Is anybody aware of a gearbox with a ratio on the order of 70:1 that is compatible with 2 cim motors and has a hex output shaft? I can’t seem to find something that fits all three characteristics anywhere.
What could possibly require a reduction that low that needs two CIM motors??
I feel like there may be more compact motor options that will serve your purpose well. Depending on what that is. VexPro versa planetaries can be customized up to 100:1 reduction, and have a one CIM input adapter. You could also use the dual 550 adapter and gear down a lot.
If you’re talking within the realm of COTS FRC, something has to give here.
With single CIM, the GEM500 from Andymark is reasonable.
If you provide another reduction after it, the plastic 3 stage gearbox from Vex would work too. Even if you paid to get some plates cut to build another stage it wouldn’t be very expensive.
COTS gearing from other industries is unlikely to accept dual motor input, so that will rule out a good deal (you’ll also find that gearing for pretty much any other industry will be more expensive).
This breaks the versaplanetary.
EDIT: Worth pointing out that 2 CIMs at 70:1 is enough torque under stall to break many things that sometimes are assumed to be bulletproof in FRC. You’ll want to verify the material for the output shaft can handle the torque (along with whatever the output gearing is) and/or limit the systems ability to maintain full stull at full voltage.
Two VersaPlanetaries with CIM adapters, one CIM on each, plugged into any one of VexPro’s single speed gearboxes that can accommodate two CIMs. Fabricating your own would be lighter and probably less expensive, but I don’t know what your machining capabilities are.
Might break stuff, though.
The hex output is the hard part. The GEM line doesn’t have hex output, only 1/2 cylindrical. We twisted the output steel shafts on the GEM boxes this past year with our arm… Granted, we were putting a dynamic ~25lb load on the end of a 8ft lever… but still. Also, you can’t put two CIMs on one GEM.
You might want to look at doing something really silly with a drivetrain gearbox and connecting it to a RAW box. That could also result in stripping the worm gear on the RAW box though… we had some experience with that too this past year.
Seriously, the AM folks were running away from us towards the end of CMP this year.
VexPro advertises the use of two 10:1 reduction stages to achieve the 100:1 reduction. Are the gearboxes actually useable at this reduction? What are some experiences with them breaking?
It was stupid of me not to provide more background info. The motor/gearbox combination if not for an FRC application. It’s for a NASA Robotic Mining Competiton robot that my club is building. We were hoping to avoid chains/belts by direct driving our 24" wheels… hence why we need so much torque. The more I think about it, the more sense it makes to just give up on trying to direct drive the wheels. The reasoning behind trying to use cim motors and FRC components is because that’s what we’re familiar with, and we intend to use the FRC control system on our robot.
If anybody has experience with non-FRC motors that may work for this application I’d love to hear it.
Edit: Hex shaft is not a requirement, just easier to work with imo.
It is usable at this ratio with certain motors, please refer to the Load Rating Guide for your particular case.
4 CIMs is overkill for every robot I’ve seen at that competition, the 2013 ISU bot was 2 Mini-CIMs, one per side.
Good luck, it’s a fun competition (I was with the Iowa State team 2012 and 2013)
-Aren
1 Like
I was watching the highlights from the 2013 competition to try to get an idea of what to expect and they did a good segment on your team. Nice interview (starts at the 21 minute mark) by the way 
Your robot was incredible that year.
If you’re good with one CIM per side with per what Aren said, Andymark has the 3-stage Toughbox (am-0977](http://www.andymark.com/TB3-p/am-0977.htm)) that gets up to 51:1. The output shaft also has a long hex replacement (am-0801).
OK, a CIMhas a stall torque of 343.4 in-oz. Using two and a 70:1 reduction means that you anticipate a stall torque around 48,100 in-oz, or 250+ ft-lb. On the AndyMark PG gearbox pages, they note that 39 ft-lb will shear a (round keyed) 10mm steel shaft. A bit of hind-brain physics tells me that the shear torque of a shaft is proportional to the cube of its diameter (someone please correct me if this is wrong! – verified by equation 2b of engineering toolbox website). Assuming the same shaft material, the diameter would have to be about 18.6 mm, or essentially 3/4" diameter. I’ve never heard of anything with a 3/4" hex shaft. I’m not aware of anything that’s “FIRST standard” with a 3/4" shaft, keyed or hex. If you need this sort of torque and power from an electric drive, my best guess for sourcing affordable motors and gears would be from the golf cart industry.
Everybody should be allowed one stupid post per month…
I haven’t hit the search yet but does anyone know of any teams that have used golf cart gearbox parts in the past? I’m genuinely curious about what else is out there. The fact is that 900 does derpy stuff from time to time and I think the AndyMark folks might appreciate it if we quit ruining their pretty gearboxes.
EDIT: Anyone looked at RC Cart gearboxes? Any of those worthwhile?
As long as you keep buying them, I don’t think AndyMark minds. 
At one point they started referring to Austin and I as “loyal customers”. 
Once school starts back we have a bag of busted parts and some analysis to be done to them that should result in a white paper about how NOT to build a giant arm. We’re learning how to fail magnificently though.
SDSM&T, 2012, for me. Try 775-18s for a direct drive via BaneBots gearboxes, maybe–that’s what we used that year, 4 of 'em. Just make sure your circuit protection–and speed controllers–can handle the current. I don’t think we ever blew a fuse (yep, we used inline fuses) but we also had speed controllers that would shut off under high-current conditions (say, about 3-10 seconds into a run) and not reboot. So we got some without that safety and… well… It worked until we dug ourselves into a small hole and got stuck on our last run. We could smell the smoke when we opened the case.
In other words…avoid doing that. Gotcha. Thanks for the input everybody! The current plan is to go with a single cim Gem500 with 3 or 4 stages- still deciding. If anybody is interested in seeing the CAD (or just a picture) PM me!
Also, can anybody speak to how well the Gem500 (and mounting bracket) can handle side loads to the output shaft (30-50lbs)?
One issue we ran into with the GEMs was that the pressed in pinions that the planet gears sit on can wobble out if there is a lot of side loading. It depends a great deal on the type of GEM used as AM has switched out the parts over the years. In theory, there are large washers inside the GEM to help with this. We went so far as to weld them at one point and even had the shop attempt to harden them… no luck, they snapped like twigs.
Also, with 4 stages, it is possible to corkscrew the output shaft so run your calculations carefully before running it on a robot.
I believe the BaneBots P80’s used to be able to give the kind of reduction and motor input that you want, but sadly the dual input has been discontinued.
http://www.chiefdelphi.com/media/papers/2990
Edit: Doh, I don’t think they offer a hex shaft regardless.
Just the speed controllers–the 775s worked like a charm if the controllers weren’t taking a siesta. They were also easy to dustproof…