ABS-114 is a shifting gearbox with a high gear reduction of 5.82:1 and a low gear of 19.78:1; a total spread of 3.4. This results in output speeds of 16.8fps and 4.9fps on a 4" wheel.
Both reductions first undergo a 64:11 reduction from the CIMs. The high gear is linked to the carrier plate for the planetary gearset, which directly drives the output shaft. The low gear goes to the sun gear of the planetary for an additional 3.4:1 (72/30 + 1) reduction.
The PTO is not powered while the gearbox is in drivetrain mode, and vice-versa. This is accomplished by using the differential aspect of a planetary gearset; either the ring gear is locked or the carrier plate is.
Weighs 2.8lbs without CIMs, and occupies a space smaller than a WCP 2-cim shifter. I think it is possible to get it lighter if one machines the planetary gearset out of aluminum, but this requires so much machining as-is I would not try that.
That’s a lot going on in a little space; very impressive, especially if it works.
The CAD or at least a few renders from other angles would be helpful to understand it all, but let’s see how much I can get it from the one render:
Is there a second pancake (or other) cylinder hiding under those CIMs, probably coaxial with the drive shaft that accomplishes the drive train shifing, or is the cylinder on top some sort of 3-state device?
Where is the PTO output located? I understand it’s linked to the ring gear, so the most obvious place would be behind the drive shaft, where I thought the shifting cylinder would be located. Doing this would not only move the shifter out of the obvious place, but it would be difficult to pass all the different coaxial shafts through the large gear, so I’m guessing that the annular gear has teeth on both the inside and outside of the annulus, and the PTO is another spur gear that takes off of that. Doing bevels would require a lot of extra work and thrust bearings for no discernible benefit.
If I understand this right, the PTO output would have to have the shifter in high speed/low torque (single stage) mode, and the carrier is locked, so the output gear ratio would be (64 / 11) * (72 / 30) = 14.0:1 at the ring gear, possibly higher or lower if the PTO is not directly taken from the ring gear.
This is incredible… After half an hour of looking at ABS-111 (the latest one I can find online) and googling planetary gear ratios, I think I sort of understand now. Time for questions:
[li]which one is the carrier plate?[/li][li]how do you lock the ring gear/carrier plate? I’m assuming there’s a custom piece of metal the render is hiding… I want to see it.[/li][li]checking my understanding: so there is a “suicide mode” (high gear/PTO) where the carrier plate is linked to the input but the carrier plate is locked?[/li][li]consider using Andymark’s new CIMcoder? I hear that it’s thinner than the 3D printed one[/li][li]how does one machine their own ballshifter? (looking at ABS-111)[/li][li] Is that 60t gear hitting the bolt next to it? ::safety:: [/li][li] how close are the CIM gears to hitting the gearbox plate? ::safety:: [/li][li] that’s a cool way to attach something normal to a gearbox. I might steal one one day. [/li][li] what are those 6 holes near the output for? I assume 2 is for mounting a piston on the other side, but 6?[/li][/ul]
[strike]I think the PTO is the front-most gear we can see, and it’s connected to the gear on the left, which would mean that the PTO shaft is to the left of the 2nd pancake piston. Looking at the 2 PTO gears it actually look like there is some sort of reverse-reduction there. [/strike]This is all conjecture though, until we get more renders/CAD.
Edit: can we at least get a render of the internals?
The new AM Cim moter encoder is thinner than the original CIMcoder however it is twice the costs and you reclaim only ~.25" of shaft. Plus the original CIMcoder is designed for an extremely robust encoder with programmable resolution. Of course I am a little biased towards my own design.
I only see two gears, the 11-tooth pinion of the near CIM and the 60-tooth gear it engages. The two structures behind the front plate do not look like gears; the spacing between “teeth” is uneven, and the tooth profile doesn’t make sense for gears.
I also don’t see a second pancake piston; are you assuming it’s below the CIMs coaxial with the 60-tooth gear?
ohhhh I see now. What I assumed to be a gear for the PTO, the thing right behind the front plate, is the carrier plate (answers one of my questions). What I assumed was “gear teeth” now look like the geometry that locks the carrier plate when the piston pushed down. Everything I said earlier is nonsense then :o. My revised guess is that the PTO is the piece of metal connected to the ring gear with a 1.875’’ bolt circle on it, connected to nothing yet.
