I believe I've finally hit on the correct solution for this existing
machine. I wouldn't do this if I were designing from scratch, but I
ended up with a whole lot less room inside the spool-up area that I first
thought I'd have.

First, the issue of constant tension: The process will work OK over a
fairly narrow range of tensions, but not over the range the machine
provided. Part of the problem was hysteresis in the friction clutch, but
part of it was due to the large differential between "empty" and "full"
spool sizes.

It turns out that a "full" spool holds thousands of feet of product, when
a standard run is in the low (3-4) hundreds of feet. So the solution
here is to bulk up the spool hub until the difference between empty and
full is only about 10% of the spool diameter. That means a 10%
difference in empty-to-full tension if constant torque is provided -- and
that's well within the range that will work properly for the process.

The other problem - hysteresis in the clutch - turned out to be a lot
simpler to solve than I first thought. The spool rests on a driving hub
that has pins which engage holes in the spool end, and is supported by a
cantelevered bearing in the driving hub. The other end of the spool is
supported by a bearing on a swing arm so the spool can be removed/re-
inserted in the machine.

The driving hub is on a simple two-row bearing, so when the clutch is not
slipping, the spool takes up a lot of tension in the form of stretch in
the product. When the clutch finally slips, some of that tension is
released by the spool by rotating backwards against the friction clutch
until the tension is at/below the slipping torque of the clutch.

Simple fix. Replace the drive hub bearing with a pair of sprag
bearing/clutches. VBX (and others) make some nice combination deep groove
ball bearing/sprag clutch combinations in standard 6200 bearing profiles.
In the 25mm i.d. sizes, they all withstand several times the required
back-off torque, and I'm using two in the same hub.

Now when the clutch slips, the spool cannot reverse. Within a half-
second or so of starting, the clutch reverts to a constant-slip mode, and
never gets back to the stiction mode. Backlash is virtually zero on the
VBX sprags.

And hey! It WORKS! G

LLoyd

On Mon, 21 Jun 2010 08:37:09 -0500, "Lloyd E. Sponenburgh"
lloydspinsidemindspring.com wrote:

I believe I've finally hit on the correct solution for this existing
machine. I wouldn't do this if I were designing from scratch, but I
ended up with a whole lot less room inside the spool-up area that I first
thought I'd have.

First, the issue of constant tension: The process will work OK over a
fairly narrow range of tensions, but not over the range the machine
provided. Part of the problem was hysteresis in the friction clutch, but
part of it was due to the large differential between "empty" and "full"
spool sizes.

It turns out that a "full" spool holds thousands of feet of product, when
a standard run is in the low (3-4) hundreds of feet. So the solution
here is to bulk up the spool hub until the difference between empty and
full is only about 10% of the spool diameter. That means a 10%
difference in empty-to-full tension if constant torque is provided -- and
that's well within the range that will work properly for the process.

The other problem - hysteresis in the clutch - turned out to be a lot
simpler to solve than I first thought. The spool rests on a driving hub
that has pins which engage holes in the spool end, and is supported by a
cantelevered bearing in the driving hub. The other end of the spool is
supported by a bearing on a swing arm so the spool can be removed/re-
inserted in the machine.

The driving hub is on a simple two-row bearing, so when the clutch is not
slipping, the spool takes up a lot of tension in the form of stretch in
the product. When the clutch finally slips, some of that tension is
released by the spool by rotating backwards against the friction clutch
until the tension is at/below the slipping torque of the clutch.

Simple fix. Replace the drive hub bearing with a pair of sprag
bearing/clutches. VBX (and others) make some nice combination deep groove
ball bearing/sprag clutch combinations in standard 6200 bearing profiles.
In the 25mm i.d. sizes, they all withstand several times the required
back-off torque, and I'm using two in the same hub.

Now when the clutch slips, the spool cannot reverse. Within a half-
second or so of starting, the clutch reverts to a constant-slip mode, and
never gets back to the stiction mode. Backlash is virtually zero on the
VBX sprags.

And hey! It WORKS! G

LLoyd
========
Thanks for the "rest of the story." Good job!


-- Unka George (George McDuffee)
...............................
The past is a foreign country;
they do things differently there.
L. P. Hartley (1895-1972), British author.
The Go-Between, Prologue (1953).

Lloyd E. Sponenburgh wrote:
I believe I've finally hit on the correct solution for this existing
machine. I wouldn't do this if I were designing from scratch, but I
ended up with a whole lot less room inside the spool-up area that I first
thought I'd have.

First, the issue of constant tension: The process will work OK over a
fairly narrow range of tensions, but not over the range the machine
provided. Part of the problem was hysteresis in the friction clutch, but
part of it was due to the large differential between "empty" and "full"
spool sizes.

It turns out that a "full" spool holds thousands of feet of product, when
a standard run is in the low (3-4) hundreds of feet. So the solution
here is to bulk up the spool hub until the difference between empty and
full is only about 10% of the spool diameter. That means a 10%
difference in empty-to-full tension if constant torque is provided -- and
that's well within the range that will work properly for the process.

The other problem - hysteresis in the clutch - turned out to be a lot
simpler to solve than I first thought. The spool rests on a driving hub
that has pins which engage holes in the spool end, and is supported by a
cantelevered bearing in the driving hub. The other end of the spool is
supported by a bearing on a swing arm so the spool can be removed/re-
inserted in the machine.

The driving hub is on a simple two-row bearing, so when the clutch is not
slipping, the spool takes up a lot of tension in the form of stretch in
the product. When the clutch finally slips, some of that tension is
released by the spool by rotating backwards against the friction clutch
until the tension is at/below the slipping torque of the clutch.

Simple fix. Replace the drive hub bearing with a pair of sprag
bearing/clutches. VBX (and others) make some nice combination deep groove
ball bearing/sprag clutch combinations in standard 6200 bearing profiles.
In the 25mm i.d. sizes, they all withstand several times the required
back-off torque, and I'm using two in the same hub.

Now when the clutch slips, the spool cannot reverse. Within a half-
second or so of starting, the clutch reverts to a constant-slip mode, and
never gets back to the stiction mode. Backlash is virtually zero on the
VBX sprags.

And hey! It WORKS!G

LLoyd




A torque motor would do the job too.


http://machinedesign.com/article/tor...the-trick-0403



John

John fired this volley in
news
A torque motor would do the job too.


But, John, it would have required "redesign" on a major basis. This
required only two things -- wrapping the takeup spool with bulking (vinyl
flooring sheet goods), and replacing ONE two-groove bearing with two
sprag clutch/bearings. Free for the sheet goods, and $50 for the two
bearings. No machine work -- none, period.

NO mechanical re-design of the machine, at all. Just re-fitting of
existing parts with new parts that fit in the same bores, same profiles.

I don't have a lot of experience re-working old machines to new specs (I
generally design to spec for new stuff), but it seems that replacing a
bearing is a LOT better than re-belting (or re-chaining) the whole
apparatus (along with building motor mounts, jackshafts, and power supply
for the motor).

All the constant-torque motors I've seen that would physically fit in
this space would require at a minimum 10:1 reduction to get to the 20ft-
lb torque I was looking for. That'd be two stages of belting or chaining
jackshafts in this case. And it wouldn't have gained me _any_ different
performance than I got with this method. Still, it would be constant
torque, not constant tension. ??? why would I want to DO that? To do it
in a "elegant" way? Elegance is in simplicity, I think.


LLoyd