Just an quick reality check but using 2 drive wheels in the back and 2 casters in the front is full legal corrected? Looked over the rules and it seems fine, but just want to make 100% sure.
if they fall within the list of allowable parts (all tetrix wheels or made out of the allowed materials) then you should not have a problem
Using caster wheels is legal. However I suggest if you can, invest in some omni wheels if you want a 2 wd, 4 wheel design. In 2008, our rookie year we had a horrible experience with caster wheels. If you are using the kit base you could easily add the Omni wheels.
I believe this advice is intended for the minibot, while the question (I assume) refers to the robot, or hostbot, or whatever the big machine is called.
Casters are certainly legal, but you may wish to consider omniwheels instead. I’m sure if you search CD (or google) for discussions of casters and omniwheels you will find that omnis are often a preferred solution for a number of very good reasons.
Jason
P.S. If you are doing a four wheeled robot with two wheel drive, remember to keep as much weight as possible over the two driven wheels.
We also had a bad experience with caster wheels in 2008. Won’t ever use them again. I too would suggest omni wheels.
Mk,
Casters are perfectly acceptable but have drawbacks. When moving from a forward to reverse direction the cater must rotate about it’s vertical axis which requires some side to side motion to accomplish. If the robot is in a position that it cannot move sideways, the casters will actually fight your commands. Other robots can direct your robot actions by pushing on the side of the robot where the casters are located. Omni wheels react the same way.
On tilted surfaces, they have a mind of their own. While they may work this year, in a year where climbing, ramps, or bumps are part of the game, casters are by far from the best choice.
We have a few rules on team 829.
Drive train rule number 1. No casters.
816 has the same rule.
MK.32, have you looked into using a simple omni-directional wheel in place of a caster? Here’s a link: http://www.andymark.com/searchresults.asp?cat=69
A pair of Omni wheels should give you similar turning ability but they will be much more predictable and have a slightly higher resistance to pushing than a caster, and you can power them if you wanted to.
From experience, casters are flat out terrible for FRC.
Omni wheels are the way to go for what you want to do.
-Nick
Slippery skid plates might even be preferable to caster wheels. At least they don’t put up a fight when you want to change direction.
A well designed ball caster might vaguely be acceptable, but not really. Casters are NOT your friend.
Also, if you choose to ignore the general advice given here about not using casters, put your drive wheels in front if you at all can – it makes it a lot easier to line up on things properly if the front of the robot isn’t pivoting around the back.
Thanks for feedback.
We will be switching to something else for sure.
Are the plastic omni-directional wheel fine or are the steel ones that much better?
Either one should work.
Also, consider adding a chain run to power the omniwheels. You get more forward power, without affecting sideways motion of the omnis.
Not to hijack this thread, but…
Our team was planning on using ball casters with CWD. We have a drive platform and its incredibly maneuverable.
3 years ago, we used omniwheels on a 4-wheeled robot for slippage, and it worked pretty well. However, it was a huge pain when they broke, considering how expensive they were.
Apparently, (before my time, and there’s not a whole lot of recollection about them, which could be good or bad) about 5-6 years ago we used a similar design to now. It was claimed to have great maneuverability, but if you dropped the robot or anything, the casters could dent, which would prevent them from rolling correctly. However, they’re cheap enough where you can easily have extras on hand.
This year, when we designed the robot, we chose ball casters based mostly on these experiences. We assume that omnis would not provide any significant prevention from being pushed and rotated around our center. We were also being cost-conscious, being a pretty low-budget team. Plastic omnis would have been about 4x as much as a heavy-duty ball casters, which are also readily available at a hardware store.
Since we’re using direct drive, powering omnis wouldn’t be worth the hassle, either.
Basically, I want to make sure that our decision was sound. Based on our pretty limited test, it seems to work pretty well, but that means nothing in FIRST 
Has anyone had any trouble with ball casters?
If you must use a caster, ball casters are about the best to use. As long as they aren’t acting like swivel casters, that is.
I believe the AM omnis have had a strengthening redesign since about 3 years ago, too.
Use AM Plastic omnis if possible. Coolie Dualies are wheels I’ve never seen used successfully, unfortunately.
What about ball caster, how well do they work?
If they don’t really work i think we probably just go with a 6 wheel drive system.
They can work. They’re better than swivel casters. That just isn’t saying much, though.
Honestly, skid pads would work quite well to replace casters, if it weren’t for the bumps in the floor. (Though that could be dealt with–see if you can get a look at 1266’s rookie robot.)
FYI, I don’t know of a single problem with 330’s omnis last year, and they took a heavy pounding. AM plastics, I want to say 8", dual configuration.
The trick with omni wheels is to put two wheels per wheel, placed so that the rollers on one wheel fill in the between-roller gaps on the other wheel (the dual configuration I mentioned above). It provides a smoother ride, reducing wear and tear on the wheel overall.
Why would skid pads work better than ball casters? They have much more friction. Do you have experience with either you’re basing this on?
I’m also thoroughly confused…
Also, not to be a downer…it is week…4(?) now. You might want to get a move on things if you are still deciding on your drivetrain this far into build.