Custom Gearboxes

[stevefang]

Hello All,


I’m Steve Fang from 5700, the SOTA Cyberdragons Robotics Team. We are a relatively new team, now going into our third year of FRC. Last year, our robot was very simple mechanically and we have very little experience with gearboxes, sprockets, belts, pneumatics etc. To try to learn as much as we can, we have divided our hardware team up into groups to take on a few offseason projects in preparation for the upcoming build season.

I’m working with a team of three on designing a custom gearbox for the 2016 Stronghold scaling challenge. Since I have little experience on this subject, I’d like to see if other teams or mentors have any suggestions or input on how to properly do this.

Our Goal:
Create a custom gearbox for a scaling mechanism to lift our 50kg robot up 2ft at a rate of 0.5m/s. We plan to do this with a winch.

From my calculations using 2x RS775 motors,(following the instructables gearbox guide http://www.instructables.com/id/Unde...earbox-Design/ ), our gear ratio would be about 21.4:1. Here is the spreadsheet for the calculations: https://docs.google.com/spreadsheets...it?usp=sharing

I also know the basic components that are needed for a custom gearbox, but we are still not sure about the specifics (material, gear types, mounting methods, spacers etc.).

Am I approaching this problem correctly? What should be my next steps? Any sites I can look at to find more information/buy parts for this project? Any comments or suggestions? Any advice on the subject would be a huge help.

We have access to basic power and shop tools (bandsaw, tablesaw, bench grinder, etc.). We have a mill and a lathe, both without DRO. We also might have access to use a waterjet at techshop.

If you have any materials or references or sources that you think would be helpful, please email them to sotacyberdragons@gmail.com.

Thank You!

Steve Fang

P.S. This is my first time posting on Chief Delphi, so please tell me if I’m not following some Chief Delphi etiquette that I’m not aware of!

[Sam Skoglund]

Hi Steve Fang. This is a great first post!

The RS775 motors are not very friendly to use in custom applications because they have a very small, short, and round shaft. If you would like to use these motors, they do come with prefabricated planetary gearboxes like this. Another good and similar option for a winch would be the Vex Pro Versa-planetary gearboxes. These can be configured to fit the various small motors, but their main advantage is that you can attach the powerful CIM motor. They are available in reductions from 3:1 to 100:1. The reason I would choose the Vex Pro option is because they have a hex output shaft that allows for easy mounting of a spool. As a young team this would be a much easier, faster, and more reliable way to obtain a winch. Custom gearboxes are great for making compact and fine tuned drivetrains, but offer little advantage when designing a winch. Oh and one note about attaching a spool to one of these gearboxes: you must support the end of the shaft in a bearing to avoid forces that could cause the gearbox to lock up or become damaged.

Hope this helps!

[BobbyVanNess]

We've used a variety of components, but I think you'll find that the majority of custom gearboxes in FRC use 1/2" hex shafts and 1.125" OD flanged hex bearings with 20dp hex broached aluminum or steel gears. It's just a super easy, reliable and effective means of transmitting power. Different teams have different preferences as far as plate material, but 1/4" aluminum gearbox plates, even with generous pocketing, provide enough support for just about any FRC application.

Generally, teams get their gearbox plates CNC milled/waterjetted. That being said, simple hole space can be done on a manual mill.

For mechanisms like a climber, I would certainly check out VEXpro VersaPlanetary gearboxes. They're pretty straightforward to work with and mount directly to CIM, miniCIM, BAG and RS-775Pro motors. That can reduce the complexity of a gear train or eliminates the need all together.

It's also worth checking out the build logs that teams have posted. Many do a great job of explaining the designs as well as the reasoning behind the decisions they make.

[GeeTwo]

I will echo the advice above, that there are currently so many COTS gearbox options from AndyMark, Vex, West Coast Products, and others which fit together without machining that you will likely improve both your competetiveness and practical learning more by exploring the COTS options prior to making your own gearboxes.

When/if you decide to take the plunge and make your own gearboxes, here are some bits of advice I've mentally compiled reading other threads (3946 has no machining resources, so I can't back most of these up with personal experience):

• Do not count on water jet for cutting with critical dimensions, such as press-fit bearing holes. Either mill these holes (best), or water jet them slightly under-size and ream to the proper size.
• Use relatively large radii on any pockets (lightening holes) in the gearbox plates. This will slightly simplify the work if done on a mill, but more importantly will distribute the stress better than a sharp inside corner.
• The most reliable solution to support an axle is to provide one support bearing at each end, with no clip rings between these bearings. (Clip ring grooves are stress risers much like sharp corners in pockets.) If you decide to cantilever a shaft (especially for a drive wheel), the connection between the two bearing plates should be significantly more than a few bolts. Most teams solve this by pressing the bearings into opposite faces of a piece of rectangular tubing. Some then put the gears inside the tube, others outside (usually inboard of the tube).
• [Also true for many COTS gearboxes] Consider well where to use steel vs aluminum. For a typical well-designed two-stage reducing gearbox, the two most stressed components are usually the output shaft and the small gear on the cluster shaft.

