pic: Spectrum Fall 2013 Drivetrain


#1



After seeing 488’s prototype all gear drivetrain, I started playing around with the concept. Thank you to Madison and everyone in those threads, they gave me a lot of ideas. We are working on a fall prototype robot and this will likely be the starting the point.

10 Wheel drive (6 VEXpro Traction Wheels & 4 VersaWheels). 1.5" wide traction wheels can fit if needed.

VEXpro Ball Shifter drives a 45 tooth gear into 60 tooth gears on the wheels. Idlers are 1 45 tooth and 2 60 tooth WCP Dog Gears. This ends up with adjusted speeds of 15.35fps & 6.76fps (JVN’s Calc). It would be possible to swap the 45 tooth gears for 35 tooth gears if you wanted to go slower; the inner wheels would have to come in but the bolts should still clear the sides of the gearbox.

All the sheet metal is .09 5052. Chassis is 28 x 27.6.

Gearboxes can drop out of the bottom, but you will have to take off two of the wheels one each side to undo the bolts.

All the dead axles are VEXpro tube stock. That means the VEXpro gears have to be drilled out to clear the axles.

I haven’t added any patterns to lighten this yet, that won’t happen till we know what is going to mount to it.


#2

A few suggestions…

  1. Consider direct driving the center wheel from the gearbox. There’s really no reason not to and it virtually guarantees that regardless of what might break elsewhere on the drive system (short of the gearbox itself) you’ll always have one working wheel. It’s also a good way to help balance the robots weight. Team 703 has done this for several years and its worked great.

  2. Consider vertically offsetting the outer wheels from the center one. Doing this greatly improved the robots ability to turn. In the past when we used large numbers of wheels in our drive, we would position the wheels on an arc of something like ~30ft, it worked well because it could turn on a dime, but in a pushing match you’d always have at least 4 wheels on the floor.

  3. Without knowing your teams experience with them, I might advise against using the VersaWheels, I’ve talked to several teams that have used them and complained that they wear down quite quickly, and unlike Traction Wheels, must be completely replaced, instead of replacing just the tread. That said, this is just what I’ve heard, your experience (if any) may very well be different.


#3

Thanks for the suggestions.

This setup allows us to get all the reduction we need without adding another gear before the drive shaft. If we direct drove we would have to do something similar to the VEX 3 stage ball shifter and it would push our gearboxes further into the chassis. This system should be nearly as reliable as a direct drive. This also requires less machining, since we don’t ever need a bearing on the far plate.

  1. Consider vertically offsetting the outer wheels from the center one. Doing this greatly improved the robots ability to turn. In the past when we used large numbers of wheels in our drive, we would position the wheels on an arc of something like ~30ft, it worked well because it could turn on a dime, but in a pushing match you’d always have at least 4 wheels on the floor.
  2. Without knowing your teams experience with them, I might advise against using the VersaWheels, I’ve talked to several teams that have used them and complained that they wear down quite quickly, and unlike Traction Wheels, must be completely replaced, instead of replacing just the tread. That said, this is just what I’ve heard, your experience (if any) may very well be different.

The two outer wheels in each set of 5 are raised 3/16" (which is a lot but it’s on purpose). The middle six shouldn’t have much of a problem turning, since it’s like a normal 8 wheel with the added benefit of part of the load on the two center wheels that don’t scrub.

We ran an 8 wheel VersaWheel setup this year and loved them when they were new. (Ask the Hawaiian Kids about trying to push us at IRI on Saturday afternoon.) However they do wear down. We have swapped them three times this season. They cost about a 1/4 of a comparable wheel, so that’s still a bargain but it does take up time. That is why we have them only in the corners and raised more than normal. Our 4 outer wheels didn’t wear nearly as badly as the others since they are only engaged in pushing matches. We also learned that you basically can use them twice if you just swap them front to back when the edge that is pushing starts to go away.


#4

This drive train is pretty awesome, we used some similar concepts on our Drive in 2013 and it ran really well.

If I remember correctly (It’s been a while) you don’t need to drill out any of Vex Pro’s 1/2" Hex Gears to clear the tube axles. One of our wheels per side was driven the same way that you’re proposing, and all we had to do was bolt the gear onto the side of the wheel and go.

We also noticed the same wear characteristics on our Versa wheels, we ran an 8WD with the outer corners raised 3/32", and got some fairly even wear on 6 of the 8 wheels, but after a while the wheels would wear enough to make the drive sit on all 8 wheels which was solved by swapping the front and rear pairs. (sounds like you did the same thing)

It might be worth looking into the viability of using a quick change gear pair in between the Transmission and wheels. There are a handful of available ratios that have a pretty nice spread for tuning the drivetrain’s final ratio.


#5

I love more and more teams doing all gear driven drives.

I would suggest that you be sure and make a cover for those trans servicing holes. The small fittings on those cylinders are very fragile and are easily broken by stuff that may end up on the field.


#6

If you guys had to pick, which gear sizes with the VersaKeys would you want with a 1.125" bearing bore?

-Aren


#7

If I had to pick? All of them :wink:


#8

Seems like 40T*, 44T*, 50T, 60T and possibly 64T would be the ‘most useful’ to have with an R8-Sized Bore - Although it would be amazing if every gear were available with some sort of bearing (bushing**) bore.

*The 40T and 44T don’t have the VersaKey, but would still be nice to have with a bearing bore.

