Jetson Module - INA Voltage Rail / PMIC_CONFIG Failure

We have two 900-13767-0000-001 modules that are experiencing INA voltage rail failures. The part numbers are exactly P3767, 180-13767-DAAA-A13. I assume they are different batches or there may be other identifying marks that would help, let me know. I’ve attached some pics of the serial output from each module.

We have a custom carrier board and at some point, during operation, the application product experienced some user interaction that led to the Jetson module being damaged (I assume, because hardware doesn’t typically randomly die when the environment remains stable). I am unable to trace back the user error that may have caused this, but the carrier custom board CAN bus transceiver is not functional on the first, and the second is not enumerating on the USB bus.

The exact error is PMIC_CONFIG reason 0x1C and “Can’t get handle to i2c device @4”. Both seem to show similar issues.

We’re wondering what voltage rail failed because two forum posts I found end at this finding with the moderator stating the hardware is damaged.

I want to help my team avoid damaging another unit. I would be surprised if this was caused by ESD, but perhaps our carrier board lacks the proper TVS protection. This is definitely a hardware failure on the carrier or module likely caused by user grounding issues or perhaps a surge event, but I’m wondering if anyone has insight into what causes this PMIC rail to die on the Jetson.

Has NVIDIA seen this issue on other Orins? Do you have specific recommendations to resolve this? Do you know what voltage rail is monitored by the NVIDIA i2c device 4 that would help me debug my carrier board further?

I basically want to know what I can do in the future to avoid damaging more Orins. If an RMA would be recommended, we can go that route but would like to get to root cause.

Hi,

The error PMIC_CONFIG reason 0x1c and Can’t get handle to i2c device @4 means MB1 firmware cannot communicate with the on module PMIC via the internal PWR_I2C bus.

To reduce the risk of future electrical damage:

  • ESD/surge protection on external facing interfaces
  • GND plane continuity at the module connector
  • Power input (VDD_IN) protection

Regarding the RMA, if the damage appears to have been caused by an external event (surge/grounding issue from the carrier board side), this would not be covered under the standard hardware warranty. The modules would need to be replaced with the new units.

If you would like to discuss further options, you can reach out to your distributor or RMASupport@nvidia.com.

Thank you,

I was mostly just curious which rail specifically, but is this a known issue? Are there other known failures on the PMIC that are caused by ESD/Grounding issues?
VDD_IN is sourced from a nice and stable 5V supply so I’d suspect the first two. Any other thoughts on where to look?