Added the ratchet/pawl and decided to go with a CIM-in-module approach. Still to come is the pulley reduction to the wheel. With the planned reduction (none), the JVN calculator is saying the free speed will be 45.35ft/sec and adjusted speed will be 36.73 ft/sec. Current weight is 6.3lbs
Check the JVN calculator for current draw with a decent weight robot. You’ll want to gear/pulley down!
Well shoot. I’m not sure how to find that on the JVN calculator (pushing current?), but I hope I can keep the 1:1 ratio. If not, I CADed two different sets of pulleys (a 1:1 and a 2:1 set). The possible reduction, however, is limited by the size of the wheel (2 inches). Don’t want pulley making ground contact instead of rubber!
I assume you’re using slip rings because you can only rotate the swerve module in one direction, so no unwinding wires.
Nope! I’m planning on making a “wire tensioner” that lets out and pulls back in wire, that way you just need to put some IGUS on the wire where it gets close to the module
Not to put too fine a point on it, but a swerve drive is a complex design challenge, and if you don’t yet have a full understanding of why a 1:1 reduction on a CIM isn’t adequate for a drivetrain, I would start with simpler, more fundamental designs. You have to walk before you can run.
Believe me, I have quite the understanding of drivetrains in the ol’ thinkeroo (hence why I thought ahead and designed 2 sets of pulleys, switching CAD to the second set today). The 1:1 reduction was not a matter of “it will be”, but instead a matter of “could it be”, you feel me?
Edit: Also, I’ve already walked here (only one ended up on chief, but I’ve designed about six different wcd/drop center/4wd setups)
According to the spec sheet, at 12V and 50A draw, a CIM will generate about 125 oz-in of torque. With no reduction, a 1" radius (2" diameter) wheel, and ignoring friction, that means you can get 125 oz of force (< 8 lbs) from each module. That’s not only not enough to win any pushing matches, that’s not enough to get across the field expeditiously.
Edit/amplification: This is AndyMark’s application note on a robot geared about 2/3 the speed of yours, with four CIMs:
what is that a 5 inch long wheel? Looks like the Mother of All Friction.
2.5" wide wheel. The 2" diameter makes it look wider.
This is a cool idea, but I have some concerns.
As a driver, that 36 fps speed seem so too big to be practical. You would cross the field in under 2 seconds. And with no way to shift down, I can’t see a practical way to move a short distance, such as to line up with a peg in this years game. And as others have mentioned you may not even have enough torque to move, let alone enough to avoid getting pushed all the way across the field.
I used I forget whose calculator, and plugged in some of the numbers you gave. See the screenshots for the results. http://imgur.com/a/GyUdD. What concerns me is that to move, you are drawing 480 A across four motors. This doesn’t seem viable unless you choose a larger* gearing than even 2:1.
*I mean Wheel side:Motor side. Make that ratio larger.
Also, just wondering, why such a wide wheel over a narrower wheel that would make turning the module easier?
How many of these designs made it to fabrication, assembly, and testing?
Hmmm. Considering I’m a broke high school student, none of them.
Can’t have a discussion about a 1:1 geared drivetrain without someone linking to this thread.
I knew it!
Aaaaactually, it’s 3:1 now
3:1 may be reasonable. I’m not 100% sure, and you certainly should ask someone with more experience than me, but I would personally gear up a little more. To the experienced folks: is pulling 20A on a drive CIM reasonable?
I’m going to have to QFT. And I also want to post another quote from your last thread that seemed to get ignored:
If I were you, I’d listen to Aren. He’s one of the most noted designers of swerve stuff ever and easily one of the smartest people I know IRL. He knows what he’s talking about, trust me.