Why does your team use belts or chains in your drivetrain?

If you’re comfortable with it, what team are you from and why does your team use belts or chains in driving your drivetrain? If your team’s made a switch from one to another, why? And finally, what type belts/chains does your team use?

Much thanks,

Designer of drivetrains

1 Like

We (1293) used belts in 2019 because we used the AM14U4 chassis, and there is no upside to trying to re-engineer it.

We used chain in 2018 because we could not hold suitable tolerances to make belts work in our (bad) custom frame.

We (5402) used belts in 2017, because AM14U3.

We (4901) used chain in 2016, because we designed a custom chain-in-tube drivetrain to let the chain hide inside the tube for protection. I was happy with this, but would need to have that kind of milling resource again to repeat it.

We (Ri3D Team Cockamamie) used belts in 2016 and 2018, both times because AM14U3.

Never used belts on a competition robot drivetrain as far as we’ve recorded (and can almost guarentee we never will) with one tiny exception*. We used #35 chain and dead axles prior to 2013 due to not having significant access to precision machinery and Vex’s clamping bearing blocks not entering the market yet. No issues there, though the chains would start to sag if not tensioned in some way.

We’ve used 25 or 25H since 2013 generally with 22t sprockets, but went with 18t this year due to our bellypan design. With the exception of this year, we’ve never had a chain fail. That incident was due to using a normal 25 half link with 25H chain (half link usage was necessitated by bottoming out our tensioning slots). Even 2016 with 6 cims and 8, 8" peneumatic wheels our only maintenance item was chain tensioning.

*2014 competition robot used 9mm belts for the practice day of our first regional to ensure maximum practice time. We knew they would need to be replaced after testing on the practice robot and they were swapped out before quals matches began.


45 has used belts every year since 2014 due to AM kitbase.

Valpo Robotics MRDC has one robot (Ernie) that uses chain. it is our oldest bot, used 35 chain with multiple tensioners and is still our most reliable bot. The chain has been a pain in the past, but we don’t want to mess Ernie up. We have two other robots, Marvin and Parker that use AM toughboxes direct driving pneumatic wheels in the back and omni wheels on unpowered dead axles in the front. We are thinking of possibly chaining Marvins front wheels for better traction, but we don’t know yet.

#25 chain-- makes it easy to do chain in tube with the 17t sprockets. We’ve found it’s plenty strong, and is lighter than #35. Looks good (chain in tube), works reliably, frees up space for other robot elements, and requires minimal maintenance after the original assembly process. Just be sure to add a small amount to c-c distance to account for chain stretch.

1 Like

997 uses #25 or #25h chain c-c. Mainly because we have a cnc mill that can make the bearing bores accurately enough. We usually use 18 tooth sprockets because we have them. We have never dropped a chain except for one instance this year at our scrimmage because I forgot to add a constant to one of the runs, It never became an issue in comp though as we never dropped that chain again.

We use belts in the kit chassis…it works…we have enough stuff to do trying to get the game pieces moving how we like, we’ll let AM make the chassis move.

We use belts on the drivetrain, but we’ve used chain on bots for other rotational power transmission stuff. We have also used chain on a few offseason drivetrains.

Edit: I forgot to put why, and I think we’re just used to it at this point. We didn’t have chain tools until 2016, so our designs just tend to revolve around it.

This year we switched to #25H chain and have had huge success with it.
The reason for the switch was due to constantly completely stripping all drivetrain belts the past 2 years.
We never plan on going back to belts in our drivetrain again.

1 Like


We use belts in tube as I described in this post. I’m pretty sure we’ve used belts since the KoP drivetrain in 2013. Design it right, and they don’t need any tensioning or other maintenance all season.

We use belts on pretty much every drive train we make. I like them because they do not stretch over use like chain, however I would not be opposed to trying it.

What do you mean, ‘add a constant to one of the runs’? thanks.

It’s a common practice to add a small distance to the calculated center-to-center distance for chain runs without a tensioner. This addition helps account for manufacturing tolerances and chain wear-in (also known as stretching).

While I haven’t done any custom drivetrains involving belt or chain, when designing mechanisms beyond what 9mm belt can handle, it has usually come down to the trade off between width and weight, with the tie going to belt because it doesn’t need to be re-tensioned.

I have not seen a 359 robot for several years, but back-in-the-day that team was well known for their use of bicycle chain in WCD set-ups. That chain is lighter that #35 and stronger that #25. It is also a bit more forgiving of misalignment.

I hope that @waialua359 can comment. Is bicycle chain still a good idea in FRC?

If you’ve got the resources to make custom sprockets, it’s certainly not my least favorite power transmission method.

1 Like

95 has used both. When we used belt it was in 2013 with the kit chassis.

Otherwise we have used #25 and #35 chain because it is strong enough and packages narrowly width-wise.

Go to in the drivebase is #25 and will be for the near future.

Driven by a few factors:

  • Inventory - we somehow ended up with a lot of 16t, 18t, and 22t sprockets and a lot of chain was purchased in 2015 (about 150ft). Right now, it’s cheaper to go into our storage closet than buy belts.

  • Easy to work with - we don’t have consistent access to precise machining for large pieces in a drivebase so Versablocks and Cams are the best option for us.

  • Reliable (for us) - Everyone will have their own ideal build: gears, #35, or belts. For us, it’s #25 and I’d rather not play around with it when we have something that works.

Generally speaking if you are throwing chain or belts or seeing other weird failures it’s a sign of errors in design/application, machining, or assembly. Hence why I said #25 is reliable for my team. It’s not THE reliable option, it’s only as good as you can build it.

Some options can save you weight or space, but go with what you can put on the field that works each match.

Many of our smaller applications are moving to belts and we’ve loved it so far. Only downside is the investment in one size applications and the expense of ordering spares in case something happens. Chain I can have a 10ft package on hand to replace any run if for some reason it fails - usually at the connection which is user error with a Dark Soul 95% of the time for our team. Belts I need to order 1-2 spares for any size we used as a precaution.

1 Like

1076 used Kitbot chassis in 2015 (i.e. belts), but every year since made a custom WCD drivetrain with #35 chain. If done right, serviceability is unparalleled; anything drivetrain-related can be fixed/replaced in under an hour.

Bike chain is awesome. Inspired by 359, we made a drivetrain using bike chain as a pre-season project a few years ago. I found it to be very smooth and light weight. We used sprockets for single speed bikes, but machining hubs to interface with them was a fair bit of work.

I reached out to Glenn from 359 and he sent me a drawing of the sprockets they make, and they are surprisingly simple:

Kind of an unconventional design, but you don’t need CNC machines to make them, and he said they have been very reliable.

In recent years we have used belts because they don’t stretch and are readily available from Vex. But bike chain is always a good choice.