Restoration of a Mason & Hamlin Ampico for David Lucas Graves & John Herrmann
130) 
131) 
132) 
130) These are the equalizers for the bass and treble.
Their function is to provide a slight cushion to the air stream inside the system.
When the roll suddenly calls for a large number of notes at a time when the system
is playing at low suction levels, these pneumatics will help provide that small amount of
extra air needed to get every note to play. Unlike the equalizers found on manually
pumped player pianos, these are very small. If they were made much larger or given
more gap in their motion, they would create too large of a cushion and keep the system
from being able to produce those crisp accents it is so good at making.
131) I thought I'd provide a few pictures of the rebuilding of the expression which
show some of the detailed work needed to make these operate properly. This photo
shows the control valves lined up next to each other. Three of the valves are meant
to control the suction level. By combining these valves in different combinations,
seven different suction levels can be produced. So that the rolls will not have to
be weakened by the inclusion of long rows of holes that signal these valves, the valves
are made as lock and cancels. One small hole on the roll will turn one of these
valves on and lock it there. Later, another valve of the three can be added to it or
even a third. To turn them off requires only that the cancel valve (far left) be
activated. When this happens, all the valves are turned off at once. The lock
and cancel requires the addition of an extra valve in each of the three expression control
valve blocks. When these were built originally, all of the valve seats for the
cancel valves were made of thick felt. I have found that this felt seat lets enough
air seep through it to cause a variation in the suction levels expected from the system.
I have found that thick buckskin has the soft characteristics of the felt and the
same thickness while being totally air tight. In this photo, you can see the
original white felt cancel seats. On the valve blocks themselves are yellow colored
buckskin seats which have been made to replace the original felt. Most rebuilders
reuse the original felt seats. I believe the thought is that, if the factory thought
it was good enough, who am I to say different. However, extensive testing and
evaluation of this type of expression has proven to me that these felt seats are a
problem. You can also see the maroon leather used to replace the leather on the
outside valves that also go into these blocks. Finally, the original cork gaskets
have been replaced with a more expensive leather gasket on each block. These gaskets
will seal much better than the cork and with less screw pressure needed to accomplish
this.
132) This is the trunk board which contains all the air channels for all the valve
and control pneumatics. This photo shows the new pouches that have been installed.
It has also been repainted and all the inner air channels have been freshly sealed.
133) 
134) 
135) 
133) This is the same trunk board from the previous photo.
This is taken from underneath. There are pouches underneath that are actuated
by the main cancel valve and operate the individual cancel valves in each block.
Almost universally, these pouches are still in perfect condition. In addition, they
need to be made of the very thinnest possible pouch leather in order to be sufficiently
sensitive at low suction levels. Because these pouches are buried deep inside the
system and get very little air exposure, they are almost universally still in perfect
condition. This unit was no different. I gave the original pouches a fresh
coat of sealant and reused them. The paper circles are replacements for the old
newspaper circles used originally. You can see the old newspaper circles at the
bottom of the photo. I always enjoy checking out the print on these circles as they
are never much yellowed and bring me back to feeling like I am reading a new paper back in
the 1920's. The purpose of this paper is to protect the pouches from any drips of
hide glue when the seal cloth is glued over them and around the corner onto the edge that
is at the upper edge of this photo.
134) These are two control pneumatics that go on the expression. These were
the ones that were covered with bellows cloth rather than motor cloth, making them very
stiff. They were also very leaky even though they had not been recovered all that
long ago. It is clear that they were never very air tight even from the moment they
were recovered. However, these must be two of the most air tight pneumatics in the
system. The one on the left is called the "re-regulator." It's
purpose is to actuate a valve inside the control box that reduces the overall volume of
the piano without stopping the expression of the music from happening. The pneumatic
on the right is called the "spring pneumatic." Its purpose is to fight a
tug of war with the three small control pneumatics that are mounted on the top of the
device and ganged together on the large wooden arm. This pneumatic is connected to
the crescendo device for that half of the system. The purpose of the crescendo is to
slowly or quickly increase or decrease the suction pressure to the spring pneumatic.
The more suction the spring pneumatic is given the harder it fights the control
pneumatics that are pushing against it. By keeping the small control pneumatics set
the same and increasing pressure to the spring pneumatic a crescendo is produced because
it is gradually pushing the assembly upward, thus opening the suction control valve
inside. If the crescendo device or this spring pneumatic leak in any way, the
expression of the system will be severely compromised.
135) This is the freshly releathered valve assembly that is attached to the
re-regulator pneumatic. When this pneumatic is activated, it pushes the maroon valve
onto its seat. This forces the air flow around that hole it passes through and
forces it through a different, smaller opening. All the expression will still occur
but because the flow is through a smaller opening, the overall level of volume is
decreased.
136) 
137) 
138) 
136) Below you see the manual expression device. When
playing standard 88 note rolls, you can express the music yourself by using a couple
buttons in the drawer. These buttons feed air into this device and thus allow air to
seep into the control pneumatics, thus effecting the volume. This particular design
of manual expression was never very good. It is very difficult to control the volume
with any amount of subtlety with this design but this is how they made them so we're sort
of stuck with it. The Duo-Art design is superior to Ampico in this aspect.
