Strange Behavior

[Gary Bonner]

We had some strange problems with our robot at Ramp Riot this weekend. In our first match, everything worked fine. However, when we put the robot on the field for our second match and turned the power on, some, but not all of the motors began randomly moving. The arm would telescope and retract and turn back and forth, and the drive wheels would jerkily move the robot back and forth. The robot should have been disabled by the field control system at this time. Resetting the RC had no effect, nor did cycling the power or disconnecting the radio. We sat out the match with the power off. Back in the pits, we tethered up and had the same problem, even with a dongle set to disable. All of the RC LEDs were blinking red. Then, all on its own, it stopped and everything worked fine. We could power off and back on and it still worked. After a long break until our next match, we put it on the field and the same thing happened when we powered up. Again, after unsuccessfully trying to diagnose it in the pits, the problem just went away. Then, a 45-minute lunch break and the same thing happens when we get back. This time we just watched it spaz for about five minutes, and it went away. After that, we kept it powered up all the time, and the problem did not re-occur. Later in the afternoon, we reloaded our code, but we didn't leave the robot turned off long enough to tell if that fixed it. There was one match in the afternoon where the robot just died in the middle of a match. The OI lit no data, but the field control system didn't report any problems.

Any idea what's going on? Should we perform an exorcism? We store our bot in a convent basement and we don't want to be bringing any evil spirits down there.

[Bharat Nain]

Check all your electrical connections. I was helping your team at Capital Clash and really could not understand your wiring(please don't take offense to that). Then check the code - somewhere, something might have gone wrong. Contact IFI because they might be able to explain this weird behavior. Thing's like that happened to our 04' bot and it was all because of sensors(potentiometers to be specific) and all we had to do was make sure everything was connected right - even grounding.

[KenWittlief]

could be:

1. an error in your SW. Were code changes made before this erratic behaviour started? can you download your last known-good code again? If the SW jumps out of its loop, then your bot will go berzerk

2. a loose wire to the power on your RC.

3. power problem with your controller? loose wires on your controller? Joysticks or custom controllers plugged into the wrong ports?

NOTE: If the bot wacks out with the tether attached then it cannot be: someone transmitting on the same frequency, or a problem with your receiver.

If you cant isolate the cause, you might have a defective robot controller, or operator interface.

[Mark McLeod]

I don't suppose the RC has any fairy dust on it, i.e., fine metal dust from working over an unprotected controller?

[Andrew Blair]

Hmmm, as said before, double check all your switch connections. Microcontrollers don't seem to like shorts, even on unused inputs. Also, make sure of two things: Battery voltage and connection. If your voltage is too low, say on a damaged battery, that throws everything awry. A loose battery connection might also be causing some of the problems. I noticed that your battery connections were shrink wrapped and (no offense), occasionally carried around by the anderson connector. A loose wire in one could easily be out without you noticing. That wouldn't account for it ignoring a disable command though. That might be bad.

Just some of Murphy's little tricks, but think it is more likely what was said before; loose electrical connections of some sort.

Originally Posted by KenWittlief

If you cant isolate the cause, you might have a defective robot controller, or operator interface.

You can use your 2004 operator interface to check if it's good, and if you want to go drastically, swap out a 2004 RC with 2005 master code to check the RC. However, that would be a little messy.

[eugenebrooks]

Originally Posted by Gary Bonner

We had some strange problems with our robot at Ramp Riot this weekend.

I would upgrade to version 11 of the master code, available from
the IFI web site. It is reputed that it addresses strange motor jumps
that might occur. It might address your problem, and it might not...

[dubious elise]

Our robot had a similar dysfunction at the Championships this past year (whenever I stepped off of the human player pad, it would scoot backwards away from the loading zone between 4 and 12") . I believe we remedied it, as a previous poster mentioned, by swapping the 2004 and 2005 controllers.

[Matt Krass]

Our robot was behaving somewhat like that at the Deer Park Invitational, it turned out we had an old version of the Master Firmware (We were using a borrowed 2004 controller, we finally got ours fixed, too bad we're not competing anymore...) and it would ignore disables and sometimes decide to do whatever it wanted. Kind of worried the field crew when our robot came alive during loading....

