Victor 888/Talon SR Weird Issue (Drivetrain)

[michaelyork]

I've been troubleshooting our robot for issues involving our drivetrain, and I just noticed an issue involving the LEDs on the motor controllers. We are running two motor controllers, going to two CIMs. When I try to rotate the robot, it works occasionally, but most of the time, it locks up, and eventually speeds forward. When it does speed forward, a light clicking sound can be heard.

Yesterday afternoon, I noticed that the right motor controller LED turns off when this issue happens. (We tested it with both a Victor 888 and a Talon SR, same issue.) It also seems to be the corresponding side of the robot that locks up. We are running a standard arcade drive program to the motor controllers, and we have never had this problem before.

We are thinking that it is either a bad CIM on the side that the motor controllers are acting funny, or something to do with the wiring as a whole. Do you have any thoughts as to why this may be happening?

[Mark McLeod]

What size breaker do you have the CIM motor controllers wired to?
I encountered a team at my last competition that wired their drive CIMs to 30a breakers (40a breakers are quite a bit better for driving).
The stall starting power draw of CIMs will usually trip 30a breakers (producing a clicking sound) before the CIMs can move against any resistance (although they might work up on blocks or on slick flooring).

[michaelyork]

Originally Posted by Mark McLeod

What size breaker do you have the CIM motor controllers wired to?
I encountered a team at my last competition that wired their drive CIMs to 30a breakers. The stall starting power draw of CIMs will usually trip 30a breakers before the CIMs can move against any resistance (although they might work up on blocks or on slick flooring).

I have the motor controllers wired to the PDP with a 40 amp breaker. They do work on surfaces with no resistance, and they can drive full speed forwards and backwards, but when I try to rotate, the problem occurs.

[Mark McLeod]

Originally Posted by michaelyork

I have the motor controllers wired to the PDP with a 40 amp breaker. They do work on surfaces with no resistance, and they can drive full speed forwards and backwards, but when I try to rotate, the problem occurs.

Then your gearing is probably too high and you are tripping the 40a breakers.
Skid steer turning, especially on carpet, requires quite a lot of power.
More power might mean two motors per side rather than one.

What type of drivetrain?
How many motors per side?
What type of gearing?
What kind of wheels/treads/belts?
Pneumatic tires are the worst for turning on carpet...

[michaelyork]

Originally Posted by Mark McLeod

Then your gearing is probably too high and you are tripping the 40a breakers.
Skid steer turning, especially on carpet, requires quite a lot of power.

What type of drivetrain?
Pneumatic tires are the worst for turning on carpet...

We currently have a 6-wheel drive, running on two CIMs (one per side). I'm not sure what kind of wheels we have, but I do know that they are flat all the way around the surface of the wheel, and are not hi-grip (they are also gray, if that helps any). We are running the KoP drivetrain, I'm not sure about the gearing used.

[Mark McLeod]

The wheels sound like hard plastic and those are okay for turning.

Turning for those kinds of wheels can be improved by several choices:
- Dropping the center wheel a little more
- Adding another motor to each side
- Changing out the gears in the gearbox for a better ratio (AndyMark has another set of recommended replacement gears - see page 4 for the 12.75:1 ratio here http://files.andymark.com/PDFs/AM14U3_UserGuide_1.1.pdf)

[michaelyork]

Originally Posted by Mark McLeod

The wheels sound like hard plastic and those are okay for turning.

Turning for those kinds of wheels can be improved by several choices:
- Dropping the center wheel a little more
- Adding another motor to each side
- Changing out the gears in the gearbox for a better ratio (AndyMark has another set of recommended replacement gears - see page 4 for the 12.75:1 ratio here http://files.andymark.com/PDFs/AM14U3_UserGuide_1.1.pdf)

As we have a competition this weekend, it seems like either adding another CIM to both sides, or dropping the center wheel a little bit would be our best options. Is there anything in the program that I could possibly do to alleviate having to disassemble our drivetrain tomorrow? We're just doing an Arcade Drive, so I thought that possibly making our own custom drive code would work (instead of relying on WPILib to cause the wheels to drag). Is this a problem that could be fixed by having the motors spin in opposite directions to be able to spin, or is it absolutely necessary to modify the drivetrain?

