Meticulous Maynard's Chain Guide

@MeticulousMaynard, I’ve put in a PR to add your pictures to the Readme. They are good shots, show exactly what the project is about. :smile:

3 Likes

No @Gero, not crap, but simple. Simple is good, too. :smile:

4 Likes

I am still jealous but to suborn to copy this wonderful design. Now I have to use metal bobbins, somehow manage to press bearing in them and create an adjustable arm.
I love how inspiring this Forum is and how openly designs are shared. Thank you @MeticulousMaynard for sharing.

4 Likes

those bobins are cute though Gero

2 Likes

delete (moved to my own thread)

As this is Meticulous Maynard’s Chain Guide, I would suggest a new topic :slight_smile:
There I would then ask on what side of your drawing the sled is. Chain jumps happened for me when the sled was moving to the top of the sheet, on the side with the shortest chain to the sled. This side has the longest loose chain and the rubber bands can’t compensate for the chains trying to go vertical. Because the frame and the motors are tilted, the loose chain is coming in from a more vertical angle and when you feed that chain out, the side of the chain will try to ride on the tip of the tooth of the sprocket. With a noticeable ‘Bang’ the chain will either skip a link or fall into its meant position. If you have chain jumps starting from the sled side of the chain, your sprockets and sled-chain-mounting-points are not parallel to the work piece. With the pantographs you add or remove mounting spacers, with the brackets you go one hole up or down. One more reason for chain jumps could also be flexing top beams, as both sprockets get turned out of alignment. I apologize if you knew already, hope it still it can help someone else.

The changing pivot point of the chain on the sprocket depending on the chain angle, is an error that is already, or can be compensated by the soft- or firmware, I think. I agree the a small idler can reduce the amount of error and would make room to experiment with different sprocket sizes.

Edit:

This is answered here: PDF of the formular

woops, i’ll move it to

:slight_smile:

EDIT:
this one can also be deleted, or keep the link as it is the same kind of topic just a slightly differnet goal?

1 Like

you don’t need to change the pivot point to experiment with different sprocket
sizes, Sprocket size, and the effects of the chain angle to the sprocket are
already accounted for in the current math (or if Bar hasn’t merged it yet, it
soon will be)

With the math already accounting for the effect, adding a different effect will
just complicate the math, not simplify it.

but if you go with a different sprocket size then the math accounting for the chain angle will also need to know the radius, if you used a idoler as your pivot point then that can remain the same and only affect the travel per revolution

1 Like

the math needs to know the radius, because that changes how much chain is fed
out in one revolution of the sprocket.

adding a different size pivot somewhere else just adds complexity, it doesn’t
remove any.

And if the pivot point is small enough, it can start to catch in the links of
the chain (at the very least, there is a difference between a regular link and a
master link going over the pivot) because the pivot works against the outside of
the link, while the sprocket works against the rollers on each link.

1 Like

I agree with Pyrosrock, though i don’t have a machine to show why i think he is right,

Anyone who want’s to lock horns on this?

I see dlang’s point but when you measure the chain length form center of bobbin then is 100% the same as center of sprocket just a smaller diameter and less error

What does @Bar think about this?

I’ll try to draw this tomorrow once i get my head around how i need to explain how i see it
maybe @pyrosrock has a simple explanation???

This has been well covered and discussed amongst a number of threads.
All i have been saying is that

  • Yes the chain leaves the sprocket at different places depending on the direction
  • The larger the sprocket size the bigger the difference
  • If we have the maths to calculate and compensate this great!
  • If people are talking about trying different size sprockets then that would have to be taken into consideration with that maths if they are used as the pivot points.
  • or if you used the known size sprocket as a idler pivot point, then you can play with motor configuration, drive sprocket size or more wraparound serpentine work without effecting the pivot point, thus without introducing another variable
2 Likes

This has been well covered and discussed amongst a number of threads.
All i have been saying is that

  • Yes the chain leaves the sprocket at different places depending on the direction
  • The larger the sprocket size the bigger the difference
  • If we have the maths to calculate and compensate this great!

we do

  • If people are talking about trying different size sprockets then that would
    have to be taken into consideration with that maths if they are used as the
    pivot points.

already done

  • or if you used the known size sprocket as a idler pivot point, then you can
    play with motor configuration, drive sprocket size or more wraparound
    serpentine work without effecting the pivot point, thus without introducing
    another variable

my point is that you are introducing another variable (the idler size) to the
existing system. No that variable won’t change if you change the other
variables, the idler doesn’t change, but adding the idler adds complexity
without elinminating any error on the stock systems

This thread went places. I was refraining from replying because I didn’t have anything to add to the conversation but here goes.

@Gero: I really like that bobbin idler pulley you’ve used. Simple is great! I feel that my solution is quite a bit over-engineered for what we need.

Does this mean that the firmware now calculates the sprocket size? Or that it simply offsets of attachment point of the chain? (or those basically the same thing?)

If so, I have some feedrate tests to do…

3 Likes

Does this mean that the firmware now calculates the sprocket size? Or that it
simply offsets of attachment point of the chain? (or those basically the same
thing?)

I believe that it was pointed out that the firmware is properly calculating
sprocket size. (at one point, this was hard-coded, but I was told in a former
thread that it’s now calculated)

I know that (at least for standard kinematics) this includes the pivot point
and the length of chain that is wrapped around the sprocket.

Bar has said that he has the code to do this fully with triangular kinematics,
but I’m not 100% sure it’s been merged, but if it hasn’t, that’s a small change
from where we are.

1 Like

The firmware uses link size and number of sprocket teeth to calculate the sprocket circumference.

Excellent! I’m going to have to update my firmware/gc and give @dlang’s 25 tooth sprockets a try. This weekend I’ll have to test feedrates further! :smiley:

3 Likes

I have been looking at the various posts re - chain length altering as it exits the sprocket due to differing angles . It occurred to me that it might solve the problem if the motor was able to rotate around its own axes by being accurately mounted to a trailer hub bearing or similar . The chain could the be fed through a lightweight channel guide which is fixed to the flange of the hub such that the action of the changing angle of the chain would also rotate the entire motor to match . I haven’t as yet built my maslow so i cant test the idea , but cant decide it the rotational torque of the powered motor might cause the guide to rotate too far momentarily or indeed how it would effect the calibration . Any thoughts greatly appreciated .

1 Like

That’s an interesting thought. Is the idea to help guide the chain on the sled-side to help compensate for material thickness? Personally, I’d prefer to adjust the top beam in and out to address that.

My biggest concern with this idea is that it would be really hard to keep the channel aligned with the chains at all times. Even if it was to spin freely, all it would take is a slight change in angle and it could pull the chain with it. This would distort the cut. The other thing is that the software currently calculates the catenary deflection, or chain sag. The channel would upset that sag profile, which could also cause distortion.

However, if this were to be on the slack side of the chain, it would do a good job of feeding the chain into the sprocket at the correct angle. You wouldn’t need the bearing, though, since the chain only really needs to feed in at one angle.

As with anything, we won’t really know how effective it is until it’s tested on a machine. I could entirely be wrong, it wouldn’t be the first time. :wink:

and take a look at the new chain layout, it greatly reduces the problem (if not
completely eliminating it)

1 Like