**I think the 30T would be the smallest gear that could reasonably fit an R8 Bore, so at some point it may need to step to an R6 bore, or even a bushing bore, I’m not sure.


#9

As things currently stand, you have to be careful when bolting gears onto Versawheels with bearings pressed in, as the boss around the bore on gears prevents the face with the versa pattern from being flush with the wheel hub. We laser cut an .063" spacer with the versa pattern to take up the space and make sure overzealous students don’t crush the bearing while bolting the gear to the wheel.

Having the option to select gears with bearing bores would prevent this; it’d also give us more options for clustering idler gears on common shafts. So, yeah – make 'em all with bearing bores. :slight_smile:


#10

Very good idea, I’ll look into this. Thanks

That’s one of the reasons I was looking at drilling out the gears, it will make it possible to use the same length spacers for idler gears and the drive gears on the inner sides of this assembly.


#11

We’ll see what we can do.


#12

Any suggestions for easily doing this with the VexPro ball shifter? We’ve been looking at multiple gearbox options and the long output hex shaft that’s sold only gives us 1.375 inches, which isn’t a lot if you’re trying to directly drive a wheel and a sprocket, or a gear.


#13

The “long” output shaft is used in the 2 stage transmission to allow for a sprocket(s) or another gear to drive another one that actually powers the wheel. Direct driving off of only the two stages internal to the ball shifter would normally make for a robot that is geared for too high a top speed. If you download the CAD for the 3 stage shifter you can see that the output shaft is noticeably longer for use with direct driving wheels.


#14

I really don’t understand the functional difference between direct driving a wheel using a 3 stage ball shifter and using a two stage ball shifter + gear reduction to drive the wheel. I’m pretty sure there isn’t one, really. You’re already filling the drive line with gears, might as well put a reduction in that drive line too.


#15

I’m a skeptic. Why should I like geared drivetrains?

At least in my mind, a timing belt drivetrain is both lighter and more efficient, and a chain drive is much easier to build and quicker to fix. I do see that chains and belts becoming de-tensioned is a disadvantage, but the solution to this is a cam or exact c-c system. Is there some specific advantage that gears offer over belts or chain that can’t be attained in either system?


#16

I can’t tell you why you should like it but I can tell you what’s attractive to us.

Off The Shelf Parts
Outside of the sheet metal which we get produced by our sponsor, every piece in this is off the shelf and can be assembled with little to no modification. Nothing requires a lathe or a mill. Seeing as we don’t have access to either (yet), this is an important feature. It’s currently near impossible to meet this requirement with belts.

Narrow
The new robot dimensions (assuming they stick around) make the standard kit bot way of doing chains seem really large, the kit bot or VEXpro drive in a day are good examples. For reference those are the types of drives we have used before. A WCD might be a bit thinner but we don’t have milling and welding resources but we have a great laser cutter/sheet metal sponsor.

No tensioning or ever throwing belts/chains
A good drivetrain has a great tensioning system and never throws belts or chains. I am not confident enough that we can build a drivetrain that will meet those requirements. The gear drive solves that problem.

No need for access from the top of the box
We now have the entire width of the robot for our game playing elements. We can cover the entire top of the drive train and not worry about having to ever reach in a fix a chain or belt. We can replace wheels extremely quickly with just one bolt each. This also might be marginally safer as it is harder to get a finger or drop something in the drive.

Can change wheel configurations easily
10, 8, or 6 wheel configurations are all available with this robot without changing much of anything except removing a wheel or two. You could build the same thing into a chain/belt system but that would be a lot of tensioning.

It’s a prototype
And the last reason is it’s a prototype and it’s different. In 3 months I might be completely on the other side of this issue and telling people to never build gear drivetrains ever. If that does end up being the case, we’ll have with a nice stock of gears to use for future gearboxes and a story to tell.


#17

Thanks for pointing that out, I didn’t even realize there was a 3rd stage available. That makes things a lot easier.

Also thanks to you and Madison for posting your prototype drive trains in the off-season. It’s been fun to read the discussion and mull over some ideas for sheet metal fab drive trains.


#18

Yep, that’s what we do. In fact, the Verskey pattern in the gears was meant to help with gear alignment for this purpose.


#19

I’m not trying to bash your drive nor am I against the concept of a gear drive. If we had the time and the drive to do something different I’m sure my team would prototype one as well. Just thought I would share some info about belts.

If you can manufacture a drivetrain with enough precision for a functional gear drive, you have more than enough precision to make a dead spaced belt drive, which does not need tensioners.

No need for access from the top of the box
We now have the entire width of the robot for our game playing elements. We can cover the entire top of the drive train and not worry about having to ever reach in a fix a chain or belt. We can replace wheels extremely quickly with just one bolt each. This also might be marginally safer as it is harder to get a finger or drop something in the drive.

A properly designed belt drivetrain should never need any maintenance. For the past three seasons we have not touched a drive belt since week 4-5 of build season (unless we had to remove an axle for another reason, but that doesn’t really count). So this same advantage can apply to belts as well.


#20

Yes but where do I get pulleys with the 6 hole pattern already machined (and preferably with versakeys)? Aren or Paul want to chime in here?

Also I did a quick cost break down of the Gear Drive vs. an 8 wheel #25 chain drive.
The gear system is about $170 more.

Here are the spreadsheets.