Anyhow, redownload the Master Code, version 11, it might have gotten borked. (Yes, that's the technical term )

[Al Skierkiewicz]

I go with Dr. Brooks here. One version of the IFI loader had some wierd code that was fixed in the later release. This was a problem in the first few regionals of 2005. Is it possible that someone had an old version?

[camtunkpa]

Gary,

We weren't having problems quite like that, but we had trouble with our controller on the NASA field. Sometimes the OI would flash no data in the middle of the match and we would be dead for 10-15 seconds. Other matches our OI would read disabled in the middle of the match for another 10-15 seconds. I don't wanna rag on the NASA field because they do alot for teams running off season events, but I'm just saying you might not have as big of problems as you think. I would try running the robot at your school or whatever and see if the problems still occur. That's just my $.02 on the situation.

[Wayne C.]

considering the robot worked fine later in the day with no apparent code modification I might consider tiny metal shard got wedged somewhere and was intermittently shorting out the processor. When the robot eventually was moved enough the piece knocked loose and the problem disappeared. When warm the machine worked OK- but after cooling down again it acted up.

It would'nt hurt to get some canned air and blow out all the dust amongst the wiring. Turning the machine on its side while doing this would also be a good idea and see what falls out.

Also- have you checked the batteries you are using? After lots of crashing they get internally broken and the plates can detach and rattle giving intermittent current. We are finding that our old batteries are on their last legs and they are from the same years as yours would be. What was the voltage readout you get on the Op interface? A low voltage could also give you strange behavior although it shouldn't activate deactivated motors.

Did you re-dump your code between rounds? Did that fix it?

Also Coca Cola dumped into the processor does the same thing.....but Bharat can fix it.

WC

[Matt Leese]

Originally Posted by Wayne C.

When warm the machine worked OK- but after cooling down again it acted up.

That may be indicitative of a cold solder joint. When warm, the joint expands and everything work fines. When cold, the joint contracts and no contact is made.

I experienced this first hand back in 2001 at the Championship. The RC worked at competition but when we got to the Championship, in the morning when it was significantly colder, the RC would no longer start up.

Obviously, you've had different symptoms than we had back then but it's still another possibility. I'd personally consider it a higher possibility than a piece of metal getting into the RC.

Matt

[Zoom]

We had the same problem and at the same place. Yes at Ramp Riot. First our robot was fine nothing wrong. But after lunch problems occured. After 10 sec of the match it would be disabled. The staff their said that they didnt disable us. we are using our 2004 controller because our 2005 controller is damage. The 2005 controller is not as strong as 2004. It broke where you connect the main battery. The pin just came out. Does any one know where we can send it to be fixed. Il put version 11 of the master code on it. But the strange thing is when we radio it with our radio it works fine.

[Al Skierkiewicz]

Originally Posted by Zoom

We had the same problem and at the same place. Yes at Ramp Riot. First our robot was fine nothing wrong. But after lunch problems occured. After 10 sec of the match it would be disabled. The staff their said that they didnt disable us. we are using our 2004 controller because our 2005 controller is damage. The 2005 controller is not as strong as 2004. It broke where you connect the main battery. The pin just came out. Does any one know where we can send it to be fixed. Il put version 11 of the master code on it. But the strange thing is when we radio it with our radio it works fine.

As I remember, 2004 modems do not work on 2005 field controls also 2004 modems cannot be used with 2005 RC or OI. Check IFI website.
The power input to the both the RC and the OI is the same design but both are subject to fail when abused. Either can be returned to IFI for repair. It is possible to damage the board if the metal was pulled out. Let IFI have a crack at it, call them for a repair authorization.

From the IFI website..."The new master code will only upgrade the Robot Controller. There is no way for you to upgrade the 2004 Operator Interface. The 2004 OI is somewhat compatible with the upgraded RC and you should not see a difference. The 2004 OI will not work correctly with the new Arena Controller and can not be used at FIRST competitions."

[EricH]

How was your battery situation? I know I'm asking the obvious here, but were both batteries (main and backup) charged and on the robot? If both were not on the robot, and the main battery was extra low, then you need to make sure that both are charged and on next time. (Please don't ask me how I know, I don't want to go into that.) Of course, your robot may have gotten Mechanic's Syndrome.