Hi everyone,
Thank you for the help in advance.
So our team(3958) has had a pretty shaky drivetrain for the past 3 years now. It is usually related to structure and the unanticipated strain that is put on the drivetrain during competition. This year, we tried a mechanum drivetrain after experimenting with it during the offseason.
We are having issues moving laterally. Our front left wheel seems to be spinning more freely than the other wheels, and everything works going forward and backward. We are pretty confident that it is not a programming or electrical problem. We just can’t seem to diagnose the issue so I was hoping someone here might have some insight. I can provide more info if needed but I wanted to keep it pretty short.
Thanks!
Where is your cg sitting? By any chance, is it closer to the back right? Your weight distribution can chance how fast each wheel spins.
Step 1 would be make sure that the proper wheels are on the proper corners. Without that the vectors for going lateral would get screwy and definitely not work.
While a picture of your robot might help, and video would also be useful, with mecanum drive problems strafing usually stem from not having equal pressure on all of the wheels.
You can fix this by using a suspension system where the wheels are spring loaded, probably not possible for your team at this point, by lowering and lifting each individual wheel, might be possible for your team at this point depending on how the wheels are mounted, or by moving weight around on the robot…Moving your battery might be the easiest option depending on how things are organized.
Lastly make sure your wheels are oriented properly…Looking down on your robot the rollers should make an X.
Hope this helps.
Edoga
It’s almost always better to provide more info than to keep it short.
post some pictures showing how your wheels are mounted.
test the rollers on each wheel. do they all spin freely?
what motors are you using?
what motor controllers are you using?
what gearbox are you using?
what is your gearbox ratio?
chains, belts, or direct drive? sprocket/pulley ratio?
put the bot up on blocks and give it full forward command. record the motor current and motor voltage for each motor for each wheel. also make note of the direction each wheel is spinning. repeat for full reverse, sideways left, and full sideways right commands.
One could get a similar result by allowing a little bit of flex in the frame, depending if that is an option.
We are using 4 2.5" direct driven cims
Talon SR
Gear ratio 12t to 72t(ratio of 6)
For a gear box, we use a single reduction clamping gear box from vex
This image was from inspection week 1. We have plates mounted to the outside to keep the wheels and their shafts stationary. The clamping gear box gives a lot of play with just the unsupported wheels
This looks like the same post from a while back???
https://www.chiefdelphi.com/forums/showpost.php?p=1651718&postcount=1
Your gear reduction is probably too low.
Those vex wheels are great but are very high friction for strafing. In strafing a mecanum wheel is fighting against the other wheels. Not enough torque (low gear redux) means you’re stuck. Not enough wheel force
This is phenomenal insight, thanks for sharing. Never thought to consider friction on the wheels and the gearing of our gearbox as an issue contributing to our lack of ability to strafe well.
Forward and reverse with mecanum does not require the rollers to spin.
In fact, if the rollers are so tight that they cannot spin at all, your forward/reverse will work even better (better traction).
The rollers must spin, however, to do strafing (sideways motion).
If the rollers do not spin freely when axially loaded, your strafing motion will be dramatically compromised.
Note that when strafing, the rollers will be heavily loaded axially. So even if the rollers spin freely when you spin them by hand, there may be excessive friction when they are axially loaded.
Ideally, every roller on a mecanum wheel should be supported with thrust bearings at both ends, so they have low friction when axially loaded. But this is cost prohibitive.
So instead, the rollers on a mecanum wheel should be designed so that when axially loaded, they are loaded against a low-friction surface… like a Teflon washer or some such surface. Rubber rubbing on metal? Not so good.
We are using a similar setup.
The clamping gear box from Vex, 11 to 72 ratio and 4" wheels.
My team’s bot is light this year, likely 100 lbs total.
Our Vex parts are working great!
While your bot is geared slightly faster and may be heavier, I don’t think that is the issue.
The clamping gearbox is designed to clamp onto a 1" x 2" extruded piece of aluminum. The aluminum gives the gearbox its strength. Looking at your picture, I don’t see how your transmission is mounted. According the Vex pictures, use the 1 x 2 rectangle aluminum as your frame. Your wheels will be outside the frame in a west coast style.
Note: Use flanged bearings with the flange against the alum to stay locked in.
I would put your robot an a very flat hard surface and insure each wheel is touching the surface. Then put a bar under the bot to figure out where it balances front to back and then right to left to find the center of balance.
If one wheel is higher or is not supporting enough weight, fix that and test again.
If the wheels are good, put the bot on blocks and test the controller to verify
the wheels are spinning correctly.
Dave
*Just to be clear: It is roller friction which causes problems with strafing motion. In an ideal mecanum (no roller friction etc) the wheels do not fight each other while strafing any more than they fight each other while going forward/reverse: For a given wheel torque, the pushing force (and speed) of the vehicle is the same in the fore/aft and sideways directions.