ultimate strength isn’t the issue, reducing stretch is the issue that we would
be fighting. (that and keeping it enough lighter than the chain we are
currently using to make it worthwhile)
Yes I know that belts can run massive things and handle hundreds of HP, but
those belts are very heavy and not that flexible. We need something that can
bend around a small diameter pulley, yet can be extended 10’ (3m) and be close
to perfectly straight under just a few pounds of tension.
David Lang
Notice the tension adjustment. When set loose intentionally they can allow
you to pull wire by hand, or slip to avoid a rats nest if there is a kink
upstream. But the steel V groove pinch rollers can also be adjusted very
snugly. Likely more than we can achieve with 3d printed pinch rollers. There
are serrated 4 roller wire feeders that could lift a small person off the
ground. They are about $60. Some of the common ones have 4 amp, 24volt
motors, more power than maslow. Why would they use larger motors if they
couldn’t get grip enough to use it?
using it is one thing, using it without jamming is another. feeding back and
forth repeatedly is another thing we need
The jamming comes when the user bends the
output welding cable too tight with the wire being fed down the middle. It
creates a kink and the feeder pushing wire behind it keeps on pushing. Not a
factor for us. We do not require the wire to be any more flexible than it is.
after the feed rollers, we need it to go around a pulley and then be straight
under low tension. All the mig wire I have ever used would get a bend in it and
not pull straight again without significant tension.
David Lang
But as I’ve said repeatedly, I’ve been wrong before and will be wrong again,
working hardware beats theory
There is a belt size and style for nearly every application. If small teeth
don’t work for your application, choose a larger pitch.
we also need something readily available and cheap 
Several teeth always work in unison creating a very large shear plane. There
should be a minimum of six teeth in mesh and at least 60 degrees of belt wrap.
with the maslow, the departure angle will vary a lot, from 10 degrees to 80
degrees, but we should be able to keep ~100 degrees of belt wrap, which will
ensure a good number of teeth engaged.
We are also not particularly sensitive to the pulley size, so we can also go
with a large diameter pulley to give many more teeth room to engage (a bit more
of a problem with everything on the sled, but not too big a problem)
The reason I did not suggest them for Maslow is the large potential for
misalignment. That does cause belts to fail.
misalignment also causes problems with the chains (causing them to jump), so I
wouldn’t worry too much about that.
We have large spans, uncontrolled tension, a non recirculating, non static
belt path.
tension varies, but it varies fairly smoothly, the belt path isn’t completely
static, but it’s pretty close.
my biggest concern is being able to model the stretch of the belt under the
different conditions.
David Lang
Optical calibration 
yep, or more precisely, let’s call it zone based calibration (it doesn’t need to
be optical as we’ve discussed)
that’s why I want it to be a standard option that we can select.
David Lang
If zone based calibration works properly, there will be no need to model things like stretch. Opposing factors will cancel and we only need to record and offset those portions that do not cancel.
I’m just guessing, but would bet elongation of T belts is less than chain because the belt is a monolithic tension structure. Chain is many joints, and each joint has clearance that gets reduced under tension.
If zone based calibration works properly, there will be no need to model things like stretch.
zone based calibration makes the math very simple, in theory, you can throw out
all computations, just record the actual chain lengths (or even better, encoder
position) at each calibration point and interpolate between those measurements
on the other hand, how many places do you have to calibrate? and how do you get
accurate enough positions to measure against.
I’m just guessing, but would bet elongation of T belts is less than chain
because the belt is a monolithic tension structure. Chain is many joints, and
each joint has clearance that gets reduced under tension.
on the other hand, the total of all those clearances may be less than the
stretch of the belt
on all of this, experimentation needed.
Optical calibration!
So, I’m really smitten with the capstan/windlass idea and with the rope that was discussed having been used for fishing/shipping/hammocks it seems like a really good idea to me. Am I missing some large drawback?
It would seem to me that we could wrap the rope around a knurled driven pulley of known diameter (with the feed coming out at the bottom closest to the mounting points) multiple times to ensure it doesn’t slip, with pinch rollers to ensure consistent friction regardless of where it is in the travel. We then mount the setup to the roller carrier on our ring system and the feed “always” has a back azimuth pointed at the router bit. Seems like the 13" loop could go into a cup or something to keep it neat. I’m sure that I’m a simpleton and I’m missing a whole lot, but that’s all I could think of as I read this. In my experience defusing bombs and learning to machine it seems like it would work to me lol.
EDIT: We would definitely need something to keep the slack portion of the rope higher than the feed portion, but that seems fairly simple to me with an eyelet.
remember that you have to pass a very significant amount of rope over your
capstan, with a 4" diameter capstan, you get about 1’ per full rotation, so you
will need to go about 9 full rotations. This means that the rope will need to
slip or roll horizontally or it will climb off the end of the capstan. how do
you make sure it doesn’t slip lengthwise as it’s doing so (remember, slippage of
1/10 of a mm will be a problem after it happens a few times)
also, if you have a lot of weight hanging off of the roller or linkage, you have
the problem that the weight at that point will cause deflection, so the rope no
longer is directly pointing at the bit.
I would think that you could have grooves in your pinch rollers that keep the rope from moving much, you could also have both directions from the roller going through an eyelet to keep them from them moving up and down on the driven roller. You could prevent deflection of the rollers by having them inside the roller carriage, I would think. I’m no engineer by any means, this is all just flying by the seat of my pants off of things I have witnessed. Hopefully the pictures brings more clarity to my idea. I’m a bad artist but I’m currently worse at CAD, so I hope the picture is clear.
