Mecanum Wheel Setup Help

[Dylan5019]

We decided to go with a mecanum wheel drive system this year. However, when we got everything set up with the Holonomic VI we were able to spin and strafe but not actually move forwards and backwards.

Right now we are converting our X and Y axis to radians with ATAN2 then to degrees and using that for direction, and getting our magnitude by taking sqrt(X^2+Y^2). We wired a straight 0 to rotation. We hoped to see simple translational motion working. It's also possible that the orientation of the mecanum wheels is incorrect. whe have them set so that the rollers in the front are to the left and the rollers in the back are to the right. We tried the "X" pattern and the results were worse then curently


IDK if this is a mecanical or programming error, but hopefully someone will know. If you have experience with mecanum set up you can contact us by text messaging 918-406-8420. Thank You!

[jhersh]

So you got rotation when you had wired 0 into it?

Sounds like on the 4wheel open you need to invert the motors. We inverted the 2 motors on the rights side.

[Alan Anderson]

Originally Posted by Dylan5019

whe have them set so that the rollers in the front are to the left and the rollers in the back are to the right.

You won't be able to drive straight that way. Your robot will want to move the front diagonally in one direction and the rear diagonally in the other direction, making it drive in a tight circle.

The wheels on opposite sides of the robot need to have opposite orientation. The contact between the rollers and the floor needs to be in an "O" configuration. If you have it in an "X", the robot will not be able to spin well (or at all) under its own power, and it will be easy for an external force to spin it.

[Ether]

Originally Posted by Alan Anderson

You won't be able to drive straight that way. Your robot will want to move the front diagonally in one direction and the rear diagonally in the other direction, making it drive in a tight circle.

The wheels on opposite sides of the robot need to have opposite orientation. The contact between the rollers and the floor needs to be in an "O" configuration. If you have it in an "X", the robot will not be able to spin well (or at all) under its own power, and it will be easy for an external force to spin it.

Piggybacking on Alan's post:

Just to be clear, if you view the wheels FROM BELOW (like looking at the bottom of the robot when it's tipped on the side) it should be an "O", as Alan said:

/...........\
/...........\
/...........\
.............
.............
\.........../
\.........../
\.........../


However, if you view the same robot FROM ABOVE, it will appear as an "X" configuration:

\.........../
\.........../
\.........../
.............
.............
/...........\
/...........\
/...........\


Any other arrangement of the wheels will cause problems or compromise performance.


~

[Dylan5019]

Thanks guys. I'm going to try to fix it today, and if I'm still having problems I will come back, but you did a great job of making everything really clear.

[MrForbes]

if you want to drive straight forward, all the wheels need to rotate in the same direction at the same speed. The rollers need to be in opposite directions on the two wheels on each "axle", ie. you need a right and left wheel on each end of the robot.

Check your code and see if it is set up to do that.

[MrForbes]

One more thing to check is how you wired the motors and speed controllers. If you have the same polarity wiring for all the motor connections, then you need to have the left motors have reversed direction compared to the right motors. This is because the motors on one side are mounted to the robot facing the opposite direction than the ones on the other side.

[DirectXMan12]

Hi,
We have our wheels set up according to the omnidirectional.pdf doc @ firstroboticscanada.org. The configuration that they show seems to be opposite of what you have (i.e. we an O on top). However, we are having problems spinning in circles (namely, we can't spin in circles - we can only skate, not turn). Which is the correct config, and could our config be preventing us from turning?

Best Regards,
Solly (Team 1740 Lead programmer)

[jhersh]

Originally Posted by DirectXMan12

Hi,
We have our wheels set up according to the omnidirectional.pdf doc @ firstroboticscanada.org. The configuration that they show seems to be opposite of what you have (i.e. we an O on top). However, we are having problems spinning in circles (namely, we can't spin in circles - we can only skate, not turn). Which is the correct config, and could our config be preventing us from turning?

You must be misunderstanding the document you found. The behavior you described (can't spin) is exactly the expected behavior from putting the wheel on with an X on the bottom. Rotate your tires and try again.

