Constriction Insert on Ampico B Regulator Needle Valve

The needle-valve 'assembly' associated with the two Ampico B expression
regulators is an unusual design.  The implementation allows one to
adjust air flow within each expression regulator to realize a particular
condition.  That condition is the setting of the quiescent stack vacuum
level.  It must be just right (sufficiently low, say about 5" H2O) to
reliably obtain the lowest intensity of softly speaking notes.  Dr.
Clarence N. Hickman, the designer of the Ampico B system (1924 - 1929),
devoted substantial effort to enhance the performance characteristics of
the previously existing Ampico A reproducing piano, originally developed
(1913 - 1926) by C. F. Stoddard.

The needle valve itself is physically sandwiched between two pneumatic
constrictions.  One constriction is a circular orifice about 1/8"
diameter (about 1/2" long) drilled into metal and is built into one
arm of the needle valve assembly.  The shank of the threaded needle
intersects this airway a bit past the midpoint.  The other orifice
constriction (when present) is commonly formed by an annular phenolic
insert whose central circular opening is also about 1/8" in diameter,
but is about 3/16" deep (thick).

When the phenolic insert is not present, the resulting de facto
constriction (if one can call it that) is the interior cylindrical
volume of a perpendicularly mounted section of metallic tubing (O.D.
of 3/8", 3/4" long), which is the other arm of the valve assembly.
In addition to being a rubber tubing connection, the metallic tubing
serves as a container within which the phenolic insert can be placed,
if and when it is needed.  Because the 1/8" diameter 'tunnel' through
one arm is about 3 times longer than that through the other arm (when
the phenolic insert is present), the effective resistance to air flow
through one arm of the assembly is about three times the resistance of
the other arm.  The need for the phenolic insert constriction is
discussed subsequently.

For each individual Ampico B piano, the degree of airtightness of the
bass and treble sections of the stack will vary, depending largely
on the pneumatic variability (porosity) of the 83 inner valve poppet
leathers, one for each note valve.  The air flow characteristics of
the two stack spill valves (bass and treble), which are fully open
at low vacuum levels (less than #6 intensity setting), may also differ.
The relative airtightness and air flow characteristics of the two
expression regulators (each containing one pouch diaphragm that 'rolls'
conformally onto a perforated celluloid grid) and the 'end-of-the-line'
connection to the 'first intensity adjuster' also enter the pneumatic
variability picture.

In the presence of these variable air flow and leakage factors, there is
a need to establish a minimum vacuum setting appropriate for reliable
soft playing in that piano installation.  This is accomplished by
carefully adjusting the bass and treble needle valves, but certain
conditions must be satisfied first.

The needle valve is made in such a way that it cannot close completely.
That is, some air can still flow around the needle structure, even
though rotation of the needle is blocked at the end of its forward
travel.  At the end of its axial travel path, the rounded spear-like
tip of the tapered needle runs into a metal wall.  This prevents any
additional forward-moving turning of the threaded rod which constitutes
the bulk of the body of the needle itself.  This implementation is not
accidental; it is done intentionally.

When the tapered needle is screwed all the way in (as far as it can go),
air can still flow through the associated internal passageway of the
valve.  It is easy to verify this assertion with a simple mouth test,
where blowing through the assembly (after removing it from the regulator
structure) is noticeably impeded as the needle is turned into the
assembly.  Curiously, this effect is not as pronounced as one might
imagine for a conventional needle valve configuration, where one would
expect substantial resistance to air flow.  Clearly, at the end of the
needle's full forward travel, air can still pass through the residual
opening.  Why was the needle valve assembly constructed in this manner?

If the needle valve were fabricated to block the air flow completely,
at least two undesirable situations could arise.  One, the needle
could be turned in 'too far', resulting in a pneumatically sealed and
potentially mechanically jammed configuration.  Even if the needle is
not turned in too far for jamming to occur, the residual opening could
still be small enough to collect dust and/or circulatory debris so
that a blockage could ultimately occur.  So, how does one design a
(needle) valve to prevent the occurrence of either or both of these
two undesirable situations and still be able to set up appropriately
reliable pneumatic conditions for minimum intensity piano playing?

One can do this by 'distributing' two separate and independently
fixed constrictions in series (in a row) with the needle valve
(one constriction on either side of the needle).  Then, the end-to-end
air flow through the 'composite' valve can be impeded sufficiently to
produce an approximation to the pneumatic effect that would occur if
just a single conventional, fully closable, needle-valve of standard
design were used.  However, with this alternative construction, the
constrictive choking effect could still occur, but only over a more
limited dynamic range.  Of course, with the multiple (two additional)
constrictions in the Hickman design, jamming of the needle valve would
be precluded and debris-induced blockages would certainly be much less
likely to occur.

