FTC - Mecanum Wheels

Hi, I was wondering if there are differences between diverse brands of mecanum wheels. For example Gobilda, Rev, Tetrix or even cheaper ones you can find on Amazon.

Thank you!

Definitely. There are various characteristics of each of these wheels that affect how you can mount them and how smoothly they roll.

For mounting, you’ll find various methods-- hole patterns that match various build systems, or that may not exist (in the case of wheels designed to ride directly on 1/2" hex shaft).

For smoothness, the main factors that affect this are whether the rollers are riding on bushings or ball bearings, and the ratio of rollers to the diameter. In general, you would prefer ball bearings and more rollers relative to the diameter. For example, standard AM 4" mecanums have 6 rollers per 4", and are riding on bushings, whereas the Gobilda mecanum wheels have 9 rollers for a 4" wheel and ride on ball bearings.

These differences can significantly change how you use and interface with the wheels, and in general, teams prefer wheels that roll smoother. Personally, I like the new REV 75mm wheels or the roller bearing nexus clones (Nexus, Gobilda, et cetera). I also have used the Andymark 4" HD wheels, and while they are nice, they probably aren’t worth it for an FTC team when you can get really solid, lighter weight wheels for half the price or less.

1 Like

So, basically our team has a really low budget, so they prefer to get the cheap 60mm Amazon ones (40dlls). Are we going to face any problem with that?

The concerns with any unknown-to-the-FTC-world mecanums are going to come in one of two flavors:

  1. Making or finding a compatible hub to get it to your 6mm D shaft, 5mm hex shaft, or whatever you’re using.
  2. Is it actually a good wheel? (Rolling smoothness, durability, capacity, support, etc.)

If you’re that low-budget, I would give serious thought to skipping the mecanums and going with a traditional skid-steer system using 20:1 gearboxes. I know a lot of teams are going for height in their quick builds, but I think there is a lot of room for a skid-steer robot racing stones back and forth (and, as necessary, into the foundation).

3 Likes

I assume you’re talking about something like these?

If so, I see a few things that are red flags to me. Their load capacity seems like barely enough for an FTC robot, it doesn’t look like they have ball bearings in them, and I am not familiar with the supplier. I think you’d also need to do something custom to get them on a standard shaft for FTC.

Unfortunately it looks like the REV 75mm wheels are out of stock-- they’re just $5 or so more for a lot more trustworthy wheel.

I would take a strong look at Billfred’s suggestion of skipping mecanum altogether, but if you really, really want to do mecanum, you have a few options. You could adapt the wheels above, or you could use either the AndyMark Standard Mecanum wheels or the VEX EDR mecanum wheels. If you’re patient/don’t need them immediately, you might be able to wait until the REV wheels are back in stock (which is probably my recommended option if you can wait that long). In the interim, driving with traction wheels is probably acceptable.

1 Like

I highly recommend against this one. the andymark standard mecanum wheels are not good at all for dealing with load. They are good for intakes, but not much else. This is because the standard mecanums do not have bearings inside them, making their strafing at an FTC weight very inconsistent, and the rollers do tend to break. I highly recommend the gobilda mecanums. they aren’t as cheap at face value, but gobilda offers a 25% discount to all teams, and these mecanum wheels are very high quality, and regarded as the best of the best when it comes to mecanum wheels in FTC.

1 Like

Yeah, honestly, I hate those things, but if your criteria is “really cheap mecanum wheels” and you’re building a light robot… I still wouldn’t prefer them, but they exist.

I second the idea that you can always build something with your current kit (that presumably contains traction and/or omni wheels) and purchase something else later when you have the funds for decent mecanum wheels.

Mecanum seems to be all the rage in FTC, but you can also make a great robot with a west coast or other tank-style drivetrain. Plus, with the potential for lots of traffic in this year’s game, something with decent traction is going to win any pushing battles.

