Hello everyone, I have a few questions regarding chain tensioning.
First of all I understand that when it comes to chains, there is always going to be a need for some kind of a tensioner since the roller chain stretches under the tension they are under over time, however I have read 1114’s presentation in which they explain that when properly calculating the center distance between 2 sprockets chain tensioning can be avoided, but I still see teams who are calculating the center distance correctly use tensioners, is it just me thinking they calculated the distance correctly when they didn’t or am I missing something? Plus I was wondering how sprocket size, sprocket teeth and chain size can help reduce the need for tensioning.
In addition, chain tensioning in frc is generally talked about under the subject of drivetrains, however chains on robots are used on many different types of mechanisms like arms, elevators, intakes, climbers and so on, but when searching information about those mechanisms, chain tensioning is overlooked. Is chain tensioners necessary on mechanisms other then the drivetrain?
Our team uses chain tensioners for elevators. I think using larger sprockets along with a well calculated center distance can also remove the need for chain tensioning. But for something like an intake rollers, the backlash doesn’t make a difference in how the roller works so chain tensioning wouldn’t be beneficial.
First, I’d like to correct the notion that change stretches. It doesn’t (until you get close to it’s yield point). What most people call stretch is really the wear-in of the pins, links, and rollers as well as the wearing-off of any flash, burrs, or high points.
You can buy “pre-stretched” chain to prevent this - I know 254 does that.
Next, you can use the center-to-center distance to stay away from chain tensioners. There have been numerous threads here about how to do that, and there are literally a dozen or more chain-distance/link calculators on the net.
However, the exact math from the calculators may not work because of the chain tolerance. As a result many chains are too loose. Experienced teams add a bit of length to their center to center, and it varies depending on the designer.
In a Michigan State Seminar Paul Copioli recommended adding .018" for #35 chain and .012" for #25 chain to your exact center to center distance (it doesn’t matter how long the chain is).
The higher the load and the smaller the sprocket the more important tensioning becomes. You’ll see some people using dual chains in high load long distance applications like an elevator to avoid chain skipping.
All that said, we generally have a wait to tighten the chain on every run we use, whether it’s shim stock behind a mounting plate or versa blocks and a cam tensioner. That covers a lot of problems that can be caused when things aren’t made perfectly, and we’re professionals at not making things perfectly.
I’ve always run #25 .018 long and #35 .012 longer, based on Triple Helix (?) testing. But tensioning is only a requirement for things like arms where small amounts of backlash hurt your performance.
In many situations, it’s not just about the size of the sprocket, but also the “chain wrap” around the sprocket. If your chain is only interacting with 90degrees of the sprocket, it’s obviously much more likely to skip than if you route it around 180+degrees of the sprocket. A good rule of thumb is to plan for 120+ degrees of contact between the chain and sprocket. In some situation chain idlers or tensioners are used to adjust chain routing such that you get more wrap around your drive and driven sprockets.
Wouldn’t you say backlash in the drivetrain can cause autonomous performance issues?
Other than small sprockets, what other elements make the chain system more prone to chain slip and chain wear? Do load and chain length have an effect?
Hypothetically, sure. In practice, not a whole lot. A large portion of autonomous modes don’t actually end up changing the direction that their motors are running, and that means you can “pre-load” your chain when setting up your robot before the match to account for the slack in the chain. In situations that involve turning in place (and thus changing motor direction), if you’re using a gyro to help control your turn, that is measuring “after” the backlash. Additionally, the error produced by chain backlash is likely to be less than other errors in your autonomous system (wheel slip, gyro drift, etc). Finally, any vision targetting or other sensors are often used to make the final adjustments towards the target (or the target is large enough to not care about slight misalignment, such as in 2018).
You forgot one item - if you place your encoder directly on the driven wheel shaft, then any backlash in the chain run going to that shaft won’t matter as it happens before the sensor. So long as that wheel has solid contact with the ground the entire time, it’ll be as accurate as anything (wheel slip would still impact it, though!)
That’s not necessarily true; backlash/chain slip can cause errors with integral control, if the system is logging control effort that isn’t actually doing anything, but you don’t really need integral control for most mechanisms on your robot.
Also, the torque is transferred on a larger moment arm, so the same type of chain doing the same work on larger sprockets has less tension differential between the runs.
Say I use large sprockets, say I calculated the center distance, would you still say I need a tensioner? Or does it depend on the load the arm needs to handle, do lighter mechanisms still require tensioners?
Plus if it’s not too much trouble can you link examples for teams using tensioners on their subsystems?
Even if you calculate the center distance, run a tensioner on arm chain. 1072 had to add one to our arm mid-season and it’s still giving us problems.
I don’t have many good examples of tensioners, but 1072 and 1678 have our CAD public on OnShape.
If you didn’t build a tensioner into your chain run, but have enough space, you can insert a floating sprocket a few sizes larger than the sprockets on your axles near the middle of the run to take up some slack. To increase tension, move the floating sprocket towards one end, or replace with an even larger sprocket.
So let me see if I understand everything correctly.
Generally, every chain need a tensioner, however I can minimize the need for a tensioner with proper sprocket to sprocket distance, big sprockets, and enough chain wrap, infect I can minimize it to the point where in systems where backlash doesn’t really effect the performance (like a drivetrain or an intake) tensioners become unnecessary. But, in systems where backlash can cause damage (like an elevator or an arm), it’s good to have tensioners so the system will perform well and not get damaged.