pic: Manually machinable one-speed gearbox

That’s a lot of weight you’re gaining though (~2-4lbs depending on drivetrain), and you would have to settle for 1/8" wall probably. If teams say the chain doesn’t break, then it doesn’t break. That might be okay if you have the space and weight and mounting issues solved, because then you get a decent efficiency bonus.

This season, 1477 used a Texas Tube style chassis. On a 4WD with all omni wheels, we used VexPro 16t #25 sprockets. We had tensioners on both our practice bot and our competition bot, and never used them on either robot. We had no issues with the 16t sprockets, and our chain runs were more than 200 links long. Game permitting, we will likely be running a Texas Tube style drive chassis again. Our experience with #25 chain in 2x1 tube has been outstanding thus far.

This is something I’ve been thinking about a lot lately. If using 1/16th wall 2x1 tube, it is theoretically possible to run 24t pulleys and 9mm belts from VexPro inside the 2x1. This could result in an extremely lightweight and efficient drive, at the cost of some structural integrity in the drive chassis. In my experience, #25 chain is less likely to break than 9mm timing belts, but is heavier and less efficient, assuming both the belt/pulley and chain/sprocket are assembled correctly.

Here is a picture of 2791 running timing belts inside tube extrusion.

2791 has continued iterating on their belt-in-tube drive since then. Chris Picone would be the guy to talk to about that.

192 has also used belt in tube for the past three years. 15mm wide GT2 in 2x2 .125 wall. With our set up and access points it’s incredibly fast to assemble and tear down even though you’ll never need to.

It is possible to run 15mm wide belts in 2x1.5 but you’d have to bore out the pulleys around the bearings or move the bearings out. I wouldn’t recommend 9mm belt in the DT unless you had a proper tensioning system like 971 and even then they have had issues breaking belts.

How are your access points for belt in tube set up? How is it assembled?

Been running belt in a tube structure formed from sheet metal for the past three years with no issues. The sheet metal works great because we can start with a readily available belt/sprocket combo and size the tube accordingly, rather than the other way around. Access is achieved by opening up the entire side of the tube, which makes installation a lot easier than the through-the-end approach or small access pockets required by extrusion.

It is a fair bit larger than the texas tube approach, due to the increased sprocket diameter/width, and the need to add additional clearance for the rivets which hold the tube together, but it works well with our resources, and we like the insurance it provides.

One thing I want to make clear is that with the Texas Tube approach, it is almost physically impossible for the chain to jump off the sprocket if assembled correctly, because of the constraints of the tube. For failure to occur, the chain would have to essentially break, the risk of which can be minimized by not using masterlinks. Even if the chain loosens and rubs against the tube after a lot of usage, it will still probably be safe (as the other teams have noted, all these robots have been through multiple regionals and championships, and in the case of 118 some very defense heavy games like 2013 and 2014).

Other benefits include:
-Low maintenance (basically assemble and never touch again for the rest of the season)
-More space and weight efficient
-Relatively simple, no special equipment besides a mill required
-Keeps things clean and out of harm’s way (the number of times I saw noodles getting stuck in the chains of other teams was too much to count, because of Texas Tube, we had virtually no problems driving over and around the noodles).
-This has more applications than just the drive train!

Above is an image of 192’s 2014 base about to be welded up. There are two rectangular holes on the tops of each side as well as 4 holes in the side in addition to the bearing holes. First the belts go in the rectangular holes in the top, then the pulleys are slid into the side holes and rolled down the tube to their bearings holes. The gearbox is put on, axles are put through the pulleys, bearings popped in, and wheels are put on. Assembly time is about 5 minutes a side.

Abhishek, I’m not doubting the effectiveness of Texas Tube using chain. The results speak for themselves. I’m merely exploring other options. My students expressed interest in trying out timing belts for the drive train, so I just wanted to see what experiences other teams had with it.

I don’t understand. This thread talks about 2x1x.125 tube being used for Texas Tube. Why would we have to “settle” for 1/8" wall?