Mecanum has also lent itself especially well to the last three games (Rover Ruckus, FIRST Relic Recovery, Velocity Vortex), which had sprint distances well under half the field (sometimes, barely 1/3 of it) and a benefit to sideways alignment. With the sprint distance 3-5x higher this year, I think teams that don’t approach their drivetrain with fresh eyes will be in for a rude awakening.

2 Likes

Very true. Or, in the case of last year, just park and never move at all. I am glad to finally see an FTC game that has some additional strategy to it and a higher potential for diverse bots.

Why do you see sprint distances as a prime reason to reconsider mecanum? In my mind there are much, much better reasons to rethink it— namely, programming requirement/drivers practice, precision requirements for placement vs delivery, and probability of/susceptibility to defense.

It is likely colored by the joy we felt at being able to outrun most defenders on the FRC field this year. In our case, by throwing more power at the problem and gearing accordingly. It would be a very rare FTC robot to benefit from more than four motors, but pointing all the force vectors in the same direction would have a meaningful effect all the same.

(To be clear, I view the sprint argument as a game-specific kicker over and above the evergreen reasons you listed.)

Mecanum drivetrains can still get quite speedy, even if they aren’t quite as fast as a tank drive. I don’t think the speed of going back and forth between the depot and the foundation will be the defining factor this year, especially at the higher levels.

Let’s assume you moved your foundation in autonomous, as you should, and let’s assume you’re using the alliance bridge.

  • Stones 1-6 (minus any Skystones) are going to be about 5-7 (eyeballed, without the aid of the Pythagorean Theorem) feet away as you grab them. Which isn’t crazy compared to crater-to-lander last year, if you weren’t doing some crazy expanding robot.
  • Stones 7-12 (the opposing alliance’s quarry) are 8-10 feet away. That’s not unheard of, since you had depot claiming last year, but you didn’t cycle that in any of the last three years.
  • Your Loading Zone is about 12-14 feet away from the nearest spot on the foundation, further if you want to hit the far corner. You might’ve seen some of that in Cascade Effect, but I can’t think of it in any more recent game than that.

Will speed alone knock off a RedNek or Gluten Free? Nah. But only two dozen teams a year wind up under those bright lights–thousands more are just trying to advance to a state tournament or get off of the bottom of their league standings. If this discussion gives them enough pause to really think about their resource allocation this season (and maybe that allocation still includes mecanums, but they thought about it rather than just seeing what worked last year), I think it’s a win.

Our prototype drivetrain was speedy enough that we decided that the extra agility with mecanum wheels was beneficial. I definitely agree with you though, just blindly choosing mecanum because it worked last year would be a mistake. And when I said higher levels, I was referring to state comps lol. Unless your robot is very slow, I don’t think it will be to much of a problem.

Autonomous has been the main reason most of my teams have stuck with a tank drive. It’s a lot easier to program for accurate location and there are always a ton of points tied up in autonomous. In our region, it can win you the state championship or at least get you close every year. If you advance past that then all bets are off, but it’s a good starting point.

One of our teams is going mecanum this year so I am curious about how they will deal with it.

Check out ACME Robotics’s Road Runner.

Thanks, I will pass that along to the team.

I’m not saying that it’s impossible to make an accurate and fast mecanum drive, just that the level of sophistication is not as high for a tank drive – your use you drive train wheels and maybe a gyro for your odometry and any kid who has programmed an FLL robot can make that work accurately without much training because that is how FLL teams program robots. There have definitely been advantages for mecanum bots in the last couple of games, though, especially in teleOp.

I won’t deny that it’s harder to make a good mecanum autonomous, but I’m of the opinion that “harder to program” is not a good reason to not do something that has potential benefits. Then again, some of the stuff I do in robotics is basically masochistic so :man_shrugging:. However, there are lots of fantastic tools and libraries out there that can do most of the heavy lifting in the code, programmers just need to find them and use them. No point in reinventing the wheel.

Building robots is an exercise in resource allocation. You are free to prioritize differently from others in this thread.