Part Number: DRV10987
Other Parts Discussed in Thread: DRV8308, DRV8308EVM, DRV10970

Hi,Support team!!

I am using the DRV10987 with a fan with four pole pair, Kt =  9.2mV/Hz, Rph = 1.08 Ohm, Imax = 1.6A, RPM max = 30000.

"DRV10987 0x90 0xC33E"
"DRV10987 0x91 0xA3A"
"DRV10987 0x92 0x20F1"
"DRV10987 0x93 0x2AF8"
"DRV10987 0x94 0x3F8F"
"DRV10987 0x95 0x703F"
"DRV10987 0x96 0x47"

Q1)BEMF constant(Kt)(mV/Hz) value in display tab of GUI should be not changed,but it is changing according to frequency.

      does it mean the Ktc is displayed?

Q2)in my application,As frequency increases ,the BEMF constant(Kt)(mV/Hz) value in display tab of GUI decreases.

     if the rotation speed increase more than 10000rpm ,the BEMF constant(Kt)(mV/Hz) value became less than 0mV/Hz and the motor was stopped because Abnormal Kt is detected.

     I would like to know how to calculate the value and whether this behavior is correct or not.

    in the datasheet,   the "motorKt" is integrated periodically while motor is spinning

                                  Ktc (V/Hz) = ({MotorKt} / 2) / 1090

Q3) the datasheet recommends maximum operating closed loop speed is 1000Hz.(electrical frequency)

     as I wrote before, I would like to use DRV10987 under 30000rpm,4pole pairs BLDC. 

    is it Out of scope of specification?
Thanks and best regards.
  • A1)The BEMF Constant in the display tab is the BEMF constant measured by the DRV01987 when its doing its sensorless calculations. As long as your tuning settings are correct, the BEMf constant should stay around the value of the BEMF constant you entered
    A2) The measured BEMF constant lowering as you increased motor speed cmd indicates there is something incorrect in your tuning, or the DRV10987 cant spin the motor at the higher speed you want. Please refer to the DRV10987 tuning guide.
    A3) Hz = RPM*#OfPolePairs/60 = 30000rpm*4polePairs/60 = 2000Hz. So this rpm would be out of specification.
  • Thank you for your reply.

    A3) It is not reccomended value, is it?

          it's very unfortunate ,but I have to choose another solution.

          thanks.

         

  • I think it is recommended value, so it may still work, but I am not sure. 

    I will also transfer you to our C2000 support to see if we can recommend any other devices to spin your motor at your desired 2000 Hz electrical frequency.

  • Hi,Sanmesh Udhayakumar -san

    Thank you for your quickly response.
    Can I ask two question?

    Q1)I would like to know the specification of maximum operating closed loop speed (Hz).
    I found that approximately 50000RPM are supported with motor constructions of up to 16poles in DRV8308 datasheet

    Q2)What is the bottleneck that determines maximum operating closed loop speed(Hz)?

    Thanks.
  • 1) The maximum operating speed for the DRV10987 as mentioned previously is 1000Hz. Just as a note: DRV8308 is a hall-sensored device, so if your motor has hall sensors, then you can use that.

    2) The things that determine maximum operating closed loop speed are:

    a) The motor control algorithm. This also includes whether the algorithm is sensored, or sensorless. Usually, sensored algorithms (used with motors with sensors like halls or encoders) can spin motors at faster speeds. also the accuracy of control by the motor control algorithm. This is usually the bottleneck, and thus you probably just want to see on the datasheet what the recommended max operating speed of the motor is

    b) Possibly the pwm frequency and the adc sampling if it is used. But these are not usually the limited factors.

  • Hi,Sanmesh Udhayakumar -san
    Thanks you for always quickly response.

    1) I think there are differences between recommended value and absolute maximum rating value,
    you mean that the 1000Hz is defined as absolute maximum rating value ,not recommended value.
    is it correct?
    if it is recommended value ,I would like to know the absolute maximum rating value of the speed(Hz)
    This is important for our decision.

    2)-a) if the 1000Hz depends on the sensorless algorithm,our application can monitor the hole sensor with CPU.
    if the cpu monitor the rotation speed,can we use drv10987 in the condition of more than 2000Hz?
    2)-b) I also think PWM freq. is not bottleneck,because our current system use PWM 50kHz and it's enough fast against the rotation speed.
    about adc sampling, I'm not sure because the inner structure is not opened from you.

    Thanks and best regards.
  • 1) 1000Hz is recommended value. I  do not know what the absolute maximum rating value is. 

    2) When you say "hole sensor", did you mean your motor hall-effect sensor? If so, we can recommend parts other than the DRV10987, since DRV10987 is meant for sensorless motor driving. We can spin at higher speeds for motors with sensors like Halls. For example, the DRV8308 can spin motors up to 6.7kHz like you said before, and we would recommend you look at DRV8308evm http://www.ti.com/tool/DRV8308EVM. Can you confirm if your motor has hall effect sensors (halls)?

  • Thank you .
    1)I got 1000Hz is recommended value.
    Do we have the possibility of using the DRV10987 at higher speeds more than 2000Hz?

    2)DRV8308 needs another CPU to control the FET.It doesn't match our requirement because our current solution is same as your suggestion and we don't want to change the IC without rewards such as cost reduction,space saving.

    Thanks and best regards.
  • 1) I am not sure if DRV10987 can spin motor up to 2000 Hz. I would think it would be unlikely since it is 2x the max frequency we recommend. I mentioned in the previous posts what causes the limitations of the max speed motor can be driven.

    If you want to try seeing if you can spin up to the speed, please try the settings attached below that you can load into the GUI or you can open as a csv to see what the register settings are.

    user5945607e2eSettings.csv

    Also picture of the settings are below. They are conservative settings, to give you the most likely probability of spinnign properly, but they may need adjustments. Also make sure the phase resistance entered is the phase to center tap resistance, not the phase to phase resistance.

    2) I do not think the DRV8308 needs an external MCU. The external MCU is optional and I think the device has internal speed-loop control. So this solution may work. A reference design is here http://www.ti.com/tool/TIDA-00197 

    3) If you are looking for another cost-competitive solution with DRV10987 solution that has hall sensors to enable faster, I would recommend DRV10970 ().

  • Hello,

    The DRV8308 can be used without an MCU if your motor has halls, does your motor have halls?

    Regards,

    -Adam
  • Hello,

    Thank you for your reply.

    Yes, our motor has halls sensors.

    I found DRV8308 has  clock frequency mode function.

    According to the datasheet,it drives the motor to match frequency between CLKIN and FG frequency.

    Q1) if I would like to drive the motor of 4poles pairs at 30000rpm,should I input the signal of 500Hz to CLKIN terminal? 

    Q2) if the torque changes,Does DRV8308 adjust the motor rotation speed to keep at 30000rpm?

    Q3) in our application,We need to control the motor speed with rapid deceleration.(ex)30000rpm to 10000rpm within 0.1sec)

          For rapid deceleration ,I think I should use the DIR pin to reverse while motor is still spinning.

          I know this may cause excessive current flow,but I would like to know the behavior.

         if you have a better idea ,could you give me your suggestion.

    Thanks and best regards.