pic: 2 speed, 2 CIM custom transmission (1st attempt) [ISO]

This is our first attempt at designing a custom transmission.

The transmission has 2 speeds, powered by 2 CIMs. Shifter is based off the AndyMark Super Shifter design (HiTech servo with dog gear)

High gear: 15.41 ft/s, Ratio: 7.33:1. Stage 1: 12 to 40. Stage 2: 40 to 44
Low gear: 6.78 ft/s, Ratio: 16.67:1. Stage 1: 12 to 24. Stage 2: 24 to 60

We really would like your feedback. We are building these soon to test in offseason events before we compete in a real tournament. We have no experience with designing transmissions, but have been around FRC since 2009.

Additional views:

Your gear ratio calculations are off - should be something like this?

High Gear - 12:40, 40:44, 44:44? - 3:67:1
Low Gear - 12:40, 24:60, 44:44? - 8.33:1

Is the 3rd stage just a 1:1 ratio? You’re going to be running very fast with that setup. Check out the WCP ratios - you can make the 2nd stage have a more substantial reduction to give you a more reasonable final ratio.

Why a servo instead of pneumatics?

At first glance it looks pretty good (especially at a first attempt). A few questions/comments.

I might make the webs around the CIMs a little thicker. Is there a bearing on your first stage? I can’t see one in renders. Also your fillets on the inside corners seems rather small. What radius are they and how are you planning to manufacture these?

Very curious as to why you’ve offset the output shaft? It doesn’t appear to be doing anything for your speeds (1:1), plus is just adding weight and more inefficiency (extra stages). Also, although I have zero first hand experience with it, from what I gather using a servo for shifting can introduce several issues with shifting. There’s a reason pneumatics are the standard these days. So I’ve read anyhow. Perhaps someone else can chime in on that giving a more detailed explanation.

Yes, the 3rd stage is just a 1:1 ratio. If that is the case, then why would that 3rd stage change the overall gear ratio?

We were scared of adding extra loads to the shaft where the shifting occurs if we used that shaft to drive the wheels. Therefore we put the 1:1 ratio to drive the wheels off a separate axle.

We are using a servo because we do not plan on having pneumatics on the robot at all. It would be a waste to add pneumatics just for shifting…

Yes there is - I didn’t set up my render properly therefore it makes things a little hard to see…

The current radius of all of our fillets is 1/16th inch. That was just an arbitrary value though so we can change it. We plan on water jetting the plates therefore we don’t have to worry about any machining

I would personally heavily recommend against placing the snap ring where you have it located on the output shaft.

Looking at your Design/MFG capabilities I would recommend:

  1. Turning the Shaft down at one end (backside/gearbox side) to 1/2" round
  2. Place the snap ring on the round side
  3. Constrain the bearing so it doesn’t pop out.

Something like this, that way you don’t need to place a snap ring or e clip mid way on a shaft. I’d be more worried of placing a groove midshaft on a drive gearbox than the shifting output shaft breaking. YMMV.

Good work on this gearbox, I like the offset shaft design. That could let you use belts in the first stage too for quieter operation.

You should use pneumatics for more than just shifting IMO, but even barring that the servo shifters have been plagued with problems over the years. Using a pneumatic shifter is much safer, unless you plan on using a fairly beefy servo.

The shifter shaft will still see the same loads that the wheel shaft does if it’s a 1:1 ratio, theoretically. That being said having the 3rd stage means you can add some extra reduction (which it looks like you’ll need) - Scott’s calculations are correct. A 3.67:1 is not nearly enough for even a 4" wheel.
Water jetting is great. 1/16" radius is fine for that. Watch out for tapers and tolerances though - you may need to bore out the holes later.

What RC said about using round bearings is what I would do as well. Circlips are a pain. :stuck_out_tongue:

Since it’s a 1:1 ratio, the shifting shaft will see the exact same torque as the output shaft. How does this help?

Anyhow, anecdotally, using shifting shaft as an output has never been a problem.

Pneumatics are more snappy and forceful, at least from what I’ve heard on here. Servo rules make legal servos less powerful than some legal pneumatic cylinders. It’ll probably work anyways, but I think it’s a tradeoff to reduce complexity for lower performance.

I don’t anticipate the drivers shifting super often, so I am not super concerned about the timing or ‘snapiness’ the shifting takes.

We reviewed our math and we were off quite a bit ::rtm:: :eek: . We are going to make that 3rd stage a 26 tooth to 54 tooth gear which will add a 2.1:1 ratio. Now our speeds should match up!

Fortunately, the only thing we have to change are those gears!

Depending on your waterjet you will likely find there is too much draft on your cuts (unless you have access to a 5 axis) and the gear centers need to be machined or at least reamed.

Can’t tell what size snap rings you’re running but they look small to go over the hex. if you run a 1/2" ID snap ring and try to install over 1/2" hex you over stretch the ring. May consider E-clips. I’ve heard of teams running 9/16" snap rings to go over 1/2" hex, I haven’t tried it. 14mm might work as well. We generally turn down the ends and avoid snap rings when we can, but either approach works.

Second what RC said about snap ring groves as a stress concentration in an area where torque is being transmitted.

The applied torque is not the only loading on that shaft.

It’s perfectly fine to shift on your output shaft in an FRC drive - almost all of the COTS shifters do this.