I believe that rolls that tear on rewind are caused by a multitude
of things, and it's seldom the geometry of the spool box or its
components.

I do not recommend modifying the spool box because you are tearing
rolls.  I also do not recommend taking off the rewind brake -- on
any but the worst players with the worst tracking mechanisms.  And
I definitely do not recommend shortening the takeup spool.

Comments about coin pianos:

 - Since end plates of the spool frame are extra heavy and precision
drilled, often with bushings or bearings installed as well,
 - and since rolls are not reliant on small, thin cores with wobbly
flanges,
 - and since the rolls are not changed by the pianolist each time it
plays, and are not subjected at all to the same wear and tear of rolls
which have been played on perhaps a half-dozen old players through
their lifetime, losing tabs and being damaged in the process,
 - that's why rolls played on automatic players do not tear up, even
though most of them have been played far more times than any ordinary
home player roll.

I have a few questions that these points bring to mind:

Nelson-Wiggen pianos used a wooden spool box, and they track well.
Is it the precision geometry of the spool box or the larger holes in
the roll that makes the Nelson-Wiggen not need a tracker mechanism?

Early Wurlitzer instruments used a wooden spool box with a small
five-tune roll wound on a wooden spool with wood spool ends (flanges).
They also have a small take-up spool with adjustable flanges on the
take-up spools.  These instruments have small holes in the music roll.

By the way, it was a common math error that gave them their unusual
hole spacing.  These machines have no tracker and played the rolls
more times than any home player and were often used on many different
machines as the route operator would take the rolls from one location
to another.  They tracked well and are usually found with the roll
edges intact.  Was it the precision of the spool box that prevented
these rolls from tearing and tacking well?  I know it wasn't the
quality of the paper, as it varied from poor to okay.

Every coin piano I have ever seen has flanges on the take-up spool that
are adjustable.  There were many cutting errors in the coin piano rolls
yet they track well, even Coinola rolls with their nine-to-the-inch
spacing.

The Reproduco had a model that used small rolls much like a home player
piano with no tracker mechanism.  I know of no original coin pianos
with bearings, other than the Mills machines which were the most
precision made of all the coin piano spool frames.  There is little
play in them even after 90 years of use.

My question of the day is why not try to improve the geometry of an old
player that tears up the rolls?  Were they all perfect the day they
left the factory?

Don Teach

 [ I feel that the crude Bakelite and plastic flanges are a big part
 [ of the problem, because they allow the spool to wobble on the chucks.
 [ Systems using pin-end spools, such as 58- and 65-note organs and
 [ pianos and most European player pianos and orchestrions, seem to
 [ be gentler.  -- Robbie