Communication lost

[Skehtb]

Every time we drive our robot, it suddenly stops moving and the communication sign on the computer (the one that tells if there is communication between the robot and the computer) turns red. It only happens when we are activating the drive train - "moving" (4 CIM motors). When we shoot, collect balls and any other thing except for drive - it doesn't happen!
I hope you understand the problem, we realy need your help (off-sesason event in 3 weeks!)

Thanks!!

[dellagd]

Hey-

Are you using a 2CAN? We were, and we traced the problem to our 2CAN causing us very similar problems. Try using CAN over the serial port on the crio (bypassing the 2CAN).

This happened another time and was solved by this

[Al Skierkiewicz]

Skehtb,
You don't elaborate on what other issues you are seeing so I will make some general statements.
If the robots stops functioning for something like 20-50 seconds, then you are drawing the battery voltage down to the point that the Crio is rebooting.
If the robot stops for ~50 seconds then the power supply to the radio is causing it to reboot.
If the robot just stops for about 3-4 seconds and you are using Jaguars, then you are faulting the Jags (probably to high current). If that is the case, the LED on the Jags will flash red during the 3-4 seconds.
As to what could be the cause, you could have batteries that are not fully charging or if they are old, may not have the capacity to drive the robot any longer. If you rule out the battery, then loose connections to the battery, the PD or the main breaker can also cause these issues. If the robot just stops but you can start driving almost immediately, the Crio has an inhibit function based on the battery connection you make to the analog module in slot 1. If that voltage falls to 5.5 volts, the Crio inhibits all output. Since you are in Israel, can we assume you made it to an event outside of Israel? If so, your drive train may have been damaged during transit and now has high friction due to the damage. Check the drivetrain by lifting the robot and moving the wheels by hand. If you are using a transmission that you can slip into neutral, the wheels should move very easy. Depending on your design, the friction could be caused by misaligned drive shafts, tight chain, bent wheel shafts or contact between drive components and robot chassis. In some case, stretched chain could be causing the drive train to bind or hang up on other parts.

[IndySam]

Great advice on this thread.

We also had a similar problem a couple of year back and traced it to the cRio not being isolated from the frame, a screw under it had worked loose. When we removed this screw the problem went away.

Since this event we have always made a large effort to make sure the cRio stays isolated.

[Brandon Zalinsky]

We're having the exact same problem. Could it be the wireless bridge? We've already broken one this year.

[Al Skierkiewicz]

Brandon,
It is really difficult to diagnose problems without more data. In your case, it is possible but watch the lights on the Dlink. The robot is talking to you, you must listen to find out what might be wrong.

[Skehtb]

The Crio is isolated, but i'll check it again.

Originally Posted by Al Skierkiewicz

Skehtb,
You don't elaborate on what other issues you are seeing so I will make some general statements.
If the robots stops functioning for something like 20-50 seconds, then you are drawing the battery voltage down to the point that the Crio is rebooting.
If the robot stops for ~50 seconds then the power supply to the radio is causing it to reboot.
If the robot just stops for about 3-4 seconds and you are using Jaguars, then you are faulting the Jags (probably to high current). If that is the case, the LED on the Jags will flash red during the 3-4 seconds.
As to what could be the cause, you could have batteries that are not fully charging or if they are old, may not have the capacity to drive the robot any longer. If you rule out the battery, then loose connections to the battery, the PD or the main breaker can also cause these issues. If the robot just stops but you can start driving almost immediately, the Crio has an inhibit function based on the battery connection you make to the analog module in slot 1. If that voltage falls to 5.5 volts, the Crio inhibits all output. Since you are in Israel, can we assume you made it to an event outside of Israel? If so, your drive train may have been damaged during transit and now has high friction due to the damage. Check the drivetrain by lifting the robot and moving the wheels by hand. If you are using a transmission that you can slip into neutral, the wheels should move very easy. Depending on your design, the friction could be caused by misaligned drive shafts, tight chain, bent wheel shafts or contact between drive components and robot chassis. In some case, stretched chain could be causing the drive train to bind or hang up on other parts.