-Joe

[Ether]

Originally Posted by DirectXMan12

Hi,
We have our wheels set up according to the omnidirectional.pdf doc @ firstroboticscanada.org.

The picture shown on page 8 of that document is INCORRECT, and would cause the behavior you are experiencing.

Originally Posted by DirectXMan12

we are having problems spinning in circles (namely, we can't spin in circles - we can only skate, not turn). Which is the correct config, and could our config be preventing us from turning?

The correct configuration is show here:

http://lib.store.yahoo.net/lib/yhst-...lSpecSheet.pdf

If you have mounted your mecanum wheels so that the rollers form a "diamond" (sometimes called an "O") when viewed from the top, then that is wrong, and it would cause the "no spin" behavior you are observing:

http://www.chiefdelphi.com/forums/showpost.php?p=907174

~

[Ian Mackenzie]

I apologize for any confusion from the presentation I posted - the picture is meant to show the pattern of contact between the wheels and the ground, as if you were looking from above but the top half of each wheel was transparent. I discussed this at length when I actually gave the presentation at the FIRST Conference, but perhaps I should have put more detail in the posted version...

UPDATED: I've edited the presentation to try to clarify the mecanum wheel orientation, and I've uploaded the revised version to both http://www.firstroboticscanada.org/s...irectional.pdf and http://www.chiefdelphi.com/media/papers/1836. (If you try to download the new version and it doesn't have an 'Updated' note on the first page, then it's probably the copy from your browser's cache - try using a different browser to get the updated copy.) Thanks to Ether for letting me know about this!

[mjgard]

So we are trying to figure out which is the correct way to put on the Mecanum wheels. We have done some research and I have attached a link for a research paper on them. This link is the complete opposite from what many of you say and what it shown on the previously attached andymark pdf.

Which is correct and why:

http://www.scipub.org/fulltext/ajas/ajas351831-1835.pdf

or

http://lib.store.yahoo.net/lib/yhst-...lSpecSheet.pdf

[Ether]

Originally Posted by mjgard

So we are trying to figure out which is the correct way to put on the Mecanum wheels. We have done some research and I have attached a link for a research paper on them. This link is the complete opposite from what many of you say and what it shown on the previously attached andymark pdf.

Which is correct and why:

http://www.scipub.org/fulltext/ajas/ajas351831-1835.pdf

or

http://lib.store.yahoo.net/lib/yhst-...lSpecSheet.pdf

Figure 2 and Figure 4 of the first one are incorrect.

The Table and Diagram on the second one are correct.

Explanations are given here:

http://www.chiefdelphi.com/media/papers/download/2826

http://www.chiefdelphi.com/media/papers/download/2722

http://www.chiefdelphi.com/media/papers/download/2739

[jawebste]

Originally Posted by Ether

Figure 2 and Figure 4 of the first one are incorrect.

The Table and Diagram on the second one are correct.

Michael,
I agree with Ether. And Ether you are a great reference for all of us out here. Thanks for posting.
We also found the research confusing and contradicting, but we have had very little problem this year with our first year with Mecanum wheels using the AndyMark diagram. When we numbered and oriented our wheels and the connections to the PWM to match the AndyMark diagram we had no problem getting motion using the Labview Holonmic Drive with Cartesian coordinates. We did have to invert one of the motors, but everything else was great.
This is for Ether, would the Figure 2 be correct if we were looking at it "from the carpet?"
Thanks,
Jean-Ann

[Ether]

Originally Posted by jawebste

This is for Ether, would the Figure 2 be correct if we were looking at it "from the carpet?"
Thanks,
Jean-Ann

Yes and no.

Yes, that is what the rollers should look like "from the carpet", but...

No, the figure must be a "top view" not a "carpet view" because of the way the force vectors are drawn. If the figure were intended to be a "carpet view" then the force vectors would be wrong.

So, if it's a top view then the rollers are shown incorrectly; if it's a carpet-view then the force vectors are drawn incorrectly.