Is it only us or do other teams see random controller disconnects with V5 too?
We fight quite frequent disconnects, both in the lab as well as on the tournaments. It came to the point that the drivers are taking a screw with them to the field so they could press the reset button on the controller in case the disconnect happens mid-match.
Symptoms: Robot stops taking input and won’t reconnect on its own. If you reset the controller, it would reconnect in about 15s, so you can keep driving for the rest of the match. We have had disconnect like this in 2 qualification matches during the last tournament and one during the skills run, so you can imagine the impact on an otherwise very experienced and highly scoring MS team…
Are there any diagnostic tools or techniques that would allow us to eliminate or at least better investigate the issue?
I recall that for VexNET2, there apparently was a way for field technicians to observe the radio channels and gather some telemetry to judge and explain dropouts. Is there anything like that available for V5 radio?
We don’t see disconnects, we see full on crashes. Timer on the brain stops, cannot turn off the brain without unplugging the battery, controller can’t connect.
For us, both the brain and the controller keep working, just the radio LED turns solid red, the controller switches to red backlight/connection lost message.
The two wouldn’t reconnect, but a reset of the controller (by the hidden button) will get them reconnect quickly enough to be practical in during the match.
It has happened to the team again during the state championship, in at least 3 matches.
That sucks. We only had that once, but it was blamed on the field by the refs so we got a rerun. (That field was used for the quarterfinals, and it happened again, so the 8th ranked alliance after loosing the first match actually got to worlds since 1st rank tipped over during auton.)
@nenik and anyone else who has experienced disconnects, I’m looking for more feedback to see if we have a reproducible pattern. If you could provide a few details.
Did disconnect happen during autonomous or driver control, did it happen during the transition from disabled to enabled or just randomly at any time during the match. ?
What was the robot programmed with, VCS, PROS etc. ?
Were you displaying any text on the controller LCD ?
Did the controller screen go red during the disconnect ? was the radio searching, were the controller controls still responsive to changing screens (ie. could you try and run the program again).
Was the controller battery fully charged ?
What color was the radio led on the v5 brain when disconnected, red, green, was it flashing ?
@jpearman We’ve had the same thing happen. Ours disconnected usually around 1 minute after enabling, but it sometimes happened during auton. We used PROS. No text on the controller LCD, but there was some rumbling. As far as I remember, the controller sometimes went red, or sometimes just totally froze. It was not searching and did not reconnect. Battery was fully charged.
After one of our competitions where this happened, I tried diffing it with our other program that worked perfectly fine. It finally got down to the point where the only difference was the program name. If I had a space in it, it would disconnect. If I didn’t, it was fine. Weirdly our other team had a space and they had no issues.
We tried swapping out the controller, got a new antenna, new cables, and nothing helped. It was really weird, but for some reason the space thing helps, as crazy as it sounds.
Back in an earlier competition we had a TON of disconnect issues. At that time:
VCS programming
Name of programs were on controller LCD but no other text I think
Happened during driving control at random points in the match (anything from 1 - 5+ times for anywhere from 3 - 10 seconds at a time before reconnecting)
They didn’t do anything to try and run the program again, it would just randomly come back. Don’t remember a red controller screen
Controller battery was charged (maybe not fully?)
Not sure about radio color
During the competition they did change out the cable to the radio (towards the end) and it seemed to improve somewhat (although still did it a bit). After the competition, someone official looked over the robot and said that most everything seemed fine but recommended we try putting the cable in a different port to avoid any bending at all between it and the radio, and also maybe a shock absorber on the radio (we never tested these things - moved away from V5 instead). He also was able to see the disconnects on the brain records.
@jpearman We’ve been having disconnect issues at different times. It happens whenever we are connected to a competition switch or a field control. The v5 controller suddenly flashes red and say robot link lost. We have to reset the controller every time it happens. Usually it happens more often in driver but it has happened in Auton before. Either way it’s really annoying and keeps ruining our matches.
At one point we thought is was a faulty controller switch in practice. We tried another switch and Made sure only one controller was connected to it at a time. It worked for a little bit but then we disconnected again.
We are using VCS. We have text on the LCD of both the controller and v5 brain. The battery is usually charged. The radio is always still green.
Any help would be greatly appreciated. We really don’t want to lose at states because of a disconnect.
Edit: the time we were disconnecting every match was maybe the first time our controller was fully charged in a week. Whenever we kept the controller plugged in it would lessen our chances at a disconnect.
VCS. (For the sake of completeness: using simple and small VCS source which includes all the actual implementation in separate versionable files, but it’s still VexOS, VCS APIs and no 3rd party libraries either)
Not anymore. We did before, but for the last tournament, all the controller screen API usage was removed, including the rumble. I have reviewed their code and nothing was left.
AFAIK the controller went red, won’t reconnect on its own and the drivers usually quickly reset the controller using the hidden reset button. What they do next to resume driving, I don’t know, but I will ask for more details on our next practice.
Hopefully we could follow better diagnostic steps next time it happens in the lab.
Not necessary fully, but well charged at least. They charge before the tournament and also between matches when possible. Note: The team uses partner joysticks.
Before the disconnect, the radio was flashing green. During the disconnect, it was solid red, after the controller reset, it started blinking green again.
Since we had troubles with cables before (port disconnects based on cable angle), I have also tried quickly unplugging and replugging the radio cable, but the symtomps of a disconnect forced this way were different and the controller reconnected on its own.
Across 2 clubs, 7 teams (8 brains), we have seen severe radio disconnects in two bots during auton / driver control, and 1 bot disconnect during pre-auton (while waiting for match to start) in at least 3 matches at the google event.