EDIT: I would think that you could add a groove, similar to threading, after the knurling was done that the rope would naturally want to fall into. You could still have a lot of purchase with the knurling…I’m beginning to think that you would still need the rope to be pulled from the looped end though…this idea doesn’t solve that problem at all.
withaspiral groove, the rope will move 9 rope diameters in 9 rotations, tht’s
significnt
I think I see what you are saying, but as long as it is following a consistent path we can calculate for that can’t we?
If you are pulling from too low (referencing the Z axis), you pull the sled away
from the workpiece. If you pull from too high, you tend to tip more (especially
at the edges).
also, what diameter rope are you talking about? what is 9-10x this diameter? how
much will the pulling point move? (note, if you can accept that much movement,
you don’t need to worry about slack, just let it wind in the grooves)
I’m not saying that it can’t be done, just pointing out the problems to be
solved.
also, how much stretch is there in ‘low stretch’ rope? when we looked at this
last time, we found that there was a significant amount of stretch in even
braided steel rope.
You could definitely have it so that the rope is coming out at the same level as the rollers for the ring…unless I’m misunderstanding what you’re saying. This would still require a frame that has hard points, I didn’t see any mention of having no frame. I am deviating as little from the current Maslow as possible in my ideas of how this would be done.
Someone mentioned a rope of 7/64" Amsteel blue earlier on. This rope stretches 0.46% at 10% load, with a max load of no less than 1,400lbs according to their site. With a sled of 45lbs/1400lbs we are looking at using 3.214% of the max load. You get an elastic elongation of 0.96% at 30% load. The pulling point should be held stationary by an eyelet so that it is consistent. The eyelet would basically be the same as the hole that the chain goes through on the current Maslow roller carriages.
0.0046 x 12ft = 0.0552’ x 12 = 0.6624" x 25.4 = 16.825mm of stretch that the machine would pull in if we used the motors to pull 140lbs of force on the ropes. That’s probably an absurd amount to think we would pull, but it was easy math for my simple brain.
I can try to plot the points in FreeCAD for that rope stretch with the % of stretch on the Y and % of load on the X to make sure they’re linear. I don’t see a spec for the diameter of the rope changing as it stretches, but I would think that we could hold it under a consistent tension so that the elongation and diameter stay consistent. Wouldn’t that be part of the beauty of having the four motor Maslow?
EDIT: I would think the preload on the ropes would actually allow us to move a little bit quicker, like having a sturdy base on a machine.
You could definitely have it so that the rope is coming out at the same level as the rollers for the ring…unless I’m misunderstanding what you’re saying. This would still require a frame that has hard points, I didn’t see any mention of having no frame. I am deviating as little from the current Maslow as possible in my ideas of how this would be done.
not if the height of the rope is dependent on how much has been wound
Someone mentioned a rope of 7/64" Amsteel blue earlier on. This rope stretches
0.46% at 10% load, with a max load of no less than 1,400lbs according to their
site. With a sled
of 45lbs/1400lbs we are looking at using 3.214% of the max load. You get an
elastic elongation of 0.96% at 30% load. The pulling point should be held
stationary by an eyelet so that it is consistent. The eyelet would basically
be the same as the hole that the chain goes through on the current Maslow
roller carriages.0.0046 x 12ft = 0.0552’ x 12 = 0.6624" x 25.4 = 16.825mm of stretch that the
machine would pull in if we used the motors to pull 140lbs of force on the
ropes. That’s probably an absurd amount to think we would pull, but it was
easy math for my simple brain.
each motor can pull 30Kg, so it’s not off by much.
I think things will get very interesting if you wind under high tension (with
the rope stretching) and then release under low tension (cut across the top
towards the motor, high tension, then cut down the side, low tension)
I can try to plot the points in FreeCAD for that rope stretch with the % of
stretch on the Y and % of load on the X to make sure they’re linear. I don’t
see a spec for the diameter of the rope changing as it stretches, but I would
think that we could hold it under a consistent tension so that the elongation
and diameter stay consistent. Wouldn’t that be part of the beauty of having
the four motor Maslow?
you can’t keep all lines at equal tension as the sled moves around the
workpiece, the angles of the lines will vary enough that you will run into
problems. think about being at the bottom center. If you have all lines at equal
tension, the sled will move up because the tension of the top lines will have
more of a vertical component than the tension of the bottom lines.
The rope isn’t being wound around the roller as it winds, the rope would be passing through. I am thinking of pillageTHENburn’s idea of the sled climbing rope, I don’t think I was clear on that before.
If the ring is keeping all lines coincident at the router bit I don’t see why they can’t be under equal tension that is in excess of the weight of the sled. I will go chew on that for a while and see if I can figure out what I’m missing, carry on while I sit over here with my dunce cap trying to learn lol
EDIT: I think I’m really missing how this sled of mine would stay vertically oriented also…I’m sure I’m overlooking many aspects in my excitement for the idea.
My initial design for a four motor sled used this approach. However, the rope would wrap up on itself and get bound as the machine pulled in and let out rope. My implementation could have been flawed, but it was a nightmare. I tried my best to control the wrap and ended up discovering that the friction of the rope pressing against the spool for even a small amount of wrap (45 degrees) was enough to turn the spool consistently. I found no need to wrap at all.
That means instead of incremental encoders we can use a commodity 10 turn potentiometer and have an absolute encoder! Yeah.
Were you trying to turn the spool or move something by turning the spool? I’m thinking that the servo motor would drive the spool as it climbed the rope. I was just thinking of wrapping the rope around the spool for added friction so that it doesn’t slip and lose position. I’ve used the idea for a corner in a long pull line, but I have never tried to drive the central pulley. At first I thought that having rollers on each side of the driven pulley would allow the excess to go slack but the more I think about it the more I think that wouldn’t work at all.
