Starting out with Swerve Drive.
 

I am a student on team 3620, and now that champs are over, we are interested in getting started with learning swerve drive. The problem is we don't know where to start. Does anyone have any advice or suggestions on getting started with a cost-effective swerve drive? We aren't looking for a competition-ready system yet, we just want a lightweight system that we can put on our practice bot to get familiar with the programming. Thanks for any help or assistance you can offer.

Re: Starting out with Swerve Drive.
 

Andymark sells a swerve drive. You could always look at that for ideas on how to build your own. Or buy one, take it apart and figure out anything you want to change from there. Then build 4 of your own.

Re: Starting out with Swerve Drive.
 

I too have been searching around the net looking for swerve drive systems, I found this (http://youtu.be/VRcOsjI2gkE)

It looks like a nice setup and is very simple (even the gear switching) I am looking into using this for summer projects/experiments aswell

Re: Starting out with Swerve Drive.
 

Oh dear. And I thought ball shifters were bad. Please don't build a shifter that unmeshes and remeshes gears like that video shows. You'll be stripping teeth by match 10!

Re: Starting out with Swerve Drive.
 

I can't speak highly enough for 221 Systems' stuff. They are amazing. We used the revolution pro modules this year and they were great. http://www.team221.com

Re: Starting out with Swerve Drive.
 

I had the same doubts. They later completed the swerve without problems (there's a video somewhere), although I don't know if they went thtough build with it.
Ballshifters work fine for the life we expect of them.

AndyMark sells the Swerve and Steer, which is basically everything you need to mechanically build a swerve drive. They were out of stock of assembled kits and shipped some unassembled, but apparently even if they were assembled you had to take them apart and reassemble due to some manufacturing errors and tolerances. Still, to the teams I talked to at champs, they were very easy to implement once you assembled them (buy a couple extra modules).

A lot of people like the Revolution swerve modules sold by team221. They are pretty light, but require you to set up the steering motors and drive motors to transfer power to the module.

If you go custom, be prepared to manufacture all the parts. Design with cost in mind and it won't cost much, but will probably need to manufacture a couple parts. Look around chief delphi; there have been many designs posted over the years.

Go to team 1640's Swerve Central wiki page to learn about their swerves and stuff. It's a great page with lots of resources.

Re: Starting out with Swerve Drive.
 

We've built some heavy-duty swerve this year, if you want a CAD render PM me!

Re: Starting out with Swerve Drive.
 

Before starting you'll need to understand the machining capabilities of your team and other available local sources. Do you have CNC or manual
Mills. Skilled machinists on both lathes and mills. Welding or Water Jet? Swerve requires some close machining tolerances to pull off. If you have the some of the capabilities then you can design for what your strengths are. If you have limited capabilities, a purchased solution may be the way to go.

Shifting swerve adds considerable complexity and cost that may not be justified. It's also one more thing that can go wrong. Depending on the game and your strategy it might make sense.

Re: Starting out with Swerve Drive.
 

We just finished our fourth year, have run swerve in competition for the last 2 years, and are on our 3rd fully custom design iteration. It is truly a great off season project and I'm glad you're going after it.

In addition to the good hardware- and manufacturing-centric advice so far... don't forget the software. I know we got a lot of help from Bomb Squad getting started on the software and have run with it since then.

We have posted CAD software from (I think) Rev 2 of our swerve on CD. I don't think Rev 3 is posted anywhere yet. We're happy to talk with teams about it. Although I'm posting here I am not thw mentor on our team who's worked with the students on this, and I'd have the students talk to you anyway. PM me if you'd like to talk to our students on the modeling, mechanical, or software side. I've got a similar request from another team and maybe we could coordinate (after AP exams and finals).

Re: Starting out with Swerve Drive.
 

Like I was saying "Experiment" with similar kinds of drive shifting, not exactly what he was doing. Something that would be a better implementation (I didn't really like the gear switching myself) would be nice.
On a side note:
Making a custom swerve drive over the summer would be a fun project, not as fun as just buying it.

Re: Starting out with Swerve Drive.
 

I actually just saw your 4143 swerve drive in Curie and spent some time talking with your students. From my notes you had the six wire rotary switch with the motor/gearbox on the pivot. Nice approach to eliminate the bevel/miter gears. I think the only thing missing was any speed encoder feedback.

Yes, we are all about #teamsliprings, but sincerely wish FIRST would reinstate the mercotac model as legal. We had double, arguably triple, redundancy in the slip rings we used this year, but they were... not elegant. Until we put them in the 3d printed towers to get some clearance we needed.

We also like the simplicity of the straight gearing (including for encoders). The bag motors and vex planetary gearboxes for steering have worked well for us. One hitch in the gearboxes required a custom fix to overcome a gear alignnent/retention problem (we talked to vex, they endorsed the fix and seemed to be familiar with the problem)

I should add, no chain. At least not on the swerve. We had enough elsewhere.

Our "motor-in-wheel" variant is interesting too. I know we had a module out in the pit on display bur did not run that one in competition.

