Creating a Tiny Three-Motor Swerve Drive

TL;DR: New swerve design! Download CAD here: https://grabcad.com/library/super-compact-three-motor-swerve-drive-1

After seeing the extraordinary success of a certain teal-colored team with a three-motor swerve drive, I decided to try my own hand at it. My goal for the project was to make a 3-motor swerve which fits in roughly the same profile as the SDS MK4i. To do this, I switched to using a NEO 550 for the pivot, and then squeezed the 550 in between the two drive motors.

The module is not too bad for weight - SOLIDWORKS gives it a mass of 5.85lb, which is actually not much heavier than my last attempt at a swerve.

I couldn’t really find any issues with the SDS swerve design below the pivot bearing, so I just copied that more or less wholesale. I suppose you can pretend that it reduces the number of bespoke parts on the system.

The packaging on this design is pretty tight - “clearance is clearance” was king when I did the layout. If I made this in practice, I would do a little bit of tolerance analysis to make sure that things don’t interfere. However, I’m not the one building this, so I’ve decided it’s not my problem :>.

Here’s the master sketch I used when designing this module. I tried out using the master sketch workflow (instead of multibody) for this project and I liked it - I may be a new convert.

Download CAD here: https://grabcad.com/library/super-compact-three-motor-swerve-drive-1

49 Likes

I too absolutely hate my batteries

54 Likes

Looks good! This is one of the designs that I hadn’t seen before. It took me a moment to realize that the neo 550 is used for turning (im too tired to think rn).

2 Likes

But do you like blue banners? :blue_square:


I appreciate this. It’s a lot more accessible for teams to modify a few parts and test than to do an entire redesign.

8 Likes

As a result of some of the issue this year with the flat head screw in the intermediate shaft backing out, it may be worth looking into using the new SDS Intermediate Shaft V2 with a e-clip instead. Otherwise, I love the design, great work!

Edit:
Added Link: Shaft, Intermediate, V2 (MK4/4i) – Swerve Drive Specialties

3 Likes

credit where credit’s due, the use of a neo 550 as the pivot motor is pretty smart. I always thought the use of a full-size brushless motor as the pivot on SDS modules was a little excessive.

7 Likes

[boomer]
I miss the Spotlight feature.
[/boomer]

11 Likes

I believe the main reason for SDS and WCP using a full size motor for azimuth was to keep motor plates symmetrical.

2 Likes

Whats the free speed of this thing?

Rev and just about everyone prior to SDS modules rn looking at this take.

27 Likes

I really like it, but you just went from occupying 8 ports on your PDP to 12. That’s a whole lotta drive motors…

That may be part of it, but I’d think it also has a lot to do with designing the module to work with “the only motor you’ll ever need”.

1 Like

The SDS MK2 module was designed well before the Falcon 500’s release.

Fair.

Very cool looking!!!

Looks like the overall drive ratio is 6.9:1, which is slower than the SDS L2. I guess I’m trying to figure out the point if it’s not going to be faster. Did I misread the ratios?

For the second stage could you swap the 24T to be the driving gear and the 20T to be the driven gear?

2 Likes

It’s a bit of that - though it also has a lot to do with not requiring a team to mix motor controller ecosystems since there’s only (1) Brushless Option Available in the Sub-CIM class of motors and it requires a Spark Max, which itself only works with (as of current) NEOs. In theory, making the Modules CIM Class Motor Compatible means that teams are free to run a bunch of different options and/or dual Falcons which eliminate the controller entirely.

Back on topic, this module is super cool since it basically hits the “All of the Drive Power” objective without taking up a bunch of extra space.

The cost difference from a neo and a neo 550 is only 18 bucks though :frowning: but you do save space, and the high speed can be useful.

Would be nice if the motor controllers were cheaper.

1 Like

How would a bag or 775 motor work instead of a CIM, just looking at documentation the 775 is already used as an azimuth, and the Versaplanetary system can adapt it to a CIM. Both can use brushed controllers, would this not work due to spacing or something?

1 Like

I’d like if there was a brushless motor with integrated controller with the form factor of a 550 or 775 and isn’t underpowered like the Dynamo. Falcons can be very convenient because of their integrated controllers but there are a few situations where the power they offer is excessive like intakes or indexers.

3 Likes

I don’t get all the hype around it, yea high tide had a three-motor module but it came with a lot of disadvantages, keep in mind that this year the number of subsystems you needed was quite low vs recent years, and adding another 4 motors to the already 8 motors that a swerve needs is just ludicrous in my opinion. and that’s just the PDH slots we also have the problem of how you supply that amount of power to those modules.
that what are the solutions?

  1. better motors use falcons instead of NEOs and use FOC to have more power from them.
  2. create a differential swerve, although it’s tough, especially the control aspect of it gives you the ability to have “two motors that contribute to the velocity”.
  3. don’t do anything focus on another aspect that can help you be faster on the field for example have a lighter robot.
1 Like