We’re currently looking at trying out swerve drive for the first time this summer. We are looking at getting a cnc and one of our coaches has a 3d printer, what would be a good way of manufacturing it ourselves.
1 Like
Here are some resources that cover the development of a couple other successful swerve drives:
1640’s wiki:
https://team1640.com/wiki/index.php/Swerve_Central
2767’s swerve design document:
https://www.chiefdelphi.com/uploads/default/original/3X/4/e/4ed7bb78a109bbcb9e93d01df5f3258228d26e93.pdf
1 Like
What is your reasoning behind not using sds?
4 Likes
The cost, a full kit is around 1500+, and we have other things we are trying to take care of. BOM limit is still like 4k I think.
The CAD for our module design is public.
In 2019 we have replaced the miniCIMs shown in the CAD with NEOs and in 2020 we replaced the RS550 steering motor with a miNEO and reduced the VP-Lite gearbox from 2 stages to 1 stage. In 2022, we replaced the VP drive gearbox with a Rev Maxplanetary.
We also iterated off of this design very slightly in 2020 and again in 2022 to change how the absolute encoder is installed. If you think you might want to develop a design using ours as a starting point and want more details on our current absolute encoder design, let me know.
We also have an earlier dead axle version that would be easy to adapt to treaded wheels (such as the SDS wheels).
I’m not sure if the CNC router can machine the main module gear (we use waterjet or laser cut parts). But, the CNC should be good for everything else. You could probably adapt this design to use a belt if you changed the gear to a pulley profile, added a secondary pulley plate with a bottom flange or 3D printed a pulley to attach to the main module gear plate (without a gear profile) and then changed the steering gear to a pulley.
The BOM is no longer required.
4 Likes
So there’s no limit on build cost anymore?
Nope. Unless something changes for next year.
There’s an individual item limit
1 Like
R301:
No individual, non-KOP item or software shall have a Fair Market Value (FMV) that exceeds $600 USD. The total cost of COMPONENTS purchased in bulk may exceed $600 USD as long as the cost of an individual COMPONENT does not exceed $600 USD.
Our team developed swerve drive with the same resources you describe. There’s a build thread covering the entire development process here: Reasonable Swerve by Team 7561
1 Like
A lesson we learned the hard way: the CODE is as much or more work as the hardware. If this is your first try, you should bias your selection toward something where you can get your hands on working code to go with it…
For a summer learning experience, don’t be afraid to swap milled/routed Aluminum parts to polycarb or similar plastic. Print most of your stuff; ideally using a filled fiber if your printer can handle it (hardened steel or ruby nozzle). Once you have it functioning you can think about swapping plastic to metal parts as you build time driving it.
I believe all the COTS models have publicly available CAD if you’re looking for more references.
There is no BOM.
An important rule that still exists is R302 for this year, “FABRICATED ITEMS created before Kickoff are not permitted.” For teams making their own swerve modules, the parts of those modules that are not COTS must be re-manufactured after kickoff each season. This rule might be enough to tip the balance toward COTS swerve modules vs. custom modules for some teams that would otherwise consider the custom route.
2 Likes
I wonder where in the “ship of Theseus” spectrum a team-manufactured replacement part identical to a COTS part would fall?
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.