![]() |
Robot looses comms when hit/jolted
Our robot, for the last 2 or 3 years, seems to loose connection whenever its hit pretty hard by another robot. It was never that big of a deal, but today we got our bot driving and attempted the bump. It did make it over, but we lost communication when it got to the other side.
The only thing I can think of is something inside the cRIO could be damaged, seeing as this happens year after year. Or possibly the D-link, but I somewhat doubt that. Anyone have this happen, or any ideas? |
Re: Robot looses comms when hit/jolted
D-link power plug or Ethernet connection, maybe?
If you have the wrong size of power plug, or the connection is finicky, that could be the issue. |
Re: Robot looses comms when hit/jolted
What you said is possible, however it's also possible that something is becoming disconnected when you bump things. I'd look into it further at your next meeting and get is figured out quick. If your robot looses communications every time it gets bumped, an accidental robot-to-robot collision will leave you helpless on the field. From me to you, fixing this problem should be a number 1 priority for your team now.
|
Re: Robot looses comms when hit/jolted
it was usually a harder hit, not every robot-robot hit. but going over the bump seems to trigger it. I'll check the dlink wires tomorrow, but it is the right type. Its the one that came with the dlink.
|
Re: Robot looses comms when hit/jolted
Quote:
|
Re: Robot looses comms when hit/jolted
Loose wire or possibly a cracked solder joint inside the AP.
|
Re: Robot looses comms when hit/jolted
Tripple check every electrical power connection. A loose connection could easily cause what you're describing, and if in the right place (like your battery terminals) could occur year after year. If you can wiggle the wires at all, tighten them down.
|
Re: Robot looses comms when hit/jolted
And check for little metal shavings in places they shouldn't be (like in between digital sidecar pins). When knocked around, they have a tendency to short things out and cause problems.
|
Re: Robot looses comms when hit/jolted
We had a similar issue back in 2010 whenever we would kick the ball (caused the entire bot to shake). Apparently the power connector to the cRio was lose. Check that out.
- Sunny G. |
Re: Robot looses comms when hit/jolted
Quote:
|
Re: Robot looses comms when hit/jolted
Quote:
- Sunny G. |
Re: Robot looses comms when hit/jolted
You need to find out what exactly is making you lose connection.
If you regain connection after a few seconds, the problem probably lies in the cRIO. Make sure you have a solid connection to the PD board If it takes some time, its the DAP. If its the DAP, make sure you have a solid connection to one of the dedicated power lines on the PD board. |
Re: Robot looses comms when hit/jolted
Brandon,
There are a few things that could lead to your problem. 1. The power connector on the DAP has a spring clip for one side of the power input. If you tried to insert a large diameter plug at some point or if the connector was bumped in the past, this spring could be deformed enough to cause intermittent connection. The plug should be snug when inserted. Sometimes a small probe can be used to bend the spring back into position. 2. The power cable to the radio is not supported near the radio. My recommendation to all teams is to use a wire tie near the radio to secure the wire. During violent robot action, the wire and connector will not move. 3. Loose wires or stray strands of wire can easily move to an adjacent contact during violent robot action. Make sure the screws holding the wire are tight, tug test the wire and make sure there are no strands visible. 4. Some PD boards were manufactured with a small, surface mount capacitor near the edge of the board where the radio and crio power supplies are located. It is possible that this cap has come loose and is floating around in the vicinity of the radio power supply. I am of course, expecting that you are using the 5 volt regulator, connected to the +12 volt radio connector on the PD and that all regulator connections are tight/soldered and insulated. |
Re: Robot looses comms when hit/jolted
Quote:
What exactly do you mean by the DAP? that black connector in the cRIO and distribution board? |
Re: Robot looses comms when hit/jolted
Quote:
|
Re: Robot looses comms when hit/jolted
A couple seasons ago, this was very common. The easiest diagnostic was to time how long it took to regain power on the field, or to walk to the robot and observe the LEDs to see which devices are still powered and which are not. Finally, try if you have it narrowed down, wiggle and yank on various wires to simulate the shock of the hit. Keep in mind that it could also be a short, so look for bare wires that could be hitting anything, frayed wires, and messy wiring connections entering the cRIO or other devices.
Greg McKaskle |
Re: Robot looses comms when hit/jolted
We had the same problem and finally traced it to a loose ground. We fixed the ground and that solved it.
