RoboRIO version 2 not connecting to Driver Station, Phoenix Tuner v1 or v2

[mshafer]

> Is there a way, a procedure, process that we can do to thoroughly test this specific roboRIO

Only option I know of is to run the robot through it's courses (lots of driver practice).

But I reached out to some internal contacts to ask if there's anything else I can pass along.

[mshafer]

<whoops, double post>

[mshafer]

One other question -> are you seeing this issue establishing communication with just this roboRIO and laptop, and you are able to communicate correctly with your older roboRIO on that laptop just fine?

[Mr.G-Team20]

No. Regardless of laptop or PC this roboRIO 2 behaves erratically. At school students each use their own laptops or "driver station" laptop. Right now I am at home, using a PC running Windows 10. We probably tested 7 different laptops, with various configurations (CPU & RAM). Made sure they all had latest Java and Microsoft updates.

In the past days, I uninstalled everything from my PC - VSCode, Java, NI software (and all the tools) cleaned up everything, TEMP folders, windows cache and storage sense. Re-downloaded everything and re-installed everything. Today I left the roboRIO powered the whole day connected to a lab power supply, voltage set at 12.000, current limited to 4A. Attached on the CANbus is a Talon SRX and a small 775 motor.

The max current drawn by the 775 is 3.5A.

Through-out the day I was enabling, moving the motor forward/backward for couple minutes.

After the 4 or 5 time enabling the robot code, the "Robot Code" indicator turned red. No matter what I did, restarted code, rebooted the roboRIO, deployed the code again I couldn't get it to turn that thing green.

The only way I was able to was by re-imaging the SD card (used another new one), setting the team number then deploy the code. 

I cannot in good faith tell the students go ahead and use this roboRIO on the competition robot. 

[mshafer]

I heard back from my contact.

"If the robot code goes red and never comes back until the image is reinstalled, it makes me think it's a quite bad system crash. "

And you stated that simply rebooting the roboRIO was not enough (which should restart all system processes and the deployed startup app)...

Two things we can try:
* Open a Java example template, deploy it to the roboRIO without any modifications, let it run for a day and see if it ever disconnects (it should not).

  If it never disconnects, then try just adding motors and test again.

* There's also a new release of CTRE from 1/13/23 (v5.30.3 which fixed caniv calls on windows) 

[Mr.G-Team20]

I did leave it over the week-end continuously powered. It did not disconnect at all. I will push an empty Java example template and let it sit, see what happens.

Downloaded the new firmware from CTRE and installed on the controllers. 

Will run the code with no motors and see what happens.

Thank you Matt. 

[Mr.G-Team20]

I did leave it running with an empty robot code, no motors, connected to driver station, since yesterday. Did not disconnect, did not lose connection at all , robot code , joysticks.

What should I do next? 

[mshafer · Accepted solution]

Try slowly adding in the rest of the team code (maybe start with just declaring motors) and repeat the test.

Either, the phoenix tuner update resolved the issue, or (by slowly building up the team code) we'll be able to identify what call is causing the crash (and then hopefully diagnose where the issue lies and hopefully find a workaround and/or report to owning group).