As part of an exercise in making the smallest possible FTC drivetrain (cubic volume-wise), I’ve been working with a few people on the FRC and FTC Discord servers to reduce size, and I’ve come up with a design roughly like this:
(1.625" circles are Colsons, 1.25" circles are sprockets, green lines are the chain run)
While this has a long way to go (pocketing, standoff holes, etc), I’m first trying to find a way to make it a lot cheaper. By using one chain run instead of the standard two side-by-side on the center axle, I’ve gained .5". However, this uses 4 idler sprockets, which makes this option a lot more expensive since I’d need the extra sprockets as well as two additional bearings per idler axle.
What alternatives to idler sprockets are there?
So far, I’ve thought of:
Delrin rod stock, lathed down at the ends and sandwiched between the two side plates
A true idler sprocket with included bearing (this is too expensive)
An axle (same configuration as the Delrin idea) with a bronze bushing or bearing running over it.
How well would these ideas work, or are there better or cheaper alternatives?
Full disclosure: my degree is not in mechanical engineering, but Delrin should work. There is no need to use a full slice of round stock, you only need a bit more than the section of the circumference that engages with the chain plus enough material behind it for support and mounting. Some Delrin on the sides to keep the chain from wearing against the side plates may or may not be needed.
If the goal is to minimize the volume of the drive, I strongly suggest reducing the height of the chain run. You should minimize the radius of the “idlers” used to get sufficient engagement on the center sprocket and move them much closer to the center sprocket both vertically and horizontally. once you do that, the top segment of the chain can be lowered considerably. If you really want to push it to the limit, a Delrin sheet supporting the top section can be used to prevent the chain from colliding with itself. Just radius the edges so the chain doesn’t snag.
One last thing, the lower left chain segment appears to be tangent to the wheel rather than the sprocket, If that is being used to calculate distances, it could throw your result off.
While you’ve saved some width in that you only have one belt, you’ve added significant height, parts, complexity, and opportunity for friction to steal the power your motors are generating. If you want to keep down to a single belt/chain without idlers, there is always the option of four wheels. If you keep your track wider than your wheelbase length, you should still be able to drive skid-steer, at some cost in stability falling forwards/backwards. That’s what we did with our FRC robot this year - kept the CoG really low and built a 4 wheel drive train that was wider than it was long. It does short wheelies on fast starts and stops, but has never fallen over.
If you really feel the need to do 6 wheels but reduce to a single belt, you can achieve a reduction in parts and height by using larger wheels on the corners than on the center wheels, and eliminate your outer idle pulleys. If you scale the wheel pulleys by the same proportion as the wheel diameters, the wheels will still all move at the same speed against the carpet. Note that center wheels already tend to wear faster than corner wheels because they bear most of the weight of the robot; using smaller wheels in the center will exaggerate this.
The hardest part when dealing with this drivetrain was that if the wheel sprockets were not concentric with the wheels, then the chain will be tight or loose depending on the wheel rotation. This is true with a simple two wheel chain run, but is amplified when you have more wheels on the same chain. Using a more precisely machined sprocket/wheel setup would fix this problem.
Note the green delrin idler (it doesn’t rotate, it’s bolted in place). It was machined on a lathe to have a raised portion that contacts the roller part of the chain, instead of the plate links that connect each roller together. This kind of setup gives less friction than a setup where the plates of the roller chain rub against the idler.
As I recall, we started with a similar chain path to what you have pictured, but changed it to what you see in the photo to get more chain wrap around the center wheel.
This drivetrain worked great for FRC in 2008. I’m not sure how much the added friction of this kind of drivetrain would affect an FTC robot though.
