pic: 6WD FLIPPED 8x775 SS


Seeing as our team already has a couple go to standard 8x775 gearboxes done, I figured I would try something different.

This drive utilizes double sided GT2 timing belts from Gates/SDP and the GT2 pulleys from WCP for the first stage of reduction. 12T GT2 pulleys from WCP are pressed onto the 775 Pros, and they drive two 60T GT2 pulleys also from WCP. The 1/2"-3/8" hex adapter from WCP is utilized on the bore of the 60T pulleys, so that a 14T 20DP gear from VEX can run on the same shaft, on the opposite side of the tube. That gear drives a 46T 20DP gear on the wheel shafts (used a 44T gear in the CAD model as the 46T gear file seems to be missing from the VEX page). 22T sprockets from VEX are used to transfer the drive from the middle wheels to front wheels. The belt system with the gears powers both the middle and back wheels. A Grayhill 63R encoder is mounted to the tube and has an 18T printed gear pressed onto it and driven off the 46T gear on the wheel shaft.

Overall weight is 19.83lbs. This is the first drive I have successfully been able to get under 20lbs realistically. The frame perimeter is 30.5"x25.5" (adhering to the 112in frame perimeter used in past seasons). The wheel base is 24" L x 22" W, with a 0.06 drop on the center wheels. All the wheels are 4" diameter 1.5" wide colsons with 1/2" hex hub. The hubs on them have been machined down to the center width to save weight, something teams have done in the past on the 4" wheels. I used 4" wheels so that they could easily be swapped for omnis, tread wheels, etc. (would love to see either a 3.5" r 3" omni, or a 3.25" colson for more crossover between wheel types).

Overall, I’m very happy with how this turned out. The big question mark would be whether the GT2 belts can hold up to the load on a drive train. I need to delve into the Gates documentation some more and this would probably need to be tested to know for sure.

Things missing: not much. Every fastener, bolt, rivet, etc. has been included in the model. I would probably put a 3d printed cover around the belt system for protection and utilize the 3d printed motor covers for the 775 that some teams have used.

As always, comments/questions welcome.

Do you have a tutorial on how you did that lightning on the boxtube? its really nice and allows for easy bolting unlike our teams current style of the triangles.

Wow, this is incredible. I plugged in your description of all the pulley sizes and gearings into the JVN calculator and free speed seems to 19.90 ft/sec (adjusted to be ~16.12 ft/sec), which seems a little on the faster side for a single speed drivetrain.

Of course the nascent question is whether those belts would be able to hold up and endure running the drivetrain for a day’s worth of qualification matches, but I’m also concerned about the positioning of those 775pros on the outside of the driverails. How would you propose to ensure that they (and their wires) are properly protected from outside forces?

Congrats on creating an incredible <20 lb drivetrain! I always look forward to seeing your CAD posts on this site.

How are bumpers being mounted?
Any worries about the 775pro wires being so close to the frame perimeter?
Are there tensioners on those belts?
Are you worried about that lightening being a bit aggressive?
Should you need to do maintenance on any of the belts in the tube (for whatever reason), is there a way to access them without disassembling the whole chassis?

Other than those few points, it looks very nice.

305T belts? 15mm wide?

Working tension will be 114 x (15/25.4) = 67 lbf.

4x stall torque is ~26 lbf-in on 12T pulley, or about 130 lbf-in at one 60T pulley (with the other momentarily unloaded). That is about 115 lbf belt tension.

Not a lot of margin. :rolleyes:

This is one of the best uses of space on a drivetrain I have seen in a while! Great job!

300T belt. Parthere.

  • Not too worried about it. As I said I would probably put the printed 775 covers on them to protect the terminals, and the bumper would cover the motors as well.

-There are not tensioners on the belt. The positions of all the pulleys were done with the correct path length for the 900mm belt. If it proved to need some tensioners, some radial bearings could be bolted to the tube to accomplish this.

-I am not worried about the lightening being too aggressive. I’ve found that once the entire robot is build the other mechanisms and structure will help tie together and strengthen the chassis more. Not every subsystem has to hold up on its own, as long as they all hold up together.

-There aren’t any belts in the tube. There is one run of 25 chain going from the middle wheel to the front wheel. We and many other teams have run chain in tube before with no issues.

I agree it is a bit fast, but I figured if I’m only running a single speed I’m not pushing my way through any defense, so might as well be nimble. This drive would definitely require some good controls when it comes to current, etc. so that if you do get hung up by defense the motors don’t burn up.

I do not have a tutorial, but it is pretty basic. I basically draw a hexagon using the sketch tool centered between the two wheel bearing holes. I dimension one of the sides of the hexagon, and then draw more lines to connect all the points to the center. Draw a 1/4" hole in the center of the hexagon. All those lines then become construction liens. From there you do offsets of all the lines to create the actual cutout profiles. After that you just linear pattern to each side as far as you can before hitting the wheel bearing hole. Extrude cut. Fillet all the corners and done.

[strike]Unless you were planning on building another complete superstructure a few inches above the top of the current chassis that extends beyond the wheels, I don’t see how you could legally mount bumpers (according to 2017 rules). 4x 775pros and 2x 4" colsons side-by-side are 15" wide minimum. That greatly exceeds the max gap distance for bumper mounting. The only possible way that I could think that follows the rules is if you can squeeze a bumper mount underneath the middle 775pros, but that doesn’t look very feasible from the render. Is this something that you considered?[/strike]

Completely missed an obvious solution. This is what happens when you decide it’s a good idea to post at 4am. Good night everyone

Could just add a 2x1 that is cantilevered off the edge of the frame.

Maybe I’m missing something. I’m not a mechie, so feel free to point out the obvious. But, how does the belt engage the pulleys on the outer two motors? Looks like the back side of the belts would be wrapped around them, do you have a belt with teeth on both sides?

Yes. Click the links in posts 5 and 7.

Double sided belt.

Just a friendly reminder that unpocketed 1/16 wall tube is significantly stronger and nearly comparable in weight to pocketed 1/8 wall tube…and vastly faster to manufacture.

Really cool cad. If only we could properly power it without popping the breaker. Adding more motors essentially just adds thermal mass at one point :(.

Harder to put bearings and screws into though.

As was said I would need to add plates for all the bearings and such, which really isn’t too big of a deal.

Do you have a good source for 3x1x1/16? McMaster doesn’t have it and previously suggested sites such as online metals, etc. do not either.

Luckily it’s not a huge deal for us to get something like this machined.

Really impressed with the packaging of the solution, terrified of impacts. We have bumpers covering every inch of tubing on our last two robots and even with unpocketed .090 and .125 we still had some sizable dings.

What makes you say that?

Assuming the tube is 6061T6, 35ksi YS, the bearing strength for a typical flanged bearing installation would be:

.063in1.125in35,000psi = ~2500lbf

PEM nuts and riv-nuts install just fine into .063in wall tubing, as do rivets.

I see no barrier to using 1/.16th wall tubing, assuming it was available. I would bet $1 that it would be substantially stronger than the pocketing you’ve put in the current design.

I do really like the design though! Quite creative placements of the motors, and excellent use of double-sided belts to make it happen.

I distinctly remember a bearing in a 2016 drivetrain on 1/16" wall tubing failing catastrophically. The picture was posted here on CD. For that reason I like to use 1/8".
It’s also a lot easier not to crush 1/8" wall tubing when putting screws through it.

Yes, I think this happened to 230 in 2016.

Additionally, using 1/16 wall material to support a bearing directly can be problematic due to the undercut near the bearing flange - if your material is too thin, you can “fall in” to that undercut when you press the bearing in, losing support / alignment / precision.