Above, you can see that the valve that activates the re-regulator has had most of the
restoration done and is ready to be glued back together. To the upper right are the
cleaned up control pneumatics which are ready to be recovered.
137) The control pneumatics have been recovered and new leather gaskets installed.
These had been badly damaged by the previous rebuilder so the job of making them
work was harder. However, once finished, they were as good as new. The old bad
work had gone so far as to find that they never cleaned the old pneumatic cloth and glue
off the edges of these boards before gluing the too thick motor cloth onto them.
138) The expression is now totally reassembled.
139) 
140) 
141) 
139) Another view of the restored expression. It only
remains to bench test it to find any leaks or malfunctions before putting it aside to move
onto the next expression. Since they are essentially the same, I will not be doing a
large photo layout of the second expression.
140) I hadn't gotten around to placing the new labels on the valves until I had
finished rebuilding the second expression. You can see in this photo that new labels
have been glued on to indicate how to tube up the mechanism later.
141) While working on the second unit, I thought it would be informative to show
what happens to these expression pneumatics when you forget to keep them oriented exactly
as they were from the factory. Not only did this second expression have the
pneumatics out of order as the first expression did, but the installation of the screws
had been done without using any pilot holes. One of the pneumatics had broken in two
when they screwed it down yet they never went back and repaired that. I had to make
a new pneumatic to replace the one that was so badly damaged. The others were
salvageable. The big problem with this is that these pneumatics and the air channels
that lead to them as well as the gaskets that they seat onto must all be absolutely air
tight. If they are not, the expression will not operate properly. In order for
the music to play in such a way that each note is exactly the volume that the maker wanted
it to be, everything must be so tight that it can respond in a millisecond to any signal
from the roll. Anything slower than that due to stiff pneumatic cloth, stiff pouches
or leaky parts will effect everything. Let me explain a little bit about how the
roll makes the music expressive. It is possible for these systems to literally pick
a melody out of the texture of the chords being played. To do this, the system is
set to a low pressure and all the background notes are sent off toward the strings.
In the next millisecond, the suction pressure is increased and the melody note is
activated. Since the melody note is struck with greater pressure, it will fly up to
the strings faster than the rest of the notes. With perfect timing, all the notes
strike the strings at the same instant but the melody note is louder. So you see
that pin point accuracy and precision rebuilding is critical to get the mechanism to
operate in this way. With the way it was rebuilt before, the system would only be
capable of general settings of loud and soft varying slowly from time to time. The
subtle work of picking out melodies and making very intricate expression patterns would
have been completely impossible. What's worse, the timing of the notes being played
and the expression setting volume would be off so the wrong notes would get pressures that
were intended for notes to follow or notes just previous.
142) 
143) 
144) 
142) This photo was taken to show a problem we ran into with
the second expression. The valve wire that holds the most important valve in the
piano had been badly bent long ago. When I attempted to straighten it very carefully
so I could get it apart, it broke. This wire is tricky to make because it has to fit
snugly into the metal wire guide shown on the black box at the top center of this photo.
Yet the wire must be loose enough in the guide to move freely. Also, the
threads must be a certain length but no longer than that. Since I did have a couple
of Ampico A expressions that I keep for parts, I was able to steal the valve wire from the
other expression.
143) Both expressions are finished and it is time to rebuild the crescendos.
If you look at the picture #139 you can sort of see how the expression works. The
blonde arm is connected to three square pneumatics. These pneumatics are normally
supplied with suction at all times. On the other end of the wire which holds the
large expression valve inside the box, is the wine colored pneumatic. This is the
"spring pneumatic." Its job is to push the wire the other way. The
blonde arm pushes the wire down and the spring pneumatic pushes the wire up. By
altering the strength of the pneumatics top and bottom, you can control with great
precision the exact location of the expression valve. The job of the crescendo is to
dictate the suction level to be supplied to the spring pneumatic. The crescendo can
be set to slowly increase or decrease the suction level to the spring pneumatic or it can
do so quickly. Even if you make no changes to the upper square pneumatics, you can
change the location of the valve and how much air it permits to pass through the
expression just by altering the suction levels in the spring pneumatic. It is one
of the most critical set ups in the Ampico A design that both the spring pneumatic and the
crescendo be absolutely air tight and also be exceedingly flexible in their
operation. The last rebuilder did not even try to get the crescendo air tight and
they used pneumatic cloth which was far too thick as well. These crescendos were
sluggish indeed in their operation.
144) Here is a photo of a the first crescendo I rebuilt after it was taken apart.
145) 
146) 
147) 
145) This photo shows part of the reason why these
crescendos were leaking so badly. The last rebuilder did not even try to clean off
the old pneumatic cloth left behind when they tore off the original cloth. That is a
mistake that can only be attributed to laziness or ignorance.
