In our testing for the past few days, our mecanum drive has been exhibiting some weird behaviors when strafing and rotating. First, when I strafe, the robot basically acts as if it is rotating around a single wheel. The robot will go diagonal and then rotate more and more until the robot is doing circles. When correcting this with counter rotation (with gyro and manual joystick) in order to get the robot to strafe straight the top wheels are almost stalled out while the rear wheels are going full speed. For rotating, the robot moves in an arc-like fashion instead of rotating in place. We checked on blocks that the wheels are moving correctly. Also, even though our robot looks to be very front heavy, we found the center of mass is very close to the middle of the bot. If anyone has any possible solutions please let me know. I want to see if there are any solutions that we have not tried that are not hard to do before we start replacing any wheels, gearboxes, motors, etc.
Try printing the motor commands that are being sent to each motor. Even though the wheels are moving in the right direction, they may be moving at different speeds. You could also try recalibrating motor controllers?
Have you checked the wheels are the correct orientation? They should form an “X” shape when viewed from above.
We have a similar problem on one of our demo bots. We plan on fixing it by adding weight to one side and balancing it out.
We have run into a similar issue where the wheels aren’t all contacting the ground, if you are able test this quickly by loosening up the frame to give the wheels a little bit of play.
We also experienced similar problems and just converted to traction wheels. I believe the biggest reason is probably poor weight distribution. Try correcting the weight and trying again to see how it performs. In addition, the wheels should be as square as possible to one another.
With mecanum drive you have to be very careful that weight is equally distributed.Uneven weight distribution can give you massive issues of not being able to strafe. If it were me I would do the paper test by putting paper underneath all wheels and make sure there is resistance when pulling the paper out from under the wheel.
I don’t think one pound will make or break functionality, but weight distribution is quite important. Another thing to check here: the frame has not been distorted or warped such that one wheel is not fully in contact with the ground.
We have used Mecanums 5 out of the past 7 years and never had the problems you are talking about. It does sound like you have one wheel that might not be fully level. We use a NavX gyro corrected drive which looks at not only direction but corrects for rotational rate as well. This year we we purchased the following and the performance increase the drivers have noticed was remarkable.
This is less of an issue than it is made out to be, provided that the robot has a gyro and has good software to correct for any misalignment. It especially helps to have some way to verify wheels are not slipping / breaking traction. A mecanum drive in FRC is going to need good software to prop it up and can’t usually rely on mechanical perfection.
The Past 3 years that we ran Mecanums we use the drive code from 1519’s 2015 robot. It uses a gyro to maintain heading as well as command the TalonSRX in a closed loop velocity mode using the SRX Magnetic encoders vs percent output. Last year our cross field auto the one side hit the scale platform everytime but never went off course. We could have tried to shorten the first move to stop sooner and stay on the carpet but it hit the goal every time so we left it.
Also check that your rollers are spinning freely. It sounds like one or more (front?) wheels may have sticky rollers.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.