I have been working, literally for years, to collect all this
information, as I rebuild the spool frame on my Caliola.  When completed
I will write this up as dimensioned drawings with photographs for
inclusion on the MMD.

My suggestion, unless you want to spend years and have an endless
budget, would be to order one of the custom roll frames (catalog no.
2100) from Player Piano Co, Wichita, Kan.

That said, let me share some of what I have gleaned from many hours
spent over the last five years investigating, learning, collecting
hundreds of photographs from many organ rallies, tracking down parts.
Much of what I discovered is already in the MMD archives.

The Graf patent is a schematic version.  I have scanned some of it into
an older version of Ashlar's computer-aided design and modeling program
called Vellum.  While the patent is complete, there are a lot of extra
features used to expand the claims, to make them as broad as possible.
On the other hand, there are a number of essential items missing.

Duplicating a Wurlitzer roll frame is not for the faint of heart.  I am
still working on the bill of materials (a feature in the Vellum
program).  There are hundreds of parts in the complete frame.  I would
not be surprised to find nearly 1000, when all is said and done.

I am more than happy to share the IGES and DXF files of the designs when
complete. This may be a few months.  The late Ken Smith did draw up a
set of dimensioned plans.  Unfortunately he has it set up as a pressure
system.  I made some parts using Ken's plans, and they did not fit or
match the dimensions of original parts I later acquired.  Possibly Ken
took the dimensions from an Artizan roll frame.

There are six major variations of the frame.  These divide into three
subsections, each with two major variations.

I call the first of these subsections power transmission and speed
regulation.  Early versions use a leather belt and pulley.  Another
variation uses a bevel gear and transmission rod.  This system is shown
in the patent.  Somewhere along the line Wurlitzer removed the speed
regulator from the frame, but used the same castings.  There are
variations and revisions of the speed regulator.  I call these the "A"
frame and the "V" frame.  Photographs of these can be seen in Bill
Black's excellent articles on Bruce Zubee's carousels.com website.

The second subsection is the rewind controller.  Bill Black documents on
the carousels.com website a system which takes its power from the
pressure pump.  This is more typical of Wurlitzer organs.  Its pianos
and later instruments use the improved version shown in the patent with
the Block pneumatic.  "A" frame speed regulator versions use a different
method of tripping the rewind.

The third subsection is the frame proper.  There are two major
variations, simplex and duplex roll frames.   Simplex frames can have
the same gearing as a duplex, when used with A or V speed subsections.
Or simplex frames can have the speed governor attached, as noted in the
first subsection.

Tracker bars add another level of complexity.  In the MMD archives are
instructions for making a tracker from a solid piece of stock.  But that
is not the way Wurlitzer made tracker bars.  Wurlitzer used a design
which was easy to mass produce, but was cutting-edge technology in the
1920s: a stamped sheet-metal oragami that folds into a tracker bar.  The
stamping machines alone cost tens of thousands of dollars.  I looked
into this a few years back.  It would cost the same to make the dies,
whether I produced one or 100,000 tracker bars using the Wurlitzer
method.  The price break really starts to come when one makes stamped
metal items in the millions.  A few cents each, but I am not sure that
there is currently a demand for a few million tracker bars.

A generation ago, Pittsburgher Steve Lanick was able to make complete
roll frame assemblies on a small scale.  This was when some of the
infrastructure still existed.  Now EPA and OSHA regulations have all but
shut down such activities.  Basically one needs to get together a few
friends and go into the shop late at night for some moonlighting.  In
this day an age I find that most employers tend to frown on such
activity.

Marking the holes in the bar and drilling them is not at all difficult.
The .1227" spacing is 1/40th subtracted from pi-millimeters.  So any
metric screw-cutting lathe can be used with a 40:1 dividing head.  Each
crank of the head advances the stock 1 millimeter.  There are 40
divisions in the dividing circle, so the last advance is 1 pin less or
pi/40.

Casting a few sets of roll frame side plates is something I would love
to do.  I have all the measurements and can create the patterns with my
CNC mill.  But are there 10 or 25 people on the MMD willing to fund the
actual foundry work?  I am also short a few take-up spool flanges
myself.  So I would probably want to have those cast first.  They tend
to break.

I should note that Wurlitzer was casting some of the frame parts in the
1920's from aluminum.  This is the equivalent in this day and age of
working with nanotech carbon composites.  Some of the heavy wear parts
(shifter clutches) are made from pot metal.  After 80 years the aluminum
and other alloys break down.

As I recall the diameter on the supply spool flange is about 5" and on
the take-up spool flange 7".  They are 1/8" thick.  The machinist who
turned and trued up the 7-inch take-up flanges for me told me, "Never
again."

One dimension of interest.  The size of the cardboard roll cores on
which rolls are spooled is the same as the "Standard" mailing tube
available at U.S. post offices.  These fit perfectly onto Wurlitzer
spool chucks.

The tracker bar rides on two rods in the roll frame and is equipped with
a thumb-screw/wing-nut adjustment, so that it can be moved a half-inch
or so right or left, to properly track the holes in whatever roll is put
on the roll frame.  There are also other alignment adjustments provided,
not to mention leather braking pads to adjust tension on the roll paper
during play and rewind.  There are so many details that it is amazing
that the design works at all.  That said, the whole of the system is
made to an incredible tolerance that only makes one admire the old
methods all the more.

Julie Porter