Drive Train/mechanim wheels

[Cow Bell Solo]

Our team was wondering what you need and how you create a drive train with mechanim wheels?? If anybody does know that would be nice to be worked on for next year.

[Kelly]

There have been lots of threads on this before. Search for mecanum. Also consider holonomic.

[Lil' Lavery]

Read this whitepaper:
http://www.chiefdelphi.com/media/papers/1836

Read these on Mecanums:
http://www.araa.asn.au/acra/acra2002...eter-Tlale.pdf
http://www.usfirst.org/community/frc...t.aspx?id=1112 (scroll down to the Mecanum presentation by Hine)

Read some of these threads:
http://www.chiefdelphi.com/forums/sh...ad.php?t=51321
http://www.chiefdelphi.com/forums/sh...ad.php?t=49625
http://www.chiefdelphi.com/forums/sh...ad.php?t=57282
http://www.chiefdelphi.com/forums/sh...ad.php?t=48637

[tseres]

http://andymark.biz/mecanumwheels.html
is this what you're looking for?

mecanum wheels mount the same as regular wheels, so if something goes horribly wrong you can switch them out. personally, i prefer omni wheels, but mecanums have their benefits.

[Andrew Schuetze]

Lil' Lavery - The third and fourth links are not loading for me.
My question for those experienced with mecanum set ups is can they be used on a rectangular base? I don't see an issue with motion in the X & Y directions and maybe the diagonals might be a bit off, but what happens to a zero degree spin in such a situation??

[Madison]

Originally Posted by Andrew Schuetze

Lil' Lavery - The third and fourth links are not loading for me.
My question for those experienced with mecanum set ups is can they be used on a rectangular base? I don't see an issue with motion in the X & Y directions and maybe the diagonals might be a bit off, but what happens to a zero degree spin in such a situation??

If you build a closed-loop system, you can compensate for the misalignment of the mecanum wheel rollers on a rectangular base. We built an open-loop system and the error created in diagonal movement because of our rectangular base was insignificant compared to the overall error due to variances in motor operation and gearbox efficiency at each wheel.

[dtengineering]

Closing the loop... either with encoders on each wheel, or a gyro, or some other form of control... is necessary to enable the robot to travel in a straight line.

A good mecanum drive is much more challenging than a good tank drive. Read carefully the many threads that have discussed the topic, and then go for it. Very few people have ever seen mecanum wheels in action... and they are usually very impressed by what they can do.

Jason

[AdamHeard]

Originally Posted by dtengineering

Closing the loop... either with encoders on each wheel, or a gyro, or some other form of control... is necessary to enable the robot to travel in a straight line.

A good mecanum drive is much more challenging than a good tank drive. Read carefully the many threads that have discussed the topic, and then go for it. Very few people have ever seen mecanum wheels in action... and they are usually very impressed by what they can do.

Jason

Exactly... they is a huge difference between a team that has awesome closed loop control & programming (team 40 in 2006) and a team that just puts mecanums on the robot.

[EricH]

Originally Posted by Andrew Schuetze

My question for those experienced with mecanum set ups is can they be used on a rectangular base? I don't see an issue with motion in the X & Y directions and maybe the diagonals might be a bit off, but what happens to a zero degree spin in such a situation??

Yes. Many teams use them on rectangular, at least that I've seen. X/Y motion, no problem; diagonal, you might want to tweak the angle of the rollers, but standard will work, zero degree spin I don't know about. You could do it, but you might get the four-wheel bounce or a really smooth turn, depending which way the rollers are lined up. You might want to buld a base and test it to make sure you get everything roughly down before you build the real thing.

[tseres]

if anyone needs it, i have omni code that can use a gyro, but doesn;t have too. it's not tested, but mathmatically it works. just thought i'd say that because apparently mecanums are controlled the same as omnis.

[fimmel]

Originally Posted by tseres

if anyone needs it, i have omni code that can use a gyro, but doesn;t have too. it's not tested, but mathmatically it works. just thought i'd say that because apparently mecanums are controlled the same as omnis.

could you either post the code or pm me? thanks

[EricH]

Originally Posted by tseres

if anyone needs it, i have omni code that can use a gyro, but doesn;t have too. it's not tested, but mathmatically it works. just thought i'd say that because apparently mecanums are controlled the same as omnis.

Sorry, but it's not the same.

Mecanum setup:

///.....\\\ or \\\.....///

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

Omni setup:
...---
l.........l
l.........l
l.........l
...---

Translation: It could be adapted, but as written, I don't think it will work. A mecanum system is a little more complicated in terms of motor direction than an omni system.

[Jeff K.]

Forward and backward speed are faster than horizontal speed, because even though the rollers are at 45 degree angles(or whatever you want them to be), they tend to spin more than actually move the robot. Each wheel has to be independently controlled, simplest way is having each wheel powered by its own gearbox and its own motor, although that may not be the lightest. We had a full closed loop program with gyro and encoders. We took out the gyro because it wasn't too reliable. I think by end of competitions though, we also took out the encoders because it was just easier for the drivers, and each transmission also went at relatively the same speed, so it translated fairly uniformly.

Definitely recommend coding before season to play around with it. I think Michael might come upon this thread sometime and post about the code. The code for this and a holonomic platform, with omni wheels at 45 degrees to the sides, are fairly similar as they work in the same way, but the speed of a holonomic platform is faster moving diagonally, but equal forward, back, and side to side.

Worked on for next year, you mean offseason, as maybe next year mecanum might be the worst choice for strategy, depending. We went with mecanum this year and it was an interesting experience. Learned a lot, but realized that we didn't really need the translation features, and 6wd would have been a lot easier.

Hope that helped.

[Alan Anderson]

Originally Posted by EricH

A mecanum system is a little more complicated in terms of motor direction than an omni system.

Mathematically, it turns out that they are the same. I wrote some very simple control software for a mecanum drivebase a couple of years ago. It is exactly like controlling a four-omniwheel drivebase. The relevant factor is the angle at which the small rollers spin freely.

[JesseK]

I too believe the controls for omni vs. mecanum is the same for robot-centric driving. Simply align the omni wheels in an "O" pattern offset from the chassis and you have it.

Your biggest problem in building a mecanum/holonomic drive train will not be manufacturing it and getting the mechanics to work properly. In fact the hardest thing will be keeping the team energized, focused, and motivated when they're creating probably the most complex system they've faced thus far.

Take it in steps. There are several teams I saw (and have photos of) in Atlanta who have a small CIM attached to the 56mm 12:1 Banebots transmissions that go directly into AM Mecanum wheels. Some of these teams won their regioinals, so that tells me it's a simple and effective enough setup. This setup costs ~$1100 (avg. chassis price included), so be prepared.