pic: FRC3880 - Tiki Techs - 2015 Prototype Drive

[asid61]

If you could consolidate your gearboxes into two 3 cim gearboxes, you could save a ton of weight and motors. I don't see the point in adding the power above 6 cims, unless you're planning on hitting 20 fps+ with 0.01s less time(not really math, just an order of magnitude calculation).

[Aren Siekmeier]

As others have pointed out, there isn't much reason to go with 8 motors over 6, or even 4.

But even if you insist on 8, or at least having the option to mount more motors, you can still save a considerable amount of weight and improve performance by combining 4 on a side into a single transmission, with two output shafts. You can still do so with the WCP DS components.

Duplicating the shifting mechanism on both sides adds quite a bit of weight and cost and means that shifting may not be synchronized. Separating powertrains between front and back means that you could lose half your power if you're wedged or tipped up on your back four wheels. And there's more friction and inertia with all those stages.

Also, mounting the encoder on the input cluster like you've shown will introduce quite a bit more backlash between the encoder and your wheels, and will require knowledge of the precise shift state to schedule gains properly. Combining everything into one gearbox should give you some options to keep the encoder on the gearbox if you need to, while putting it much closer to the wheel, preferably downstream from the shift mechanism.

[Michael Corsetto]

Apparently, hex/octagonal, chain-in-tube, CIM-over-center-wheel WCD is the new fad in DT Design.

Don't miss the forest for the trees here people.

This is coming from a team that rocked an octo-WCD in 2014.

Consider performing a honest and thorough weighted design table to determine if some of the innovations you are proposing actually outweigh the costs.

I would encourage all teams to practice this evaluation method in the pre-season. Just like there is no better time to learn CAD and DT Design, there's no better time to learn how to objectively evaluate designs based off a set list of requirements/criteria (trust me, this is no easy task. It is arguably harder than actually designing the mechanism!)

Beyond big-picture, love the effort being put forward. Winners are made in the off-season!

-Mike

[themccannman]

I detect a serious lack of power and weight in this drivetrain. Far too little power, you'll get pushed all over the place, and not nearly enough material, your robot will float away in a light breeze. I would also certainly increase your gear count, you can never have enough reductions, efficiency can be dangerous. Additional gears certainly help your weight situation too, 4 individual gearboxes certainly capitalizes on this.

[nathannfm]

Originally Posted by Andrew Lawrence

I really have a hard time believing that this drive system even without the electronics is 25 lbs.

He says in the previous image that EACH gearbox weighs 15lb. Something is very wrong somewhere with weights in this CAD model as the WCP 3 CIM shifter this looks similar to would weigh 8.5lb with 1 CIM and 1 MiniCIM. I suppose the 15lb could be referring to each SIDE of the drive, ie. 2 of these put together, that gives a much more reasonable number. Still thats 30lb in just gearboxes, so, going under the stretch assumption that 25lb is this drive WITHOUT gearboxes thats a 55lb drive, which sounds much more reasonable.

All of that is a moot point though because OP has stated that the only reasons that there are 2 gearboxes per side is because they don't currently have the ability to make large enough plates to combine them, and that they want to be able to use the gearboxes elsewhere when the project is done.
A side note to that, MOE has had some success making plates that are larger than the travel distance of our mill by splitting the drawing into 2 parts with 1 shared hole. The shared hole is used to align the second drawing to the plate. You will also need 2 flat reference sides. This may not be possible in your situation and it may not be accurate enough for gear spacing but it worked well for chain.

[TikiTech]

Originally Posted by Gregor

I would not feel comfortable putting half the weight of my robot into gearboxes. This is even ignoring the breaker issues of an 8 motor high traction drive.

Ok my bad. I am new to the iProperties thing.

This should have been caught by me, knowing that the CIMs together weigh about 5lbs.

There is no way that transmission weighs 9 lbs!!! More like 3.

The transmission is basically a hybrid of a WCP 2 CIM DS, 2.15lbs. and 3 CIM DS, 3.5lbs. Weight will be between that since it is basically a stretched DS if you will. A 2CIM DS with one extra idler. We went with an idler gear instead of belts. Much easier for us, we had it on hand.

The entire transmission was designed using the existing gears from our 3CIM DS making it easy to test by just replacing the frames. Future orders of different gear ratios from WCP will just work.

