Roller claw: Belt or Chain?

What do you like best when designing a roller claw for FRC applications?

What is your preference for roller claws?
  • 35 chain
  • 25 chain
  • Timing belt
  • Round belt
  • Mated gears
  • Other (please explain)

0 voters


Also, potentially Flat Belts on a crowned pulley. Anything that is light weight and can take some distortion and still operate.

My first selection was “timing belts”, but I wanted to make sure “other” was represented above so I changed my vote.


I clicked “round belt” in the poll, because I missed the “other” option, but this is what I actually endorse.

I selected 25 chain primarily because of the flexibility it offers in changing the geometry of the arm during prototyping. Adding or removing a few links from a chain when you change your C-C distance is a lot quicker than waiting for a new belt length to arrive from your favorite belt source.

Once you have a final geometry that you like, converting that design to a timing belt configuration would be great as the timing belt is smoother. But I’m not sure that in most applications timing belts would be easier to package or result in lower weight than a 25 chain on small aluminum sprockets. So, in most cases, it probably makes sense to stick with the 25 chain as your final design.

1 Like

Depends on the game piece and the manipulator I’m attaching it to. I selected #25 chain since that’s what we used in the two most recent roller claws we’ve built, but round belt and flat belt are also options if we want to have our power transmission also directly interact with game pieces, and #35 is an option for heavier duty applications (like Recycle Rush totes).

For the rollers, direct drive (possibly through a gearbox) when weight at the end is not an issue, belt (usually timing belt 'cause that’s what we have) when it is.
For the claw, mated gears or separate controls or one fixed or pneumatics or just flexible arms (polycarb tubing) depending on the application.

Timing belt for claws that shouldn’t slip, flat belt in the somewhat rare cases that you want your claw to slip. Flat belt makes for a cheap, useful friction clutch in certain situations.

I don’t like to use roller chain so close to game pieces. Seems like it could throw grease or damage the game pieces if something ends up where it’s not supposed to be, and tensioning a short run of chain is annoying.


Flat belt, followed by round belt. Chain and timing bents do not slip (or get damaged if they do), and a roller claw can get jammed, stalling the motors. A belt that can slip is like a fuse, preventing further damage up the line.

1072 did a crossed round belt in 2019 for making our roller claw spin in opposite direction. It was several rows of compliant wheels driven by a timing belt from a VP.
#25 chain driving compliant wheels would be my choice in many situations because of its robustness and lower precision. Fusing the round belts had to be done carefully to make sure there were no splits.

I’ve done all of the above and direct drive off the motor.
It really depends on what the needs for the game are.
How much power?
How Fast?
How much space?
How much weight?
Ask all those, optimize, and add couple iterations finding out what breaks then you have the right answer.

Benefits of Belt is that you can use it both for power transmission and as a conveyor to move game pieces.


If I had to pick one without fully understanding the beating and torque requirements, timing belt.
If the loads are sufficiently low or I anticipate needing to futz with the geometry, belting (flat or round… preference towards round because they track easier).

All belts can be figure-8-ed to provide reversing which is particularly useful for some roller claws.

You see, you say that…

2018 roller claw, chain intentionally exposed to kick cubes out of diagonal dead-zones:


I’ve done the same thing on a few smol robots to get out of high-centering on bumps.