In the quest to measure valve leakage in a Duo-Art, Julian Dyer set
forth an interesting proposal, to measure the vacuum drop at the bleed
strip. I still do not understand what problems rebuilders are running
into with Duo-Art valves, frankly.
The Duo-Art valve is much less critical than the Ampico B single valve,
for example, because of the relatively short note tubing runs going
directly to the valve itself.
The very slight differences measured in fractions of an inch of water
are actually immaterial, when you add up all the other reasons the note
may be playing slightly differently. If Julian will also listen to the
note being played as well as measuring the differences, he will notice
(as I have) that there is barely a correlation, except on the clearly
That is because of things like striker rod friction, the tightness of
the leather nuts on the pneumatic finger, the degree of lost motion
still present, the resistance of the key, the differences in regulation
between the capstan and the hammer (always present), the different
resistances of the damper line, etc. etc. Then you also have the
"perceived" loudness differences between hammers voiced softer and
These tiny fluctuations from "perfect" are not the reason Duo-Arts have
performance problems. Some of the finest sounding and playing Duo-Arts
you will ever hear will not pass these tests, and the factory never
required such exquisite effort to properly set one up! If we start to
believe that only such meticulousness will ever again produce a decent
Duo-Art, then we are going to discourage a lot of people from even
trying, and blaming it on the piano's bad design. That is absolutely
NOT the case!
For years now, people have been writing articles about how to set
valves with a micrometer, for example. Others use a shim gauge. None
of those methods work very well when you are shooting for 1/32" gap
with a suede leather face. As you go down in gap length, you greatly
exacerbate your valve problems, and that is why these newly-valved
pianos are giving trouble, I suspect.
If rebuilders would just stop trying to gap their valves so short, and
make sure that no valve has less travel than .035-.040, then they will
be on the right track, and suddenly, the Duo-Arts will spring to life
Blow through the valve plate hole and feel the resistance of that hole
to your breath. Now put that valve plate down with two screws and the
valve inside and blow again. There will be a valve position where you
can start to feel it throttle the passage of air. If you set the valve
with a gauge to .032 and blow through it, you will feel a tremendous
throttling. Take just one paper shim out, replace the plate, and blow
again. What do you know! You can actually feel the difference that
only ONE 2-1/2 thousandths paper shim makes! What does this tell you?
It says that you can actually measure the air passage difference
between those two "settings."
It also tells you something else that is very valuable. As you remove
punchings and the difference becomes less and less, likewise does the
difference become less to the player. So where I set these valves is
just at the point that I can feel a certain amount of "throttle"
resistance under the full bore of the valve plate hole. That sets me
to an optimum setting. Forget valve travel times. Set the valves
optimally this way, and the speed of response takes care of itself.
Anytime you set valves close and precision, you also need a precision
valve to seal properly. The longer a valve takes to get sealed off,
the more time you have to add to its travel time. And since a valve
with a very short gap requires about .005-.010 "squiggle" to seat
firmly on its seats, by denying it proper travel, you also subtract
from its ability to firm itself on the seat by the same amount, in
I also suggest that after you rough in the valves, using a valve plate
with the stem guide and a paper gasket glued to it, that you then re-
regulate them again when you have two of the four screws down on the
Don't gauge Duo-Art valves! Blow them. And once you do that, mike
[measure with a micrometer caliper] a bunch of them, and you will see
then why I don't mike them! I could go into a big dissertation on my
method of valve setting and why mikeing doesn't work, but this should
suffice for now. Just remember, there's no mystery to it!