Team 4159 Dual Cim Two Speed Swerve
 

This is the fourth iteration of our swerve module that we have designed, and the first one that we are planning to build and test during the off season.

Specs:
(JVN calculator’s) theoretical speed: (with 1 CIM + 1 miniCIM)
4.93 feet per second low, 14.78 feet per second high.
with a simple gear swap, we can get 6.08 feet per second low, 18.25 feet per second high.
Wheel Diameter:
2.5 inch colson default, can swap for a 3 inch colson.
Size:
5.125 in x 5.125 in x 12.6 in
2.25 in ground clearance
Weight:
approximately 10-11lbs
Steering:
bag motor versaplanetary at 1-50 ratio plus 30-60 gear reduction for a total of 1-100 reduction for a theoretical 90 degree turn speed of .12 seconds (courtesy of JVN Calculator)
Encoders:
-Steering: US Digital MA3 absolute encoder from Andymark, geared 1-1 to wheel rotation
-Wheel: US Digital S4 incremental encoder, currently randomly geared on the side to make life hell for our programmers.


We thank the internet for inspiration:
Andymark’s Swerve and Steer,
Team 221’s Wild Swerve and Revolution Pro,
Every Swerve on Chief Delphi,

Re: Team 4159 Dual Cim Two Speed Swerve
 

As a programmer, yeah, I know that feel

Re: Team 4159 Dual Cim Two Speed Swerve
 

Looks really good, and that weight is a very healthy one.
Why are you using standoffs to attach teh palte to the 2x1?
Why dog shifter over ball shifter?

Re: Team 4159 Dual Cim Two Speed Swerve
 

We're using standoffs/spacers to attach to the 2x1 frame because
- the steering gear goes between the spacers on top of the 2x1 below the bottom gearbox plate. You can see it on the third render, bottom right corner.
- It gives us 2.25'' ground clearance instead of 3.25''
-It allows us to put nuts between the bottom gearbox plate and the 2x1, which lets us remove the top gearbox plate independently for gearbox repair

We're using dog shifter simply because we have some parts already in stock, and because we haven't discovered a definite answer on CD about whether ball or dog shifters are better. Aaron has tried another version using ball shifter, but only saved 0.5'' in height, which we agreed was not worth buying new parts.

Re: Team 4159 Dual Cim Two Speed Swerve
 

Please tell me we can print red and blue spacers and switch them for different colored alliances. Please :)

Re: Team 4159 Dual Cim Two Speed Swerve
 

Is your steering encoder gear aluminium? If you have access to a 3D-Printer you could make it Nylon or ABS (not that it saves much weight).

Re: Team 4159 Dual Cim Two Speed Swerve
 

Looks good.

Two biggest things:
1. Can you make the side plate modules thicker than 1/8"? I'd be concerned about how rigid those plates would be relative to each other, as any flex is going to lead to the bevel gears misaligning, and dramatically increasing tooth load as a result. It's not fun.

2. I think you could probably incorporate the 2x1 tubing into the gearbox design. It seems a bit weird to me that you have a gearbox full of 2" standoffs mounted right above a 2" tall tube. With some design tweaks, I bet you could use the 2" tube itself as the gearbox standoff, which will cut down on part count and weight (particularly bolt weight).

Re: Team 4159 Dual Cim Two Speed Swerve
 

The absolute encoder gear is aluminum from sdp-si, the incremental encoder gear is steel from andymark. I'd rather not use steel, but the andymark one is convenient.
I suppose we can 3D-print gears if we tried, but as you said, it's not really that much lighter considering they're 1/8'' thick. Besides, the sdp-si gears we're using have something called clamp type hub, and it looks nice and easy for encoders. You can kind of see it on the 5th render.

So this is a common problem? Thicker plates it is. 3/16'' maybe. What would you suggest?
If that's a real concern, we could also extend the 3/8'' shaft and put retaining rings on both sides, so the shaft becomes structural and helps the plate out.

:yikes: hmmmm...
I guess this would be a alternative trade-off we would make during competition season: less weight and simplicity versus more machining and less ease of gearbox repair.
Thanks for the idea! :) We didn't even realize that the spacers are almost exactly 2'' long, since we let them grow with the gearbox.:p

Only if you're willing to switch them every other match, Dylan.;)
Besides, most likely the spacers will be black/white delrin/nylon/aluminum standoffs. The spacers are taking more of a load this time around. The spacers are just blue on the CAD to differentiate them from all the other white/black parts.

Re: Team 4159 Dual Cim Two Speed Swerve
 

It happens sometimes. It's one of the mechanical "gotchas" that makes swerve way harder than it looks; along with the coaxial shaft bearing support, path planning, etc.

My gut (note: my gut isn't a substitute for actual engineering) says 3/16" would be fine. I believe that is what the Robonauts Revolution module uses among others. It helps that you've added standoffs between your two reductions; you're clearly thinking about this sort of thing.

I don't think switching to retaining rings on both sides of the 3/8 shaft would make it much more rigid. You can't really use a live axle structurally without loading up your bearings. That said, it's probably more rigid than not doing it, but also probably not worth the effort.

The easiest way to add in rigidity for "free" is to make the output shaft a dead axle, but obviously with very small wheels a dead axle isn't really an option.

Yeah, this certainly isn't something you HAVE to do, and it would probably require a lot of rearrangement to make it work, but it would result in a very rigid, solid gearbox for even less weight than your current setup. You could use a few standoffs on the other two sides to hold the gearbox loosely together on the shelf, then clamp it down around (and bolt through) the tube when you're mounting it.

Another thing: if you tuck your steering gear totally underneath the 2x1 and have the VP output drop down from above, you can use a steering gear much larger than your module and it can overlap the 2x1 tube somewhat, maybe letting you push the module a smidge further out the frame. This is getting a bit ridiculous though, and isn't worth a whole redesign.