Our first swerve drive

Edit: Changed title, we worked out a naming convention for our swerve designs. This is our Slip Ring Edition version 1.0

Strange Swerve SR
Evan G and I have been working on a swerve drive the past month that improves on some of the downfalls of some more standard swerve drives. This is a beta design, and we are still working and improving on it. Please let us know if you have any questions, comments, or suggestions we are open to fresh ideas.

We are using a 775 pro for our drive motor, and an andymark 9015 motor as our rotation.

The rotation motor gearing should be around 50:1 in the gearbox, even though this CAD only shows 1 stage. The drive motor has a 21.8:1 gear ratio

Our calculated weight without nuts and bolts is 5.0046 lbs

All parts are designed to be waterjet or lasercut

We are using a custom 108 tooth gear with a bearing groove cut out for 1/4" DELRIN ball bearings to allow for rotation

Our design doesn’t have bevel gears or a planetary gearbox on our drive motor, which reduces friction and power loss. Instead, our drive motor on the rotating part of the mechanism, powered by a slip ring.

The gearing gives us 15.00ft/s free load drive speed.

Here is the GrabCAD for those wanting to peruse our files:
https://grabcad.com/library/2017-summer-swerve-1

Here’s a pic:

The Empire Swerves Again (This mounting plate looks familiar…)

http://i.imgur.com/A0jBGWs.png

This is awesome. Is it the newest Strange Swerve? Will it be built?

Your design looks very nice.

You should be careful about running swerve with a 775pro drive motor. It has been done (see 16 this year), but it takes a good amount of programming to keep it from smoking. IIRC 16 said they had problems with smoking their drive motors all season (or at least for a good part of it). 775pros simply cannot give the same power output as a CIM for an entire match. If you are okay with the lower power on your drive, that’s fine, but it’s something you should take into account.

You said that this design “improves on some of the downfalls of some more standard swerve drives”. I’m interested in what downfalls you were specifically trying to improve, and what you did to improve on them.

Also, see 1533 this year :stuck_out_tongue: Note that the title of this post is a bit misleading…

Running 775pros on the drivetrain does take more upkeep, but if it’s easily serviceable then it’s really not a huge deal to switch out any smoked motors. The bad part about using CIM motors is that comparatively they are huge and heavy. We definitely wouldn’t have nearly the same weight or form factor if we tried to put the CIM on the rotating mechanism. Pros are pretty good if you take the time to check them.

Our main design goals for the swerve were: decent form factor, light weight, and a rotating drive motor. Putting the motor on the rotation mechanism gets rid of needing any planetary gearboxes, or bevel gears on the drivetrain both of which cause a decent amount of energy loss in most swerve systems.

Changed. This is now officially Strange Swerve SR 1.0 our slip ring edition of the Strange Swerve modules.

Potentially, as there isn’t a Strange Swerve 5.0 yet, this is the best one we have currently and could very well be built.

With the Talon on the rotating part of the module, you’ll need a 4-wire slip ring, correct? 2 of those contacts will have to be rated for 40A. Is that something you have a source for?

Ours didn’t have a planetary gearbox.
Putting more mass in things that rotate is not good.

More importantly,
How are you getting power to your drive motor?

This is a fun question. The only slip rings we could find for a good price were 6 wire 30 amp. That means that we have the split the 40 amp wires into 2 positive and 2 negative, go through the slip ring, and reconnect the wires to get our 40 amps back. That leaves us with the two wires needed for our CAN bus.

Be prepared for plenty of questions about this during inspection (and check vigorously to ensure compliance with electrical rules).

Have you considered dropping the split ring and simply not making the module continuous rotation? It could slightly inhibit how swiftly you re-orient your drive modules, but it should never require more than 180 degrees of rotation to have your wheels pointed the heading required. This is a sacrifice from the never more than 90 degrees for a continuous rotation module, but avoids any potential issues from a slip ring.

I would get to testing CAN through that slip ring early if you think you might use it. My gut says it won’t be happy with the noise of going through a brushed connection.

Also i’m curious where your data is to justify the statement:
“planetary gearboxes, or bevel gears on the drivetrain both of which cause a decent amount of energy loss in most swerve systems.”

Properly setup bevel gears can hit the same efficiency as spurs, yes, it is harder to do, but it’s doable, and Planetaries are typically just used for turning modules.

Also clearly the 775pro should be in the wheel.

-Aren

141 didn’t use a slip ring in one of there swerve designs when they put the motor in the hub. They set it up in a way that allowed the swerve to make a few rotations (2ish?) before the wires had issues. The software kept track and when the limit was hit, would “unwind” them.

We know CAN will probably need testing and some work to use correctly, but I believe it has it’s own error checking to an extent that should help with the noise. We’ll see.

Part of us mentioning efficiency is because, we had a lot of issues with our versa planetary gearboxes this past season on our drive motor. The gears would heat up and expand after heavy use and lockup which ended in some spectacular flames from our pros.

Axial load on the bevel gears is another concern with swerves, if careful it can be avoided pretty easily. We didn’t have any issues this past season, but we just wanted to get rid of that concern.

I completely agree, we’ll get right on that.

I’m a fan of your not-using a bevel gear.

They waste five percent of your energy and they’re hard to get or make.

Kudos.

I’ve been trained that straight bevel gears always have more sliding, hence more inefficiency.

Here’s a reference I just found that says 93-97 percent.

http://www.meadinfo.org/2008/11/gear-efficiency-spur-helical-bevel-worm.html

Yes, spiral bevel gears can theoretically reach 99% efficient, but I’ve yet to see a FIRST robot with them.

http://mwands.com/store/180-amp-6-wire-slip-ring-kit If that is the slip ring they are talking about (this is 6 contacts, 180A total = 30A each), it would only have <10 milliohms of noise. Seems small enough to me. However I think inspectors will be hesitant with the split wire approach (unless you’ve done it before I guess).

Stop that.

1 Like

I have one concern with the drive and that is you have the motor turning with the module. How are you going to wire the motor and not have it tangle up when spinning several times around?

Slip ring.