pic: 1-Stage 2-Speed WCD Gearbox



This is my attempt to make a simple single stage dual speed gearbox for a WCD.

Low gear: 5.45:1 (60/11)
High gear: 3:1 (54/18)

Since the gearing is pretty aggressive this design would be best paired with small wheels. A 3.5” Colson is about as small as could be used without the 60 tooth gear being too close to the ground. A 3.5” wheel corresponds to 12.09FPS and 21.98FPS adjusted (81%) top speeds.
The design uses an automation direct single acting spring loaded pneumatic cylinder. This has the benefit of defaulting into low gear if the system isn’t pressurized. This could be beneficial for autonomous if we forget to pressurize the tanks before the start of a match.

Two 9mm width belts would be used with the VexPro toothed pulley to transfer motion to the front and rear wheels. Having the belts on the outside of the frame rail gives us less space between the frame rails, but has the benefit of making the belts easily accessible if they need to be changed.

Some concerns I have with the design are the wheel being cantilevered father than normal from the frame rail, the 60 tooth gear being so close to the ground, and the gearing being so aggressive. However, I still think the design could work well for a game with a relevantly flat ground where high top speeds are important.

1 Like

What a cool design!! I’d love to take a more in-depth look at the CAD, but its a really cool concept and props on the execution! Belts or chains on the wheel side of the tube has been done, but I’d recommend running your chain/belts inside the tube or using standoffs to push the whole gearbox back to fit it on the gearbox side. Even if they were free speeds, those numbers seem a little high. Adjusted for 81% “real life speed” that sounds absurdly fast and I don’t even think switching to 3" wheels (might not even fit) would help you there.

This is a really cool concept, and with a bit more work on the execution/tuning it could end up a really cool and conveniently packaged way to do shifting! Smaller scale than many COTS shifting gearboxes for sure!

Good stuff!!

Love the gearbox, I have been thinking about a gearbox like this ever since Rob posted this: https://www.chiefdelphi.com/media/photos/42611.

I was thinking of using a chain reduction after the gear stage instead of direct drive, it gets the speeds a bit lower. Also, if you use Andymark’s EVO shifter cylinder: http://www.andymark.com/product-p/am-3494.htm you can get a better face mount than a normal nose mounted cylinder. (I believe it packages in there).

It just packages so nicely, big fan of this.

Since you’re already making your own dog gears, increase your diameters to make the speeds half feasible.

I suggest the following ratios for 3.5" wheel:

14:62 high gear for a free speed of 18.4 ft/s (adjusted 14.9 ft/s)
11:64 low gear for a free speed of 14 ft/s (adjusted 11.3)

Your limiting factor here is your near uncontrollable high gear speed. The reason such a transmission isn’t commonly used is because it pushes the high gear so high that it’s not practical for most teams and most games and/or makes the spread too small to the point that it’s not even worth shifting (such as the suggested ratios above).

It’s definitely a fun design, but it’s not more commonplace for good reasons.

Looking more closely at math in the JVN Calc, there isn’t quite a good solution with 20dp gears and stock commonly available to FRC teams to bring this down to a more common/realistic speed (your low gear is faster than some team’s high gear speeds historically). Any more reduction or smaller wheel and the driven gear would be bigger than the wheel, rendering you unable to drive.

With this concept design in mind, I may consider using the AM 32dp cim pinion, but that also probably isn’t a good idea for reasons that should be obvious.

TL;DR, short of some excessive design changes like custom gears in a potentially stronger material, this gearbox layout just doesn’t realistically fit the bill for most teams, if any.

Props for coming together with this cool concept!!

I had a post written up that said basically the same thing, but I got scooped!

Even with 32DP custom dog/ball gears, I think you could only reasonably get down to about 9fps low gear with something like a 15:120 reduction on a 4" wheel. I haven’t done the math to see if steel 32DP teeth could handle final stage drivetrain loads. A 25:110 high gear reduction would make it a 1.82 spread, final adjusted speeds about 9.5fps/17fps, which could be decent, certainly much closer to traditional speeds and spreads. But now you’re talking about custom steel 32dp dog gears and I just don’t think it’s worth the effort for what you get.

As far as putting belts on the outside, I’d go #25 chain if it was me. 9mm HTD belts are really pushing it, and the cantilever gets absurd if you go to double up 15mm belts. Plus, if you put belts/chain on the outside of the drive rail you’re effectively pushing the chassis rail inwards by that amount relative to the max robot dimensions. It might be worth the tradeoff if it doesn’t impact other mechanisms, but if you use chain you can slim that down substantially.

That looks like an amazing concept. Maybe run the belt/chain inside the tube to get the wheel a bit less cantilevered.

Also, you’ve got three bearings pretty close together on the output shaft. I’d get rid of that largest bearing.

I am confused on how these kinds of gearboxes work. Can someone help better explain it?

Are both gears in the middle running at once?
I am a little lost if someone happens to have a write-up or video that shows how this kind of shifter works.