pic: Another Controversial Gearbox



Here is my take on a 775pro gearbox. If we were to actually to use this on a competition robot, we would most likely current limit all of the motors to 25-30amps. The gear ratio is 17.65:1, uses 12 to 100 32dp, and 34 to 72 20dp. This ratio is a little fast but It was just a trial to see how 4 motors would fit in a gearbox. This setup uses chain in tube. Feel free to ask any questions or leave suggestions.

I like this design quite a bit. It seems it might be a bit tough to install the chain-in-tube, but other than that it’s very clean.

Having done chain in tube before, it actually isn’t as hard as you might think. Yeah it can be annoying to do at first but the mentality I have with chain in tube is set it and forget it.

This looks good, although I would worry about the wires possibly extending outside of the frame perimeter, assuming that’s a rule next year. (and the possibility of another robot damaging the wiring)

I plan to use the Ozzy Boards 775 Inline Powerpole Connecter in this setup. There would be no wires on the ends of the motors using those.

http://ozzyboards.com/775-inline-powerpole-connector/

I have always been curious. Do you use the team221 double sprockets like robonauts used to use or do you use two vex pro sprockets? Never doing an intube chassis before I didn’t think the vex sprockets would fit

Sent from my SM-G950U using Tapatalk

When I did it before I used an 18t with no drop. Because there was no drop, an 18t worked. In this setup I have 16t in there at the moment because the Versablocks do not have a centered bearing hole (for drop center setups), but I haven’t checked to see if an 18t will work. Theoretically the 17t that 221 makes would be best for chain in tube, but I have not used them before.

Nice design Cooper!

How is the meshing of the pinions on each motor adjusted? Do you rely on really accurate machining on the plates?

How is the gearbox attached to the tube? The two larger diameter spacers?

It sounds like this design has been designed for zero drop. Would it be better to have a minimal amount of drop, say 1/8" in case the floor/carpet is not flat?

Versaframe is designed to allow 18t sprockets in a chain-in-tube configuration when using the versablocks IIRC, so you don’t need to go to Team221 unless you are running 1/8" wall tubes (as opposed to 0.1" VF walls).

There is no meshing adjustment. This plate would ideally be water jet or laser cut, and if not CNC routered.

I have run drop center drives in the past and to be honest, I just don’t like them. They provide inconsistent results in autonomous, and turning scrub and wheel wear is really annoying. Our 2017 drive had no drop and we ran 4 omni 2 colson and loved it other than turning being a bit quick, so this setup should help slow it down. Our team will most likely NOT run drop center in the future opting for a 4 colson 2 omni, or 2 colson 2 omni setup. This drivetrain I have created reflects the 4 colson 2 omni setup with no drop which I believe a lot of teams had success with this year.

The gearbox is attached to the tube by two 10-32 bolts through standoffs that you see at the bottom as well as clamped onto the tube with half of a Versablock. There are 2 tapped holes on the wheel side that allow the Versablock to hold the gearbox. I have attached an image.

https://imgur.com/a/k1p2f

Out of curiosity, what kinds of autonomous routines have you run that led you to this conclusion? We found a 1/8’’ drop-center to be remarkably consistent this year when following motion profiles generated from Jaci’s pathfinder; wheel scrub was very well-approximated by simply using a wider “effective trackwidth” than the geometrical trackwidth.

We have had more of an issue with the robot rocking different with large changes in acceleration. Not so much turning scrub. Turning scrub only becomes an issue later on when the wheels have worn from scrubbing too much causing turns to be different, as friction increases with wheel wear. In all it just eliminates one more variable and gives us one less thing to worry about that can cause problems in keeping auto routines consistent regional to regional.

Ah. We use trapezoidal acceleration, so this hasn’t been such a problem for us.

2882’s 2017 robot used a kit chassis and by the end of the season, the middle wheels had worn so much there was essentially no drop.

Will the 10-32’s through the spacers crush the tube or will the heads be inside the tube? Crushing the tube would mean your gearbox would start flopping around :frowning:

How bad this gets will depend on where your center of gravity/mass is. One of the first things I teach the FLL teams is to increase their accuracy and repeatability by not starting and maneuvering at 100% power all the time. This past season, I was watching the programmer for a 3 digit team tuning the turn in their auto routine out at the KISD STEM Center. She was pretty frustrated because she kept getting about 7-8" of variation in where their robot hit the side peg. When I told her to cut the power when making the turn, the amount of variation dropped to something like 2-3".

The heads of the bolts are inside of the tube.

With motion profiling this year we were able to do some pretty incredible things.