Problem with losing communication while on the field.

[GLDomination]

We are team 3145 and we won the rookie all-star award after losing in the quaterfinals at the Portland regional.

During our quaterfinal matches, we lost communication twice. I was wondering if there was any way that this could happen due to a physical problem. We were thinking that the camera may have been hit and that the cable may have popped out. Could this cause us to lose comm?(We lost comm for two matches and it happened around halfway through during each).

Any help would be greatly appreciated. We are trying to figure this out before we get to Atlanta, but it is hard to do without a robot. Thanks for any info you can give.

[EricH]

Check your cables, especially between the cRIO and the gaming adapter. That's the most likely culprit. It probably wouldn't be the camera cable (why would not being able to see where the target is shut down communications?).

[TubaMorg]

Your camera coming unplugged COULD cause you lose communication in a way. I cannot speak for LabView, but if you are using Java and your camera code isn't set up correctly, an attempt to read the data from a disconnected camera could cause an exception that might crash the cRio.

[whackedwatchdog]

If you're using the flat (Ribbon?) cable that came in the KOP, that might be your problem. When we were at Peachtree, and during practice, we had a couple of games where we lost communications for seemingly no reason. When we swapped the ribbon cable out to regular, bulky Ethernet cable, we had no problems. I don't know if it was absolute luck, or if there actually is a problem with those cables. Either way, it might be worth a shot!

[rspurlin]

First two places I would look would be the ethernet cable from the cRio to the gaming adapter and the power cable from the power distribution board to the gaming adapter. Loose connections on either of those would be a likely cause.

Do you regain communications? IF so, hoe long does it take? It could be the cRio rebooting after a power fail or other condition.

[Russ Beavis]

Look for intermittent shorts between the robot chassis and the cRIO or camera chassis. Use a multimeter to determine whether your chassis is electrically connected to any of the robot's wiring.

If the cRIO or camera momentarily touch a chassis that is connected to a voltage other than ground, the power supplies will probably reset.

Russ

[Chris_Elston]

Invest in a HOT GLUE gun, and hot glue the Ethernet connectors into the radio and crio. Also hot glue the power connector in the radio. For that matter hot glue EVERYTHING, the PWM connectors into the victors, CAT5 cables into camera, etc. Hot glue all the way around each connector. Pot-it-in like the military people do it.

[Nuttle89]

We have had problems like this at both regionals we have attended so far. We tried tons of troubleshooting methods. From having our wiring looked at by an electrical engineer, to switching from 2CAN to CAN, to fully switching our code over from Jave to C++, we frequently would just lose communication mid-match. We even considered replacing our C-Rio at the regional..

As it turns out, we grabbed a spare radio from the parts desk, had it programmed, and we ran fine for the rest of the day. The issue is, there isn't a way to detect a faulty radio assuming it is getting power, not that I'm aware of.


It was extremely frustrating because we knew it wasn't our code or our hardware. We were assuming it was some error with the FMS. We never fixed the problem at the DC regional, and were scrambling to find the problem with the condensed schedule at Pittsburgh.

[gordonsolo13]

It could have also been a problem with your classmate. If you allow it to go into hibernate after turning it on, it will have problems with connecting. We had problems on the first three matches at our regionals because of such a problem (along with our the fact that our during our first match our mechanics forgot to reconnect the speed controllers.)

[Chris_Elston]

Originally Posted by gordonsolo13

It could have also been a problem with your classmate. If you allow it to go into hibernate after turning it on, it will have problems with connecting. We had problems on the first three matches at our regionals because of such a problem (along with our the fact that our during our first match our mechanics forgot to reconnect the speed controllers.)

Oh, this is a GOOD point. This happened to us once and also when you close the lid on the classmate too. I know...ask my drive team :-) I closed the lid on them during drive practice one day, to ween them away from the camera feedback. I said "You just lost the live feedback from the camera during the match, so keep on driving..." Right when I closed the lid, the whole robot died...guess you had to be there. It was a funny moment.... Moral of the story, don't close your Classmate lid in the middle of a match, instead put a piece of paper over the screen to cover up the video feed...

[TubaMorg]

Originally Posted by Chris_Elston

Oh, this is a GOOD point. This happened to us once and also when you close the lid on the classmate too. I know...ask my drive team :-) I closed the lid on them during drive practice one day, to ween them away from the camera feedback. I said "You just lost the live feedback from the camera during the match, so keep on driving..." Right when I closed the lid, the whole robot died...guess you had to be there. It was a funny moment.... Moral of the story, don't close your Classmate lid in the middle of a match, instead put a piece of paper over the screen to cover up the video feed...

I know this isn't really part of thread, but why is it bad to have video feedback? We worked hard to get ours working well, now I'm wondering if there is a consequence we haven't thought of. Are the drivers staring at it too much?

[Chris_Elston]

Originally Posted by TubaMorg

I know this isn't really part of thread, but why is it bad to have video feedback? We worked hard to get ours working well, now I'm wondering if there is a consequence we haven't thought of. Are the drivers staring at it too much?

We trained them both ways. But not totally relying on the camera only. Definitely if they can look at the robot to get an AIM or if you have "auto aim" in tele-op is the best way to drive but to look at the camera to aim seems like it takes the driver longer to make the decision to kick verses looking at the robot and making the alignment shot.

[Mark McLeod]

Originally Posted by Chris_Elston

Moral of the story, don't close your Classmate lid in the middle of a match, instead put a piece of paper over the screen to cover up the video feed...

It's easier to just minimize or close the Dashboard window.

[topgun]

Also, check the connection between your Classmate and the ethernet cable. Our plastic retention clip on the Classmate (not the ethernet cable) is broken off and any jostling of the cable will cause it to temporarily lose connection (hence, communication to the robot).

- T

[robself705]

We were having this same problem A LOT at the Wisconsin regional. We were about to win the entire regional when our middle zone bot lost communications in the last two matches. (we won the first match of the final round 10-4). From what I saw almost everyone had these problems all day. I am trying to consolidate the issue so we don't have it again.

Before every match we:
log out and log in on the class mate
DO NOT close the lid on the class mate
make sure everyone's class mate wireless is off
check all Ethernet cables are plugged in
check for FMS connection and voltage before we give the judges the ok

Twice during our matches we lost connection in the middle of the match
Our alliance in the final rounds lost communications twice in a row AFTER autonomous mode. The judges told us it was NOT the field or FMS, it was the team's robot. Well then what is it? The bridge? Let me know if any of you have had this issue and what solved it. It sounds like people are replacing their wireless bridges and that solves the problem.


Thanks!