Re: mecanum vs. omni
 

Thank you for the kind comment. I think there are a lot of people who "get it"! And thank goodness... that's what makes FIRST so cool. As for the drive train, I don't really think there is much point in putting a lot of torque to the wheels in either a mecanum or omni system... after all, if you are in a pushing match with one of these drive systems, you are probably losing. So I expect you will see them geared higher... and, as there are four gearboxes, rather than two, you will be less likely to see them have (or need) multi-speed gearboxes. We used 12:1 banebots with a direct drive system. For the first year in four years we NEVER lost a chain! The Andymark mecanums are 8" diameter, so we're moving pretty fast at high speed, and don't have much poop to push with... but when you build omni-drive, you're building a wide receiver, not a linebacker. If you are pushing... you're losing!

As for control, on board feedback and programming is much more crucial, simply because you have an extra degree of freedom. We went with a PID speed control loop set by encoders on each axle... but I know teams have also had success with a gyro. So long as all the wheels are on the ground (almost all of the time on a carpeted floor) we have good pretty good control of what we are doing. So so long as you have good feedback on-board, and an intutive control system, and time to practice... you really don't give much up in the way of control.

There are a few examples on youtube... just search up mecanum or meccanum or mechanum. Wikipedia also has a link to some video from a company that builds mecanum forklifts... which are really rather cool.

Jason

Re: mecanum vs. omni
 

Can any one verify this? (Not that he said it at a meeting but that it is true) I have heard this a few times but Mecanum wheels with 45 degree rollers (Swiss 45 wheels) should have the same speed strafing as driving forward. If this isn't the case could some one explain why (my relatively basic understanding of vectors tells me same speed Forward and back wards as side to side, no different than traditional omnis in the 4 corners at 45* angles.)

James

Re: mecanum vs. omni
 

:cool: jamesbrown
what is ment is 4 wheel drive forward and back as only 2 wheel drive while strafing. As with an omni its 2 wheel drive in all directions.

can1

Re: mecanum vs. omni
 

It depends on how stiff the diagonal rollers are. If you have them nice and loose and freewheeling, a mecanum wheel will respond just like an omniwheel with similarly loose rollers. If you have them tightened down a little so they don't roll quite as easily, you'll get better forward traction but lose power to roller friction in sideways motion.

Re: mecanum vs. omni
 

I've enjoyed using the omni's for the past few years just because they are "cool". Admittedly not as cool as mecanums but still...

To clarify what I said earlier (why I don't like mecanums), each wheel creates a vector of force in the direction that it moves but regardless of what direction the robot moves in it also creates a vector perpendicular to the wheel. When driving forward two of the wheels have this vector to the left and two have it to the right so the net is no lateral vector. When strafing these vectors point all in the same direction and the forward/reverse vectors cancel. so no matter how you build or program you robot, by the nature of mecanums there is extra force on the frame and extra sidewards force on the bearings.

In regard to the gyro. I really like to gyro because our bot would be unusable without it. Since there can be errors in the gyro code we have a way to turn that function off and the machine becomes completely uncontrollable just because of differences in the motors. With the gyro on it is so stable that one of the drive motors can be unplugged but the machine is still very much drivable.

For me I have never even begun to try and delude myself that an omni or mecanum could compete with a tank drive in a pushing match. We tried once in 2005 and got pushed around the field, however, the team that won our regional didn't actually get to win or even score against us simply because they couldn't get around us since we were too maneuverable. Traction has never been a factor in my reasoning. However, I do like the idea of the ability to change a mecanum into something else more easily.

Re: mecanum vs. omni
 

I would love to try a mecanum drive sometime. The only problem I see is that unless there is a huge change in the current game dynamics, the robot might do well in qualifying, but would be completely dominated by a good defensive driver in the finals.

Re: mecanum vs. omni
 

What's missing in hardware can always be fixed in software. :D

Re: mecanum vs. omni
 

This is true (gee... funny how those basic physical concepts like vectors tend to be, eh?) and something we briefly worried about during the design stage this year, but it turned out to be only a minor issue with our mecanum drive. We machined our own hubs to do a direct drive of the AM mecanum off a BB tranny. We ended up putting set screws into each side of the hub, and actually mounted a small allen key right on the robot so that we didn't have to look for one when it was time to adjust the wheels and/or tighten up the set screws that had wiggled loose due, in part, to the changing lateral forces. This only needed to be done a few times each regional. If we were really concerned about it we could have simply pinned or loc-tited the hubs to the shaft and they would have stayed put permanently.

During build we did have to manufacture a couple of spacers to go between the bearings of two of the the BB trannies due to some issues with the location of the shoulder on the drive shaft... the drive shaft actually had about 1/16" of end play... but that would have been necessary with tank drive wheels, as well. After all, although mecanums put lateral forces onto the drive shaft, so do traction wheels when you are pushed from the side.

So yes, mecanums generate lateral forces on the drive axle, but it isn't really that big of a deal. Regular bearings can handle it, and if your axle mounts/drive system can handle the lateral forces imposed by traction wheels being pushed sideways, then your mecanums should not present a real challenge based on our experience.

