LAUNCHXL-F28379D: Abnormal heating up issues

Hi Ibrahim,

What system are these boards connected to? When the C2000 chip heats up, it's usually an indication that the device has been damaged, even if it seems the device is working and executing firmware correctly. Are you ensuring that no voltages exceeding 3.3V are being supplied to any of the device pins on the board?

The board can be fixed by swapping out the C2000 chip with an available working device.

Regards,

Peter

Hi Peter,

We use it for standard inverter control. (PWM to isolated Gate drivers, and Current/voltage measurements to ADC). I also use it to drive a contactor coil via darlington pairs (ULN2003)

Yes, I only use the power from the on board FTDI chip. Unless there is a problem with its 3.3V output, I do not know what can cause this for multiple devices. Although, I use the same program on different devices successfully.

Regards.

Hi Ibrahim,

I don't anticipate that there would be any issues with the power coming from the PC, and it is unlikely to be the PWM signals connected to the gate drivers. Can you review all inputs going to the ADC and ensure there are no voltage spikes? Are you using external op-amp or are your signals being clamped to ensure voltages do not exceed 3.3V?

Regards,

Peter

Hmm, I see. For current sensing we are using the TMCS1133 which takes power from the launchpad. I do not use any clamping or op-amps.

Best,

Ibrahim

Hi Ibrahim,

Are you using the TMCS1133EVM or a custom design PCB board. If using custom PCB, are you following the layout guidance provided in the datasheet? The TMCS1133 is a high bandwidth current-sensor which may lead to noisy output. Does your system require such a high bandwidth or can lower bandwidth part be used? Also, what is the range of your voltage sense inputs?

Regards,

Peter

Hi Peter,

We use both the EVM and a custom PCB following the same layout as the EVM.

Yes, we could not find a lower bandwidth EVM to test with so we chose what is available.

Thanks,

Ibrahim

Hi Ibrahim,

Since this is custom hardware on top of our boards, it is difficult to know the exact root cause of the damage. In these situations where there the board is heating up and the device is likely damaged, the source of the damage is often through the ADC where a signal exceeding the device voltage is received through the ADC pin. Please review the hardware and add clamping as necessary. 

Regards,

Peter

Thank you for your insight, Peter. I still do not understand how the TI current sensor can output an overvoltage signal even though it is powered by the same 3.3V of the launchpad. I will look further into the issue and try to find a way to clamp the signal somehow.

Best,

Ibrahim