Much has been written and claimed about cross valve and round valve
configurations. I built a tester once with a cross valve plate on one
side and a round valve on the other. Instead of just calculating and
modeling the performance in theory (which I also did), I tested the
actual performance in a situation which would allow me to have
identical valve travel, since it was the  same poppet, captured betwen
the two different valve plates under test (No sense guessing, right?).
The answer was so demonstrable that I demonstrated the tester at the
home of Mr. and Mrs. Paul Stelzner who at that time lived in
Washington, Mo. and were hosting an AMICA meeting at their home.  It
was inarguable. Round valve plates also have better volumetric
eficiency and are quieter at the same time. That is because a cross
valve has a larger circumferential length for the amount of area it
uncorks, and creates a much greater turbulence.

 The article referenced here was one which I wrote for the AMICA
Bulletin some years ago called, "The Cross Valve, Was It a Plus?"

Regarding "special considerations in valve settings" when cross valves
are changed out, however, I must demure. Let me suggest a test to
anyone who really wants to know.

Don't take anybody's word for it! Find out for yourself.

 Take a plastic film can, insert a  3/8" nipple through the bottom and
seal around it with a hot glue gun. Bore a 1-1/4" hole in a strip of
wood so you can clamp the board to your bench and rest the assembly in
the hole, upside-down (The poppet will be underneath and your supply
nipple pointing up, everything resting on the valve plate). Set a cross
valve plate down on the open end of the can, sealing around it
perfectly with whatever you want to use, then taping the hole in the
valve plate temporarily and testing the can nipple with your tongue to
make sure it's cartridge tight and can't leak around the valve plate's
top. Remove the tape.

Tape a wire hook to the valve poppet stem threads that you intend to
set upon the cross valve plate. Rig up a guide so the stem cannot be
tipped or wrenched off the plate. It has one direction only-- up and
down. Your test fixture is now supported upside-down by the strip of
wood with the hole, sitting on the valve plate itself. Turn on the
vacuum, snap the poppet up to its hole in the home-made stem guides,
measure vacuum accurately. Now take a dixie cup, make a string bail for
it, hang it on the poppet stem hook, and start gently filling the dixie
cup with shot through a straw. When the poppet breaks free, weigh the
cup on an accurate postage meter scale. Repeat the experiment 5 or more
times for reliability. Then do the same experiment using the round hole
plate.

To calculate, P x .036 = lbs/sq. in. (where P is inches of water vacuum
in the can). You will then be able to accurately estimate the effective
sq. in. area (not a theoretical one) as long as your trials are within
about 5% of each other, or closer. Both tests must use the same poppet,
of course. And what do you know-- the difference is barely noticeable.
Your assignment, should you choose to accept, is to gap Duo-Art valves
such that you have an optimum travel to allow the piano to achieve its
full power without having too large a transit loss or slow repetition.
Duo-Art pouch size is plenty adequate to handle the small difference in
opening pressure (if any), as clearly demonstrated by other pianos
having roughly the same size valve footprint and pouch diameter.

Part of the reason that there isn't hardly any noticeable difference
between valve plates is (my opinion, now) because the cross valve plate
allows the leather to "wrap itself" or conform to the sharp edges
around the sealing surface of a cross valve, caused by grinding it
after being punched. And what do we see after 70 years? A deep imprint
in the cross valve leather, shaped like a cross. Conversely, you do not
find round imprints of round valves, deeply imbossed, in any player
piano! I have dozens of old, used valves and cannot find even one of
them with an imbossed imprint! Isn't that interesting?  And isn't it
also interesting that no other valve plate in the business was ever
ground with sharp edges like that one, after having been stamped out?
Do we suppose there may be a connection?  I do.

As you might also guess, this can vary quite a bit according to the
leather you use. But you have to add the extra sealing area for each
little bit of leather nap that is attracted to the plate. The force is
dependent upon the total area which the leather poppet contacts, not
just the area of the hole.  This also varies greatly by the depth of
leather nap and the cushion or softness and thickness of new garment
leather. You will probably also discover that if anything, the cross
valve requires more-- not less- force to remove itself from the valve
plate at any given amount of pressure.

I don't want anyone to believe that we have a secret way of calculating
things that they cannot do for themselves. The real secret to player
rebuilding has always been patience, consistency, and the determination
to tear something down and do it again as many times as it is not quite
right, and to not begrudge yourself that time, as long as you can do it
better. That's about the only thing we can't teach, but it's also the
hardest of all to learn. So just hang in there and remember, things
aren't as mysterious as they may sound.

Craig B.