Sharing a design for a 1.5 stage shifting gearbox. It uses Falcons, but it could easily use NEOs or CIMs. Inspiration came from revisiting Patrick W’s 1-Stage Shifter from 2017. Also many thanks to @PatrickW for design input, this gearbox is more of a collab than my own design. I think it turned out pretty sweet!
High Gear is 10:54 (JVN adjusted speed of 16.7 fps with Falcons)
Low Gear is 10:16, 9:60 (JVN adjusted speed of 8.5 fps with Falcons)
Weight of gearbox, pneumatic and motors <4lbs (I don’t have 100% accurate weights on all the shifter components)
Current version shown is Chain In Tube, first version had chain on the outside. There’s also some earlier 3 motor versions that use normal pneumatic cylinders.
Wow, it took me a minute or two to figure out what was going on here. Very cool concept, and incredibly compact. Has anyone here actuated the shifter with a cantilevered bar like that before? I know I’ve seen it done with one piston on each side.
Also, here’s a cross section for anyone else struggling to see how this works.
Oops, I meant to include a cross section in the imgur album. Thanks for grabbing it. One of the really clever bits about using the cluster shaft to put a ratio between the two ball shifter gears is it totally solves any issues with gear clocking.
The best way to describe it is both motors directly drive the 54T ball shifter only. In low gear, the 54T gear is just acting as an idler between the motors and the 16T gear on the cluster shaft.
The 33 Killer Bees 2004 4-speed transmission used a similar connecting rod on both shifting shafts (#throwback). The original whitepaper link is broken, but here’s a photo from 971’s copy in 2006 (That robot had a lot of problems, but shifting wasn’t one of them).
The non-rotating cylinder theoretically helps resist binding. There’s also no reason you couldn’t do a normal air cylinder acting in line with the ball shifter shaft, but it makes the whole thing a lot longer.
I was actually looking for info on this to try it with the NEO between wheel ballshifting gearbox, and happened to find my conceptualized setup on 118’s bot this year.
Shifter Pneumatic position is opposite the shifter shaft plunger, so it won’t ever move. It’s a CAD error, Patrick is right that if you make the screw that attaches to the plunger 1/2” longer that fixes it. I’ll update the CAD later tonight.
Really cool design on this Evan! I attempted to do something very similar with 3 cims a few years back, but I was never very happy with how it tuned out. I love the use of a cots shifter shaft.
Just from the pictures above, I have a few minor structural concerns that I think could be easily remedied.
The bearings on the small cluster shaft seem very small to me. Maybe they would be fine, but I dont see a reason they need to be so small.
Looks like only two bolts to attach the gearbox to the frame. You might want more.
I would be slightly concerned about the double sprocket having a large counterbore you could face down the end of the vex shaft instead.
You could definitely make the bearings bigger in this version. They were sized based off of the 3 motor version. They are the same size bearings that go on the Falcon shaft so I’m not too concerned about it. Extra bolts is probably a good idea.
I personally would run the chain outside the tube and not bother with modifying either the sprocket or the ball shifter output since I like having access to my chains, but it looks pretty in CAD and it’s a popular option. Facing down the shaft is certainly feasible.
Very nice clean compact design, couple of questions. What is being used for the cluster shaft? It looks like the 16 tooth spur is 3/8 hex and the 9 tooth spur would need to be Falcon spline, would this be a custom part? Also, when doing a chain it tube, what tube size is used, and is that the 17 tooth AndyMark sprocket? We at 236 have never done CIT due to the chain clearance when using a .125 center drop in .125 wall 1X2.
The 9T gear is the CIM version (217-6335) that VEXpro released last year. I’d assume they’d press it onto the tiny custom shaft.
They’re using the 17T AndyMark sprocket with what looks like 2x1x.125" tube. 319 ran the same setup in 2019 and ran a .100" drop by moving the middle wheel down .050" from center, and the outer wheels up .050" from center.
I don’t think more than 2 bolts is needed. 319 has attached our gearboxes with 2 bolts since 2016. The bolts only need to keep the gearbox against the drive rail, and to counteract the torque applied by the motors via the wheel. Two bolts is plenty enough for that.
I won’t speak too much on why you should do this, but it makes it easier to justify, especially as a safety net, when it’s negligibly larger than the single speed option.
In fact, I’d argue the safety net reasoning is a large part of the reason that 254, for example, has continued to run 2 speed gearboxes, despite running in a single speed the past couple of years. You would have to talk to them to get a definitive answer though, and maybe that will change this year.
This might be a good spot to do some load calculations. My intuition tells me that the bearing next to the 9t gear could see about 4x the load on either falcon shaft. However, because of limited traction, the Falcons will probably see more load in high gear.
I think you are right actually. If there were 3 cims in a conventional wcd configuration, or if the tubes had a thinner wall, I would be concerned about the weight of the gearbox deforming the tube. But that’s probably not a problem in this setup.
I’m not sure there’s any intent to actually build this, it was mostly just a design exercise to flex the brain muscle a bit before the season, and because I ran out of single speed gearbox ideas to try and it seemed like a neat design challenge.
I’m a huge proponent of single speed drive gearboxes in general, but there’s still use cases for shifters and this particular design is nice for packaging if a 3 motor gearbox doesn’t fit or you need to use only 4 motors total on drive for some reason and don’t want to sacrifice on top speed.