[Mark McLeod]

It would be easier to train your human drivers to only attempt to turn while also traveling forward or backwards rather than trying to turn in place.
That takes less energy.

Reducing your wheel diameter would also work, however that would probably mean you couldn't get over some defenses.
The problem started when you probably sized the KOP drivetrain up to 8" diameter wheels thereby increasing your overall drive ratio.

[michaelyork]

Originally Posted by Mark McLeod

It would be easier to train your human drivers to only attempt to turn while also traveling forward or backwards rather than trying to turn in place.
That takes less energy.

I'll try to tell them that at our next meeting, but over time, I'll be sure to tell them to try anything they can to reduce the load on the CIMs (adding more, etc.) Thank you for your suggestions, I hope we'll be able to get this worked out before this weekend! I'll check back in tomorrow, when we have another meeting, and see how these work out.

[rlance]

Originally Posted by michaelyork

I'll try to tell them that at our next meeting, but over time, I'll be sure to tell them to try anything they can to reduce the load on the CIMs (adding more, etc.) Thank you for your suggestions, I hope we'll be able to get this worked out before this weekend! I'll check back in tomorrow, when we have another meeting, and see how these work out.

Were you able to get it worked out. We had this same issue at our regional and ended up changing the program so that when we wanted to rotate it would alternate approx. every half second between the two motors since we couldn't turn them in opposite directions at the same time. One motor forward...other one backwards...other one forward...other one backwards. We did consider adding motors, but time was not on our side between matches.

The lesson we learned is that we should have driven on carpet in practice.

[Al Skierkiewicz]

Originally Posted by rlance

We had this same issue at our regional and ended up changing the program so that when we wanted to rotate it would alternate approx. every half second between the two motors since we couldn't turn them in opposite directions at the same time.

This is common with tank drives and lot's of sticky wheels on the floor. If you must turn in place, have your software reduce the throttle command so that you are never running full power in turns. If you don't have dropped center wheels, a six wheel and many times a four wheel drive will put drive motors in near stall conditions when turning in place.

[Alan Anderson]

Originally Posted by michaelyork

I have the motor controllers wired to the PDP with a 40 amp breaker. They do work on surfaces with no resistance, and they can drive full speed forwards and backwards, but when I try to rotate, the problem occurs.

Along with the other good suggestions, make sure to double-check your wiring. I've seen a robot do exactly what you describe when the motors were accidentally cross-wired to the two speed controllers. Look closely and be absolutely certain each Victor/Talon has the output going to exactly one CIM.

[Al Skierkiewicz]

Originally Posted by Alan Anderson

Along with the other good suggestions, make sure to double-check your wiring. I've seen a robot do exactly what you describe when the motors were accidentally cross-wired to the two speed controllers. Look closely and be absolutely certain each Victor/Talon has the output going to exactly one CIM.

This occurs more than you think. You can check by pulling the breakers for two of the motors. Then check direction and then repeat for each motor in your drive. I suspect you will find one of them going the wrong direction. That event will cause a breaker trip.

[pilleya]

Originally Posted by michaelyork

Is there anything in the program that I could possibly do to alleviate having to disassemble our drivetrain tomorrow?

If you are using Toughbox Mini's and can get you hands on two 14tooth 3/8 hex bore Andymark gears and two 50 tooth 1/2 hex bore Andymark gears. You'll be able to change the gear ratio from 10.71 to 12.75:1 which should improve things, this takes about 15-20 minutes to do each gearbox. At the same time, you can also add the extra CIM motor on each side. It should take about 30 minutes maximium to do this. You can put the CIM pinions on the motors before the event, and then count this towards your withholding allowance, as it will save you some time.

[Ether]

Originally Posted by michaelyork

We currently have a 6-wheel drive, running on two CIMs (one per side).

I know it's too late to help now, but going forward you should always do some motor calculations before building your drivetrain.

There are a number of drivetrain calculation spreadsheets in the "white papers" section here on CD.