I would like to offer an idea concerning the performance characteristics
of the Duo-Art Cross Valve.  I have only recently thought of this idea,
so anything I say here is only theory.

Here goes: Start with a sheet of suitable material (such as Teflon)
that is perhaps 3/32 of an inch thick.  Make up 88 tiny blocks that
measure approximately 1/4 inch by 1/8 inch.  A small hole is drilled
through the block in the center of the 1/4" by 1/8" face.

That small hole is made to have a tight fit on the valve stem.  The
proportions of the block are such that it fits loose enough inside
the cross valve plate to not interfere with the valve's motion, yet
the valve can no longer rotate because the block is at the center of
the cross.  The precision necessary to achieve these goals means that
the making of these blocks is not something to be undertaken lightly.

I have been friends with Bob Streicher, the expert machinist, for
many years, and he says that the blocks are no problem to make up.
He didn't quote a price, but Bob's prices are always too low anyway.
If you are going to call, note that he has a new area code, and the
old number will not redirect you.  The new number is 570 559-7403.

One good thing is that the blocks are totally "benign" and do not alter
the originality of the piano in any way.  I originally thought of them
as going on the bottom of the valve (the stem has a cut thread and not
a rolled thread, so there is no problem for the friction fit of the
block).  Later on, I realized they could go on the top as well, and,
in fact, would be easier on a retrofit.  They would also be analogous
to the top collar found on many valves, including Aeolians.

If anyone has any comments about my idea, I would like to hear from
them.  As for now, we are all in the theoretical stage.  While it may
be true that ten tons of theory is no match for one ounce of reality,
you have to start with theory.

At this point, I thought I might add some personal tips on working with
cross valves.  Before I even started my first stack, I had heard of their
reputation.  The first thing I did was to get a large, spring-loaded
reservoir (such as from a set of pumping pedals) and rebuild it so it was
as tight as possible.

Then I needed a method to measure how fast it would open when connected
to something.  Originally, I had a pointer-type indicator mounted on
one leaf of the reservoir, to pass over a stationary grid.  I used a
stop-watch to time the travel of the indicator between two points.
Later on, I used two micro switches to turn a clock on and off
automatically.

It didn't matter to me what the pressures were, or what the volumes
were.  I knew that if one side of a stack opens the reservoir in a
significantly different time from the other, then I had a problem.

All grand stacks have three tiers, and are divided into bass and
treble.  This gave me six sections to compare.  In rebuilding, I had
removed the burnt shellac from the lower valve wells and re-sealed them
with two applications of PVC-E glue from a hobbyist syringe.

I began to re-assemble the rebuilt stack.  I connected the pouch boards
to the upper boards, but I did not put the valves in yet-I sealed each
lower cross plate with duct tape.  I then saw that all six sections
were quite tight.

I then dropped in the valves, and tested again.  Although not as tight
as the duct tape, the six sections were pretty tight, and more
importantly, equal.  I assumed that after giving each valve its travel,
I would be through.  Wrong!  The six sections were leaking different
amounts.

After examination, I realized the problem: It was not until this last
step that the valves were locked in place by the top diagonal brace.
It seemed that slightly bent diagonal braces and/or slightly bent valve
stems could tilt the valve off its seat, and create a nightmare for
these stacks.

And valves held off their seat at rest were only part of the problem.
There were valves that seemed perfect at rest, but in repetition tests
at low vacuum would somehow stick.  The only explanation I could think
of was that bent stems and guides would not bind until the valve rose,
and this would be helped by the fact that the vacuum was very low, so
the valve didn't get sucked down again.

I sent the valves to Bob Streicher so he could true them upon his
lathe.  Upon reassembly, I paid close attention to top or bottom braces
that might not be perfect.  When I reassembled the stack, and tested
the vacuum timing, the bass half was very close to the treble half.
The stack played great.  The piano was shipped out of town, and the
owner never complained.

Randolph Herr