pic: High Reduction 2-Stage Gearbox

[Ari423]

loading photo ...

CAD is here (piston not displaying properly, not sure why).

[Cothron Theiss]

Do you mind explaining or posting your Lewis Gear calcs? The formulas and calculations themselves are kinda configurable, and I've run into issues not including all the necessary factors. My opinion on 32DP gears is that they're great for the first stage, but a little iffier on second stage, especially with 8" wheels. I would also caution against an aluminum pinion on the second stage since you're running 8" wheels. I've never actually run a 32DP gearbox so I'm hesitant to use it throughout a high-torque gearbox. But if you think the calculations are all good, I'd love to see how it turns out.

[Bryce2471]

Great design! I have thought about making a very similar ball shifting gearbox.

Originally Posted by Cothron Theiss

Do you mind explaining or posting your Lewis Gear calcs? The formulas and calculations themselves are kinda configurable, and I've run into issues not including all the necessary factors. My opinion on 32DP gears is that they're great for the first stage, but a little iffier on second stage, especially with 8" wheels. I would also caution against an aluminum pinion on the second stage since you're running 8" wheels. I've never actually run a 32DP gearbox so I'm hesitant to use it throughout a high-torque gearbox. But if you think the calculations are all good, I'd love to see how it turns out.

It looks like the first stage is 32dp with steel pinion gears from andymark (Just a guess). But the second stage looks like standard 20dp gears.

[Cothron Theiss]

Originally Posted by Bryce2471

It looks like the first stage is 32dp with steel pinion gears from andymark (Just a guess). But the second stage looks like standard 20dp gears.

Took another look at it. I think you're right. I'm desperately trying to zoom in on a cell phone screen, so that's my fault for missing it initially.

[Richard Wallace]

The low-speed drive gear (smaller one on the jack-shaft) looks like a failure point to me. All the power and multiplied torque of two CIMs concentrates there.

[Bryce2471]

Originally Posted by Richard Wallace

The low-speed drive gear (smaller one on the jack-shaft) looks like a failure point to me. All the power and multiplied torque of two CIMs concentrates there.

Agreed. This is a common break point on many gearboxes.
However, I think that if a 4140 steel gear from vex pro is used it should hold up well. (Guessing it's a 14 tooth)

To OP: Why did you put the CIM motor mounts so high relative to the rest of the gearbox? It seems like you could have lowered the COG and made the gearbox smaller by lowering the CIMs to just above the pneumatic cylinder.

[Andrew_L]

This is fantastic for your first DOG shifter - be proud of yourself for what you're accomplishing.

To touch on your points:

-"5.5 ft/s low, 16 ft/s high" - this is a good spread to work with. I've personally found that you'll rarely need to go lower than 6 ft/s (or 5.5, 0.5 ft/s doesn't make much of a difference), and you rarely need to go higher than 16 ft/s. For future games, sometimes you'll need your high gear to be a little lower, and sometimes you can have your low gear a little higher. Either way, that's a good design decision you've made.

- 2.0 lbs - compared to 3.1 lbs for comparable COTS gearbox - Good target weight, though I think the reason your transmission is a pound lighter than the COTS alternative is more due to some cut corners than it is to a more weight-friendly design. Make sure your plates cover the circumference of your CIMs almost completely, otherwise there's room for bending and awkward loading. Also, make sure all of your material and mass properties are assigned correctly. It's easy to realize later that some non-custom parts, like the motors, shifting shaft, and shifting cylinder, were vastly underweighted in CAD. The spacing also looks a bit off on your shifting shaft, so unless that's custom, I'd double check with the WCP DS CAD model to see if you're using it right.

- Will the 32dp gears hold up? (They should according to the Lewis Formula) - We have done 32dp gears for our initial stage on a gearbox before on 1323, though I wasn't on the team at the time (maybe RC can comment on it). I do know that we don't actively do it anymore, so I'm guessing we stopped for a reason. Where most of the problem is going to arise is when you're in high gear, and the 32dp gears are the weak link in the chain of power transmission. Those pinions look pretty thin as they are, and I'm not sure how well they'd hold up, even if they were steel. You can achieve a lightweight 2 stage gearbox for an 8" wheel using 20dp gears, you just need to get creative.

- Am I doing dog shifting correctly? (This is my first dog shifter) - Like I said earlier, double check with the WCP DS model. Your shifter shaft looks a bit off to me, and it looks like you have an extra shifting cylinder mounting plate attached to your gearbox plate (you only need one, and none on the plate itself). Also you shouldn't need a shaft collar to hold your bearings in. Look into either geometrically retaining them (aka the model's geometry physically prevents the bearings from coming out), or by using the head of a #10 button head cap screw to retain the bearing flange (as seen on the WCP DS).


Overall, this is a fantastic first shifter. Keep up the great work and continue iterating and you'll only get better and better. As my boy Macklemore once said, "The greats weren't great because at birth they could paint, the greats were great because they paint a lot".

[Andrew_L]

Originally Posted by Bryce2471

put the CIM motor mounts so high relative to the rest of the gearbox? It seems like you could have lowered the COG and made the gearbox smaller by lowering the CIMs to just above the pneumatic cylinder.

It appears as if the driving dimension there was an attempt to get the CIMs as close together as possible for a thinner gearbox.

[Ari423]

I realized I never attached the CAD, so it is in the original description.

