4146 Swerve CAD

This is a pet project I’ve been working on. Fully independent, infinitely rotating, two speed swerve modules complete with pneumatics, encoders, and custom pocketing. The gearbox is the guts of a BallShifter repackaged into the frame, and the miter gears are my interpretation of the data given by McMaster for 6843K12. Wheels are the new 3.25" DT Versa Wheels from VexPro. The thrust bearing is actually a tapered roller bearing also supplied by McMaster. Please post comments, questions, concerns, or criticisms. I’ll try to respond to them as soon as possible. Thanks



I like the design, but I have a consern and and recommendation.
1st: have you made a BOM For the design? I think you might find it would be very expensive to run four of these modules on a robot.
2nd: I would recommend using smaller miter gears, and trying to optimize the size of the lower wheel module. I would try to get the outside diameter closer to that of your wheel.
Another way to do this would be to this would be to get rid of the thrust bearing. Many people would tell you that it’s necessary on the miter, but I didn’t use one a swerve drive i used this off-season. It Didn’t cause any issues.

Overall, seems like a nice design. A suggestion and a couple of questions:

The keyed miter gears that you’re using are very expensive relative to those with an unfinished bore. I suggest that you check out the gears that I used in our design, from SDP-SI (Part A 1C 4-Y20020), which only cost $11 each. Instead of using a key or a set screw, we’re attaching the gears to the shaft with a taper pin (http://www.mcmaster.com/#taper-dowel-pins/=pzpvtu.

How’s the yellow box with the steering sensor made? What’s the sensor and how’s it attached to the shaft?

I’d just use these: http://www.wcproducts.net/217-3344/

They’re already broached in 3/8" hex.

The support for your steering shaft is not going to cut it.

For reference, we did two bearings 1.5" apart (and 2" apart on another rev) that were 25x37x7mm.

Some teams do a large ring at the base of the module.

In addition to the need for more bearing support on the steering shaft, the bottoms of the CIM and BAG motors are awfully close to the ground. Remember that the wire leads of the CIM come out of the back, effectively decreasing your ground clearance further. You don’t want to snag the CIM motor leads on the carpet or something else on the field and have it rip out the wires.

I would suggest when you add a second bearing support to the swerve module steering that you do so by increasing the height of the module slightly to get additional ground clearance for the motors.

Wow lots of good criticism and questions! Ill try to cover everything.

  1. I just redesigned with the WCP/VexPro Miter gears (see attachment) which cut a quarter inch out of the width and length of the footprint of the bottom module. I initially hesitated to do this because it required me to make custom shafts, but it worked out in the end.
  2. I have done a BOM, and yes it is expensive. It comes out to about $600 per module, not including custom parts and hardware. We have a sponsor who would be willing to machine the parts for us though.
  3. The yellow box is an encoder cover, presumably 3d printed. However it is not necessary, just a luxury. The encoders I have designed for are the US Digital S4s, and they are attached to the shaft via the VexPro Ball Shifter Encoder Hardware Kit from VexPro, which includes the gears.
  4. The ground clearance is currently ~3/8", which I deemed satisfactory for my design purposes. However, it could be increased very easily by increasing the thickness of one of the black spacers above the wheel housing.
  5. Adam, are you talking about up-down/radial support or lateral/angular support? Because if you are referring to up-down/radial, it is much more than satisfactory I assure you. Here is the link for the taper bearing Taper Bearing. If you are referring to lateral/angular, how would you recommend I do that? Another plate positioned above the current one housing another bearing?

Thanks for all the criticism, compliments and questions!

Lateral and angular as you say.

Another plate is a good option, check out the CAD files in my signature if you want to see how we did it. 1717 does the same as us (but with 30mm bearings) in a really similar dual plate setup.

Custom machined shafts aren’t necessary if you are using gears with a hex bore. Take a look at some of the products vexpro offers, in particular, these shafts and the hex-bore bearings listed at the top of this page.

I would echo the point that the bearing supporting the wheel module wont cut it. If you are familiar with free body diagrams, draw one (for the wheel module) in the state where the robot is pushing against a wall to see why. Most teams with successful coaxial swerve drives have two bearings to support the wheel module to handle the load.

Another point that hasn’t been mentioned yet: For the CIM powering the drive wheel, you have two chain stages, both appear to be geared 1:1. If you use the chain stages of the transmission to provide more reduction, you will have less stress on other components. A big sprocket on the wheel and a small sprocket driving it will make the bevel gears see less stress, and decrease the likelihood of those parts failing. If you put a reduction in both of your chain stages, you may even be able to remove a gear stage from the CIM gearbox (in most cases, the fewer stages a gearbox has, the more efficient it is).

After lots of hard work, taking into consideration all the things said here on Chief, I’ve come up with this. After adding the second plate to support the shaft, I flipped the BAG and VersaPlanetary 180 degrees to give more ground clearance. I also added a 10 tooth 25 sprocket underneath the top plate in tandum with the VexPro Ball Shifter encoder gears to provide more accurate position control. There is one other option I’ve thought of for mounting the BAG and VersaPlanetary Gearbox. I could get rid one of the planetary stages, and add some gears between the plates to provide the desired reduction for steering the module. Is that a good idea? Is the extra weight worth the reduction in height? (in this situation I would flip the BAG back down to its original position) Please post any comments, questions, compliments, concerns or criticisms; I will respond as soon as possible.

Swerve 2.0.png

Swerve 2.0.png