Jason

P.S. It has just occured to me that traction wheels may actually generate higher peak lateral forces on the drive train (although I know your concern focuses mostly on the cycling of lateral forces whenever the mecanum changes speed as opposed to peak). When the robot is hit... hard... from the side the mecanums will translate some of that impact force into rotational force... or, at the very least... slide and skid... while the traction wheels will direct 100% of the impact as a lateral force (okay... with a bit of it coming in as torque, too).

Re: mecanum vs. omni
 

Sure, that part made sense from the beginning. What I'm trying to find out is why you believe this is any more of an issue with mecanums than with omniwheels.

(It's an issue with regular wheels too, by the way. There's a lot of lateral force involved when a tank-drive robot turns.)

Re: mecanum vs. omni
 

Maybe that would be a very stupid question... but I need to make it, so first of all i will explain the situation

Here in Brazil, no team has used Omni or Mechanum wheels at all... and i must say that my team has an old obsession about using it... but since no team here in brazil has used, is very hard to get them here, and we didn't talk so much about it before, some things are "dark" for us...

I would like to know the diference between Holonomic drive, Omnidrive and as i saw in some posts here... the Kiwi drive...
i mean about the movements of the bot... is there any difference or it's just the way the wheels and the system is designed??

We never heard about it before... for us using these kind of wheels would create just one more way to drive the bot... and now we see that we have 3 at least... we are confused...

Re: mecanum vs. omni
 

The difference between holonomic and kiwi is the number of wheels. A kiwi drive uses three wheels that are set up 120 degrees apart. A holonomic drive uses four wheels set 90 degrees apart. A kiwi drive is easier because of control and the physics of a holonomic but a kiwi inherently unstable because it is a tripod (how many three wheel cars do you see? They tend to be too tippy).There was another team that used a kiwi drive the first year that we used a holonomic but they where very tippy and didn't drive very well.

The reason that side forces aren't a problem with omni is that in an ideal omni wheel you can't have side force since there are rollers in that direction. People use them in tank drives to turn the robot easier because they don't create any lateral force in almost every situation. They only case that creates side force on the wheel is when the side of the wheel contacts something without contacting the roller. I could use 1/16th in acrylic to captivate my wheel because it didn't create side force. That is what I like about them, I only needed one strong piece for mounting my wheels (hub).

Re: mecanum vs. omni
 

Our team has been using Mecanum Wheels for the past three competition seasons. While we do not have the same experience with omni wheels, we have found Mecanums to be very advantageous. For starters, the myth that Mecanums cannot push or defend all depends on the specific style and type mecanum wheel being. In 2005 and 2007, we used a six inch mecanum wheel design, which ment we had to use a small roller. This gave us little contact with the floor at any given time. However, we were fast and maneuverable enough that we did not need to defend and we had very offensive robots. However, in 2006 when we used a larger wheel and large rollers, we were a very defensive bot. At our final competitions we could stop one or ever two robots at a time very easily from scoring in the high goal as well as push other robots when necissary. The amount of friction and pushing power you get from mecanum wheels depends on the size and material of the wheel as well as the roller design. We also found that we could climb the 30 degree rame at the end of the match without any problems.

Re: mecanum vs. omni
 

Tippiness is just a matter of where the center of gravity ends up. A Killough ("kiwi") base is actually a very good holonomic platform because it's a tripod. All three omniwheels are always contacting the ground, so it won't go off in odd directions because of losing traction at one corner like a four-wheeled platform can.
Ah, I think I understand now. When you say "side forces", you're apparently talking about thrust forces on the axle. Agreed, an ideal omniwheel won't try to push its axle in or out. And since you just now mentioned tank drive, I think we might have been talking past each other. With a pair of omniwheels instead of traction wheels on a tank-drive platform, the in/out force on the axle is indeed a side force on the robot. I've been focusing on holonomic drives, and I was tripped up by your mention of extra forces on the frame. Since a traditional four-omni drivebase has the axles at 45-degree angles, it puts just as much side force on the frame as a traditional mecanum drivebase.

Re: mecanum vs. omni
 

It sounds like there is an ideal opportunity for you to introduce some cool new wheels to Brazil. The Omni wheels are great for much more than just an omni directional drive... as mentioned above when you put omni wheels on one end of your traditional "tank" drive robot it becomes much, much easier to turn it.

However in the context of an introduction to the various omni-directional drive systems, you might find http://www.youtube.com/watch?v=r5WKgQJtToM helpful. If you go to http://www.andymark.biz/wheels.html you will see some good photos (and pricing) for some very effective wheel solutions.

There is much more available by searching the forums here and finding the FIRST workshop presentations at the USFIRST website. Hope that helps clear things up, and thanks to MVRT for a great video resource on youtube.

Jason

Re: mecanum vs. omni
 

yeah... that's our plan... but first of all we need to understand it consistently, after we need to convince our sponsor... and then go to the "dirty work"...

It seems that this year we'll finally test some new kind of drive system... and for just one reason it looks like mechanum whells are a little better than omniwheels (for a person that never worked with any of them):

If something goes wrong... it's easy to put a "normal" wheel in one of the end and go for a simple tank drive...

This year we experienced great problems with your drivetrain system... because our bot wasn't able to turn well... so we always lost much time turning... maybe that made us "far away" from greater success...

thanks... the video was very nice... it cleared many things here... now we are studying some topics lil' avery listed here... nice ones, by the way...

Good bye, at least for now :D