Output signal failing problem fixed!

[technowizard12]

My team's been trying to solve a problem with the output signals switching rapidly on and off many times per second. We tested every component that we thought could possibly be causing this problem, replacing some of it which was faulty. Even after replacing our faulty bridge with a new one (for testing, unless FIRST deigns to let us use it for competition), switching between wired and wireless connections, disabling most every bit of the code going on the robot (to make sure that the cRIO wasn't being overtaxed, and thereby freaking out), checking the priorities of the VIs we were running (to see if any of them were hogging the processor), making sure that the watchdog wasn't malfunctioning, and was being properly fed (assuming proper operation), and checking to make sure that the issue wasn't caused by power supply, our eyes landed on the one thing left which could have gone wrong: the Classmate PC. We think that the problem was (well, still is. it just happens less often) caused by the low amount of RAM on the Classmate PC. The signal to the outputs failing, henceforth known as "flickering", was caused by the Classmate PC having senior moments, not being able to keep up with the data stream, and therefore placing farting in our data. Occasionally we get an error about the watchdog not being fed, which is still caused by the same thing. Attached is a graph of the current going to a single spike's signal wire. You'll see that the current drops to about once every twenty seconds. usually, this only lasts a couple milliseconds, although one burst of three events lasted three seconds.

[Mark McLeod]

On the Driver Station Diagnostics tab error message screen, do you get Watchdog error messages that look like:

Watchdog Expiration: System 4, User 2

See if the count for System corresponds to your plotted spike dropouts. Reboot the cRIO to reset the counts.
Some of these Expiration messages are normal. You'll get them sometimes when you change states from Disabled to Enable or from Autonomous to Teleop or when a practice match comes to an end.

Are you able to replace the Classmate with a more powerful personal PC to test your theory?
The DS Update 1.0 will install everything you need to make any PC become the Driver Station.

[technowizard12]

The watchdog errors only occasionally correspond with the ceasing of the signal. We figured that the watchdog error when we first enable was normal, because it was easily explained. Thanks for the advice on installing the station on another computer, we'll try it.

[Mark McLeod]

We don't get messages for every System Watchdog dropout, so it's the System count that might or might not match up.

The count in the error message isn't always sequential, implying that it can't always communicate back to the DS to tell us about them when they occur.

[technowizard12]

I installed the driver's station on my 64 bit 5 gig ram Windows Vista computer and ran the robot from it while we were graphing. We discovered that the flickering has nothing to do with the computer. There's something going on with the cRIO, or something else electronic, that we can't fix. We hooked up our E-Box to a robot today and ran it. We tested it before this, and the flickering was clearly there, but it didn't effect or make itself evident in the robot's driving. It's a problem without a fix which doesn't effect the robot noticeably.