Custom Gearbox Review

[Owen Busler]

Good morning CD! Ive been experimenting with the idea of a custom gearbox now that we are much more confident with our shopbot skills. Here is my first shot. Right now it is a 15:1 reduction (12:60 and 24:72) and weighs 2.8 lbs without cims. What I like about this approach is that we can change the way it is mounded based off our drivetrain and we can easily swap ratios mid season. I would love comments and criticism especially in the lightening pattern. thanks all!

https://drive.google.com/open?id=0B6...DJ5V0tqN3h0TXM

[Chak]

The lightening pattern looks fine, if a bit conservative. Is there a reason the gearbox profile is a square? If you mess with the profile a bit and get it to fit closer to the internals, you could save more weight and get rid of the tiny lightening holes on the sides. One way to do this is to extend the CIM mounting screws and also have them hold the gearbox together, getting rid of the corner bolts and a good chunk of metal on the plate.

Your output shaft looks a little bit long. Are you planning on making this gearbox part of a larger drivetrain design? Or is this a just for fun thing?

[Owen Busler]

Thanks! I plan to mess around with getting it smaller and inverted cims. Yeah this is going through a drive pod which is why it's so long. Thank you!

[Cothron Theiss]

First off, what wheel diameter are you planning for? With 4" wheels, this is verrrrry slow, which may be what you want. Also, I would play around with Dillon Carey and Aren Hill's wonderful new design calculator to decide if you want to use a 3 CIM or a 2 CIM+MiniCIM configuration. (I'm a fan of 2 CIM+MiniCIM.)

How are you planning on making these plates? CNC mill, router, water/laser jet cutter? If it's water/laser jet, your fillet diameter should be fine.

You have stand-offs on your top two bolts; I would put these on your bottom four as well. This stiffens the plates and makes assembly quicker/easier since you don't have to worry about over-tightening those bottom screws.

Your lightening pattern looks very good to my mostly untrained eye, but I would run some simulation on that just to confirm. I will say that your mounting holes and bearing holes do not allow for much tolerance in your motor or gear spacing. In a perfect world, this is fine and in the real world, this is almost always fine. But if you want to be able to change gear ratios or motor combinations easily and often with this gearbox, you may want to add another set of CIM mounting holes with a slightly different spacing that are rotated around the shaft clearance hole you already have. Just some thoughts, you could probably leave it as-is and be fine.

Oh, and make sure you consider how you're going to assemble/disassemble this on the fly. You may want to tap the first plate (the one mounted to the frame), change the tapped holes in the second plate (the one the motors are attached to) to clearance holes, and use lock-nuts to fasten the second plate. This would let you pull off the second plate with the motors and at least one stage of the gearing, while leaving the rest of the gearbox mounted to the frame.

With relatively easy access to the gearbox and a quick pit crew, you could use this design to change your drive ratios between matches during competition.

[Owen Busler]

Thank you for all the feedback! Right now we plan to go with the 8" pneumatic wheels from Andy mark hence the really high reduction. We plan to cut this on a router with an 1/8th inch bit and evey corner is filleted to a 1/16 radius.

I will look into the power and weight tradeoffs with two cim one mini cim and add the standoffs to the lower.

Thanks again!

[RoboChair]

Originally Posted by Owen Busler

We plan to cut this on a router with an 1/8th inch bit and every corner is filleted to a 1/16 radius.

This is a bad strategy. You should always aim to have your fillets larger in radius than your bit radius unless you really have no work around. The reason for this is because endmills do not like going into a corner stopping, then changing direction. They would be much happier and will give a better finish and less likely to break in corners if they arc through it a little.

I would suggest 0.075"-0.090" radii for an 1/8" endmill.

[Owen Busler]

Wow thank you! this is our first time cutting with our 1/8th inch bit. Thanks for the tip.

[RoboChair]

Originally Posted by Owen Busler

Wow thank you! this is our first time cutting with our 1/8th inch bit. Thanks for the tip.

It's not really something most people realize until someone points out those issues to them and they think a moment and have it make perfect sense. But having done both through my prior ignorance and present enlightenment on the subject, it is a very observable problem once you understand what's going on. I've lost a 3/8" carbide ball endmill to tooling foam before on my work's CNC router, that was one heck of a wtf moment. Generally tooling has a great dislike for sudden stops and changes in direction, and that's why most CAM software has many features for trying to reduce them.

Machining is a game of tricks, the more you know the better you can be at it. I love teaching students all the tricks I have learned over the years and find the moment of realization in their faces as they think "Wow, why didn't I think of that? That's so simple. Now I feel stupid." I then make it a point to the student that there are plenty of people out there that can just as easily make me feel that way. It would take several lifetimes of hard work and mistakes to learn all there is to machining. I'd be happy to die with a 1/10th of what's out there! I would consider that a huge achievement.

I get oddly philosophical about making/building/machining...