Weirdest issue with our robot!

[voteforandy1]

I'm on team 3131 and as we were doing driving tryouts our robot suddenly developed an odd problem. Whenever we turn to the left, the jaguars seem to lose signal because they just go blinky lights. What's weird is that it's only when we turn left! At first the problem was only when yu turned maybe 45 degrees hard. And we kept repeating the problem to troubleshoot and it got so sensitive that now it loses signal really fast. We have narrowed out loose pwms, faulty jags, losing power to jags, and the obvious. Is this maybe a proggramming problem or a electrical problem. Before the problem started we were being fairly gentle with it. At leat we weren't shaking the electronics very bad. We left the meeting tonight with no answer and were coming back tomorrow to trouble shoot. Any advice? If there are any questions then just ask and I'd be happy to answer. We are desperate!!! Thanks!
-Andy Hunt (Gladstone, team 3131)

[reversed_rocker]

we had a similar problem last year, turned out to be a faulty jag. sometimes when they go bad they will only run in one direction. if you wrote it in C++ or java you should post your code for people to take a gander at

[voteforandy1]

I'll check that, we didn't think about the jag only running 1 way. We will voltimeter it tomorrow. Any other advice anybody?

[Tom Line]

1. Have you changed to a new battery?

2. Have you checked your drivetrain on the right side (assuming tank drive) for any unwanted drag that might be stalling the motors and causing the jags to go into self-protect mode for 3 seconds?

3. What is your drive configuration?

4. Are you running dual cims on both sides and have you confirmed each individual motor is running?

[Al Skierkiewicz]

Andy,
It could be a variety of problems including a low battery. Were the jaguars slowly flashing red? This is an indication of a fault, usually over current. It lasts 3-4 seconds before the jaguars return to function. Were they flashing yellow like they do when the robot is disabled, then the battery voltage was so low that the Crio went into voltage protect mode. Certain robot designs draw very high currents in turns. It is also possible that there is something wrong mechanically on one side or the other of the robot drive system.

[rsegrest]

What code/language are you using? If it is LabView are you using scalars in your programming?

[voteforandy1]

1. Have you changed to a new battery?

2. Have you checked your drivetrain on the right side (assuming tank drive) for any unwanted drag that might be stalling the motors and causing the jags to go into self-protect mode for 3 seconds?

3. What is your drive configuration?

4. Are you running dual cims on both sides and have you confirmed each individual motor is running?

1. Yes, and it actually seemed to delay make the problem more tolerant(good) but then it quickly got worse and worse.
2. We did and there was a slight difference, but it actually seemed to be easier to turn on the right side, (the side that's pushing and causes the problem) so I think that's probably not the issue.
3. We are doing four cim motor tank drive.
4. Yes, and yes however our test consisted of looking whether both the lights changed to the right color but we should test the actual electrical outputs with a voltimeter. And about programming, I am on electrical team so I know very little about our proggramming however I will ask our head programmer that and ask for the file to post. I do know we are using labview however.

[J_Beth]

If the problem started when you were turning hard to the left and it got worse as you kept turning to the left, could it be that your joysticks have gone bad, or is that even possible.

[voteforandy1]

Andy,
It could be a variety of problems including a low battery. Were the jaguars slowly flashing red? This is an indication of a fault, usually over current. It lasts 3-4 seconds before the jaguars return to function. Were they flashing yellow like they do when the robot is disabled, then the battery voltage was so low that the Crio went into voltage protect mode. Certain robot designs draw very high currents in turns. It is also possible that there is something wrong mechanically on one side or the other of the robot drive system.

SMART!!! I didn't think about the voltage turning off the crio or putting it in protection mode. I think that is the most likely problem considering what I've seen. It makes perfect sense for what's happening. And it would make sense that it would only happen turning to the right because we sanded our wheels down and I bet one side is still much rougher than the other which makes it tough to turn which take a lot of energy. I will definetly look into that tomorrow! Thank you a billion times, chief Delphi rocks!

[reversed_rocker]

the joysticks can become uncalibrated. they center themselves when you plug them into the driver station, if you accidentally bump them while you're starting up you can have a joystick thats not centered properly. i dont think thats the issue but its a problem thats good to be aware of

[J_Beth]

the joysticks can become uncalibrated. they center themselves when you plug them into the driver station, if you accidentally bump them while you're starting up you can have a joystick thats not centered properly. i dont think thats the issue but its a problem thats good to be aware of

well, i am not quite sure of how joysticks operate, if they use pots or something else to measure their location, but could a sensor have gone bad so that it no longer reads that the stick is going left?

[Al Skierkiewicz]

Andy,
The Crio goes into protect when the battery voltage falls below 5.5 volts for some length of time. You can get a little insight to this if you watch the voltage reading on the drivers station. Voltage samples are not continuous but if you notice voltage running up and down in turns and getting as low as say 7 or 8 volts, the dips are likely much lower. Reboot of the Crio can take place if the battery voltage should dip to around 4.5 volts at which time the 24 volt regulator in the PD starts to cut out. You know how long the Crio takes to boot so this is a risky issue for competition. Tank turns with high friction drives and wheels, cause CIM motors to draw near stall current. That is 129 amps per motor.

[voteforandy1]

For some reason our classmate shows our battery voltage is at 0.01 at all times when enabled, but that can't be true. How do I fix that?

[Zac Pilarski]

how are you getting your voltage value

[weinertj]

I'm on the same team as Andy. Yes, we are using Labview, and we have scalars.

[voteforandy1]

We think the problem is with the cRio voltage protection mode. How can we be sure? What is the indicator of voltage protection mode other then low voltage? Is it a light flashing on the cRio? And if this is the problem, then how do we lower the voltage we're using when we turn? Thanks to anybody who responds!

[PAR_WIG1350]

I once had a bike that did something very similar. Every time I turned left, I fell off the bike. I turned out the brake cable was in an unusual position so when I turned left, the front brakes were applied.

Sorry for being off topic

try zip-ties, they fix everything.:D

ON-TOPIC EDIT: I think the cRIO actually power cycles in voltage protection mode so I would look for signs of the cRIO rebooting. (Flashing lights, erratic starting-stopping of the motors, etc.) Also, If any thing, I think you would be drawing too much current which should trip the breakers in the PD board; the PD board limits the voltage already. Are the transmissions lubricated sufficiently? Are the wheels too soft? Remember, the CIMs can only output so much power, to get more torque you could try a different gear ratio or you could run the jaguars (assuming you are using them) in torque (current) control mode.

[J_Beth]

For some reason our classmate shows our battery voltage is at 0.01 at all times when enabled, but that can't be true. How do I fix that?

Look at <R66> it tells you to install a jumper on the breakout board that is installed in slot 1 of the CRio. We had this same issue last year, and i believe this was how we solved it.

[Al Skierkiewicz]

Justin,
The jumper is needed for the Crio to monitor battery voltage. When the voltage falls to about 5.5 volts, the Crio disables PWM outputs. This is done in anticipation of the voltage falling to 4.5 volts at which point the 24 volt power supply becomes unstable. It is possible to draw the battery down when you have multiple CIM motors in stall.