The 1920's Aeolian wind motor is one of the best-designed in the
industry.  With double-acting pneumatics sharing the hinges, crank-arms
and slide valves -- plus a crankshaft design that keeps slide valve
travel at a minimum -- the friction is greatly reduced with maximum
efficiency the result.

Unfortunately, some measure of this remarkable efficiency is commonly
lost by using an improper rebuilding technique.  The tendency, especially
among first-time rebuilders, is to hold one end fully closed with
rubber bands while applying covering material to the other end.  This
results in less power and lower efficiency associated with excessive
flexing of the covering.

If the motor still has its original cloth, you will notice that with
the crankshaft assembly removed the cloth will stretch tight at one
end before the other end closes completely.  The gap is, or should be,
one-quarter inch.  It may be a bit more on some of the earlier motors
without the slide valve travel reduction.

Most rebuilt motors do not run at a uniform speed throughout each
revolution.  One reason is that the amount of material used varies from
one pneumatic to another, or from one half to the other on the double
type.  Another reason is that the covering on one or both sides of one
or more pneumatics, especially toward the hinge, will stretch tight
during the normal operation of the motor, even though the span is
correct.

If you use a spacer to hold one end of a double pneumatic open the
quarter inch or so while recovering the other end, use a thinner spacer
to compensate for the covering thickness when applying the second
covering.  That may seem like splitting hairs, but to one with a good
sense of rhythm, the slightest variation from the steady beat of dance
music can be quite an irritation.

I've never done this, but I've thought of building a tester using
a strobe disc connected to a wind motor to produce a tone which would
of course waver according to fluctuations in the motor's speed.  The
results would be most interesting, as the steadier the speed, the less
wow and flutter would be audible through the speaker of the tester.

At this writing I am wondering what covering material I should use on
my next motor.  One possibility would be soft glove leather, somewhat
thicker than that used for pouches, with a sixteenth-inch overlap at
the hinge and sealed airtight.  The thinnest double-thickness motor
cloth would also be a good choice, either extended to the hinge with
a slight overlap, or done the original way with pouch leather at the
hinge.  Of the two, the all-leather covering would undoubtedly be the
longest-wearing and probably best for the most frequently played
instruments.

Jeffrey Wood

 [ How can the improved performance after rebuilding be tested and
 [ demonstrated (with numerical data)?  -- Robbie