Hi,
we have built some swerve drive modules, but the motors are turning as well in this design, so we thought that since the rest of the build team doesn’t want to make the motors stationary, we needed another solution.
I welcome other solutions, but I think a slip ring might work best.
Any suggestions / experience?
Slip rings are tricky–not necessarily because they’re tough to work with, but because the GDC may or may not allow them. Last year, they were legal, and 973 used them on an offseason swerve, IIRC. Before that, the last time I remember them being legal was 2008 (190 used a custom one at the base of their mast).
ditto.
the other concern i would have is wear, which would depend more on your implementation and your driver’s driving habits. and by wear, i mean loss of good contact, causing unusually high resistance, a loss in motor power, and a noticeable amount of heat.
If your design is not going to be greater than 360 degrees of rotation, you should be able to get along with just wire going down the center of your pivot. Slips rings add weight and size to the module and provide one more place where series resistance creeps up on your efficiency.
From R40 2011
The branch circuit may include intermediate elements such as COTS connectors, splices, COTS flexible/rolling/sliding contacts, and COTS slip rings, as long as the entire electrical pathway is via appropriately gauged conductors.
Emphasis mine, you may need to prove the slip rings you use are capable of the stall current in the motor you use to satisfy that last part. Assuming the rules remain the same.
I should also have included a caveat to not use mercury filled electrical contacts. This fits under the ‘hazardous materials’ sections of the robot rules.
No slip rings there; just bevel gears.
We had planned on using slip rings in our 2010 robot before they were deemed illegal.
They weren’t all that necessary anyways; our coding was set up to automatically untwist the wires if they got wrapped up too much from turning (three revolutions, I think). Here’s a video of it in action. I think you can see it “correcting” itself at 2:12.
I think the potential for power loss is too great. I agree with Al and Paul that you can get away with having enough slack to allow the module to rotate, and just code the robot to avoid entanglement. Maybe you could use a reel of some sort if you don’t like loose wires.
If you weren’t worried about the power losses. They can be very noisy. So I would rule out using one in any application involving sensors or analog components.
Sliprings are one of those devices that are a great solution if nothing else will work. It’s exciting to see a design working well, but, the design can be very time consuming to do right.
It doesn’t help that First is very conservative with wire sizing. If COTS slip rings are required they may be expensive and heave for the current ratings needed.
One possible solution is to hack the commutator and bell housing to an adequately sized DC motor. “Commutators” are segmented slip rings.
We did a slip ring design last year, but it was a short term exercise on the way to a better solution.
If your worried about size and weight, then you should probably look into a co-axial module anyway.
The wire size comes right from the NEC for open frame wiring.
As an alternative to slip rings, take a cue from the auto industry, where they need to wire an airbag on the steering wheel.
Clock-spring wiring.
Basically, you wrap some wire in a controlled space, loosely, allowing perhaps 2 or 3 turns in each direction easily. Use software to ensure you don’t exceed that.
After having spent 26 years running a planetarium which contained dozens of solid silver slip-rings, I would rather gnaw off my right arm than try to use slip-rings on a FIRST robot. Those things are nothing but trouble!
You mostly likely will trade the solution to one problem for about a dozen other problems.
If you insist on going the slip-ring route, consider solutions to the following:
Sounds like a good engineering project!
Dr. Bob
ditto
Doc,
Someday I would love to get a look inside one of those!
As EricH mentioned, 190 used them in 2008 (there were actually two in that design: the large, custom built one for motor power, and a small signal one nested inside for sensor feedback). Having been involved with the design of that robot, and with the coaxial swerve we used on the 2009 robot, I would heavily recommend against a slip-ring swerve for FIRST.
As has been mentioned, slip-rings can be really fussy, but they are also expensive (especially as you get to the size needed to meet motor wire gauge requirements) and have some fairly strict packaging requirements. Most commercial slip-rings are not designed to handle side-loads or heavy impacts, so you’re going to end up with a lot of added weight, size and mechanical complexity to get them aligned and well protected. If you need infinite turning, a coaxial mechanical system may seem more complex but in practice is going to be no worse than with slip-rings, and much more robust than modules.
There is also a wealth of information on both coaxial design and modularized designs and programming. If you venture into the slip-ring swerve realm, there’s a lot of uncharted waters that is ripe with unresolved issues and untested methods.
tl;dr - Don’t do a slip-ring swerve, it’s not worth it. If you have to have infinite turning, go coaxial. Otherwise solve it with clever wiring and code, as others have mentioned in the thread.