Yes, I’m assuming that there is a second piston there based on ABS-111, the closest iteration of this I can find online, plus I can’t think of a way that the top piston can possibly reach and engage the sun gear even with a third state. Edit: Also, the holes in the gearbox plate near the output, assuming identical gearbox plates, suggest a mount for the Vexpro pancake piston.
I forgot to update the Partner Space recently it looks like. I just finished, so you can take a look at the CAD here. As a side note, ABS-110 (the PTO shifter with the 3-position cylinder) is 0.7lbs lighter than this and far easier to machine. It’s also about as compact. The only disadvantage it has compared to this one is the lower spread; 2.3 versus 3.4.
Yes, there is another Vex pancake cylinder underneath the CIMs.
The PTO output is a 32 tooth #25 sprocket that sits in between the 60t gear and the ring gear. It is linked to the ring gear via an aluminum hub/plate. Only one shaft; no double-tooth things involved. None of the gear teeth have to be machined fortunately.
That’s right, the carrier is locked and the sun gear is engaged to make the PTO work.
The carrier plate is the frontmost aluminum plate, in front of the brown steel ring gear.
The odd tooth-looking dealies on the carrier and ring engage with 1/4" steel pins with rounded ends to lock them. The cylinder simultaneously disengages/engages at the same time.
There is indeed a “suicide mode”. Hopefully one would check to make sure the code is working right before activating that endgame mechanism. :eek:
Oddly enough, the 3D printed version is better because I actually want less CIM shaft protruding out. It can intersect with the PTO chain if it’s too long. I have the second one to act as a spacer.
Usually it’s very tricky to machine, but this one is easier than most. Start with a rod, turn down the outside dimensions leaving 13.74mm for thunderhex, bore out the cylinder for the ball part, drill through with a 1/4" drill bit. Then take it over to the mill, stick it on a dividing head, drill the 6 ball holes and mill down the hex. The gears you just stick on a fixture and CNC.
The linkage thingy between the shifting cylinder and the shifting rod is something of my own design, utilizing a 3/8"-16 outer 10-32 inner threaded insert and a couple bearings. Very easy to machine and more compact then the Vex option.
Yes, that bolt is. Good catch, I must have moved something. I just replaced it with a countersunk screw so it no longer touches.
The CIM gears are about 0.6" in diameter, and the hole in the plate to accept the CIM is 0.75", so no danger there.
Thank you, I liked it too.
2 for mounting the cylinder, 4 for mounting. This has a couple options for mounting; you can use the bottom 2 screws that keep the gearbox together or just the 4 holes in the front plate.
There is no real “PTO shaft”; there is a single shaft. The PTO output is attached to the ring gear, which is possible due to epicyclic (planetary)/differential gearing shenanigans.
I would try for an internal render, but the internals are rather convoluted so it’s hard to get everything with one picture. Let me give it a try though.
I hope I’m misreading this, but I can’t figure out how to make sense of this. If the PTO output comes down the same shaft as the wheel drive, what makes the PTO mechanism move in one case but the wheels in the other?
Whoops, I didn’t mean it that way. The PTO output is a 32t #25 sprocket with a versakey. It is bolted to a plate that is bolted to the ring gear. The only interface it has with the main shaft is a bearing it rides on to keep it centered. Basically, the PTO output is a “dead axle” sprocket, not a shaft (which I dislike but is hard to work around).
So that thing that looks in the render like a round disk between the ring gear and the 64-tooth gear is actually a #25 sprocket? This would certainly be convenient in that PTO could be geared up or down by a factor of about 2 through selection of a PTO receptor sprocket.
The round disk is a hub for the #25 sprocket, which is hidden behind it in the render. The GrabCAD model should clear it up.
You can use pretty much any sprocket with a versakey pattern as long as it’s not large enough to hit the standoffs.
Took me a bit to figure out how the carrier plate ever gets unlocked, that is neat Just don’t apply power while actuating the PTO cylinder!
It’s hard to imagine you’re using the CIM encoder to line up the locking notches with the cylinder. Can the rounded end of the locking pin ride up the outer contour to get into the notch? Why not make the locking notch the only groove, so the locking pin rides the OD until it engages? Like the locking plate on 118’s arm in 2014: http://s22.postimg.org/3sk056lxt/Screenshot_from_2016_01_03_19_25_33.png
Also, for the locking pins, I wonder if you could use something like mcmaster 84835A22 to reduce friction if you want to actuate the PTO cylinder and slide it in into place. You’d have to turn the gearbox slowly to avoid wrecking the lock after PTO engage and before drive disengage.