pic: Flipped NEO AM14U3



This is a design I had been working on over the last month or so with Mini-CIMs, but was never happy with the resulting width. The NEO is short enough to allow the drive plates to be .5" closer together (while using chain instead of belts) than the stock AM14U3.

Besides the two red plates, this can be manufactured with hand tools. The larger plate carries the motors and bearings, and mounts to existing holes on the AM14U3. Then all that is needed is for clearance holes to be drilled for the bearing flanges, motor shafts, and standoffs. The tops of the wheels get quite close to the upper flanges, but those can be easily filed back.

GrabCad here: https://grabcad.com/library/flipped-neo-am14u3-1
Some details are missing, such as shafts, bolts, some spacers, and the longer chain runs, but I wanted to get some feedback.

Looks cool!

The long chain runs are a challenge. C-C distances are 11.617", requiring an odd number of links (137). They are practically on the floor. Tensioning them might need some creativity also – wrap some Telfon around the Neos?

AndyMark makes a chain upgrade kit for the AM14Us, so I assumed:ahh: the existing axle holes would work. I couldn’t find any directions or assembly instructions that would help with the tension, nor could I find any team that had used one. I realize now I don’t need to use the bigger sprockets on the non-motor side.

This is a really cool concept!

As a solution to the potential issues mentioned by Richard, it seems to be pretty easy to change the other wheel pictured to also be live axle. From some quick GrabCAD measurements there seems to be enough room for a bearing hole. Just run another chain from the left driven gear to the wheel, just like for the center wheel.

To keep with the thread of simple manufacturing, teams could make the added bearing holes using the techniques outlined in this thread.

The last wheel could remain dead axle and use smaller sprockets while maintaining the tiny width profile this design focuses on.

EDIT: Partial snipe

25 (0.25" pin pitch) chain is strong enough for FRC drive trains, but ONLY if it is properly aligned and tensioned. If your drive train design does not include good mechanical features to control those, then you should stick with #35 chain – heavier, and more forgiving.

If you decide to go with chain on the AM14U3, you should explore your options when choosing axle holes. While all of them will work (if the chain is properly aligned and tensioned), only some of them will allow you to use a chain with an even number of links. Doing that, and avoiding the use of master links as well, will help you dodge significant frustration in your team’s pit somewhere along the line.

Chain can and should be assembled as one continuous strand using a pin-press tool. This one is pretty good for 25. There are other options available for #35.

A center distance calculator is also a big help. I like this one.

I love love love the concept. But was just having a look at the cad and it looks like the bearings for the gearbox in the kit plate are interfering with the NEO’s. Is that just an issue with the step file I’m looking at?

No, you’re right, I had that plate 1/4" thick before, then changed it to 1/8" without checking everything. I don’t have enough room to change it back to 1/4" without moving the entire drivetrain plates another .5" apart. I could have the bearing go through both the drivetrain and the gearbox plate, but that’s exactly the kind of precision I’m trying to avoid with this design.

One of the reasons I didn’t initially do this is because there is very little room before the inward bellypan flange on the bottom. I was able to barely fit in an 18t sprocket, and after changing the CIM pinion to the 11t version, it sits at an adjusted 20fps, based off the published free speed of 5760 rpm.

https://i.imgur.com/pUJqXZ6l.jpg

GrabCadis updated to this picture, but the idler bearing still intersect the motors, as I haven’t decided on how I want to deal with that yet.

As much as I like the idea of putting NEOs in the wheel well, 3946 tried using stock KoP chassis holes with chain in 2016. Do NOT recommend without some additional tensioning; endless headaches. Also did the math with belts, didn’t seem any better.

Generally we talk about drivetrains in free speed, to prevent introduction of an arbitrary (albeit well sourced) constant which creates confusion. Assuming you’re using the JVN speed constant of 81%, that is an incredibly fast single speed drive. Is it too fast for the NEO? Maybe not. Is it too fast for the driver? Probably.

Part of the frame can always be cut out by an angle grinder to allow for a larger final chain reduction. That’s what I would do to get it down to about 15fps since the rest of it looks absolutely stunning.

Yeah its a lot faster than I would ever build, that’s just where it is at the moment. I’ve had to make quite a few changes to gear/sprocket sizes without adjusting the the rest to keep it at a sane speed. It just wasn’t a priority at the time. I was only able get it down to 19.5 free before starting to redo most of the gearbox.

https://i.imgur.com/Y43wDARl.jpg

GrabCadupdated

I was able to accidentally solve many problems at once, although it took a long time to get there.
The free speed is now 14.3 fps (was ~25).
Short chain runs are properly C-C’d and chain clears everything easily.
Moved the outer live axle to the furthest stock hole, yielding in an extra inch of wheelbase and room for motors.
Idler bearings no longer intersect motors due to larger idler gears.
Larger driven sprockets, does require angle grinder or the hacksaw and pliers trick.
Unfortunately the idler gears now stick up above the frame by .5"

I’ll have to figure something out IRL with one of the extra chassis we have in our shop, unless someone else has pictures of it done successfully.

These new NEO’s are conducive for some great new and innovative designs.

My only question would be having enough room to route the wires from the Neo’s back inside the robot. There does not seem to be any space behind the motors to get the wires out.

The NEO’s wires don’t come out the back, they come out near the front:

https://cdn8.bigcommerce.com/s-t3eo8vwp22/images/stencil/1280x1280/products/361/1213/NEO_HERO__39078.1542226252.png?c=2&imbypass=on

The wires on the NEO don’t come out the back, they come out the side near the mounting surface.

Thanks my bad. Still seems a tight fit.

Agreed, I am worried that there isn’t enough room to replace the motors without taking off the whole outer plate (and the outer wheels because their bolts hold it on).

The AM14U3 is not designed for easy maintenance like that. If you want to replace motors on the stock AM14U3, you have to take the gearbox apart, which is inside the chassis and usually not very accessible once superstructure is on. So only having to remove the side rail is an improvement in that sense.

Obviously, a different drivetrain design could mitigate those issues, but I do appreciate what OP is trying to accomplish within the confines of the existing (or at least most recent) kit chassis and some out-of-the-box thinking.