If you designed to package your drive in a 2x1 tube you could use the .1" wall tubing from Vexpro to shave a little weight down. 2x2 doesn’t come in .1 wall and it wouldn’t be wise to press bearings into .0625" wall 2x2 in a drivebase.

Drawing a quick 36" tube in Inventor I had the following weights:

.1" wall 2x1 Versa Tube: 1.97lbs. (0.65lbs. per foot)
.125" wall 2x1 tube: 2.4lbs. (.8lbs. per foot)
.125" wall 2x2 tube: 3.3lbs. (1.1lbs. per foot)

It all depends on what you have planned and are building on top. We had 2x1 Versa tube uprights coming off of our setup so 2x2 tubing worked very well for us. With the extra space it also can make mounting off of the tube a little easier and you can lighten the tube to bring the weight down a little. My personal preference would be a lightened 2x2 tube but that’s just me.

I know, I wasn’t referring to the belt system. I just wanted to say that in case people who haven’t seen the method are reading this thread, because when talking to people in the pits at competition, they ask if we are ever afraid of the chain failing inside the tube where we don’t have easy access to it and if we should cut access holes.

Thanks for doing this. Honestly, I’m not overly concerned with the extra weight. You’re likely to save a little weight by using belts instead of chain, so it may cancel itself out.

Okay, it just seemed like you were on the defensive a little bit. Thank you for spelling out the benefits. I love the idea of Texas Tube just because I really like the efficient use of space, but there are always those on every team that question its ease of use.

I do have more questions though. How would you get a continuous loop of chain without a master link? My team has only ever used master links with our chain.

Buy one of these for 25 chain (also available at bike shops if you google enough):
http://www.team221.com/viewproduct.php?id=70

Follow these steps:
http://www.billetboard.com/showthread.php/7013-How-Too-Use-a-Dark-Soul-Chain-Tool

First year using one and it was one of the best things we purchased this year. :slight_smile: We had 50+ feet of 25 chain on our robot and only had four master links where our elevator platform connected to the chains. Using lighter chain (25 over 35), no master links, and C-C distances has really make working with chains an easy process and eliminated a lot of my worries in the past two seasons in our mechanisms.

We did #25 chain in 2x2x.125 tube on our drop-center 6WD this season and had a good experience. No maintenance all season. There was some chain slap against the tube initially (16T sprockets and long chain runs; not a real problem but noisy), so we slid in a strip of UHMW to dampen that and now you’d never know.

We used a 2x2 because the frame rails in our design had minimal cross bracing due to the large opening for acquiring game pieces.

So it sounds like the consensus here is that to do a Texas Tube design, you must keep all of the axles horizontally inline, and just put smaller diameter wheels on the outside to create a “drop center.” I’ve been throwing together designs in CAD for different drivetrain styles, and I wonder if someone can explain to me if this would or wouldn’t work:


(sorry for the poor render quality)

Rather than dropping the center or raising the outsides, I did both. I lowered the middle 1/16" and raised the outsides by 1/16", to end up with a
“drop” of 1/8". From everything I can see and read, this would fit very snugly into a 1x2x.125 tube. It’s completely COTS and would let you use 4", 6", or really any other size wheel anywhere on the drivetrain – which is big for teams with few resources.

General specs:
2"x1"x.125" aluminum tube
AndyMark Toughbox-Mini gearbox (alum plate cut down slightly)
17t #25 Team 221 dual sprocket

You can also use omni wheels on the outside wheels, especially in a game with little or no pushing and defense like this year.

Yes, we’ve done this. The 1/8" total drop is “shared” equally between the outside and inside axles.

In fact, our entire setup is very similar to this, with only a couple differences:
– we used 2x1.5x.125 tube
– we used VexPro 16t #25 sprockets

Did you have the issue that 118 mentioned earlier of the chain grinding on the outer bearing race?

No, in fact we deliberately chose the 2x1.5" tube to avoid this issue with the 16t sprockets.

Here is a technical drawing of (an early version of) our drive tube.