Other teams would have occasional issues, but appear to be mostly caused by harder impacts. Responses to the questions:
Bot1
Autonomous or Driver control? Disconnected midway through autonomous; students would reconnect by doing a hard reset of controlle (using a screw). Driver control disconnects frequent, but random. This brain had problems at first 2 tournaments - tried changing ports, wires, radios, radio position for the bot with no success. Finally swapped the brain and radio before eliminations at 2nd tourney with a spare – problem largely resolved. Replacement brain was ok at last 2 tournaments, with only occasional disconnects (ie hard impact). Using original controller
Robot Program? VCS, C++
Controller LCD Text? No. Some lines of Brain text
Controller Screen On Disconnect? Went RED, screen still responsive. Event log verified radio link lost. Team would reconnect by doing a hard-reset of controller (with screw)
Controller Battery Fully Charged? perhaps not 100% always, but at least 50%.
Radio LED On disconnect? Red (was green prior to disconnect)
Bot2
Autonomous or Driver control? Both. Disconnected at the end of auton routine nearly every match at Google Signature, frequent disconnects during driver control. Had similar issues 2 weeks before, but thought that perhaps issues were mitigated at first tourney with wiring/port/radio changes. Turned out to not be correct.
Robot Program? VCS, C++
Controller LCD Text? Yes - controller prints text at begin of main, begin of driver control (before while loop), They also had controller prints in 2 of 4 auton routines, but believe they were running the 2 routines without the controller prints.
Controller Screen On Disconnect? Controller display went RED, but still responsive. Event log verified radio link lost. This team would use controller to restart the program.
Controller Battery Fully Charged? Not verified, but believe so - and this had been emphasized
Radio LED On disconnect? Red (was green prior to disconnect)
Bot3
Autonomous or Driver control? Neither - only in pre-auton awaiting start of match. Disconnected during pre-auton prior to match start in at least 3 matches at Google.
Robot Program? VCS, C++
Controller LCD Text? Yes - has more complex pre-auton function, and prints to controller during this time. main, driver control (prior to while loop), and several autons have the print code as well. However, at the google event, the team did not run any of the autons with print code.
Controller Screen On Disconnect? Went RED, screen still responsive. Event log verified radio link lost. This team would use controller to restart the program.
Controller Battery Fully Charged? Yes
Radio LED On disconnect? Red (was green prior to disconnect)
Started happing all last night during driver practice - no code changes what so ever
Programming language/tool: PROS
Batteries: both remote and robot 100% charged
Remotes: two master and partner
Master looses connection, lcd turns red can not turn it off, nor can we turn V5 off without disconnecting power from battery. Master remote requires reset via hidden button, partner never fails in the situation and when reconnected to master after it is reset and reconnected works as expected. Lights on V5 radio have solid RED led’s.
Did not upgrade PROS to newest kernel released a bit ago we are on PROS kernel 3.1.5 and NO OKAPI library
We have DC several times this year. It typically happens more in matches rather in driver practice. It would usually disconnect before auton, and then again during driver control. It was super bad and threw off our driving a lot. However, here’s a solution (I don’t know if this was stated before here, but here goes nothing)
Essentially, take one of the old wires (or any wire for that matter). Strip both sides of the wire, and attach one to the metal of the robot, and the other end so that it touches the tile of the field. It grounds the robot, and overall it worked extremely well at state. We only disconnected once at state, and that was because we didn’t check the wire touching the tiles.
@jpearman I have a correction on the post above on bot 3 details (the one with the pre-auton disconnects). The team confirmed today that they did run their autons with the controller text display. So controller prints in pre-auton ( auton selection code) and simple controller prints in the beginning of auton and user control.
Another addition to this thread, we were at a scrimmage yesterday using the electronic game controllers as usual. None of the other teams present were using dual controllers; two V5 controllers to control their bots, and yet none of them experienced disconnect issues. We were, and we experienced disconnects in almost every match.
Details:
all disconnects were in driver
programmed with vcs
text displayed to controller and brain on both controllers
both controller screens turned red
controller batteries were reasonably high
radio led sequence: green, solid red, green when reconnected
Other notes:
our radio is mounted low, and we plan to raise it to the top and test it more once we set up the electronic game controllers at our own field
we use two controllers, i believe there is a chance that this is a contributing factor, our problems began at the first competition where we began using 2 controllers
collisions did not seem to contribute to the disconnects
at the first competition that we experienced this at, we were told that the order in which you connect to the control towers and power your robot on mattered, but I don’t believe this is the case
** we were able to reconnect by simply powering off the primary controller, dismissing the error code on the second controller, powering the primary controller back on and it would reconnect within a few seconds to resume driver**
We had this happen to us many times too. The best way to not ruin your controller is to turn off the brain. If that doesn’t work, then there is a little hole on the back of the controller. Stick something into it and it will reset. That is not a good method but if you have to you should.
Connections get lost at random points. It happened before the match a few times. It has happened going from auto to driver, it happens immediately after they start driving, it also occasionally happens when they hit a wall (but not always and not the only time it disconnects).
We have sprayed static spray on the tires and at the last meet the fields were sprayed.
No text displayed on the remote
When it disconnects the radio light goes from green flash to solid red and the remotes go to red screen. The kids reset the remotes with the button on the back and go back into their program and resume driving.
This happens almost solely at competitions, so its hard for us to replicate. It happens on red side, blue side, field 1, field 2, skills, etc. So it doesn’t seem to be related to a field issue.