Three Motor Swerve Module

Hello CD,

Over the past month or so, I have begun the process of adapting 1678’s 2021 three motor swerve and SDS’s MK4 Modules into a three motor swerve as a kind of side project for an independent studies class at my school. As we are all aware three motors per module is OVERKILL to say the least (and IMO was only viable in like 3 games 2008,14,21) but I thought it would be a fun project to do this offseason (and in my heart of hearts hope that one day it could be used some day).

First off, in the around 30-40ish minutes of searching on CD I’ve found little to none in the way of documentation for SPECIFICALLY a three motor swerve module. There is the 1678 2021 Skills Competition Submission that encompasses 1678’s small, three motor swerve drive robot. This became the basis of the project. How do I, a largely self taught FRC student, translate a complicated module used by a top team into something that 1. can be understood by me and others (remember this is a school project lol), and 2. that I can feasibly build in a semester of class.

This led into the CAD phase of the project. Starting off, 1678 uses the SDS MK3s + their own custom parts to adapt the module for three motors. This led to “updating” the modules by pretty much plugging the 1678 extra parts onto a MK4 module. This was not a very big feat as the MK3 and MK4 are in fact very similar. I also felt as if I could be a bit more lenient on the spacing/sizing of the motor plates as 1678 had a very small package to fit them into, while I do not. This led to “beefing up” the plates (adding surface area) as well as “sandwich-ing” them (idk the correct term) and connecting via standoffs for extra stability.
|317px;x224px; ^top plates are the original 1678 plates, bottom is the new plates.

image
^right is the CAD orignal of the 1678 module, and left is the “sandwiched(?)” version


^isometric view of the module

After these steps were completed, we come to the part of the project I am currently on, fabrication. Currently our lab has two MakerBot Method X 3D printers that I’m (trying and sometimes failing) to use to print the two 3D printed pulleys that are seated under the original drive double gear and the original steering shaft. I’ve gotten these prints to work successfully as PLA, but I am struggling to get them to not “string” as carbon fiber nylon prints. I’ve managed to get the driving pulley to print out of the CFN but I’m still working on getting the steering pulley done. I have faith it will work after some extruder temperature changes. Additionally we were gifted a second hand Glowforge at the beginning of this school year so I plan to laser the motor plates out of some acrylic (that isn’t dangerous to inhale when done). I do have doubts about if the acrylic is going to be strong enough to withstand the forces in play. If this becomes an issue, I could contact a local CNC-equipped shop that would be willing to machine some motor plates.


^left 3D printed steering pulley in PLA, right 3D printed drive pulley also in PLA

This is all I have at the moment, but I would be happy to keep updating if there is interest, if not no worries, but it might be a while as I’m waiting on some MK4s (hurry up SDS) (only joking of course). I can also post the CAD link if need be. Lmk what y’all think! Any and all input is welcome, have a good one!

-Alex

22 Likes

This looks incredible!!!

3 Likes

Well this explains the carpet shortage.

25 Likes

You may find that you can print plates and get something good enough for form and fit, especially with the filled plastics. Gentle function, maybe. CNC routed polycarb will get you decent levels of functional plates. That’s what I did with Nick Cousin’s Winter Swerve… CNC routed fiberglass will get fully functional, but its kind of icky to route; lots of “don’t even think about breathing this stuff” dust…

So, your goal here is maximizing coolness, right? With a side order of making it easier to break the tires free?

4 Likes

The gods of symmetry are angry! Could you move the second drive motor so it’s equidistant between the first drive motor and the azimuth motor?

If the goal is to make a pro street, funny car swerve, then mission accomplished, otherwise 3 NEO550s would do the trick and still have high coolness factor.

2 Likes

That’s the hope… :eyes:

That’s interesting, I didn’t think about printing the plates, I thought they would break easy, definitely going to see if I can do it now.

And yes, coolness and outright performance is the goal here. Our team is from the PCH district, a somewhat underdeveloped district on the technical side so it might be cool as a quirk of the team doing something that few have done soo…

I thought about moving the secondary drive motor to the middle like 148 did… BUT this was easier… if it turns out as a problem having the motor far away I definitely plan on moving it…

I love the analogy of a funny car swerve, as a fan of drag racing I’m honored, but also at the same time if we can get a performance advantage from this module I’m all for it

1 Like

It is coming along quite nicely :slight_smile:

9 Likes

You might want to buy the mineral rights below the venue just in case you hit oil when those things drill down through the floor.

19 Likes

ITS DONE!! Well V1.0ish that is. Had to modify the gearing a little bit because the main driving pulley was rubbing up against the steering pulley. Re-printed a smaller middle drive pulley so that constituted different belts so on and so forth. Another point of issue would be flex in the plates, but I’m working on a fix for that.

-Alex

5 Likes

I was bored waiting for Champs to start…

So I did this in like an hour-ish…

  • It’s just a modified top plate to the MK4i that allows another Falcon to be slotted in that drives the same gear as the primary drive Falcon.
  • Theres no changes to the ratio so it kinda defeats the purpose of adding more power, but hey it exists now…

Link to CAD:

3 Likes

Love it. I think more and more teams will do something similar in open-field games. A little overkill but doesn’t the i in MK4i stand for incredibly OP.

1 Like

The ground clearance on those Falcons is a bit terrifying to me. For actual game-play, if your robot goes over something it’s not supposed to (cable protector, another robot, field debris, a cone that doesn’t like pushy-pushy, what-have-you), you might end up with some interesting lateral forces on those poor little fellas.

1 Like

Yea… there’s a lot that isn’t optimal about the first iteration of the module. I think using the MK4i for the second iteration should solve this… I’m currently looking to re-CAD the three motor MK4i plate to have all of the motors inverted… but we shall see

Of course, not having your motors inverted means your modules are taller, so they’re in the way of upper mechanical mechanisms…

Ahhh, trade-offs.

1 Like

what about changing the gears to L1 than you can have the L1 speed with a heavy robot plus you don’t need to do a lot more moding

Yeah, that’s probably the most optimal way to run it without a bunch of geometry changes.

Hightide did something like this Upgrade Swerve MK4i Increase acceleration and speed w 2X drive motors - #6 by Pompano

I think 148 made a 3-motor swerve module in 2021

1 Like