The problem appears with different batteries. We found out that the robot doesn't stop functioning when we drive it forward. It happens only when try to turn it (and I'm pretty sure it happens only when it truns to the left), drive it backwards, or stop it after driving forward. I'm not sure about how long the robot stops functioning, but its definetly more than 20 seconds. I'll check it.
I don't understand how high friction or bent wheel shafts can cause the problem. And if it changes anything - our drivetrain is 8wd and this season we did not participated in an event outside of Israel.

[mikets]

Originally Posted by Skehtb

I don't understand how high friction or bent wheel shafts can cause the problem.

High friction or bent wheel will cause the motor(s) to stall. Stall draws tremendous current which could draw down the battery voltage or cause the Jaguars to cut off due to over current.

[Al Skierkiewicz]

Mike is correct. However when the condition exists only on turns, I am going to guess you have sticky tires and are using a tank turning for steering. When this type of drive mechanism is used with high friction tires, the motors do tend to stall or get very near. A CIM motor stalls at 131 amps so with all four motors in stall, you are drawing so much current from the battery that the terminal voltage falls below the point that the power supplies in the PD can keep running. If both the Crio and the Dlink reboot, the Dlink takes longer to come back. Also, and I have to say this, if the Dlink is not powered from the dedicated +12 volt radio output on the PD, then small fluctuations in battery voltage will cause the Dlink to reboot.

[Alan Anderson]

Originally Posted by Skehtb

We found out that the robot doesn't stop functioning when we drive it forward. It happens only when try to turn it (and I'm pretty sure it happens only when it truns to the left), drive it backwards, or stop it after driving forward. I'm not sure about how long the robot stops functioning, but its definetly more than 20 seconds. I'll check it.

Look for a shorted wire between one of the left side drive motors and its speed controller. You might be getting voltage on the frame when a motor drives backwards, and another shorted frame connection somewhere (likely a power return or "ground" wire) could be causing your D-Link or cRIO to reset. Or you might have cross-wired motor leads, with one motor connecting to two different speed controllers. That results in very weird symptoms that can sometimes match what you're describing.

You can start troubleshooting electrical issues by removing the circuit breaker from the power branch to one speed controller at a time and seeing if the problem can be narrowed down to one motor's circuitry.

[philso]

An easy way to test what Al is suggesting is to put some blocks under your robot so that none of the wheels touch the ground then "drive" the same manouvers that cause the problem. You now have no traction so none of the motors should stall due to the choice of wheel tread. You may also find certain wheels that don't turn, or do not turn smoothly, due to high friction in gearboxes or some other issue.

Good luck.

Phil

[Skehtb]

When we tried to "drive" the robot while the wheels were "in the air" (not touching the ground), nothing happened - it worked smoothly. We lost communication only when we drove the robot on the ground.

[Alan Anderson]

Originally Posted by Skehtb

When we tried to "drive" the robot while the wheels were "in the air" (not touching the ground), nothing happened - it worked smoothly. We lost communication only when we drove the robot on the ground.

Watch the lights: the big orange Robot Signal Light, the ones on the D-Link router, the ones on the cRIO. They'll give you good clues as to what is going on with the robot when you lose control.

Your symptoms are starting to point more clearly to a power issue. One of the common problems teams have is connecting the robot radio incorrectly. Can you describe exactly how the D-Link is being powered? Specifically, where is it connected to the Power Distribution board, and what else is connected to its power input?

[Al Skierkiewicz]

Originally Posted by Skehtb

When we tried to "drive" the robot while the wheels were "in the air" (not touching the ground), nothing happened - it worked smoothly. We lost communication only when we drove the robot on the ground.

This is pointing more and more to drivetrain friction. Either misalignment or turning frictions are high.

[Alan Anderson]

Originally Posted by Alan Anderson

Your symptoms are starting to point more clearly to a power issue.

Originally Posted by Al Skierkiewicz

This is pointing more and more to drivetrain friction.

Al and I are not really disagreeing here. Excessive drivetrain friction will drag the battery voltage down, but that alone will only cause a momentary loss of control. It shouldn't cause a complete loss of communication. If things are properly wired, the control system is designed to shut down the motors before the voltage gets too low to maintain its operation. A robot that reboots when you try to turn on the ground, but works okay when up on blocks, has more than one thing wrong with it.