Originally Posted by Bryce2471

Agreed. This is a common break point on many gearboxes. However, I think that if a 4140 steel gear from vex pro is used it should hold up well. (Guessing it's a 14 tooth)

To OP: Why did you put the CIM motor mounts so high relative to the rest of the gearbox? It seems like you could have lowered the COG and made the gearbox smaller by lowering the CIMs to just above the pneumatic cylinder.

Good idea on the steel 14t gear. I'll make that change. I didn't really consider the placement of the CIMs very much. I could push them closer together if I need a thin gearbox or pull them down if I want to lower CoG. It shouldn't affect the design much.

Originally Posted by Andrew_L

- 2.0 lbs - compared to 3.1 lbs for comparable COTS gearbox - Good target weight, though I think the reason your transmission is a pound lighter than the COTS alternative is more due to some cut corners than it is to a more weight-friendly design. Make sure your plates cover the circumference of your CIMs almost completely, otherwise there's room for bending and awkward loading. Also, make sure all of your material and mass properties are assigned correctly. It's easy to realize later that some non-custom parts, like the motors, shifting shaft, and shifting cylinder, were vastly underweighted in CAD. The spacing also looks a bit off on your shifting shaft, so unless that's custom, I'd double check with the WCP DS CAD model to see if you're using it right.

- Will the 32dp gears hold up? (They should according to the Lewis Formula) - We have done 32dp gears for our initial stage on a gearbox before on 1323, though I wasn't on the team at the time (maybe RC can comment on it). I do know that we don't actively do it anymore, so I'm guessing we stopped for a reason. Where most of the problem is going to arise is when you're in high gear, and the 32dp gears are the weak link in the chain of power transmission. Those pinions look pretty thin as they are, and I'm not sure how well they'd hold up, even if they were steel. You can achieve a lightweight 2 stage gearbox for an 8" wheel using 20dp gears, you just need to get creative.

- Am I doing dog shifting correctly? (This is my first dog shifter) - Like I said earlier, double check with the WCP DS model. Your shifter shaft looks a bit off to me, and it looks like you have an extra shifting cylinder mounting plate attached to your gearbox plate (you only need one, and none on the plate itself). Also you shouldn't need a shaft collar to hold your bearings in. Look into either geometrically retaining them (aka the model's geometry physically prevents the bearings from coming out), or by using the head of a #10 button head cap screw to retain the bearing flange (as seen on the WCP DS).

The reason for the big weight difference is that all-20dp COTS gearboxes need 3 stages for the same reduction, and the extra stage adds a lot of weight. I will consider adding more plate to support the CIMs from bending. The shifting shaft is custom because the dog gear is from AndyMark while the rest of the shifting mechanism is Vex. The AM dog gear is 1/2" hex and the Vex one is 5/8" hex, so I would have to use Vex's output shaft to use Vex's dog gear.

I'd be interested in hearing why you stopped using 32dp first stages. I don't have much experience with actually making custom gearboxes so I appreciate any real-world experiences I can get.

The pneumatic cylinder only has one plate, which screws into the rear plate of the gearbox. I tried to constrain the bearings on the output shaft by squeezing the flanges in between the plates. The dog gear means I can't put spacers between the gears, so the bearings could slide inward. I would use a screw and washer to constrain them, but in the front there is a shaft sticking out (i.e. it's not the end of the shaft) and in the back the shifter shaft goes through the center so I can't put a screw there. If there's something obvious I'm missing, please hit me on the side of the head with knowledge.

[Andrew_L]

Originally Posted by Ari423

The reason for the big weight difference is that all-20dp COTS gearboxes need 3 stages for the same reduction, and the extra stage adds a lot of weight. I will consider adding more plate to support the CIMs from bending. The shifting shaft is custom because the dog gear is from AndyMark while the rest of the shifting mechanism is Vex. The AM dog gear is 1/2" hex and the Vex one is 5/8" hex, so I would have to use Vex's output shaft to use Vex's dog gear.

You can accomplish the same reduction you have in two stages of 20dp gears for a similar weight. And while an extra stage adds weight, it's definitely not a whole pound's worth. Why are you using the AM DOG gear? Just use the DOG gear and shifter shaft from the WCP DS - it's easier than making your own and the larger hex on the DOG makes spacing the bearings super simple.

Originally Posted by Ari423

The pneumatic cylinder only has one plate, which screws into the rear plate of the gearbox. I tried to constrain the bearings on the output shaft by squeezing the flanges in between the plates. The dog gear means I can't put spacers between the gears, so the bearings could slide inward. I would use a screw and washer to constrain them, but in the front there is a shaft sticking out (i.e. it's not the end of the shaft) and in the back the shifter shaft goes through the center so I can't put a screw there. If there's something obvious I'm missing, please hit me on the side of the head with knowledge.

Like I said earlier, using the WCP DS output shaft and DOG gear solve the issue with the bearings sliding inwards. And check out the WCP DS model and you'll see how bolt heads retain the bearing flange.

[Cothron Theiss]

OP: How important is it to you that both plates are identical? I think you could cut out some weight by making the second plate shorter and removing the material you don't need. Also, how are you mounting the gearbox to the chassis?

[Max Boord]

As someone who learned of this the hard way, you really should make the bosses go all the way to the edge of the cims to prevent major gear grinding issues. Also, whats the purpose of the shaft collar on the output shaft? The shaft should be retained by a bearing round and a snap ring at the back meaning the shaft collar is not doing anything.