Macanum Drive Train Problem

[Thorn]

My team purchased the upgrade kit from AndyMark.

The motors are not all turning at the same speed. There is no driving the robot straight without a programming hack because maximum power on the wheels results in different RPMs for different wheels.

There was a long wait on back-order, and now as we enter the final week it is very stressful to not have a working drive train.

I think the problem is the connection to the gear boxes, but it's difficult to see exactly what is wrong and how to fix it. Has anyone seen this issue before? How can we resolve this?

[Justin m]

Hello,
Even though I was not on the team when we last did mecanum drive, I heard that encoders are necessary for the RPM, as every motor is different and will give differing RPM s at the same voltage

[ehochstein]

We had this issue, turns out somehow our students ended up using a screw that was too long between inside the gearbox and the CIM Motor. When this happens, the caps lift up slightly. Check to make sure this is not happening.

Make sure you grease the geatboxes.

What sort of motor controllers are you using?

[Electronica1]

My team's programmers were thinking the same thing with our system. The real reason is that your center of gravity is not perfectly in the center of the robot so the weight on each wheel is different. This causes certain wheels to push more than others. If you are not at the weight limit, it might be worth it to add weight to get the center of gravity in the center of the robot.

[ehfeinberg]

Would it be possible to post a picture of your set up? 449 used the toughbox nano's with 6" mecanum wheels and had no problems what so ever.

Do all of the gearboxes have the same internal gearing and the same total gear ratio? The kit seems to have a 12.75:1 ratio while if you buy each gearbox individually they could have a 10.71:1, 8.45:1, or 5.95:1 ratio.

Are all four of your drive motors normal cims and one isn't a mini-cim?

[Thorn]

Thank you all for you your feedback, but I'm afraid most of the obvious ideas have already been checked.

We are testing the wheels now with no weight on them at all. I can easily tell by looking at the wheels that they are not spinning at the same rate. All gear boxes came from the Andy Mark upgrade kit so they all should have the same ratio (all the gear boxes do look identical inside). We did grease the gear boxes.

We are using Jaguars. My feeling is that the problem is mechanical rather than electrical. We aren't using the Mecanum drive code (with the Java WPI Lib) right now, we are just running full power to the motors or whatever we get from joyStick.getY() input.

Here are a couple of pictures for reference:
top view of control system

Wheels without wires

Do Jaguars sometimes develop an issue where 100% power really means 80% power because the Jaguar is failing?

[Whippet]

Try rotating the wheels by hand. If any of them are harder to turn, you may have overtightened the bolts going into the axles. We had a similar issue caused by that.

[PVCMike]

It's asking a lot in my opinion to expect that they all output the exact same torques and speeds at a given current. We've run mechanum drives many times. In the past, we've used encoders, gyros and accelerometers to combat the issue of varying wheel speeds. They all proved pretty effective in measuring these variations and sorting them out electronically.

From a mechanical standpoint, I would definitely (with the robot off) give the wheels a spin with your hands to make sure the resistance is relatively the same on all 4 wheels. Also make sure all of your electrical connections are good and tight and that you're getting equal output voltage to each motor.

[Chris_Ely]

Have you calibrated the Jaguars?