• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 29 subscribers
  • Views 322 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

DRV8412: Please review the schematic

Chase Jeong
Mastermind 6520 points
Part Number: DRV8412

Dear TI experts,

My customer now tests their own PCB with DRV8412. Could you review the schematic below?

The problem is, they use this circuit to operate peltier.

if the duty of PWM is under 50%, it is okay. (about under 5A current consumption.)

But if the duty of PWM is slightly increased and about 80% (about 6~7A) DRV8412 "sometimes" does not work.

they tested several PCBs, all the PCBs have a problem but each PCB have different location of failed operation of DRV8412. (one PCB contains 6 DRV8412s and 6 peltiers.)

Can you guess what is the problem? Please advise me about this.

Best regards,

Chase

  • +1
    TI__Mastermind 23271 points

    Hey Chase,

    It sounds like a thermal issue to me - have they measured and monitored the device temperature under load?  

    Also can they monitor it to see if they are not working due to OCP error.  The 4.7uH inductor on the output may be too small - it may be saturating and causing an OCP error.  Recommend trying a larger value like 10uH or 22uH.  

    Is this a new design?  If so, I recommend looking into the DRV8962 per this recent FAQ - https://e2e.ti.com/support/motor-drivers-group/motor-drivers/f/motor-drivers-forum/1256603/faq-what-are-the-best-alternate-devices-for-drv8432 

    Regards,

    Jacob

  • 0 in reply to Jacob Thompson
    Mastermind 6520 points

    Dear Jacob,

    Thank you for your support.

    Customer Don't think that it is not a thermal problem because when they operate with PWM duty 80%, it is working well during 2 or 3 seconds, after it does not work with FAULT_N pin(P18) turns to low. They could not feel thermal problem from failed IC. (on the other hand they feel very hot from the IC which is working well.

    Could you tell me your opinion about it?

    And here are more test results from customer ;

    1. Customer uses maximum 3A output current buck converter for 12V GVDD input. all 6 DRV8412 are connected to 1 buck converter for 12V GVDD.

    And 24V PVDD pins are connected to 24V output, 1000W SMPS module.

    -> Could it be a problem?

    2. They also tested PWM frequency from 100KHz to 500KHz. but the results are same.

    3. Once one DRV8412 have failed, then they operate only one DRV8412 (other 5 DRV8412s are disconnecting all peltiers which is connected to PWM_A and PWM_B) the result is same. that only one DRV8412 does not work.

    4. they tested all 10 PCBs. all PCBs have a problem, but specific location of failed DRV8412 is different.

    5. All 6 DRV8412s have almost same layout. And they use 2oz copper pour for PCB. 

    I asked waveform about failed moment of DRV8412, with PWM output signal and FAULT_N pin. I will share it with you asap.

    And here is my last question of today.

     - What is the meaning of FAULT_N pin turns to low? Which signal should I check from the schematic?

  • +1 in reply to Chase Jeong
    TI__Mastermind 23271 points

    Hey Chase,

    Chase Jeong said:
     - What is the meaning of FAULT_N pin turns to low? Which signal should I check from the schematic?

    The nFAULT pin going low indicates some kind of a fault.  This could be caused by many of the protections - undervoltage, power-on reset, overtemperature, overcurrent, etc. So we need to get scope captures or other data to help narrow down which fault is causing it.  

    Chase Jeong said:
    when they operate with PWM duty 80%, it is working well during 2 or 3 seconds, after it does not work with FAULT_N pin(P18) turns to low. They could not feel thermal problem from failed IC. (on the other hand they feel very hot from the IC which is working well.

    Make sure the thermal pad is sufficiently soldered down.  I have seen similar issues coming from the thermal pad not having enough solder under it, and when the solder is melted it can ball up under one side of the device leaving the other side of the thermal pad unconnected. This can cause uneven heating in the device and false OCP trips or just rapid device heating leading to thermal trip.  

    Chase Jeong said:

    1. Customer uses maximum 3A output current buck converter for 12V GVDD input. all 6 DRV8412 are connected to 1 buck converter for 12V GVDD.

    And 24V PVDD pins are connected to 24V output, 1000W SMPS module.

    -> Could it be a problem?

    A scope capture would help to make sure these power supplies are stable when running, but I don't see any issue with this.  Make sure GVDD does not dip below 9.8V.  

    Can you monitor the OTW (Overtemperature Warning) pin as well? If this does not trip then we can maybe rule out thermal shutdown.

    Regards,

    Jacob

  • 0 in reply to Jacob Thompson
    Mastermind 6520 points

    Dear Jacob,

    Thank you for your support.

    Now I am discussing this issue with my customer. I will ask you more right after this discussion is done.

    One more thing, They did not prepare waveform but when they see the fault, FAULT goes to low but OTW remains high.

    Could you explain what it means?

    Best regards,

    Chase

  • +1 in reply to Chase Jeong
    TI__Mastermind 23271 points

    Hey Chase,

    Chase Jeong said:
    when they see the fault, FAULT goes to low but OTW remains high.

    Per Table 1. Protection Mode Signal Descriptions this indicates an Overcurrent Shutdown (likely) or a GVDD undervoltage protection ocurred (less likely)

    Best,

    Jacob

  • 0 in reply to Jacob Thompson
    Mastermind 6520 points

    Dear Jacob,

    Thank you for your support.

    My customer found that thermal pad was not soldered properly. After they rework the IC and enough solder is attached, the problem is disappeared.

    Thank you for your support again.

    Best regards,

    Chase

  • 0 in reply to Chase Jeong
    TI__Mastermind 23271 points

    Great, thank you for the update! It is always valuable to hear what fixed it. Glad to help!

    Best,

    Jacob