[John Hill]

Sorry, I originally posted this in the wrong place, I hope this is more appropriate!

I trust this provides food for thought!




This is a concept for moving a long axis. The principle is a development of the endless chain hoist. The drive element is on the right and consists of two sprockets/drums/etc that are on the same axis and locked together, they are of close but not identical diameter, all other sprockets/pulleys etc are free running. The gantry moves at a rate that is half the difference between the 'fast' and 'slow' chain/belt/cable speed.

Consider for example the case where the drives are 21 and 22 tooth sprockets and a chain of half inch pitch is used, for each revolution of the drive the gantry would move 0.25".

[Mooney201]

Look at steel or Kevlar toothed belts... Are being used on a plasma cnc...lightweight easy to work with...zero backlash..

[wizard]

Sorry, I originally posted this in the wrong place, I hope this is more appropriate!

I trust this provides food for thought!

Hey some of us have work to do today!

This is a concept for moving a long axis. The principle is a development of the endless chain hoist.

I'm not familiar with that principle. The idea of having two different size sprockets on the same shaft bother me so I will have to think on that one a bit.

The drive element is on the right and consists of two sprockets/drums/etc that are on the same axis and locked together, they are of close but not identical diameter, all other sprockets/pulleys etc are free running. The gantry moves at a rate that is half the difference between the 'fast' and 'slow' chain/belt/cable speed.

If this does work as suggested I would think practical matters would get in the way of a successful build. For one I have a strong ad version to using chains in a linear motion application where accuracy is expected. Even if you use belts I'm not convinced that good results can be easily had. The problem is that you now have multiple runs between the gantry and the machines frame. Not to be mentioned is the outside loop.

As I see it all these runs would contribute to a stiffness problem common to belt drives instead of reducing the problem. At least in my mind it does right now. Especially if the belts are unsupported which they are in this design.

Consider for example the case where the drives are 21 and 22 tooth sprockets and a chain of half inch pitch is used, for each revolution of the drive the gantry would move 0.25".

That doesn't sound right at all to me. I haven't done the math so maybe it is a gut feel that is completely off. Given that do you have a formula that describes this motion.

My thinking here is that a 22 tooth sprocket would move the chain far more that a half of an inch in one revolution. The mind is still churning here.

If the drive indeed works as you have described then you effectively have a very large "gear" reduction. This would likely mean a servi drive over a stepper drive.

So when are you going to build this, it would be interesting to see the results.

[John Hill]

Thanks for the comments...

I am sorry if my explanation is not clear. This is what Wiki has to say about the differential pulley (which I know as an endless chain hoist).....

A differential pulley, also called "Weston differential pulley", sometimes "chain hoist" or colloquially "chain fall", is used to manually lift very heavy objects like car engines. It is operated by pulling upon the slack section of a continuous chain that wraps around pulleys. The relative size of two connected pulleys determines the maximum weight that can be lifted by hand.

As you can see there is a loop of chain supporting the load and a loop which is being pulled. As the chain is pulled over the large pulley the load is lifted but because the two pulleys are connected pulling on the chain also pays out chain over the smaller pulley. The difference between the amount of chain being paid out and the amount of chain being wound up is mechanical advantage plus there is the advantage of having the chain pass around the load pulley.

Obviously, as the loop supporting the load decreases in length the loop being pulled increases by the same amount, in my concept I have taken this second loop around fixed pulleys and brought it back to the load (for example a table gantry) to make a double acting contrivance.

My intention is to build this on a router table with a two metre (about 6'6") travel with a stepper drive. Although there is a considerable mechanical reduction it is only that of a feed screw and in my example of using half inch chain the minimum equivalent pitch attainable would be 4 tpi.

Curiously, the size of the drive sprockets is not critical, only that they be different and as one tooth is the minimum feasible difference that is what I use for calculations. Larger sprockets would however mean faster chain speed which might be undesirable but I imagine very small sprockets would invite more backlash.

Although there will probably be more suitable materials I will begin with using bicycle chain so as to keep this project, which might be a total failure, well within the scope of a DIY project.

John

[talonosi]

I think its a cool idea and unique application for that mechanical system, but I'm wondering if it offers to many variables that could decrease is reliability and accuracy. How would this compare (pros and cons) with simply having a pulley reduction driving a standard belt arrangement? both could easily provide the required reduction.

[John Hill]

Hi Talonosi..

The chain is moving much faster than the gantry, more than 40:1 in the example I gave so maybe any errors coming from irregularities in the chain or sprockets would be divided by 40?

[dsquire]

Hi John

Very interesting concept that you have going on here. I'm going to be following along here and on MadModder.

Cheers

Don

[arithmetic]

Much simplified idea based on the same principle.Attachment 263740

[John Hill]

Arithmetic, yes, very interesting.....