Jim Wick |
Re: Robot looses comms when hit/jolted
The c-Rio input wires on our practice bot have come loose, our main breaker wires have come loose, pwm's have come loose. We're in the process of hot gluing everything at the moment. Strange things happen when you pretend that your robot is the General Lee. :D :D
|
Re: Robot looses comms when hit/jolted
This was happening to us early on, I narrowed it to two things...
1.) I had quickly wired the DAP-1522 to the 5v converter and that into a normal 12v output on the PD board. Well those were dipping voltage as my motors stalled working their way across the bump. So I wired that up to the 12v WAGO dedicated on the front of the PD board. ..that worked for awhile and it started happening again.... SOOO 2.) I figured out that as the cRio bumped it'd lose comms.... a simple tightening up of its mounting screws, and now the bot can bump, rock and roll all it wants, I stay connected. |
Re: Robot looses comms when hit/jolted
Don't use hot glue. Figure out why they are coming loose. To hot glue something that is incorrectly assembled means you can't fix it when you need to. Most often when wires come loose on the Crio, it is because they were improperly striped. If the wires to not fully insert into the connector they will be pushed out when you tighten the screw on the side. If the main breaker comes loose, I have to ask if you are using the locking hardware that came with the breaker? For those items that move with the robot, find a place near the connection and use a wire tie to secure them to the robot. We loop the wire going to anything and then take it straight down to the deck. Since we use perf stock to mount everything, we simply pass a tie through the perf stock and around all the wires. The loop allows some movement, and the tie restricts the movement so that it can't be pulled off.
|
Re: Robot looses comms when hit/jolted
we had the same exact problem and it ended up being that the signal light had exposed wires that were hitting the 80 20. if you havent rewired the light you might want to do that.
|
Re: Robot looses comms when hit/jolted
Is there any reason why not to use the 5 volt wago connector on the PD board for the radio and connect the camera with the 12V to 5V connector?
How can you tell if the problem is a Radio reset or a CRIO reset. |
Re: Robot looses comms when hit/jolted
We had this same problem today when trying to go over the bridge. We reproduced it by dropping the front the robot an inch or 2 against the floor. We then focused in on the cRIO. We shook it (a bit violently) and the 'Communication' and 'Robot Code' both went red on the driver's station.
We tightened down the cRIO, pulled out the modules, vacuumed it out, reseated them and tightened down the Digital ribbon cable. After doing all these it never happened again. If we had more time we would have tried one thing at a time since it would have been nice to know which it really was, but you know how it goes at crunch time a day or 2 before bagging it. |
Re: Robot looses comms when hit/jolted
Quote:
Quote:
|
Re: Robot looses comms when hit/jolted
I also remember that the converter is regulated, while the 5v supply on the PD is not. The radio needs to always have 5v and the PD's supply can't properly accomplish that. Is that correct?
|
Re: Robot looses comms when hit/jolted
Quote:
|
Re: Robot looses comms when hit/jolted
hot glue the ethernet connections
|
Re: Robot looses comms when hit/jolted
Its my understanding the correct answer is the 12v port on the PDB is highly regulated (ie if the battery voltage droops way down under super high instanteous load (8v, 6v perhaps lower) the output of the port will still be 12v). The finned 12v to 5v converter module is not so regulated (ie the 5v might collapse if the 12v it sees droops below say 10v. But that will not happen its source is the 12v port on the PDB that it is so highly regulated. Clearly FIRST wants the bridge to have 5v at all costs to reduce loss of connection glitchs.
We know this to be true because last year we accidently wired the finned converter to a standard 12v circuit breaker source and gots of loss of connection errors as the battery voltage faded. Clearly the finned regulator regulation had limits. I have no idea what the regulation specs on the 5v port on the PDB is. You would think it would need excellent voltage regulation for the camera. |
Re: Robot looses comms when hit/jolted
The 5 volt, 12 volt and 24 volt outputs on the PD will be maintained as long as the battery input remains at 4.5 volts or above. The 12 volt to 5 volt regulator requires a minimum of 7 volts input to make 5 volts output. By connecting the 5 volt convertor to the PD 12 wireless output, you will maintain communications. If you connect the convertor to a battery output on the PD ( and your inspector fails to see it), the battery voltage regularly will fall below the required input voltage for the convertor and your radio will reboot.
|
| All times are GMT -5. The time now is 09:58. |
Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi