One of our students designed this custom shifting gearbox for use at offseason events, after our aggressively geared 3 CIM single speed drivetrain tripped the main breaker a few many times this season. It is somewhat unique in that it uses a “reverted” gearing layout, in which high gear is transmitted directly to the output shaft through a single stage reduction, while low gear is a three stage reduction, routed up through the secondary shaft and then back to the central output. Advantages of this layout include theoretical efficiency/reliability/responsiveness gains in high gear that come with single stage gearing, reduced size of gearing on the secondary shaft, and the ability to lower the CIMs as far as possible for CoG reasons. A few unoptimal choices were driven by the need to retrofit the gearbox to our 2014 robot and work with as many gears/parts we have on hand as possible, but we’re all quite happy with how it came out.
This looks really good Joe! Will 1687 be running these at BattleCry later this month?
Any concerns about using lexan as your sideplates instead of aluminum? Not necessarily for cracking but also for flexibility of those plates.
Either way it looks really nice, can’t wait to hear how they work out for that defensive monster you guys made this season! 
We had many debates about running 3 CIM ss, 2 CIM two speed, or 3 CIM two speed and debated the main breaker tripping.
In the end we ran 3 CIM ballshifters (as it was cheaper than WCP 2 CIM shifters) and tripped our main breaker once during the final seconds at UNH. After having to play 5 semi final matches our robot was getting very worn out and in the finals we t-boned 176 for a solid 7 seconds before tripping with 5 seconds left on the clock.
Thanks Brendan! Yup, we plan to have two of these manufactured and in the robot within the next week.
Not too worried about the 1/4" Lexan, we’ve done a lot of similar things over the past few years with no issue, due to our inhouse CNC router which handles Lexan beautifully, but just can’t do aluminum. You have to be a bit more mindful than usual about creating a broad footprint for both mounting and supporting standoffs, but as long as that’s well done, they hold up fine. Our shooter gearbox this year used similar construction, and didn’t get the reinforcement from butting right up against the chassis sheet metal that these guys will.
Looking forward to playing with you guys at Battlecry!
How does the shifting actually work on this? I could be wrong, but it looks to me like you’d need the shifting assembly on the primary shaft to make this work, but I only see one on the secondary shaft (and I can’t figure out what it’s for, since it seems like it’d work fine with only one on the primary shaft).
The dog shift assembly is on the primary output shaft, just hidden by the CIM motors – what you’re seeing above is the much more visible encoder mount and coupling. The 70t primary reduction gear and 48t dog gear are bolted together and both free spin on the shaft, engaging with the output only when the dog is pulled towards the CIMs. Extending the shifting piston engages the 60t gear nearest the front face of the gearbox, such that power must pass through the reduction of the secondary shaft.
That makes a lot more sense, thanks.
Very clever! I like the idea of using extra reductions on a low gear.
what are the gear reductions and the rest of your drive train?
High gear is 6.4:1, low is 11.4:1, giving us speeds of about 8 and 14.5 feet per second, riding on 4" treaded colson wheels, set up in a belt driven WCD with sheet metal construction. Utilizing gears that we had on hand was a high priority for this design, as budget for offseason improvements is next to nonexistant, and this played a large part in driving these choices. Future iterations would likely have a larger reduction spread and faster high gear, as long as the game calls for it.
Any reason why you opted for a low of 8 fps instead of something like 5?
Did you make a custom output/shifting shaft for this gearbox? It seems that the stock WCP shafts don’t have enough room to fit two gears on the high side and still be secured through a bearing at the plate. We’ve been investigating a very similar concept and if the stroke of the shifting cylinder is limited to less than 1/2", it looks like it might be possible, with some creative use of spacers for the low side gear. I’m interested how you solved this issue.
We’re not comfortable making a stepped hex shaft like the WCP shafts are. The AM dogs work on 1/2" hex shafts, but the dog profile is different between WCP and AM and would introduce some backlash into the system when reversing directions.