[asid61]

If you want to use a custom box for this, I would recommend using a dual-input Versaplanetary stage to fit both 775pro motors, following by a 5:1 or 7:1 VP stage. At that point you only have to do a small 3:1 or 4:1 reduction to get close to what you want at the spool.
Fully custom gearboxes shouls be avoided IMO if you don't have CNC machining capability of some kind. You'll end up with a heavier, larger product. That being said, integrating a Versaplanetary with an extra output stage to a spool can be made very compact and with minimal machining tools.
Look into the JVN Mechanical Design calculator instead of making your own calculators. Link: https://www.chiefdelphi.com/media/papers/3188
The "linear mechanism" tab should have all you need, including projected current draw and other useful statistics.

[KyleDavidKruege]

We have made a few Custom gearboxes in the past that are always based on AndyMark ToughBoxes. They already come with replacement gear sets to make pretty much any reduction you might want (as long as it's only 2 stages). But if you want more stages or simply need to change the layout so it fits better into the bot we have make new "sideplates" with the hole spacing, which I would also suggest milling (a CNC can place them with enough accuracy, but we would always CNC them a bit small and then fit them to the bearings by hand). Then you can even use the same shafts and everything else that the ToughBoxes use. if you need more than two stages simply buy multiple middle shafts and tie those together, although then you need to make sure that your gears aren't touching each other (eg. if all three reductions use the exact same reduction then you will end up meshing the output gear of the first reduction with the input gear of the third reduction) you can also get around this by making the gear box three gears wide, but then you need to machine your own shafts, which we have also done. overall...

Avoid making parts that you can make a small tweek to your design and buy online (would a COTS gearbox work?)

Fit Bearings by hand (as stated above)

Lay everything out on paper or in CAD and have multiple people look over it. Include the gears, shafts, spacers, any mounting holes, how it will be held together, motor(s), and mounting holes for the motor(s) and any sensors or the output (I have designed and built gearboxes only to realize that it was impossible to attach the motor because the motor mounting holes were not accessible when the gearbox was assembled)

I should also probably mention I am by no means an expert in this area, I'm only sharing what I've learned (usually by failure myself)

[Seth Mallory]

I agree that you can buy very good gearboxes that perform very well. It is unlikely that you will make a gearbox that performs better then one you buy. This observation is from a team that makes gearboxes each year. There are 3 good reasons to make you own gearboxes.

1. Pride, The students get a great ego boost when they succeed and see their gearbox perform at events.

2. Space saving, A unique design that allows a gearbox to fit where no others will fit. Last year we increase the usable space by 3" in the center of our robot and used that space for our shooter.

3. Education, This is the best reason and the one that makes it worth while. The students have to come up with a goal. Then they work out a design and CAD thru many iterations. Then many design reviews with further changes. Now they build prototypes and find that their design may need further modification so things fit together. Now they must test and find what needs improvement. Then they can make they final gearboxes. During this time they are learning how to manage the project, people, and time deadlines.

[NickE]

There's nothing wrong (or particularly scary) about using 755 motors in a custom gearbox without a Versaplanetary. Consider using the West Coast Products 32DP gears for the first stage reduction. You can buy a pinion which presses onto the 775 shaft. With the 80T gear, you can get a 6.6:1 reduction in a pretty small and integrated package. Combine that with a pretty modest second stage reduction with standard 20DP gears, and you can get that 21:1 reduction you mentioned in your first post.

This general concept is pretty much what 254 did this year for our hanger winch.

[stevefang]

Thanks for all your input and suggestions.

Since we have limited machining access, we will use the VersaPlanetary Gearboxes for now. They seem to be the best for applications like this considering how little space we have. With our 21:1 ratio with 2x RS775s, we are planning to use a dual motor input, then a 7:1 kit, and a 3:1 kit above that. Our output shaft will likely be 1/2" hex since I've heard that spools can be easily attached to that.

Also I'd like to know how other other teams' methods on making/mounting winches. Any particular products proven reliable? Links to sources and certain products would help greatly as well.

Steve Fang