I2C timeout

Hi,

I have a PIC device that interacts with my Jetson Nano using I2C and GPIO. We’re not sure what caused this, but we now only get timeouts on I2C bus 1. If I run i2cdetect -y -r 1, it will take a long time for each address to return which will only give “–”. It seems to be something related to the hardware because I tested all other peripherals on another Jetson Nano and they work fine. I also switched the SD cards and got the same issue for that device.

I saw this post which says that it may have to do with regulator settings. I’m not sure what that is referring to. And we didn’t change any setting. It was running in 10W mode. The problem also still occurred when I disconnected most other peripherals.

Do you mean the PIC device cause I2C timeout on one nano, and no on another nano? If so, are both nano boards same version? Is there any rework on first nano board?

No, I mean that the PIC device works on one nano but not on the other. The nanos should be the same version. There wouldn’t be any rework on either board.

It’s possible that particular I2C bus on the Nano that doesn’t work (or the level shifter on that Nano) has become physically damaged.
Do you have access to another I2C device, like a real-time clock chip, or a thermometer, or something? If so, can you see if that is detected?
Also, do you have a logic analyzer or I2C bus monitor of some sort, that you can use to see what’s going on on the bus?
Even just an oscilloscope would be helpful.

We have a logic analyzer. When trying it out, we saw the i2c line was pulled low and then nothing happened.

Did you try other devices on this I2C? If still only fail on this nano, then it might be a hw issue which needs to run RMA for it.

So I tried a TOF eval board that we had and I can read the appropriate values from the reference registers on a working and non-working Nano. What could possibly be causing an issue when my PIC device can communicate with one Nano but has timeouts on the other? Everything between the devices is the same including the SD card.

How many nano boards did you test? Please share the S/N of them so we can check the module version. If finally the module version is same, then the failed module might need RMA.

I tested the TOF eval board with just the broken and a working Nano. The PIC device works on two Nanos. I only have the broken and one working Nano at the moment. The S/N for the working is 1423019055106. The S/N for the damaged is 1423019135375. Also, just to be clear, the damaged Nano was working previously. It just randomly stopped working recently.

How much is RMA?

Same version board, you can follow this for RMA: https://developer.nvidia.com/embedded/faq#rma-process