Robbie mentioned this in regard to my valve tester:
[ The Ampico B valve (the old leaky valves, anyhow) seem to have
[ troubles at low vacuum and slow paper speed, like Tempo 50.
[ Instead of opening suddenly, the port is gradually opened.
[ Under this condition (old B valves) a magnet valve gives a
[ noticeably better performance than the music roll, apparently
[ because it operates quickly. Could your tester be adapted to
[ simulate the lethargic signal of a hole passing slowly over the
[ tracker bar, like the rotating-plate valve produces?
First off, I think this tester could simulate about anything. The
astable multivibrator starts about 1 pps and exceeds 1000 pps with an
initial pulse width of (calculated) 220 microseconds. The actuation
time of the valve varies by a factor of roughly .8 to 1 second or so,
which means I can make the pulse width overlap the rate. Gosh, that
means I could test valves on waltzes, too! I had never thought about
Regarding the sensitivity of the valve to the tester, initially,
I would adjust one of the ports on the speaker valve to the overall
resistance in the pouch line of an Ampico and simulate that, just as
though it had about 6 feet of trackerbar tubing fed through a typical
trackerbar tube. That would be the way to test them, anyway. You can
do that with a "tuned" bubble jar easily enough. The method is on pp.
200-201 in The Orchestrion Builder's Manual And Pneumatics Handbook.
Having recently done another model B, I checked the actuation thres-
hold of several valves just to see how sensitive they can be, once
restored correctly, and I was amazed. They jump when the trackerbar
is uncovered about equal to a little over two #70 holes. Eyeballing
that, it kinda looks to the naked eye like the hole in the paper and
the tracker bar hole are tangent but not overlapping. Of course, they
leak. Put a glass on it, and you can see an overlap of about .015-.017
on a roll perforated with the original perforators. So tiny it is
uncanny, yet there is that note down solid at a first intensity of
I am constantly amazed at that valve. It is so cool! As a great friend
of mine used to say, "Now somebody thought that out!"
Regarding gradual opening of valves, that is usually the result of
stiff pouches under low vacuum. The original pouch leather is always
so stiff and leaky that the fixed #70 bleed in that block is useless.
In this particular piano from a California collection, some
professional "rebuilder" (a more polite term than what I use privately)
had this very problem, and so, by means of a toothpick, put a drop of
"space age adhesive" (also a term only for public consumption) into
about 1/3 of the valves, which dries clear and invisible, and ran
through the bleed to solidify on the back side.
Now, that bleed is very tiny. It is in a 1/8" hole (roughly), and
buried deeply in the wood, inside the pouch block, making it almost
invisible. Inspecting it, (which is all you ever do) we could see
nothing wrong, of course. It was clean brass in there. Until we got
them in the piano and tried to play it. Now with the new, almost
airtight pouches, many of these valves floated half-way up and disabled
the piano. Of course, I didn't know what could be causing the problem
for awhile (almost a day's worth of extra troubleshooting).
Once discovered, I removed all the block valves and stabbed through
all the bleeds with a #70 drill. You'd have thought that would have
solved the problem, but no. It only solved it for about a day or so.
Then the little flap of "space-age miracle adhesive" (having stuck to
the convex rear of the bleed) relaxed back over the hole in some of
them, and now we were right back where we started. It was a case of
either a lazy or an ignorant previous rebuilder whose (very) little
knowledge at the time became a "dangerous" thing. Overall, this cost
an extra week's work. The valves had to be removed and replaced about
three times in all.
In the first place, blocking the bleed with glue is truly an idiot's
game. If you're going to do something that lazy as a repair, then just
stuff the bleed insertion hole with something. It's a lot easier,
quicker, and visible to the next guy.
But in the second place, anyone who ignores the pouches on an Ampico B
has ignored the single most important expression-dependent device in
the entire piano. Ampico sealed their pouches with rubber cement,
which gets hard over time, and fairly quickly at that, because it is so
thin and vulnerable to moisture and tanning chemicals from the leather.
Once hard, it cracks microscopically, and leaks.
To reseal such a pouch becomes a pneumatic disaster. You will find
yourself cranking up the first intensity a full inch to play it, and
when it then does play, it is uneven and plays too hard. (That's a
good 20% increase, by the way). But because such a pouch is also too
slow, the ball bleed actuates at its initiation point and is already
up and waiting on the pouch, which then must try to make it the rest
of the way with the equivalent of about a #56 bleed or worse, since the
leather now is leaky too. It's a real "Catch-22".
That is why so many Ampico B's don't play as well as their older
cousins. Very very few of them yet have their new pouches installed,
and fewer still have new ball bleeds. That means that most are just
marginally repaired. It's amazing how good they can still sound, even
in this condition, but also incredible that most owners will never
realize how much better they could sound when done thoroughly.
So if anybody wonders if their Ampico B received an attempt at new
pouches or not, there are two ways to tell. First, if the valve block
has been resanded, removing all the old shellac, and a fresh new
sealant applied, then unless this was done for appearance or deception,
they have new pouches. And if all the blocks still have their old
sealant, but you can definitely see a freshly glued pouch line on all
the valves, then you have replacement pouches.
Not that that alone will help, but, it may be a good start, if the new
leather is light enough. (Ampico B pouches must be selected very
critically from a number of extra-thin skins and sealed very well in
order to have their original characteristics. This is not at all a
problem in Ampico A's block valves, which are an order of magnitude
easier to do.)
[ It sounds like pulse duty-cycle modulation or similar will do
[ the trick, Craig. The reason I want to simulate the "lethargic
[ hole" is to demonstrate the big before-vs.-after difference
[ when a bad unit valve is properly rebuilt. With your test setup
[ I should be able to measure, and record, the dynamic pressures
[ and motions in the valve and a pneumatic, and see the improvement
[ expressed in numbers. -- Robbie