Last year we used the WCP 3CIM DS and are extremely happy with it. Never had any issues with breakers tripping on our 6 wheel WCD but space was tight.

We wanted a stretched 2 CIM DS, if you will. That was set up for in-frame chain setup and over the frame motor orientation. (see hex frame).

Being able to orientate the motors in either direction and a quick and easy encoder placement was the goal.

Our competition bots have always had a post tapped into the front or rear axle that the encoder used. Not always pretty..

Something different for us..

[Brian Selle]

Originally Posted by Abhishek R

Oh, I didn't realize there was a drop. I faintly remember team 118 having to use different sized wheels because they could not enough space in the tube to have a big enough drop with the sprockets and chain. Did you have a solution to that problem or did everything seem to fit inside the tube just fine?

I think the difference is that 118 used a 1x2 tube so there no room for wheel drop. The 3880 prototype uses 1.5x3 tube so they have plenty of space.

[Abhishek R]

Originally Posted by Brian Selle

I think the difference is that 118 used a 1x2 tube so there no room for wheel drop. The 3880 prototype uses 1.5x3 tube so they have plenty of space.

Makes sense, didn't realize it was 1.5x3.

[asid61]

Originally Posted by TikiTech

Ok my bad. I am new to the iProperties thing.

This should have been caught by me, knowing that the CIMs together weigh about 5lbs.

There is no way that transmission weighs 9 lbs!!! More like 3.

The transmission is basically a hybrid of a WCP 2 CIM DS, 2.15lbs. and 3 CIM DS, 3.5lbs. Weight will be between that since it is basically a stretched DS if you will. A 2CIM DS with one extra idler. We went with an idler gear instead of belts. Much easier for us, we had it on hand.

The entire transmission was designed using the existing gears from our 3CIM DS making it easy to test by just replacing the frames. Future orders of different gear ratios from WCP will just work.

Last year we used the WCP 3CIM DS and are extremely happy with it. Never had any issues with breakers tripping on our 6 wheel WCD but space was tight.

We wanted a stretched 2 CIM DS, if you will. That was set up for in-frame chain setup and over the frame motor orientation. (see hex frame).

Being able to orientate the motors in either direction and a quick and easy encoder placement was the goal.

Our competition bots have always had a post tapped into the front or rear axle that the encoder used. Not always pretty..

Something different for us..

Ah, ok, that makes sense. A 9lb transmission would be ridiculous.
However, that being said, I still higly recommend you consolidate your side gearboxes into one gearx each. There is no purpose to having 4 IMO.

[thinker&planner]

Anyone ever hear of battery voltage drop causing cRIO to restart? ;p

I say this partly jokingly, but partly practically. Even with 6 CIM WCD we were dropping out consistently during offseason when we changed our ratio to about 9fps.
We are looking into limiting acceleration/deceleration, current limiting code, etc.

[nathannfm]

Originally Posted by thinker&planner

Anyone ever hear of battery voltage drop causing cRIO to restart? ;p

I say this partly jokingly, but partly practically. Even with 6 CIM WCD we were dropping out consistently during offseason when we changed our ratio to about 9fps.
We are looking into limiting acceleration/deceleration, current limiting code, etc.

See if you can check your driver station logs for the voltage graph during your matches, that sounds like you might have some bad batteries.

[Cash4587]

Where might I find CAD models of the colson performa wheels? I am looking for a variety of them but don't know where to look. But what you have pictured is 4x1.5 and 4x2 if I am not mistaken and would be a good starting place for me. Any help is appreciated. I have checked 3dcontentcentral with no luck. Also google hasn't come up with a match yet.

Edit:
After more looking I found what I needed. Thanks.

[Tyler2517]

Originally Posted by nathannfm

See if you can check your driver station logs for the voltage graph during your matches, that sounds like you might have some bad batteries.

It can also be a CAN brown out.

[Munchskull]

Over kill, much?

[JesseK]

Originally Posted by Cash4587

Where might I find CAD models of the colson performa wheels? I am looking for a variety of them but don't know where to look. But what you have pictured is 4x1.5 and 4x2 if I am not mistaken and would be a good starting place for me. Any help is appreciated. I have checked 3dcontentcentral with no luck. Also google hasn't come up with a match yet.

Edit:
After more looking I found what I needed. Thanks.

Where did you find them?

I wind up making a generic gray-colored cylinder with rounded edges for my models.