• State TI Thinks Resolved
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 81 views
  • Users 0 members are here
Support feedback
Options
Options
Related

DRV8231A: How to solve current test error too high for DRV8231A

Peng Zhang
Prodigy 140 points
Part Number: DRV8231A
Other Parts Discussed in Thread: INA241A, DRV8876, DRV8231

Tool/software:

Hi Team,

My customer now wants to use DRV8231A as motor driver. And they have accuracy requirement as Iout < 0.05A; error < 5mA. and Iout< 0.15A error as 10mA.

The DRV8231A absolutely cannot achieve such accuracy. Could you give us some suggestion in how to improve current test accuracy to meet customers needed.

I wonder if we could suggest customers to test current with I-V converter. But that will lead low efficacy. Do you have any better suggestions?  

And I want to figure out in what stage should we suggest customer use internal FET. Is there a power or torque threshold in distinguishing whether to use the internal FET or the External FET solution.  

BR,

Peng

  • 0
    TI__Mastermind 19860 points

    Hi Peng,

    Please allow me time to review. I'll reach out by tomorrow with an update.

    Best,

    David

  • 0
    TI__Mastermind 19860 points

    Hi Peng,

    Yes, it seems like the customer will definitely need to use an external solution if the current sense accuracy is that high. 

    Peng Zhang said:
    And I want to figure out in what stage should we suggest customer use internal FET. Is there a power or torque threshold in distinguishing whether to use the internal FET or the External FET solution.

    I'm not sure how this question relates to current sense accuracy. Integrated vs external FET devices depend on current capability and heat dissipation. For high current and best thermal performance a gate driver is recommended. 

    For best implementation of external current sense solutions, I'll point you to the current sensing team to provide some possible solutions. 

    Best,

    David

  • 0 in reply to David Medis
    TI__Mastermind 32310 points

    Hi Peng,

    If they are using an external I recommend the INA241 due to the PWM.  I would also verify that the switching is under 125kHz.  Otherwise use the INA296 and live with the transient at the output due to the VCM changes.

    Below are some information about the enhanced PWM rejection of the INA241A which is in the datasheet.

  • 0 in reply to David Medis
    TI__Prodigy 140 points

    Hi David,

    I have check with Roborock they don't want to use external solution due to the waste of Layout area. Do we have some method that can calibration error, Including HW and SW method? It is important because our customer have test TMI8129 and TMI8180G have better performance of current accuracy. 

    And I have check DRV8876 have better performance in current sensing, I want to figure what cause of the difference of DRV8876 and DRV8231. Is it because the manufacturing process or the internal structure that is different?

    BR,

    Peng

  • 0 in reply to Peng Zhang
    TI__Mastermind 19860 points

    Hi Peng,

    I understand that using an external solution isn't ideal but all current sense paramters are integrated into device. The DRV8231 is a hardware (GPIO) device so no SW changes can improve current accuracy.

    What level of accuracy are they looking for?

    The DRV8876 has a better current sense accuracy because it has a smaller current mirror scaling facotor (AIPROPI). A larger scaling factor amplifies not only the desired signal but also any errors or noise present in the input current. This can result in a larger absolute error in the output current, which can degrade the accuracy of the current sense circuit.

    Best,

    David

  • 0 in reply to David Medis
    TI__Prodigy 140 points

    Hi David,

    Thanks for your kindly reply. Roborock need 5mA accuracy when Iout = 0.05A; 10mA accuracy at Iout = 0.15A.

    I just wonder could we set zero by SW just like AMP-ADC signal chain to do.

    BR,

    Peng

  • 0 in reply to Peng Zhang
    TI__Mastermind 19860 points

    Hi Peng,

    It is recommended that the customer test the device accuracy under their specific operating conditions. The datasheet specs are across voltage and temperature so it's possible to see better accuracy under certain operating conditions. 

    Yes, it would be possible to zero by SW externally if the customer is willing. 

    Best,

    David