MCT8316A: When the positive and negative direction control motor of the driver is frequently used, the driver chip does not work, which can only be solved by power-on and power-off

We found that the motor drive chip was stuck in the IPD, and the MCU reset could not solve the problem, it would remain stuck there until the driver chip was powered back on and off

It stays in this state, so the motor is not running, can you tell me why

Meanwhile did you see any latched fault ? 

In that state is it possible to read Gate driver and Controller fault and share?

 read the latch fault, but the data in the addresses of the two registers is 0, indicating that there is no fault position 1, and only the 0xeah register is displayed in the MOTOR_IPD state. At this time, even if the microcontroller is reset, the fault cannot be solved, and the solution can only be solved by powering on and off the driver chip. However, the IIC communication with the driver chip is normal at this time. Just can't drive the motor

Looks like the driver maybe going into an unknown state here in this scenario, maybe a configuration issue.

Please give me time time till 22 Jan

I saw in another post that MCF8316C1VQRGFRQ1 has a firmware update based on MCT8316A1VQRGFRQ1, but the specific update point of MCF8316C1VQRGFRQ1 is not very clear. Can you expand it in detail? What content has been updated and what problems have been solved?
After I checked the pin definitions of the two chips, is it feasible to directly use MCF8316C1VQRGFRQ1 to replace MCT8316A1VQRGFRQ1? Do you have any suggestions to help us quickly adapt MCF8316C1VQRGFRQ1

The above post path is as follows:https://e2e.ti.com/support/motor-drivers-group/motor-drivers/f/motor-drivers-forum/1316509/mcf8316a-mcf8316a-read-i2c-while-motor-is-spinning---causes-motor-stop

We tried a few lab experiments from our side, but could not replicate this issue. We are still actively working to understand this. 

Regarding below point, the e2e mainly talks about the firmware update within the device. The erratas in MCF8316A (not MCT8316A) were resolved in MCF8316C through a firmware change in the device.

I saw in another post that MCF8316C1VQRGFRQ1 has a firmware update based on MCT8316A1VQRGFRQ1

Yes, it is possible to replace the MCT8316A with the MCF8316C, are you using TI's EVM? Then you might need a firmware change for the MCU on EVM.

You can refer to this e2e FAQ on how to update the firmware if you are using the MCT8316AEVM.

So are you looking at MCF8316C instead of the MCT8316A?

Hi Vivian，Thanks for your reply。

Based on mct8136-q1 chip, we have obtained some new information during the recent test, and we hereby add that we hope you can help us to see if there are still relevant ideas for debugging

The following are IPD alignments
When the motor starts normally, the phase current waveform is all normal (basically consistent with the data manual)
When the motor stops abnormally and is stuck in the MOTOR_IPD state, there is only one spike to the current waveform in the corresponding IPD alignment stage, and the motor cannot run. The nFAULT pin status is normal and no error is reported

When the motor is at room temperature, the probability of the occurrence of the anomaly is less than one in 1000; When the motor is in an environment of -20 ° C, the anomaly will certainly appear within 100 times;

I would like to know how the ambient temperature affects the motor, which in turn causes the drive chip to get stuck in the MOTOR_IPD state

The above picture is the test waveform of the motor when it is normal
speed represents PWM signal output by MCU to MCT8136A-Q1 to control motor rotation and speed;
FG is the output speed signal fed back to MCU by MCT8136A-Q1;
I is the phase current waveform started by one of the three phases U, V and W;
HOR is the motor feedbacks the actual speed to the MCU

The above waveform is the phase current waveform of one phase when the motor falls into the abnormal state (stuck in the MOTOR_IPD state).

Thanks a lot for the waveforms. I see that you are using STM microcontroller for communication. Is it possible for the customer to use the MSP on our EVM to do I2C transactions? Why I am asking this is we have a few debug registers which are accessible only through the GUI. We want to write a few values to the debug registers which might fix this issue. Please let me know if this is possible.

Excuse me  Vivian

We have solved the problem of the above alignment phase, but currently it will occasionally get stuck in the MOTOR_RUN state
I would like to know whether the MCT8316A-Q1 chip has a forced reset function. For example, I am currently stuck in the MOTOR_RUN state (sending PWM wave again, the chip no longer outputs the drive current), and whether I have a way to adjust the chip state to reset