Possible drive train 2018

Just got finished with this design with a drop down omni wheel in the back.(the outside plate is set too transparent). Any advice on how to improve it?https://uploads.tapatalk-cdn.com/20171119/6e915903266b680ded6d8715b197f09b.jpg

Sent from my SM-G950U using Tapatalk

I guess the first question I (and probably many others) have is why?

Yes this is a possible drivetrain for 2018 and could work but I’m not sure why you have this drop down omni, so many other wheels, and your general aim in this design.

It’s hard to advise on how to improve a relatively new design when we aren’t sure what you are trying to accomplish…

The idea is that when the omni is up then we can’t get pushed, but ounce it is down only the front wheel and omni are touching so that we can monuver the front of the but better for example, placing a gear.

Sent from my SM-G950U using Tapatalk

How are your wheels connected to your gearbox? It looks like the wheels have pulleys on either side of them, any reason for that?

Also, I can’t tell from the picture but from what you said about the omni and the outer wheel being the only ones in contact with the ground and the fact that the outer wheel (non-omni) doesn’t move I would assume that it is lower than the wheels in the middle. If it is, when you’re driving normally two of the three middle wheels wouldn’t even be in contact with the ground.

I think you might have more luck with a more traditional 6 wheel setup like this. It is all cots so it would probably by quicker to setup.

3647 used a setup like this last year. It seemed to work really well for good traction as well as quick turns.

P.S. if anyone on 3647 has the cad model of that robot and would be willing to share it I would love to get a look at it. Thanks.

The wheels are all driven except the omni. They are all level on the ground

Sent from my SM-G950U using Tapatalk

I guess my next questions are the following:

  1. What purpose does having so many wheels serve if they are all level?
  2. Would the actuation of the omni wheel not then tip your entire robot forward to achieve your desired effect?
  3. Assuming the omni is down with no obvious issues, why only power the front pair of wheels?

The idea is that we can push and play defense when the omni is level with the others, but then we would put it down when we needed to presisly line up to somthing

Sent from my SM-G950U using Tapatalk

The idea is that you don’t gain traction by having more wheels. 4 traction wheels (2 per side) plus the omni would have the same effect. Now considering turning scrub, you’ll want to make sure your track length is less than or about equal (ish) to your wheel base. Also consider that having weight on non-driven wheels decreases your tractive force and thus your capacity to play defense (from a purely weight-class perspective.)

I don’t see where this benefits over a 6 wheel drop center. With a 6W drop, you don’t get the slidey 2+2 esque motion, but you also significantly reduce complexity.

I would prototype this in the offseason and see if it offers any measurable advantage, if your drivers can adapt to the switching drive style, and if the complexity is worth it.

We built a 4 wheeled off season robot and only powered the front 2 wheels.
The back wheels had omni’s. It was fun to slide in the turns at speed, but its low speed turning was poor. Would definitely recommend powering the omni wheels.

For more control try adding mecanum wheels to the front.


I’d love to wait for the actual cad model to comment on the actual design.

That being said, your logic in this design doesn’t make sense to me…

Why spend more time adding a mechanical solution to possibly make alignment/driving easier when you can make a kitbot or otherwise simple drivetrain and use that saved time to practice your driving?

I think arguing over whether or not something should be designed is pointless. No one gets hurt when something is designed. The time to think about strategy is when people want it made during build season.

Interesting design, sort of reminds me of 118’s drop down omni in 2014 offseason. I agree that powering it seems like a good idea. Similarly using a box extrusion could be easier than parallel plates depending on your resources.

The drop down omnis we used were just for getting out of t-bones, we didn’t actually use them to move around because we didn’t design them properly so they pushed down too far so the front and middle wheels didn’t have much traction. The quick turns you saw were probably due to our robot being so small and light. (pm’d you a step file for the cad)

Also, we originally got the idea from 118 like asid mentioned.

This isn’t true. Omni wheels generally have (slightly) less of a CoF than traction wheels, so a 4 traction 2 omni drive would have a bit less forward traction than an all-traction wheel setup. Additionally, a 4 traction 2 omni drive would definitely not behave the same as a 2+2 setup - the center of rotation in a 2+2 is about one end, while a 4+2 setup skews closer to the middle.

I don’t know why several posters are attacking the idea of a drop-down omni wheel - this isn’t exactly unheard of in FRC. Top teams have used drop down omni wheels for years as a way to escape T-bone pins and other targeted defense. It’s a reasonable thing to include in a drivetrain design, though it’s not exactly mandatory to be competitive.

Again, just because you don’t understand why a decision was made, doesn’t mean it is inherently wrong or inferior. These kind of posts aren’t really helpful to people looking for design evaluation.

All of that said, I think the OP’s rationale for this drive (easier time lining up) is somewhat limited. Relatively few situations are easier to align with using a 2+2 setup versus a six wheel drive, so I think the utility is limited to specific games where it is needed. Evading defenders is still reason enough to consider something like this, though you have to consider if it is worth the added complexity.

The other thing I would suggest is, if the omni wheels are already not going to be powered, consider using a ball caster instead, or even just a Delrin skid plate. If you attach a ball caster to a cylinder mounted to your frame pointed straight down, it will achieve the same effect as the pivoting omni wheel, but it is much simpler and takes up much less space. This will allow you to push the traction wheels on your drivetrain farther out, making your wheelbase longer during regular match play, making you have more resistance to being spun. Depending on your CG, you could potentially even get away with just one ball caster in the back-middle of the robot, though two would obviously be more stable.

Thanks, I saw you guys at San Diego, Las Vegas, and Beach Blitz and you guys looked great.

We prototyped a similar, bit smaller, system to this prior to this year’s Ruckus (We also drew inspiration from 118). They were an upgrade to our already existing 2017 drive system (versus being incorporated into them from the beginning). A college mentor and two students built and installed them in about 2 days.

We used them to give our robot a bit of drifting ability which, after some practice, sped up our driver’s gear cycle team. Additionally, it’s a lot of fun for the kids to drive and they love practicing with it (and also gets student we’re demoing to hooked pretty quick).

In no way does a system like this not offer any benefit. For sure, it’s different a different style and I was also skeptical at first, but it can be flashy and fun to watch, as well as effective when used properly.

As it stands, it looks like you’re liable to tip backwards when the omni wheel is up. The back-most traction wheel is only a little bit behind the center of the robot, so if the CoG shifts to behind the center or you accelerate fast, you’ll probably end up tilting back and hitting the back of the chassis (or the omni wheel, hard to see from the picture). Same thing goes for pushing matches, where that could even be considered illegal because of the rule against metal feet touching the carpet. I’m not going to say anything against a drop-down omni (enough people have already argued both sides of that issue), but I will suggest that you put another traction wheel behind the omni to make sure you stay balanced.

Also, is there a reason you chose not to direct drive any of the wheels? It appears as if all of the torque is being transferred through one belt, which just seems like you’re asking for trouble. Direct driving would lower the gearbox (therefore lower your CoG) and ensure that at least one wheel is always powered no matter how many belts may come undone.

(P.S. you should probably consider adding a drop center to the center wheel(s) if you want to be able to turn without browning out the robot)