pic: GEAR & CHAIN IN TUBE 8WD



Here is a drive I worked on today based off the concept that Chak posted a few days ago here.

Basically this has an 11:72 gear ratio incorporated inside 3x1.5x1/8" tube, with all of the chain runs also inside. This ended up resulting in some space saving compared to standard WCD setups and also less custom parts. The only custom parts really needed is the 2 main drive rails.

I utilized a US Digital S4T encoder, which did require a coupler and custom mount plate, but you could easily use one of the CIM encoders VEX sells instead for simplicity.

Curious to know what people think. I know the lightening pattern is pretty involved, I just felt like trying out the Hex Truss pattern.

Weight is 28lbs with all set weights assigned and all other materials assigned. 4" AM Performance wheels, but could easily be switched to any 1/2" hex bore wheels.

Without the top side of the tube being able to constrain the chain, won’t the eventual chain stretch and sprocket wear cause it to derail? It was my understanding that the physical constraints of the chain inside the tube were the main factor preventing derailment.

Any chance we could get some more detailed renders of the gearbox/chain runs? What I can see looks great, but it’s hard to ascertain the specifics from this.

I understand the attraction of belt/chain in tube drivetrains as a design exercise.

The trade made is very involved and long replacement times if a belt/chain/axle/bearing failed during a competition. You could very easily miss a match for the amount of time it would take to replace any part.

Considering maintenance, and tensioning should be prime considerations in determining the best way to implement a belt/chain drive train.

+1

And using 3" tubes means your chain has room to fail. If you can use larger sprockets without increasing the weight too much, the closer you can get to the insides of the tube without interference, the better. also, using sprockets that large means it would take a monumental amount of force to cause a chain failure at the sprockets, because the load would be distributed over so many teeth. This also means less wear, or “stretch” over time.

I would try and tweak the design so that the ends of the drive rails are open, so that if you do need to get to your chain or gears, you don’t have to disassemble your chassis frame.

1 Like

Using the tube to keep chain on is sub-optimal. In my opinion, you shouldn’t use anything other than proper tension, power transmission, and tooth engagement to keep a chain on the sprocket. If you design the c-c spacing properly, the chain shouldn’t get loose enough to jump off sprockets as it wears in.

In theory, there’s no where for the chain to go if it gets loose and the sprocket teeth are close to the tube wall, but its one of those things where you shouldn’t have to take advantage of it if you do it right.

It’s just that I’ve built a bunch of chain drives at exact center distance, using sprockets of a variety of different sizes and under different loads, that have stretched to the point of jumping off sprockets. Everything is properly aligned, it’s just from wear. I’ve always been told the “magic” of chain-in-tube is the tight fit prevents a loose chain from jumping, since it has nowhere to go, but maybe I just have bad chain runs or something.

1 Like

I agree that exact c-c is a viable option, and having the chain stay on because of proper tensioning and spacing is preferable, but I think having a tight fit inside the tube adds an extra measure of safety in preventing failures. Hopefully this would be redundant, and the tube would never need to keep the chain on the sprocket, but taking advantage of this peace of mind is one of the benefits of running chain in tube.

This year for our climber we ran dual #25 chains in-tube, one within each chassis side rail. Since the chain was worked in only one direction, pulling the robot up the wall, there was only one slack side. We did contain the chain with the chassis rail on the slack side. The tensioned side of the chain is in open air, no containment. 200lbs of lifting force with four CIM’s in low gear and no failures. It utilized a Vex 22 tooth sprocket on each end to transfer the drivetrain gearbox power via a PTO.

Image of in-tube climber chain run.
https://drive.google.com/open?id=0B67Uv_S_WWjpQjRVTzYxNGFjOWM

For a drive train where the chain is worked in both directions I would agree containing the chain on both sides could be beneficial and is good insurance. I wonder if anyone with in-tube chains has had issues with broken chain or chain that gets severely wedged between the sprocket and tube.

On the other side of the coin, we ran our drivetrain sprocket spacing at .015" under theoretical spacing this year (7.985" center to center sprocket spacing for 8" of chain length). There was a very large amount of slack so that the defenses didn’t hit tensioned chain and damage it or knock if off the sprockets. Our run of chain was allowed to get pushed up past the chassis rails and the defenses would impact the chassis rails instead so therefore we never lost a chain this year.

Or you could use a hex bore encoder :slight_smile:

Sorry, couldn’t get to my computer until tonight.

