pic: Manually machinable one-speed gearbox



This gearbox has an 11:64 ratio, and is meant for a chain in a tube like drive. When designing this gearbox, I kept in mind what a low resourced team can do; thus, there are no fancy geometry or pocketing in this. In fact, this can be machined using a drill press and some good calipers. The rest can be COT. Comment if you want the CAD



This gearbox has an 11:64 ratio, and is meant for a chain in a tube like drive. When designing this gearbox, I kept in mind what a low resourced team can do; thus, there are no fancy geometry or pocketing in this. In fact, this can be machined using a drill press and some good calipers. The rest can be COT. Comment if you want the CAD

Looks solid. What are the two small holes in the center for? How is the gearbox mounted?

The two center holes are for the encoder plate, which mounts on the motor side. This is mounted to the frame by using the bolts that go through the spacers.

Thsoe bolts look like they’re more than 2" apart vertically, so it’s difficult to mount to 2x1 with that.
How is the encoder attached to the shaft?

Almost impossible to do chain in tube in 2x1 anyway. It’s probably designed for 3x1 or larger.

I meant the horizontal bolts. The encoders are mounted to the GB by an adaptor plate. The shaft is mounted to the encoder by either a shaft coupler or the surgical tubing method

I think it’s possible. We were planning on converting our competition bot to it in the offseason.
Using something like versablocks as bearing blocks allows for more than ample space in the 2x1 for 2 16t 25 sprockets side-by-side.

So this is mounted using the bolts along the bottom? I feel like that could cause the gearbox to flex back, but I could be wrong (something like that happened to our 2012 robot, although it was a slightly different situation).

The bottom bolts. I don’t flex should be an issue, because motor load isn’t that cantilevered

It looks like you’re trying to direct drive a wheel with this. 11:64 is far too fast unless your wheel diameter is significantly smaller than 4".

~16fps free speed according to JVN calc. We ran that last year.

Actually a COTS product: http://www.team221.com/viewproduct.php?id=133

Could some one explain how you get the chain in the tub. To me it seems like it would be a complex process.

tl;dr, it’s not easy.

That is one way to do it. I talked to a couple teams at champs (can’t remember which) and from what I heard from them it was a largely painless process after the initial installation. Could be something to start a thread about.
I’m sure that no matter which way you do initial installation it’s going to be harder than traditional WCD, but the upkeep should be simple.

3.5" wheels on the outside is one way to do it that I overlooked, 1/4" drop turns me away though.

Tl;Dr I’m not going to be trying it anytime soon…

Oh hey that looks familiar. One correction is that the sprockets were elevator not wheel sprockets. It wasn’t the easiest thing we’ve done but it wasn’t the hardest with the hours of CAD work paying off with a relatively painless assembly.

3467 was feeling inspired by our new manual mill in our shop we designed a robot that required our drivebase and elevator gearboxes to take up little to no space.

We packaged it all in 2" x 2" x .125" wall tubing. We could have used 1.5" x 2" but opted for the 2" since it made mounting up easier with the Vexpro tubing and a little more room to work with. We used 18T sprockets for 25 chain and spaced out the bearings to be C-C however something in the math was wrong so the chains had a little more slack compared to our C-C we ran with 35 chain the year before but since it was inside the tube and impossible to come off we didn’t bother adjusting it. We may have designed it to be less than C-C I’m not sure. As we found out the slack was a nice feature since the chain was easier to install.

Overall it was a very nice setup and we loved this assembly. Each module was responsible for driving the wheels and providing power to our elevator with a duplicate module on the opposite side and an axle higher up that connected the two sides of the elevator to keep them in sync.

IIRC after these were assembled we only touched them once on the practice bot in Week 6 due to one of the axles getting pushed out while adjusting one of the sprockets on the outside for the elevator. It took a while to get the spacers and sprocket back in line since the superglue had done its job during assembly and had broken away so we had three floating pieces inside the tube on each axle. Not ideal but it worked for what we needed and unless you pulled the axles out the inside was solid. After that minor mistake we never had an issue on two robots.

Some pictures of the assembly are available here. Weight as shown fully assembled minus wheels and encoders was 16.5lbs. We had plans to lighten the large gears but always found the weight in other places.

I would highly recommend you try this out at least once in the pre-season. Our original plan was to machine these on a manual bridgeport but we needed 5 of these made and holding those tolerances on one would have been difficult so I brought them to work and CNCed them. It was our first time trying a lot of things this year for the first time IN season which is typically frowned upon but we worked through a lot of CAD revisions until we were happy. Consider the impact running chain inside your tube has on how you can mount to your frame. Since ours was only running to our elevator it wasn’t spanning the full length of the tube so we had an easier time mounting cross members in the back and our vertical assemblies needed to be re-thought a few times.

If you have any questions feel free to PM echin the student who designed it & JamesB3467 did the gearboxes. Huge thank you to 118 for sharing some insight into their design that inspired our students to tackle something harder! Also to Vexpro for their great lineup of products making complicated mechanisms easier for teams to try.

Also for getting around the issue of adding a drop into your drive, Colsons are pretty easy to throw on a lathe and shave down to size with the right cutting tool. We did this in 2014 after we accidentally designed our drivebase with no drop.

To the OP: Looks good nice to see you trying something new, hope it works out for you guys. Something to consider, what is that outer plate doing for your drive that the tube/mounting surface can’t provide? For some designs you might need it, others its a nice way to eliminate a part and shave some weight out of the base. Just a thought.

Yup, gonna echo a lot of what Brendan said.

We ran our chain in the 2x1 tubing after being inspired by the local experts 118, and were extremely happy with the results. The drivetrain was probably one of the best we’ve built yet, and we had exactly 0 issues with it during the competition season. And the best part was we didn’t have to do any real maintenance on it.

As for the installation, we just put the chain together (no masterlink), put the sprockets on the chain (not fixed, just loosely held by the chain), then dangled it into the tube as we put the shafts in place.

Worked great, 10/10, will probably run again!

So just confirming that teams who have done this use vex pro hardware and sprockets in 2x1 .125" wall tubing?

Could you elaborate on this? We had a less than stellar time trying to shave down colsons this year.