• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 263 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.

DRV8874: DRV8874 Burn, BEMF

Donghee KIM
Prodigy 30 points
Part Number: DRV8874

Tool/software:

I am driving a bidirectional DC motor using drv8874.

The operating voltage is 28V, and the motor consumes 400mA of current when not loaded.

The motor driver was damaged during operation. (The temperature rose over 100 degrees, and it did not burn due to the current limit of the power supply.)

BEMF occurred over 60V for 250us. 

There was no problem when driving a 1A resistive load and a 300mA solenoid valve.

The data sheet says that when enable is input low, it can enter brake mode to reduce the counter electromotive force, but it seems insufficient.

I also installed a 100uF electrolytic capacitor and a 100nF ceramic capacitor on the VM side.

When I searched the community, it said to install a TVS diode on both ends of the motor or increase the bulk capacitor capacity, but I am in an environment where I cannot increase it any further.

Please suggest a good solution. Please recommend an appropriate motor driver, TVS diode, Schottky diode, etc. Thank you for reading.

  • 0
    TI__Mastermind 44685 points

    Hi Donghee,

    Thank you for posting in this forum.

    Donghee KIM said:

    The motor driver was damaged during operation. (The temperature rose over 100 degrees, and it did not burn due to the current limit of the power supply.)

    BEMF occurred over 60V for 250us. 

    BEMF exceeding the supply rail by this huge amount will definitely damage the device. Please see below highlighted absolute maximum specifications in the datasheet. If the output pins go below the GND by 0.9 V or above VM rail by 0.9 V the device will be suffer permanent damages. Resistive loads with stray inductance and solenoid coil may likely have inductive kickback but no BEMF. This inductive kickback will result with body diode VF excursion below the GND or above the VM rail but not to the extent to violate the absolute maximum ratings. 

    Donghee KIM said:

    I also installed a 100uF electrolytic capacitor and a 100nF ceramic capacitor on the VM side.

    When I searched the community, it said to install a TVS diode on both ends of the motor or increase the bulk capacitor capacity, but I am in an environment where I cannot increase it any further.

    Please confirm if the bulk capacitor was very close to the VM pin of the device? Was there a series diode or a long PCB trace between the VM and the capacitor? What is the design of the power supply providing the VM rail to the device? Please share the schematic of this portion. 

    You may have to have four TVS diodes two on OUT1 one to VM and another to GND and two on OUT2 similarly. While TVS or Schottky diodes connected in this manner can help keep the OUT1, 2 voltages within specifications it does not do anything to VM rail unless there will be another TVS diode between VM and GND to absorb and mitigate voltages greater than 30 V for example.

    Can you please share oscilloscope captures of OUT1, OUT2, VM rail and if possible the motor current waveform for review? Thank you.

    Regards, Murugavel 

  • 0 in reply to Murugavel4637
    Prodigy 30 points

      

        

    Thank you for your answer. The oscilloscope waveform is orange for voltage and blue for current.

    I couldn't capture more detailed waveforms due to component damage.

    The current is 1A@1V. It's 400mA when running, and the motor rotates quickly at 6000RPM.

  • 0 in reply to Donghee KIM
    Prodigy 30 points

    It is a tantalum capacitor on the circuit, but it has now been changed to an electrolytic 100uF/50V specification.

  • 0 in reply to Donghee KIM
    TI__Mastermind 44685 points

    Hi Donghee,

    Thanks for the additional details.

    The diode D1 installed for reverse polarity protection is the cause for isolated voltage pumping on VM due to motor BEMF. Can you please try testing the drive with a shunt connection across D1? Not sure why the TVS D2 did not clamp the voltage - can you check the voltage at the D2 / C2 junction while the motor is running, with and without the shunt across D1? Thank you.

    Regards, Murugavel

  • 0 in reply to Murugavel4637
    Prodigy 30 points

    Thank you for your reply.
    I was wondering why the TVS diode was not working as you advised, and after adding one TVS diode to the VM, it worked normally.

    The temperature of the driver was 37 degrees when it was working. When it was turned off, the temperature of the driver rose to 50 degrees and it turned off normally. I deleted D1 and tried to operate it, but there was no difference.

    I have one more question.
    Can I make the motor work in reverse by giving the ph terminal low while it is working in the forward direction? Can I do it without giving it a dead time?

    Thank you very much.

  • 0 in reply to Donghee KIM
    Prodigy 30 points

    Blue is VM Voltage when motor is off.

    I don't know if it's because of the soft decay of the motor driver or the TVS diode,

    but it doesn't rise any further around 40V and slowly goes down.

    I think I should consider a TVS with a slightly lower voltage.

  • 0 in reply to Donghee KIM
    TI__Mastermind 44685 points

    Hi Donghee,

    Donghee KIM said:
    I was wondering why the TVS diode was not working as you advised, and after adding one TVS diode to the VM, it worked normally.

    Glad this addressed the issue. Seems like TVS diode is clamping as expected. 

    Donghee KIM said:
    The temperature of the driver was 37 degrees when it was working. When it was turned off, the temperature of the driver rose to 50 degrees and it turned off normally.

    Could you please describe how the turn off was done? Was it by Low-side slow decay or bridge Hi-Z. With slow decay the BEMF energy from the motor will be dissipated by the LS FETs and could increase the temperature of the driver for a short duration which is fine. 

    Donghee KIM said:
    I deleted D1 and tried to operate it, but there was no difference.

    Because you had a functional TVS on VM removing D1 did not show any difference. If no TVS shunting D1 should address the failure. However, having a TVS would be sufficient to ensure the VM voltage did not exceed specifications. D1 will provide reverse polarity protection.

    Donghee KIM said:
    Can I make the motor work in reverse by giving the ph terminal low while it is working in the forward direction? Can I do it without giving it a dead time?

    Theoretically, you could do this. However, you must ensure the FET's are capable of handling the additional current on top of the inrush current while switching direction of the spinning motor. The BEMF energy must also be dissipated by the driver. You can characterize the current excursion from 400 mA at steady state speed in one direction to inrush current for absorbing the energy until BEMF is 0 or motor stop and then reversal staring from 0 BEMF in the opposite direction. This usually may last a few ms which may be okay for the FETs to handle. This depends on the system hence my request to characterize this behavior. 

    Donghee KIM said:

    Blue is VM Voltage when motor is off.

    I don't know if it's because of the soft decay of the motor driver or the TVS diode,

    but it doesn't rise any further around 40V and slowly goes down.

    I think I should consider a TVS with a slightly lower voltage.

    This tells me you were switching off the VM power and let the motor coast to a stop. The motor driver outputs seem to go HiZ. Because VM did not go > TVS clamp voltage TVS could not have kicked in. The BEMF slowly reduces while the motor coasts to a stop. The BEMF voltage is conducted to the VM rail via the FET body diodes. If you could physically brake the motor you may notice this voltage drops faster until VM = 0. Thank you.

    Regards, Murugavel   

  • 0 in reply to Murugavel4637
    Prodigy 30 points

    I used low ide slow decay with 'enable=0'.

    Thanks to you, I'm having fun solving difficult problems.