Connection Problems using ethernet cable at events

At our two district events we have experienced driver station connection issues that we have never seen before. I am going to concentrate on the event we attened over the weekend. We started out with our trusty classmate DS and in the pits connecting DS with ethernet cable to radio seemed quite normal even after flashing the radio. When we went to the field the PC would not connect to the field.Classmate had to be rebooted at the field. When we returned from the field the problems in the pits started. In the pits we could not reliably connect to DS with our cable. We had several more difficulties connecting the field and then switched the classmate out for a Microsoft surface assuming the classmate had problems. Our problems on the field seemed to be solved but in the pits weirdness continued.

We observed that when the rio cable was plugged into the radio port closest to the radio power port (as is recommended) the FIRST driver station would connect to the rio but our labview dashboard would not connect. Further our camera feed quit working, updating one frame every 20 seconds or so, if at all. Additionally, our programming PC would not connect to the rio AT ALL. Now if we switched the rio cable to the radio port away from radio power port and plugged our DS (and development PC) into the radio port closest to the power cable, everything worked perfectly every time. Switch the cables back and DS connects but labview dasboard, camera feed and development PC connect consistenly fail.

Note that when connected to the field with rio cable plugged into the radio port closest to power (as we were admonished by the FTA to do) the dashboard worked every match but one. No idea why only one fail but we were watching it carefully after that one fail and there was only one.

We talked to some other teams and one told us that they had found connecting in the pits with ethernet cable so unreliable they use USB exclusively, in a longer term sense than just this year. Again, before this year, we have never had problems like this. We have never had any problems connecting wirelessly in our lab. Extensive testing has produced no smoking gun but we think this is a radio firmware problem.

Update: back in the lab today with radio still in event configuration, everything works as expected. This leads us to suspect the roborio, in terms of some odd interaction with the radio/external PCs.

We had similar problems last weekend (our first event). In the pits, our DS would have difficulty connecting in some cases. Sometimes the DS would connect but the SmartDashboard would not. We have all static IPs. Our programming laptop had a hard time connecting - although that seemed to be a combination of things possibly (poor USB dongle connection). Ultimately changed to deploying code via USB.
It seems like we had better luck connecting directly to the Rio but I can’t say for certain. We just kept trying things till it worked and you were never quite sure what resolved it.
We also had it a few times on the field where the SmartDashboard wouldn’t connect without a lot of effort.
I know we had the Windows Firewall turned off in Control Panel, but I don’t recall checking if the DS status agreed with that. The connection always ended up working after a bunch of connecting/reconnecting, closing/reopening, rebooting, etc, so it wasn’t a solid issue - just really annoying when you’re in a hurry.

When having issues like this I too would recommend you try the USB cable. To get things working in the pit there are a lot of network settings that come into play. Are you using static ip address or dynamic? What about CPU utilization / packet loss?

For drivestation suggestions I would suggest you take a look at this guide

I should have pointed out that this testing was done with our clone robot so the only changed item was the roborio. The radio was one we used at our event but swapped out during trouble shooting (did not help).

We used static IP. Firewall off. Etc etc. Our roborio is located in a way that makes the USB port inaccessible but we are going to fix that and use USB if there are any problems at districts. Solving communication in the pits is important but the problems we have had on the field are disconcerting leaving us wondering each match what will happen. We are going to replace the classmates…though back at our shop both of our units work just fine though one has a sloppy ethernet port.

I would also note that the issues (from last year) regarding ports on the radio for communication are believed to be solved. So I am not sure why one port would be any different. I know as a CSA at GKC / Heartland we were plugging the cable from the RIO into the other port (far from barrel) as we were bypassing sending data over the POE port as some teams have had connection issues going through it.

You might look at replacing your radio and seeing if all of your problems don’t go away.

TLDR. You should be absolutely fine with either port and if one is working better for you IMO use it.

There is definitely a difference between the radio ports in our experience. Another team mentor strongly supported this observation. We have not been able to find the cause. Note we tried two different radios same result. When we went to the field with the rio plugged into the port away from power, and had connection to field problems, the FTA spotted this cabling and was quite adamant that we should have rio next to power. I wanted to try switching ports but the kids were spooked as we had already contributed to match delays and they did not want the FTA to see we had ignored his instructions if we had trouble again…however I like your suggestion.

I should add that we “solved” our pit communication problem by switching the rio to the port away from power for programming and DS testing and then switching it back to go to the field. Our field connection problems would seem to be between the FMS and our DS PC. But we did have at least one instance where we were connected to FMS and could drive the bot but our labview dashboard did not connect. We switched the DS to a surface tablet and that seemed to clear up the field connection issues and we did not observe problems with the dashboard after that…but the problems in the pits continued through the end of the event.

For DS suggestions that guide above is the best. I try to avoid ethernet adapters like the plague as I have had bad experiences. Furthermore as they mention in the guide I know some laptops can be a lot more durable from an Ethernet perspective.

Lastly have a way of quickly strain reliefing the Ethernet cable to your DS. So many of the broken Ethernet ports are from teams tethering on the practice field and for one reason or another the robot pulls on the cord which pulls on the computer. You want the stress to go on the cable when this happens not on your port.