System to compensate for chain sag

I’m trying to use the chain tensioner elastic to counteract chain sag.
Hello, I’m trying to use the chain tension system to compensate for chain sag. The chain tension sprocket is moving at a 2 to 1 ratio and the elastic cord at a 3 to 1 ratio so the more chain is out the more tension there is on the elastic cord attached diagonally. It looks promising.


This is absolutely brilliant.

How are the results you are seeing looking? I think this is an absolutely fantastic idea.

When you are at the bottom left or right corner where the chain is the longest does it still provide tension?

Because of the 3 to 1 ratio on the springy side the tension builds up the more chain is out ,so it provides more tension when it’s needed most. Currently my pulleys have too much resistance and I will be replacing them with ball bearing ones next week,so it needs more fine tuning to be able to evaluate the impact on accuracy. I ‘m still in the setting up phase and callibration of the whole thing.



Keep up the good work!

Thank you

I’m impressed with the ingenuity to use the chains as the mechanics to create an opposite force / pull on the sled.

My head have been thinking about the same concept, that we “just” needed to find a way to have four points of force. I don’t have my maslow yet, so thinking about concepts is a bit hard.

In the meanwhile you beat me with a solution that I believe is way smarter then what I could ever have thought off!

Just a few “brainstorm” thoughts:

  1. Should we have a full circle on the sled?
  2. Should the pulleys that pulls “down” on the sled be parallel to the motors / sprockets mounting points?
  3. Could we in some way calculate what force we need in the outer positions on the maslow, to find the right configuration of the rope design?

If we ever meet, the first few rounds of beers are on me! :sunglasses:


Full circle would be nice but I positioned the attachment point so that the pull forces rotate the sled a little but not much.
Pull in the exact opposite direction of the chain would be best but hard to achieve.
The exact force is hard too find out. I just try to apply force to eliminate chain sag as much as possible.

You got my all worked up on this idea!

I agree that it could prove hard to archive, but not impossible with all the bright minds working together. Pythagoras didn’t live in vain, and I would expect that his theorem could be of some help to this matter.

Ahh - I need to sleep and have work taking my time the next few days.

Waiting for my maslow just became even harder than before… :flushed:

1 Like

Let’s see how this works.

My gut feeling is that predictable chain sag (which we can calculate) is going to be better for accuracy than chain sag + partial correction (especially when the correction isn’t pulling in line with the chain)

but I’ve been wrong before, and you learn stuff by experimenting, so let’s experiment.

I agree, if you can calculate chain sag that would be better but this system takes away a little of the inaccuracy in the current system. It’s a work in progress!

1 Like

If the Z-axis mechanism was compact enough and didn’t get in the way, perhaps the counter tension bungees could be attached with rollers like the chains, to a full circle ring, so they were in line with the chains, all pointing at the cutter bit.

That would be best but I thought of the bungee system after I finished my sled…as long as the sled doesn’t rotate beyond the range of the upper rollers it is not too bad this way.

I had to read this one twice to comprehend it. Good idea! This is way better than anything I had come up with. In fact, the Maslow Community has impressed in general.

Here are a few thoughts. A potential next-step is to determine a process for how to calibrate the tension so that it works optimally across the full work surface. It may not be necessary to attach the ropes to a ring; due to the triangular kinematics,the angle of the sled does not affect the location of the router bit relative to the chains.


That still isn’t going to have them pulling in the right direction (although it
may avoid rotating the sled to the point that it hits the limits, which would
introduce more error)

Great job @Brr
This is exactly what I suggested in A question about adding bottom motors



I like your frame design - could you share more about how it’s made?

Thank you

the frame is nothing special, just small changes from the standard design.
most interesting is that it is hinged at the top and folds away towards the ceiling when not in use.


Hmm, If that works, I’ll have to order a second pantograph from @pillageTHENburn for the bungees

I tested it thoroughly but the results were not satisfactory.
It is hard to get the tension on the bungees right across the whole work area. Now there is the chain sag compensation in the software which takes care of the sag inaccuracies so I discarded the bungee idea.


Thanks for testing it. Every time we test something and conclude it doesn’t work, we have eliminated one suboptimal path.

1 Like

@dlang posted on a chain take-up idea a few weeks back, which I thought had a nice alignment with this.

His idea was to have a circular wheel to wrap the excess chain around, and use a stretchy-rope to create rotational tension to pull on the chain. I thought that idea supplemented this idea, because the disk for the stretchy-rope could be a slightly larger diameter than the disk for the chain; it could be larger by any fraction, so that the motion ratio was not restricted to fixed fractions, unlike the above solution. It could be just 10% bigger in diameter, such that the amount of stretchy-rope take-up was only 10% more than chain take-up. This compares to 50% with the above solution. I thought that the ability to tune the motion ratio to anything, not restricted to fixed ratios, would enable better calibration of the tension across the entire work-area.