Ray Finch wrote in MMD 020429 (1922 Story and Clark 88-note Player
Piano):

> The roll motor does not seem to have a governor.  Not that the
> governor is missing but rather it looks as if it wasn't designed to
> have a governor, or at least anything that senses the motor speed
> and throttles it back. I need to better understand how this system
> governs the roll motor speed without any mechanical feedback.

I don't know of any player that has a physical governor in the sense
that you are talking about except the players equipped with a clock
work mechanism that use large springs.

All of these motors run open ended.  The more air you pump through the
motor, the faster it runs.  If you can keep the volume of air constant,
then the motor will run at a very steady and constant speed.  Of course
I am assuming that the motor is air tight and rebuilt properly.

With your thumb covering the hose line of the motor and tuning the
crankshaft backwards, it should be very hard to turn.   The harder it
is to turn then the more air tight the motor is.  If while turning the
crank it travels faster or easier when the crack is in a certain
position, than this indicates a leak and can cause the motor to lurch
or slow down and not turn smoothly.

The governor you talk about is really the large pneumatic, spring, and
a knife valve or plunger that stops the unregulated air from passing
through the regulator when the pneumatic closes.  The regulator takes
the unregulated air from the pump and converts it to a regulated vacuum
source.  The amount of vacuum is directly related to the strength of
the spring that is pulling the large pneumatic open.  This regulated
air is feed to the tempo control slide valve that determines how much
air is sucked from the wind motor.   The larger the hole, the faster
the motor moves.

 I should say that this regulator system only works because the
regulator sees a constant load from the motor.   If there is a friction
point or air leak in the motor, then the motor will slow down for that
portion of the rotation.  I have seen very leaky motors that seem to
work fine since they leak the same amount of air for the entire
360-degree rotation of the crank.  Just because it works doesn't mean
it is desirable.  You can even affect the speed of the motor by how
much tension is on the upper feed spool break.  Having too much tension
can cause the roll to run slow at the end of the roll.  These sounds
like it could be used for tempo compensation but should never be used
as one.

The 2nd smaller pneumatic is the compensating pneumatic.  Many of the
players don't even have this in their designs.  Its purpose is to help
the regulator react to sharp, fast vacuum changes.  Just rebuild it
exactly as it is found.

Pete Knobloch
Tempe, Arizona, USA