My Take on a Swerve

After receiving a drawingfrom AM on how to mesh their bevel gears, I got to work on a swerve. This swerve is greatly inspired by 1323 and 2767. I wanted to create a swerve that was modular, and could be stuck onto the corner of any Versaframe. If there are any questions on the parts I used, feel free to ask below.

FRC 226 Swerve Renders

There are 2 things I really like about this swerve drive.

  1. I can easily visualize how this attaches to the versaframe structure we currently use.

  2. The layout seems to be spread out and not as tall as other swerve drives, which I really like.

Two quick questions: What speed (adjusted) is this geared for and what are the dimensions. This looks like a very clean design and I like how it does not take up much vertical space, though I am thinking in terms of fitting electronics into the base of the robot if there is minimal space for it elsewhere.

Looks nice! You’re gonna need some feedback on your module rotation - I’d suggest an incremental encoder to keep track of positioning and a Hall Effect or Tachometer for a zero position. It also doesn’t look like you’ve got too much reduction on your module rotation. I usually run an initial 12:100 32dp stage off a single 775pro and then a secondary 11:72 stage to the rotation module itself - we’ve had success with that this offseason, but I wouldn’t run a lower reduction than that. You really don’t need anything much faster than the 60rpm (1 rps) range, I just wanted to fit the reduction in two stages.

I also wouldn’t trust the single 1.125" OD bearing up top to rigidly support your module - try sourcing a thin bearing online, we found the 95mm OD / 75mm ID slim bearings we used this offseason on Ebay for like $7 a piece iirc.

The pocketing is really nice, and the height profile is ideal in a situation where it’s needed. We ran a flipped motor config on 1323 this year for height thinking we’d need every single inch for ball storage, but in the end I think we would have been just fine with the motor up top - it helps decrease the area the module takes up, which this model could definitely use. Personally I think the dual wheel isn’t as beneficial in this scenario, and switching to a single wheel with a smaller rotation module would help greatly reduce size and weight with little loss in performance. Either way is fine, as evidenced by 2767 this year - this is just a personal preference.

Keep up the great work!

The adjusted speed is approximately 13 fps. The top and bottom plates outer edge is a 9" arc.

Just out of curiosity, do you have an estimated weight?

I thoroughly enjoy this design especially for its extremely low profile. I definitely see the Stryke Force inspiration, and the mounting layout looks really nice.

Though it is modular, it seems like it would be a pain to drill out and rivet the replacement assembly in, but I guess with enough practice (and a few quality drill bits) it wouldn’t be too much more hassle than anything else.

I love how the rotation motor is set up. It seems to save a lot of vertical space for sure.

The drive encoder seems like a last resort kind of thing, I don’t see why it wouldn’t work, though I would probably use some locktite on those bolts.

Great stuff.

Very clean,

This is probably my favorite one I’ve seen this year

This is a repeat warning so only read if you have not seen my previous posts on this subject.

Crab/swerve/omni drives are not warranted for every FRC game. If this year’s game is such that a drive of this type would be a distinct advantage, then rethink your design. In addition to the other drawbacks (weight, volume and software) this type of drive needs an extreme amount of practice to get it right. Everything about driving with crab will need to be learned because it is counter intuitive to any experience a student driver has ever learned. In our experience, a driver needs to almost be able to drive with their eyes closed. Driver training will be similar to learning a musical instrument in 6 weeks. Plan on at least an hour of driver practice everyday.

Not yet, I’ll update as soon as I do.

This is not a crab, but I do agree with the rest of this. We have done swerve twice in the past, and we did find the need to practice more than with a tank drive. This offseason we refurbished our 2008 swerve, and that turned out to be a lot more work than expected.

Thanks so much for taking the time to respond. I’m looking into using Silverthinbearing to keep my module in place. I’ll also probably reCAD this with another stage for my module rotation. Also for a sensor for module rotation, is it okay if I just use a mag encoder on the shaft connecting a pulley to the 775? Otherwise thanks a lot.

VERY nice design. The only problem I see is that your motors would be sticking out of the bottom of the bellypan. Other than that, I would have to say that this is probably one of the slickest designs after our 2012 one where we used a CIM as an axle :stuck_out_tongue:

That works for relative increments in position, but you’ll still need a zero position on the modules. I suggest using a WCP Hall Effect Sensor or a CTRE Talon Tach for designating a position on the module rotation as zero.

EDIT: Alternatively, if you’re using Talons, you can use a single Dual Encoder from WCP to accomplish the same thing as the incremental encoder + Hall/Tach combo.