13-11-2014 20:27

Tyler2517

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

This is officially my Favorite swerve concept to date. Only a few questions...

Will this cause brown outs or systems to fail due to the massive amount of power being consumed when accelerating/turning?

You will be using almost all if not all of your 40amp slots to run this, is it that much of a advantage in speed/mobility to do this?

For field centric control can you update in time with this much acceleration/speed to keep every thing consistent?

13-11-2014 20:50

asid61

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Very nice! Is that weight with all fasteners, etc? Do you have any pictures of the module itself?

13-11-2014 20:57

Munchskull

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Looks great, but does it really need two CIMS on each model?

13-11-2014 22:13

magnets

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Looks awesome, but isn't "lightweight 8-CIM drive" an oxymoron?

13-11-2014 22:16

Bryce2471

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Thanks!
In this design, auto shifting code would be used so that any time the speed was low, it would be geared or 12fps, which I believe this low enough to the breaker.

That is a good point. With the new power distribution board, ths will use all f the available power except 4 20-30 amp slots. That is a major disadvantage of this design, but that would be more than enough to build a robot in three days type manipulator this year. My goal with this drive train desing was to make an 8 motor shifting swerve drive small and light enough to be usable for a drive train dependent game.

We are implementing a limit on rotational acceleration for control-ability, weather or not we use field-centric code.

13-11-2014 22:41

Bryce2471

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

It includes everything you see in the picture. I've still been too lazy to find CAD for small snap rings because their weight is negligible, but I think I've got everything else accounted for. I'll post a close up of the module soon.
No.

But I it would be nice! In some games, extreme amounts of drive power and maneuverability is desirable enough that a drive train like this would make sense. Although, I'll admit that it's unnecessary in most cases.
Kind of ;D
I guess this calls for the picture to be renamed...

13-11-2014 23:17

asid61

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Two cims or a cim/minicim combo per module is the new metagame. You need those to go higher than 18fps, and it's useful for lower speeds.
In any case, what were you going to use them for, it not for drive? With the abundancy of motors these days, there's no point in not doing it...

14-11-2014 00:30

Tyler2517

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

There is a point of over kill and i think this might reach it...

14-11-2014 00:42

Electronica1

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Now you just need a pto for each module. Might as well get as much use as possible out of all those motors.

14-11-2014 01:03

Dunngeon

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

I think it might barely reach it, just a tiny bit.......

Except the loss of all 40amp circuits, that's kinda a big deal.


All in all though, this lightweight packaging is a pretty sweet deal. I've really liked the iterations I've seen in the last few days. Too bad we won't see this module at Bunnybot

14-11-2014 02:04

asid61

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

True, you can't exactly just add more circuits. Even so, I still don't see the need for every motor we're allotted. It really does depend on the game, but a lot can be done with pneumatics (254's robot this year had tons and tons of pneumatic cylinders and locks instead of just motors) if you are short on motors.

Also, 1 2/3 cims per module is just a tiny bit more powerful than a 6-cim drive. If you want acceleration in a swerve drive, you can either have a multiple of 3 modules or you can go cim/minicim. I prefer the latter for a multitude of reasons. Think of it: not only can you snake and translate, but you can do it at 20fps and save momentum. The possiblities are amazing!

Now a PTO for this would be the next step... Maybe #25 chains under the chassis, to use the turning motors to climb or something?

14-11-2014 17:39

Bryce2471

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

I actually agree, in most cases. For most games I would prefer the 5.4 lb, single speed, single motor module I posted earlier. If the situaltion comes again where mobility is almost everything, this would be awesome.
Thank you! I wish we could field this at bunnybot as well...
Unfortunately, the money and resources to do that just aren't there this year.
Good point. I think their robot only used four small motors outside of their drive base. That's few enough that they could have used this drive.
or you could do this:

14-11-2014 21:30

asid61

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Not if you want a full-featured swerve. The cim in the middle will limit what kind of snake turns you can make. Each wheel in the swerve has to be able to rotate and speed up fully independently.
For crab drive, you can indeed chain them together.

15-11-2014 14:56

Bryce2471

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

With good code, a rectangular robot, and 6 CIMs of torque; I think you would be able to do any swerve movement with a little scrubbing.

15-11-2014 15:30

Ether

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

What makes you think that?

15-11-2014 17:13

Bryce2471

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Each wheel is independently steered, and if you had a 6 CIM 4WD tank that was 21" by 35", it would be able to turn in place without too much trouble.

This drive would be able to turn in place or translate without any scrubbing. It wouldn't have any resistance until it turned while driving. I think this relatively small disadvantage would be outweighed by the advantage in a pushing and driving straight.

15-11-2014 23:23

Ether

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

It sounds like you just moved the goalposts. But perhaps I am misunderstanding. Are you still saying that it "would be able to do any swerve movement with a little scrubbing" ?

15-11-2014 23:35

Bryce2471

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Yes my intent was to support that claim. The only time scrubbing would be involved is when translating arbitrarily, and rorating at the same time. I believe that in that case, the 6 CIMs would be able to power though the scrubbing, given the gear ratio and width/length of the drive base.

Obviously, this is an opinion, and you can be sure without testing.

16-11-2014 00:00

Ether

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

What wheel angles and wheel speeds would you use to make that bot pivot around, say, the front left wheel? (a common swerve maneuver).

16-11-2014 00:34

Bryce2471

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

I would not call this a "common swerve maneuver" as I have only ever seen one robot get notable benefit from using it. Turning around one wheel is the worst case scenario for a swerve with front and back wheels tied because two wheels on the same side should ideally have a large speed differential.
Regardless, It should be possible to perform this maneuver with something like this:
https://scontent-a-sea.xx.fbcdn.net/...2a&oe=54DF1511
(and a lot of scrubbing)

01-12-2014 11:50

Ether

Re: pic: Lightweight 8 CIM Shifting Swerve Drive Chassis

Rotating around a point other than the Center of Geometry of the robot is a very basic swerve maneuver.

Indeed. I choose the front left wheel as the Center of Rotation to simplify the problem statement, and to pick an extreme case to try to clarify what you meant by "do any swerve movement".

If you are interested in exploring the physics of "scrubbing swerve", I invite you to participate (or just read along) here.