Here you go (sorry for the large image file)

http://i.imgur.com/kTG3fTp.png

Well in this specific situation the top half of the chain runs inside the tube while the bottom half of the chain runs under the tube. So using masterlinks the chain would be easily accessible and easy to switch. Maneuvering around the sprockets should be fine with the access slots on the bottom of the tube.

I do agree that this chain setup is not ideal, but I kind of got worked into a corner with trying to fit everything in the tube.

I am working on a variation right now with smaller wheels, and a 2" height tube, and am attempting to have the entire chain run inside the tube. The challenge there is you have to choose between:

(1) Motor clump on the corner wheels which for everything to fit means that your wheel is pushed further in from the front/back of the chassis so the CIM can sit in front
(2) Motor clump in the middle and your tube gets wider to accommodate both sprockets on the wheel shaft and the gear.

I keep going back and forth on which I’d prefer. May end up designing both and seeing how they both look as an end result.

For a drive train where the chain is worked in both directions I would agree containing the chain on both sides could be beneficial and is good insurance. I wonder if anyone with in-tube chains has had issues with broken chain or chain that gets severely wedged between the sprocket and tube.

346 ran chain in tube for our drive this year. Yes we did experience a broken chain and yes it is “difficult” to fix. Darn the ramparts.

While I was not able to view your design in detail, I would wonder if you considered assembly? It is difficult to assemble chain in tube with fixed centers. Guess I’ll take a closer look to see how you overcome that issue.

Ohh, I had not noticed that. That’s a very interesting concept; I’d like to see somebody try that out.

So using masterlinks…

Master links with what looks to be 25 chain? That’s living dangerously.

I am working on a variation right now with smaller wheels, and a 2" height tube, and am attempting to have the entire chain run inside the tube. The challenge there is you have to choose between:

(1) Motor clump on the corner wheels which for everything to fit means that your wheel is pushed further in from the front/back of the chassis so the CIM can sit in front
(2) Motor clump in the middle and your tube gets wider to accommodate both sprockets on the wheel shaft and the gear.

I keep going back and forth on which I’d prefer. May end up designing both and seeing how they both look as an end result.

Your gearbox and sprocket integration is quite nice, and it would be very impressive to see a design with all the power transmission components enclosed in 2x1" or 2x1.5" tubing. Can’t say for sure, but I imagine for your purposes, tweaking your gearbox to make option (1) work would be pretty simple. You might even be able to just rotate the two motors forward around the driven gear and keep those corner wheels out near the edge of the frame.

Here is the updated version I ended up with.

-23.33lbs, only parts missing are chain runs, and bumper supports I suppose (all hardware besides rivets included)
-24"x32" chassis
-4"x1.5" VEX Colson wheels (1/2 hex bore)
-11:70T Single Stage reduction within the drive tube
-Drive tubes are 2x2x.125" 6061, while the front and back tubes are 2x1 Versa
-All sprockets are 16T #25 VEX sprockets with the chain run fully contained inside the tube
-The middle wheel is dropped 1/16"
-In this current configuration a US Digital S4T encoder is utilized, though the CIM encoders could easily be used instead, or a Grayhill, etc.
-All parts are COTS except for the two drive rails, the two encoder plates, and the belly pan (could be lexan, wood, etc. for low cost)
-The only other modified parts are the front and back rails (lightening pattern), shafts cut to length, and the direct drive shaft is turned down on one end to 3/8" round to couple to the s4T encoder (optional depending on encoder use)

Apologies again for the large picture:

After initially dismissing the idea as “sketchy” because of the integrated gearbox, after looking at the lower custom part count and the resulting reduction in cost, I think this is a great idea for making a custom chassis at a low cost and weight.
I don’t like your lightening patterns, but apart from that the design looks great. The dual versagussets in the corners really stand out to me as a feature of 2x2 tubes.
Do you think it’s possible to make a shifting version? Can all the gears + sprockets fit?

In the past, we used sliding bearing blocks instead of tensioners to ensure that our bike chain would never jump a sprocket, even if the chain were to get looser due to wear and tear.
For the past 4 seasons, our main machinist ensured the 3 things you mentioned and we never had an issue with chain jumping off our drive ever. We basically set it and forget it, thus eliminating the need for adjustable bearing blocks and saving weight/time to make them.

Have you run a chain below tube/drive chassis/bellypan like this before? Do you run into issues with getting carpet junk into the chain or having the chain exposed to accidentally being hit on things? I am curious because we are considering running a similar chain setup, but don’t have much relevant experience to base off of.