Alexander Zenin notes a problem with his Autopiano "standard" player
action in MMD 160213:
> ... but with pedal suction the valves are not sealing to the brass
> flanges! What's wrong?
I will make a couple of assumptions (always risky). I assume the
valve facings that Alexander replaced are the ones internal to the
valve, which seat against the vacuum supply plenum, and _not_ the
facings which seal the atmosphere port when the valve is actuated.
I also assume he used a hard-tanned leather identical or equivalent
to the original material, not a soft or fuzzy napped leather.
In my experience, these valve facings almost _never_ need replacing,
as their protected location makes replacement unnecessary unless the
piano has suffered an infestation of "Lepisma saccharina" (silverfish)
or a similar vermin. The facings on the atmosphere side of the valve
are typically not so long-lived, and typically _do_ need to be replaced
with fuzzy napped material.
An important consideration that restorers _must_ keep in mind is
that steady-state leakage (i.e., no notes activated), in this or any
similarly designed player stack, is _totally_ determined by the
internal leather, stem, and seat plate/gasket tightness to the vacuum
chest, as well as the leakage _through_ the thickness of the chest
There is _no_ contribution to steady-state leakage from the
atmosphere-side facings. In fact, if you think about it, those facings
do not even need to be installed for the stack to be "tight". It is
only when one or more notes are actuated that their purpose is served.
If Alexander did indeed replace the internal facings with the correct
material, he would have had to have removed several of the force-fit
metal rings on the valve stem that hold the facings and associated felt
and small leather washers in place.
A frequent mistake of restorers is snugging up the "inside" metal ring
too tightly against the new facing, which limits the amount of "wobble"
the facing can make. A certain amount of wobble, which is not easily
quantified, is necessary to allow the facing, aided by its hard-tanned
surface, to "find" its best sealing position against the metal seat.
On the other hand, a ring that is too loose can allow leakage around
the center hole and stem.
The third, and usually biggest contributor to steady-state leakage
is at the plate-chest interface. If no blotter paper gasket was used
originally, you MUST provide a sealant between these two surfaces.
I personally use a 50% water-thinned PVC-E glue solution for this
purpose. It has enough "body" to seal well and yet is not so tenacious
that it cannot be easily later removed, without risking distorting the
I also spray the leather contact side of the plate with an aerosol by
the name of "Permalon" (formerly "Emeralon") which is difficult to
find. PPC used to carry it. It is essentially a sprayable "Teflon"
that has excellent adhesion and wear characteristics. Permalon is
clear and dries to a transparent gloss finish while Emeralon has a
jade-green color. This treatment helps the leather facing, along with
its wobble, to slide easily and seal thoroughly against the metal
I also suggest that a gap of 0.047" [1.2 mm] is much too large and can
lead to inefficient poppet operation, since it spends a larger amount of
time in the "mid" position where there is a direct "short (pneumatic)
circuit" between the vacuum plenum and atmosphere. A gap closer to
0.035" [0.9 mm] is probably more appropriate.
Steady-state leakage through the thickness of the chest walls has
analyses and treatments beyond the scope of this discussion.