After an overlong hiatus I am reactivating my Ampico B, which has
had the valves rebuilt (among other things).  I am finding a large
variability in terms of sensitivity (apparent at low pressures) and
repeat speeds.

I am aware of most, I think, of the other factors involved (integrity
of pneumatics, freedom of movement, uniformity of piano action) and
have gone over these pretty carefully.  My thought is to build a jig
for testing the valves and I'd like to discuss the strategy.  I realize
that the proof of this pudding is whether the note plays correctly, but
it is very tedious to determine problems and adjust using the stack.

I can easily derive regulated vacuum from, e.g., the pedal source in
the piano (which is about 18") but this will have to be reduced for any
effective testing.  I have in mind simply a needle valve bleed.  This
will lead to some pressure variation as the valve operates, but if the
valve "load" (i.e., the pneumatic path) is restricted to a very small
bleed the loss should not be too great.

(One wonders if the load can't just be sealed off, but I recollect that
would not permit proper valve operation).  A meter on the input vacuum
would permit testing the valve action under a variety of heads.

Issues worth testing would involve pouch leakage, bleed leakage, ball
valve operation and speed of transition.  I am sure that many on this
forum have looked at these procedures and would have far wiser things
to say about how to go about them.

A. B. Bonds