DRV8231A: Why current was tested when no motor was connect to the OUT1 and OUT2

Hi Fengyu,

I will be working with you to answer your questions.

Is the customer using their own board one or the evaluation board from TI?

Can you provide driver part of customer's schematic. I specifically would like to see the IPROPI connection and to what other nets it connects to.

Is this leakage measured in all boards tested? 

Regards,

Pablo Armet 

Applications Engineer

Hi Pablo,

Thanks for your reply.

Customer test DRV8231A using their own board.

As shown in the screenshot, PIN 1(They marked VISEN) is IPROPI. LEFT_WHEEL_1 will be send to ADC of MCU.

As customers' feedback, not all the board had the leakage. The max current they had tested was about 150mA.

Best regards,

Fengyu 

Hi Fengyu,

Thank you for the schematic and answers.

I will need more time to look into this. I will reply within 24 hours. Thanks in advance for your patience.

Regards,

Pablo Armet

Hi Pablo,

May I get your feedback for the question?

Best regards,

Fengyu

Hi Fengyu,

There is nothing obvious in the schematic that would lead me to the root cause of the leakage. 

Can you do an ABA swap? Take one of the IC with the leakage and solder to a board that shows no leakage? Also, solder a good IC to a board with leakage and observe if the leakage follows the IC or the board.

Regards,

Pablo Armet

Hi Pablo,

Thanks for your reply. We'll test the IC according to  the method you mentioned.

Regards,

Fengyu 

Fengyu,

Looking forward to your results.

Regards,

pablo

Hi Pablo,

My customer tested DRV8231A the way you said. The leakage current is always along with device, rather than the board.

Also, I want to correct the magnitude of current. The max current they had tested was about 15mA.

May I ask how we could solve the problem? Thanks.

Regards,

Fengyu

Hi Fengyu,

Fengyu Leng said:

My customer tested DRV8231A the way you said. The leakage current is always along with device, rather than the board

It sounds like a defective driver chip.

Brian

Fengyu Leng said:

Also, I want to correct the magnitude of current. The max current they had tested was about 15mA.

15mA leakage at IPROP is too much. Consider the mirror current sense  has the scale factor of 1.5mA / A, this means with 15mA at IPROP then a huge current of 15mA * A / 1.5mA = 10A  flows through the FETs with no load connected !!!!!

Brian

Hi Brian,

Thanks for your reply.

15mA is not the current at IPROP. It is the current through FETS which our customer calculated. That means the current at IPROP is 15mA*1.5mA/A=0.0225mA. The current is small but is that normal for our product?

Regards,

Fengyu

Hi Febgyu,

Thanks for your clarification. I think the IC may have been damaged somehow. Can you ask the customer to measure the resistance of the IPROPI pin to the device GND while the IC is not powered? And compare this measurement to an IC that is working fine.

Regards,

Pablo Armet

Hi Fengyu,

Fengyu Leng said:

15mA is not the current at IPROP. It is the current through FETS which our customer calculated. That means the current at IPROP is 15mA*1.5mA/A=0.0225mA. The current is small but is that normal for our product?

I never had guessed that you're talking about the FET current as it was not mentioned anywhere in the post. Thanks for clarification. 

So for 15mA flows through the FETs with 30% PWM applied at INx inputs and no motor connected, the FETs are switching and there should be some switching current loss due to FET intrinsic capacitance -- the higher the pwm frequency the more switching loss. 

1. What is the FET current when pwm = 100%  and 0% (DC level and no switching at the FETs)?

2. What is the FET current loss when pwm frequency is 10Khz and 50Khz?

This test result can help to confirm the small FET current is due to switching loss which maybe normal.

Brian

Hi Brian & Pablo,

Brian Dang said:

This test result can help to confirm the small FET current is due to switching loss which maybe normal.

Thanks for being so quick to turn around. May I kindly ask if  the phenomenon is normal, why only individual chip has the problem?  Shouldn't  these FETs with large intrinsic capacitance be eliminated?

I have asked for the chip with leakage current from our customer. I think I could test it follow the instructions you said after our customer deliver it to me. Also, if your have other idea to find the answer, please feel free to let me know. We all really want to know why this problem happened and how to solve it. 

Regards,

Fengyu

Fengyu,

Many things could have occurred, it could be an ESD damage that caused internal leakage path on IPROPI. We won't know for sure until we do some sort of failure analysis on the part.

Let me know once you have run your tests. 

Regards,

Pablo Armet

All right. I will test once I got it. Thanks!

Regards,

Fengyu

Hi team,

I am back! I tested DRV8231A without load on DRV8251AEVM and found a strange phenomenon which need you to help analyze.  As the waveform below, there is a voltage(V1,about 20mV) after the IN2 signal‘s rise edge(The time between the rising edge of IN2 and V1 is T1).

4Green: IN1   1Yellow: IN2   3Pink: Voltage on pin IPROPI  2Blue: OUT2

Different chip has different TI and V1. You can refer to another chip's test result as below.

So could you please help explain the reason why T1 and VI generate? Why different chips have different V1 and TI?  How large is the magnitude of V1 is normal?  Look forward to your reply! Thanks!

Fengyu,

Regards

Hi Fengyu,

So the FETs have about 12mA leakage current during the output turned high. I don't know why T1 delay varies from chip to chip.

Do you really need DM1 and DM2 diode? Remove them and test again without load and see if anything change.

Brian

Hi Fengyu,

I agree with Brian. Let's remove the two diodes and see if issue goes away. 

I can't really think of an explanation. I will have to keep digging into this to understand what is happening.

Best Regards,

Pablo 

Hi Fengyu,

Any updates from your side?