We discovered when testing the robot that any large deceleration (impact with another robot) can cause the radio to reboot. We are pretty sure it is the radio and not the RoboRio as the wifi connection drops. The radio does not have any other tissues and is mounted semi-rigidly. Any ideas as to prevent this from happening?
There’s this page on the WPILib documentation that might help you.
3 Likes
Are you using POE for the radio or just the barrel jack? We had a number of issues in the past with the barrel jack power supply coming loose and causing power interruptions that lead to a reboot.
Do you know what year your radio is from. There were some issues where a slight deflection of the top of the radio case resulted in the shielding inside the radio to short out some of the internal wiring. There is a fix out there for that issue, but the newer radios would have already incorporated the fix.
Those are the two most common causes from my prior experience. I will see if I can find some more details on the internal shorting issue and will add them to this post when I find them.
We are using POE as we also experienced unreliability with the jack. The radio was given to us in our rookie season (2019) I believe, but I’m not sure of the exact make.
It certainly sounds like the issue that @MiguelAngel posted (the same issue that I was talking about). We had this issue with our radios from that timeframe. We could repeatedly produce the reset by squeezing the radio housing slightly. If you are able to reproduce the reset with a moderate squeeze of the case (pushing on the top and bottom with your fingers), then this is probably your issue.
Edit: If you are at a competition right now and having this problem during your matches, you may want to see if there are any radios that the event can loan you. Go to Pit Admin and ask. If they don’t have one, they can make an announcement and it is likely that another team will have a spare radio for you to borrow. Then you can troubleshoot this after the event.
The other thing you want to probably inspect carefully are the leads on your POE wires where they plug into the VRM. You will want to wiggle those wires and make sure they are not loose in the sockets. You will also want to verify that there are no strands of the wire that are shorting out from red to black. With stranded wires, inserting them into the VRM fittings can cause the strands to fray and they can easily be sticking out and can short with the other wire.
3 Likes
We will attept the fix MinguelAngel linked, alongside double checking the power wires. Competition is in three days so we have a little time.
In regards to the fix, does it make the radio illegal for competition as it is modification of FIRST specified electronic components?
I believe we’ve also lost communication a few times while driving it, and we can generally get it back up and running if we just wiggle our ethernet cable a few times and make sure everything is plugged in properly.
The radio will still be legal. That fix was disseminated by FIRST.
1 Like
See R66-O
1 Like
Here’s a link to a description of how to confirm you are seeing a radio reboot. In general, this is caused by wiring issues (including the barrel jack or PoE) much more often than it is by the radio being shorted out. I’d advise belt-and-suspenders with the radio – power it both ways and apply the modification. But, only attempt the modification after you have looked it over and are confident you have someone who can successfully do this.
Logs for a radio reboot: Driver Station Log File Viewer — FIRST Robotics Competition documentation.
1 Like
About this, a practice that might not be very attractive but that might work as well, it’s to put hot glue on your radio ports to ensure that there’s not being any false contact on both the PoE and the barrel jack.
1 Like
A common failure point that I haven’t seen mentioned is the little automotive fuse that goes to the VRM at the end of the PDB (in addition to the wiring on the VRM itself which was mentioned). Those can often be loose and pop out on impacts*. I would also check your driver station logs to see if the RoboRio is seeing significant voltage drop, as that could be indicative of an issue with the main wiring (battery, main breaker, or PDB) even if the RoboRio isn’t fully rebooting.
*Edit: To clarify, it won’t look like it’s popped out because it’s just a momentary loss of connection. My general rule of thumb with these at events is that you should forcing them in fairly hard with your thumb and a decent amount of weight behind it. Or if you pull it out of the board with a trivial amount of force (I’ve removed one just by lighting grabbing it) it’s not in far enough.
3 Likes
Neither the RJ-45 connector used for the Ethernet/POE nor the barrel jack connector were designed for applications with exposure to high levels of vibration and/or shock so it is possible for those connectors to go open-circuit briefly when subjected to a physical shock.
A common and very effective solution is to use BOTH the POE and barrel jack connectors. They are connected together internally inside the radio so you now have redundancy. Since the two connectors are very different in design, It is not likely that both the connectors go open-circuit at exactly the same time. The wires for both connectors can be plugged into the terminals in the VRM indicated in R47.
The modification where one adds tape inside the radio is for a totally different issue and will not solve the problem if the power connector is going open-circuit.
4 Likes
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.