0.00 Volts - Code Not Downloading - Randomly Works?

[R.C.]

Hey CD,

Our 4 year programmer left and our cRio broke when some freshman yanked on the ethernet cord. So we got our cRio back and we were getting ready for some offseason competitions and trying to get a new programmer trained. In the process we added a motor and were trying to download some code.

The robot has been working fine and has always showed the voltage. The initial problem was setting up Windriver to download to the cRIO and we solved that by -> adding new connection -> VxWorks. But today we experienced a whole new problem. The Volts showed 0.00 and new robot code was not being updated. We turned off the bot and rebooted everything a few times.

Then by the end of the day, about an hour before we left the shop. I wanted to sneak the bot up to a freshman and surprise him. ALL OF A SUDDEN, the Voltage went back to the proper 12.67 and new code was downloaded.

I am not a programmer and this has been bugging me and I was wondering if anyone could explain what happened. The robot was not touched by any other member nor tampered with. Any Ideas?

Thanks,

-RC

[Roger]

I can give you a preliminary idea, but I don't have the details or the names of the parts...

The battery "reporting gizmo" is plugged into the top of one of the blocks plugged into the cRio. If it gets loose the battery reports zero. Like all delicate electronics if a connection gets loose all sorts of things could happen, or not. You must have jiggled something when you drove the robot.

Why that has anything to do with the downloading code part I don't know. Maybe it is upstream on the power or data wires that jiggled something else? Come to think of it, without it does it report zero or nothing at all? Either way, jiggling the robot must have re-seated whatever wasn't properly set.

I remember last year when the cRio stuff was sent out pre-Kickoff there was questions about no battery reporting, because this one part wasn't included. It finally came with the KOP at Kickoff.

[Jared Russell]

There is a possibility that you were accidentally downloading your old code to the robot.

When you go to FIRST->Download... WindRiver does not necessarily download the binary for the current project that you are editing. Rather, it downloads a file specified in the Preferences->FIRST menu. If you were to copy your project and make changes to the copy (which would be a reasonable thing to do when altering someone else's code), you need to remember to specify the new output file.

[Jon236]

I agree with Roger that the 0.00 volts reading was because of a loose connection. We had issues with the analog breakout and the power connection wiggling loose.

It seems hot glue will always have a role in FRC competitions!

[Russ Beavis]

Don't forget that the Analog Breakout adapter can be screwed into the NI 9201 analog input module. This obviously won't help poorly fitting jumpers for the battery monitor but I've seen many teams that don't securely attach the Analog Breakout to the cRIO module.

And, of course, make sure to use a jumper that fits snugly on the 3-pin battery monitor header.

Russ

[R.C.]

Thanks for all the Replies, I was stumped thinking about it and most of the replies makes sense, I'll see what happens today.

Thanks,

-RC

[Roger]

I'm not sure a screwed-down box would have done us any good -- our summer robot had a good hour of driving by elementary school students, getting under lunchroom tables and banging everywhere. Without much protection, it did survive but the box was bent by the wires. Probably if it was screwed in it would have snapped off. Luckily some careful straightening of the pins got it working again.

Jared, I'll have to keep that in mind for future reference, even though we're using LabVIEW. Even when copying to a new folder I wonder if it is working on the new directory or still pointing to the old directory. I think ye old MP-Lab had that problem feature.

[Jon236]

Even if you screw in the analog board, the power connector from the PDB can come loose....that was our problem.