Yesterday John Phillips mentioned that his "Independent brand" 65-note
piano rolls tracked poorly, with the music perforations seeming to be
too far to the left.  This is a common problem with some pinned-end
rolls, especially ones made by other companies than Aeolian.

This difficulty often stems from the way that '6-to-the-inch' are
supposed to be aligned in the spool box.  65 note pianos, 58 note
organs, and other 6/" instruments had no automatic tracking device;
the wide hole spacing usually was able to allow for any normal paper
width variations.  But in order for this to work, the paper has to
start out at some constant relationship with a fixed reference point
in the player.

This point is the end-face of the drive chuck, which can be permanently
adjusted with a screw in the transmission, but can't move once set.
This should be very slightly more than 9/16" from the right edge of
the paper.

In order for all rolls to meet this requirement, the spools were made
in a particular way to insure this.  Each flange-end was made from two
pieces, the core and the flange itself.  Sometimes the core was one
long wooden stick, serving as the entire spool core, and extending the
whole length of the roll.  In later years, there were two short core
pieces, one for each end, stuck into a cardboard tube.  The flanges
were separate, disc shaped turnings, slipped over a reduced diameter
extension, or tenon, turned on the ends of the core.

The secret is the way this tenon was turned.  In the 'back-knife
machine' (the automatic device that makes turned wooden objects), the
blade that formed the end was stepped, with two faces exactly 9/16"
apart.  The outer step on the blade, cut-off the far end of the tenon,
while the same blade made the shoulder, or inner end of the tenon,
exactly 9/16" back from the end.  When the flange disc is glued on to
the tenon, its inner face, and thus the right edge of the paper, ends
up exactly 9/16" from the end of the tenon, which is firmly pressed
against the face of the drive chuck by the idler chuck spring.

The problem is that the glued joint which holds the flange turning
to the core often breaks, causing the flange to slip to the right, and
making the paper also wander rightwards.  This problem was recognized
by some of the makers.  All kinds of measures were taken, such as
driving nails through the flange into the core, or forcing a tight-
fitting eyelet over the protruding end of the tenon, to reinforce this
joint.  But in many cheaper rolls, this was just a glue joint, and is
often found to have worked loose over time.

If the roll is unwound, and found to have this problem, it is easily
fixed.  If the tenon is found to be shorter than the standard 9/16"
length, or if for some unknown reason, the roll is actually punched off-
center to the right, some paper washers or punchings can be slipped
over the drive-pin to correct the problem.

Richard Vance