(direct link: http://imgur.com/nSQHV04 )
2x 775 pros each side on ~12:1 ratio for ~20fps ridiculousness on 3.25" VEX wheels…all fit inside a 2x1.5" tube!
Traction in the back and omni in the front for predictable autonomous performance in off-season programming training.
Not posting CAD source yet because there’s still a couple interferences and more than a couple “line fits” (what are tolerances even).
Looks neat but the cutout you have for the motor causes a LOT of strength to be lost in the box, a hard impact would imho fold the box up at that joint.
Mechanical design aside, a 4x 775pro drive at 20 fps is asking for a lot of trouble. 775pros are not CIMs. For a fully weighted robot, you’ll be drawing 125A per motor at stall (assuming that you’re running them at 12V). Even on an 80lb robot, that’s 65A. If you look at the Vex motor testing, these motors can’t even take half that current for a whole match. There’s a good chance you’ll smoke the motors just accelerating from a standstill, and if not they probably won’t last a whole match of driving around at that current. You might be able to get away with 2x 775pros on each side if you gear it down a good amount more and have a very light robot, but it would be a lot safer to add another one (if not two) to each side.
Doesn’t look like there’s room for 1 of the 2 attachment fasteners for each motor.
Title is misleading, the 775pros are not in a tube. 
-Aren
This is a really neat idea, but the main issue is accessibility here. I wouldn’t particularly want to be the pit mechanic who has to realign or even replace a belt inside of some aluminum tubing. Cheeseholeing would make it better, but I don’t think it would be a perfect solution.
All else aside, I’d be very worried about using a 4 total 775pro drivetrain. Teams with 8 775pros still have to run current limiting and other software tools to make sure they don’t run into problems, and I think 2 is asking waaay too much of these motors.
With your design as is, it looks like you could just add 2 more 775’s in the same or very similar configuration to the other wheels, and then you could potentially get away without chaining the two sides, pushing your 2 by 1.5 to the much more common 2 by 1. We had better results running a similar setup with slippery pig when we chained the 2 shafts with with separate motors together, but I suppose the design could still work fine without it.
Looks like a really creative concept and I love your packaging! Keep up the good work and keep iterating cause I’d love to see some cad.
I think this needs to be highlighted. The 775pro is an good motor for FRC, but it leaves a LOT to be desired as a drivetrain motor. Just dropping one in is a recipe for disaster. It takes a lot of engineering and software that most folks won’t think about to make this work.
What you have to watch out for when not chaining the front and back wheels is if you push against something and either the front or back set of wheels lift off the ground you’ve lost half of your drive train power.
What makes belt-in-tube more problematic than chain-in-tube? Dangle the belt (or assembled chain) in on one side, pin with the axles, design so misalignment can’t happen. Even 4901 made chain-in-tube work after a brief learning curve.
That said, the concept of two-775s-on-each-end to spread the load (and the related gearing) would make this more difficult. But that doesn’t strike me as a blank-in-tube issue, really.
Belt in tube is surprisingly easy to put together. Easier than chain in tube, because the belt isn’t constantly folding over itself or otherwise doing anything other than being a loop. 228 and 2791 chain in tube drives take about 5 minutes to put together per side. It’s really not that hard, and you do not need access windows to do it.
No comment on the strength of the belt in this design or if it’s harder when there’s also a gearbox in there etc. But it is probably undersized for the application and you should not use a 9mm belt / 18t pulley in an FRC drive like this, if that is what you’re using.
Thanks everyone!
This thought experiment is for a ~50-lb, 20"x20" test bed, with calcs showing a ~50A stall current. And would still run current limiting through the Talon platform, and still needs better motor ventilation.
And I’ll still probably smoke at least one in the process of learning ::safety::
The tube width driver is the stock aluminum gear widths (and bearing widths), not the belting - I would need to run steel teeth to get the width down and get inside a 2x1. Could be kind of fun but I’d have to plumb the depths of SDPSI and Motion Industries to get there.
(yes, I definitely need to clean up the motor clearance cut! :ahh: )
@Aren: Sorry to disappoint, not all of us are ready to get on your level yet 
This is a PWN-ivation drive train right here. 775 drive motors with the drive base that 1625 likes to run (two omni and two regular wheels) 
Just for reference the 775pro drive for the 1296 robot in 2016 “Mittens” was 2x motors per side, and the final weight was 78lbs, geared for 17fps free speed. I am confident with the recent current limiting options on the Talon we would’ve been fine. The robot had a ton of troublefree drive time at the shop.
Also they were running at a full 12v peak, no silly voltage limiting.
-Aren