Re: Starting out with Swerve Drive.
 

https://grabcad.com/library/swerve-d...or-frc-robot-1
Look at this one, this was made by my twin brother Tristan, he dedicated 6 months to this and is still make improvements on it (so this isn't the final version, this is also based of Team 16's swerve (which is the godfather to all swerve drive in FIRST) if you have any questions about this file or you don't have access to a 3d CAD program i can post some screenshots, just send me a message.

Re: Starting out with Swerve Drive.
 

There are several places you should look for awesome swerves that were not yet mentioned:

1717-IMO best swerve in FRC; light, well driven 2 speed, I don't know of any failures
http://wiki.team1640.com/index.php?t...rain_Team_Page

1640-excellent documentation. You can look up the design process and I believe CAD models all the way from where they started.
http://wiki.team1640.com/index.php?title=Swerve_Central

3928-the pioneers of cim in wheel swerve
http://www.chiefdelphi.com/forums/sh...d.php?t=104504

2451-continued development on neutrinos swerve
http://www.chiefdelphi.com/media/photos/40298

696-new swerve this year, very elegant IMO
http://2015blog.team696.org/?paged=7

aren hill personal swerve
http://www.chiefdelphi.com/media/photos/41030

I didn't link pics directly but you should be able to find some good stuff based on that.

Re: Starting out with Swerve Drive.
 

Our (Team 2067, Apple PI) swerve drawings and Labview code are posted on this page:

http://www.applepirobotics.org/useful-links/

We have used a similar system the past two seasons, It has been mechanically trouble free. It is based on using the older style 221 swerve modules, with custom gearbox for drive and steering inputs.

It is heavy, which may be a focus of offseason development, but we like the robustness. The COTs components (221 modules and many gears, and motors), have run for 2 seasons of districts (over 160 matches, plus 10's of hours of offseason demos, scrimmages and messing around) with little to no visible wear. (except for the wheels themselves last year)

We have limited machining capability - so the design is crude but can be made if you can drill a hole with some precision.

The Labview code posted is our third generation, a fourth will be posted later this year, but the changes are minor. We know of a few teams that were able to use this code out of the box.

Re: Starting out with Swerve Drive.
 

On the topic of "where to start", one particular implementation deserves to be at least mentioned.
The original:
http://www.chiefdelphi.com/media/photos/12949

For development of hardware, this setup would be more cost effective, since you only need two modules. You could actually get away with one for pure hardware development. I suggest omnis in place of the casters.

Re: Starting out with Swerve Drive.
 

I would argue 2451's swerve drive here is better than 1717's:
http://www.chiefdelphi.com/media/photos/40911

It's a variation of Aren's swerve, but shifting. Plus, they released the CAD for it.

Re: Starting out with Swerve Drive.
 

i've uploaded the file. Somehow CD won't allow me to create papers
https://drive.google.com/folderview?...Vk&usp=sharing

Re: Starting out with Swerve Drive.
 

I'm having a little difficulty following here: are the arguments of "best" saying "best for competition" or "best for learning swerve on a tight budget"?

If we have a pile of CIMs and a Bridgeport (and the sharp students that know how to use it), how much much coin are we talking for swerve on a flyweight chassis (NOT a competition quality chassis)? How much machining time to go with it?

Re: Starting out with Swerve Drive.
 

we were looking into doing swerve, since we've done mecanum and love the maneuverability. The problem we have is where to start when it comes with machining, finding the parts, how to build custom gearboxes and the sorts. It was never an area touched by my team.

Also I have concerns with swerve taking a toll on the battery draw and a pushing match to cause our battery to die.(somehow we use already use a lot of power in a match.)

Re: Starting out with Swerve Drive.
 

If your looking for a custom swerve drive that can be build quickly with only manual machines, I would recommend looking at our most recent design. Here is a link:
http://www.chiefdelphi.com/media/photos/41039

The first module was almost completely designed and manufactured in one week by a couple students and a mentor. It was build with only a manual Bridgeport and a table top manual lathe.

Re: Starting out with Swerve Drive.
 

Throwing something that one can do with swerve into this tread for visibility:

https://github.com/faust1706/Smooth-Swerve

It's a swerve adaptation of @notincontrol's smooth path planner/trajectory calculator.

Re: Starting out with Swerve Drive.
 

Thanks for sharing your design. This is going to be a lot of help to us moving forward because I think my team has all the machinery and knowledge to build something like this. Do you think you could post the CAD model or link to it so that the CAD students can see how it works?

Re: Starting out with Swerve Drive.
 

Sorry I forgo that the thread I liked doesn't have the CAD. Here is a thread with a CAD link:
http://www.chiefdelphi.com/media/photos/41070

I hope it's useful. Let me know if you have trouble pulling the files (I have never tested it myself) or if you have any questions.

Re: Starting out with Swerve Drive.
 

Im also working on a custom swerve and i am not sure how to steer/mount it. It is modeled a lot like 221 systems and the wheel and turning module weighs roughly .9 lbs according to solidworks. I plan on mounting the system from the top. How do you mount coaxial swerve modules?

Re: Starting out with Swerve Drive.
 

I've seen both in person and personally prefer 1717s, it's lighter, more compact and has a longer track record of success. Also it seems easier to machine the wheels and to change the tread but I don't know that for sure. I wish the 1640 wiki had pics from their more recent iterations, they've made some nice improvements since 2012.