So what does the orifice associated with the phenolic insert (when
present) do?  Remember that the Ampico B vacuum regulator functions
with air flowing through it (across both sides of the regulator pouch
diaphragm) at all times.  The phenolic insert (with circular orifice)
adds just the needed amount of pneumatic impedance to the series air
flow circuit associated with the vacuum regulator.  This pneumatic
design modification allows the vacuum level in that section of the stack
to hover within a small but useful range near the desired quiescent
value of about 5" water gage.  Testing for this performance condition
is conducted by measuring stack vacuum with the needle screwed in all
the way.

Recall that the function of the needle valve is to make it more
difficult for air to flow through the passageway within which it is
contained.  In normal operation, when the needle is screwed in (i.e.,
more constricted opening), the quiescent vacuum level in the stack
decreases.  For confirmation of this statement, see the bottom of
page 8 in the 1929 Ampico B Service Manual, where it states: "If these
adjustments were reversed with E tightly closed and D open, pump suction
in chamber G would cause the pouch A to seal the holes in the grid and
no suction would be developed in the wind chest F."

 [ https://www.mmdigest.com/Gallery/Tech/Ampico/amp08.htm 

Clearly, one has to be assured that the 5" H2O vacuum level can be
reached within the limits of the mechanical adjustment range of the
needle.  Otherwise, sufficiently soft playing of the piano notes could
not be obtained.  Then why is it that some Ampico B systems do not
incorporate any phenolic inserts (within the needle valve assembly)
at all?

This is explained by the variable degree of residual pneumatic leakage
present in the Ampico B system as well as the flow characteristics of
the air within that system.  Accounting for note valve inner leathers,
locations and integrity of stack spill valves, airtightness of pressure
regulators, etc., some stacks and associated regulation devices are
pneumatically very tight, some less so.

The absence of a phenolic insert within the needle valve assembly
means that the 5" H2O vacuum goal was achievable for that system as
manufactured.  When a phenolic insert is present, it indicates that
the 5" H2O vacuum level could not be reached with the needle screwed
in all the way.  The essence of this proposition is contained in a
'What-does-that-mean?' statement from the 1929 Ampico Service Manual,
page 9, 4th line down, repeated here for contextual clarity: "The
opening E is adjusted at the factory to fit the scale of intensities to
the particular piano into which the Ampico is installed.  The opening E
can also be adjusted to even up the No. 1 intensity pressures, Bass and
Treble."

To summarize, if adjustment to full closure (not jammed) of the
needle valve in the vacuum regulator of the stack does not allow the
target vacuum level of about 5" of H2O to be reached, then a phenolic
constriction is inserted in the 3/8" O.D. tube.  This will increase the
series pneumatic impedance slightly, but sufficiently, in that circuit
so that the target vacuum level can be reached with appropriate setting
of the valve needle.

The diameter of the circular orifice in the center of the phenolic
insert may also need to be 'adjusted' to realize the target vacuum
level and its associated vacuum adjustment range.  This 'trimming'
operation may have been conducted by an experienced operator at the
end of the production line during final inspection of the piano.
This 'inspector' may have had access to precision sets of multiple
phenolic inserts, some with larger orifices, and some with smaller.
His job would have been to select one (or two) or none of them to
insert into the 3/8" O.D. tubing of the B expression regulator to
realize the quiescent vacuum level target of 5" H2O for each piano.

This could explain why different size orifices are found in original
needle valve assemblies (in the metal body as well as the phenolic
insert) from different pianos.  It would also explain the less
common occurrence of a phenolic insert being found in the bass section
regulator, but not found in the treble section regulator, or vice versa.
Typically, one finds the presence (or absence) of phenolic constrictions
in both the bass and treble expression regulators of a given piano.

There is an analogue of this design approach in electrical engineering.
One can put two extra fixed resistors (not variable, not necessarily
identical) in series with a potentiometer (three-terminal variable
resistor), one on either side of the device, to limit the amount of
current flowing through the entire series circuit.  Prolonged and
excessive current flow could easily damage (burn out) some or all of
the resistive element of the potentiometer.

This more complex configuration (it requires the addition of two
more components), limits (reduces) the resistance adjustment range
capability of this circuit.  However, the potentiometer itself is now
'protected' from excessive current flow, and finer control (reduced
range) of the circuit-limited current can be realized.

In pneumatic terms, this implementation reflects a practical,
cost-effective tradeoff of incorporating two additional components
(simple constrictions) in exchange for limited, but finer control of
the quiescent vacuum level target.  This clever philosophy appears to be
what C. N. Hickman utilized when he designed the needle valve assemblies
on the bass and treble expression regulators for the Model B Ampico.

Bill Koenigsberg
Concord, Massachusetts

 [ Excerpts from the Ampico "Model B" 1929 Service Manual are at
 [   https://www.mmdigest.com/Gallery/Tech/Ampico/index.html 
 [ An enlarged view of the 1st Intensity Adjuster Block
 [ and needle valve E is in the illustration at
 [   https://www.mmdigest.com/Gallery/Tech/Ampico/amp28a.gif 
 [ -- Robbie