We are facing some problems here with two Tegra X1 modules when loading the images using Jetpack 2.3 and the Jetson Development Board. I will describe the issues for the TegraX1-A and the TegraX1-B.
TegraX1-A: This one was working fine, but suddenly it started acting as it was volatile. We can put it in “Recovery Mode”, load the image, and test it without problems. We can also reboot it by software, but when we poweroff the board or the power source is down and later we try to power up again, the board doesn’t boot and it seems like all the data has gone. We have to load the image again in order to use it.
TegraX1-B: This one is on a worst state. It was also working correctly, but suddenly we were not able to load any image on it anymore. When we power on the module and put it in “Recovery Mode” we noticed that doing “lsusb” on the host computer, the NVIDIA device doesn’t appear, and this causes when loading the image, an “USB communication failed.Check if device is in recovery” error appears and the image is not loaded.
We bought those modules with others and only those 2 have failed.
Does anyone have experienced a similar behavior? Or may be someone know or have an idea about what is going on with those Tegras?
Also, does anyone know how to returned this Tegra modules to get checked and repaired?
I don’t think a Jetson needs a powered HUB for recovery mode, it has its own power, so USB power for this one device should not be an issue. There may still be something related to a connector being near failure (micro connectors are rather sensitive to torque breaking traces), or something about USB on the host which is marginal (not that it is the fault of the host, it’s just that everything in USB chain, including a HUB and cables, adds up and it all must work at the same time). If the host is a VM this is always suspect.
RMA procedure is listed near the top of this:
Well, the problem doesn’t seem to be power supply but the module itself, we have been working with a batch of boards for months and now just these two are giving problems.
We are trying to contact NVIDIA about this problem, maybe there was a batch of their modules showing these symptoms? We will try the RMA procedure but it would be nice to hear if NVIDIA engineers have seen the problem.
Hello CarlosR92 and DavidSoto, I have checked with various engineers who work on the TX1 (as well as my own recollection from this forum) and have not heard of either issue occurring on the TX1 modules previously.
I would concur with linuxdev’s advice to begin the RMA process on these modules.
Thank you both for your comments, we just started the RMA process. Thanks again.