146) This photo shows all the parts after they were restored but before they were
reassembled. I take great care to thoroughly reseal every component with thickened
shellac. Before putting this back together I satisfied myself that the crescendo
pneumatic was absolutely air tight.
147) A finished crescendo ready for final testing.
148) 
149) 
150) 
148) A full on shot of the top of the drawer before
restoration began. All that has been removed are the mahogany cover boards so that
they could be refinished with the rest of the piano.
149) This is the right side of the drawer. It contains the transmission which
transfers the energy of the motor to the roll, the play/rewind pneumatic and control
valves, the tracking mechanism, the wind motor governor, and the function valve assembly.
150) This is the left side of the drawer. On this side, you find all the
control valves, the wind motor, the connecting shaft that carries the motor's energy to
the right side of the drawer where the transmission is, the ladder chain guide spool, the
manual expression valves, a cut out block for a single signal line, the automatic shut off
pneumatic and valve and the on/off switch. This side of the drawer was in sorry
shape. A rat's nest was built under the wind motor and had been crammed into every
corner. Rat urine and feces had corroded all the metal on the wind motor and all the
original bushings (many of which can normally be expected to still be in usable condition)
were all shot. The crank shaft was so badly rusted, I feared that I would not be
able to make it work again at first. The gold colored switches control the automatic
rewind and shut off function; the expression system on/off selector; the damper pedal
on/off selector/ and a switch to adjust from normal play to subdued or brilliant as
desired. The valve assemblies beneath these switches are often bad. They are
made of pot metal which has often cracked up so badly by now that it cannot be reused.
Happily, this set of pot metal switches were still in good working order. I
did have to take a couple hours to resurface them and re-affix the nipples in them.
However, when I was finished, they were looking and working just like new. Since
replacement parts like these pot metal valves are not part of the standard fee for
rebuilding, it is good news we could reuse them. This way there will be no effect on
the final price.
151) 
152) 
153) 
151) This is another pot metal component which is usually reusable. After
getting this transmission out of the drawer so I could examine it closely and test its
functions, I found that one of the shafts had a nasty stiff spot at each turn at one give
point. I knew this was probably due to warpage in the pot metal frame.
However, I attempted to gently bend the shaft and frame to correct the problem, hoping not
to damage the frame. However, the frame was too worn out and it broke. The
problem with the lack of free travel in the shafts was caused by this so a new frame was
ordered. All the brass components will be polished and transferred to the new frame
when it comes in.
152) This spool serve the purpose of guiding the ladder chain that runs from the
wind motor to the crossover shaft. The mice that had been living in that drawer
(who, by the way, left a great deal of their own fur behind as well!) had chewed on this
spool, leaving little of the edges that help to guide the chain. I cut the damaged
edges away and removed the bushing cloth inside and out. Then I made new shoulders
and drilled them out to match the core.
153) Once the glue had dried, I pained the appropriate places black as before and
bushed the inside shaft hole and the outside track where the ladder chain rides.
154) 
155) 
156) 
154) This shows the drawer with all components out except for the transmission and
spool end hardware.
155) This is the underside of the drawer after the cover boards were removed.
This revealed the original lead tubing leads. These leads are often still in good
shape and can be reused. In fact, it is very desirable to reuse them if you
can. Putting in new lead lines is very costly and using neoprene lines is a
challenge since they take up a bit more space than the lead does. A past restorer
attempted to reuse these failing lead lines by splicing rubber tubing into areas which
were corroded clear through. My take on this is that you cannot reuse the original
lead tubing unless it is look nearly perfect still. This drawer will have to have
all new lead in it. This is unfortunate since the standard price for rebuilding
assumes all wood and metal components are in restorable/reusable condition. It will
cost a little more because we have to replace these lines. However, there's nothing
to do about it except make it right. These lines will never be trustworthy again.
156) After cutting away all the lead down as far as the nipples, the hard part comes
into play. Both at the tracker bar and at the elbows at the back of the drawer, the
lead is cemented into place with a very hard substance we have never really been able to
put a chemical analysis to. Whatever this stuff is, it is very challenging to get it
off the nipples without damaging anything. It took several hours for me to remove
the last of the lead and cement from around the nipples so as to prevent damaging
anything.
157) 
158) 
159) 
157) Here you see the tracker bar after all the old lead and cement have been
cleaned away.
158) Now the tracker bar has been polished and lightly coated with Dow 111 which
will help protect it from future tarnishing without leaving the telltale marks after use
that lacquer sealant can leave. At this same time, the tracker ears were removed,
polished inside and out and new leather installed.
159) Here are the various cover boards, controls, valves and other hardware to the
drawer. All of it has been repainted, polished, repaired and so forth. It will
all go back into the drawer once the new transmission frame comes.
160) 
161) 
162) 
160) These are all the various pneumatic components of the drawer which must be
rebuilt.
161) Another shot of the emptied drawer before the transmission was removed.
162) This is the motor governor. It maintains a smooth operation of the wind
motor and is also the means by which we set the tempo.
163) 
164) 
165) 
163) This is the tempo plate inside the motor governor. Blue chalk outlines
the orifice through which the air is metered. The shape of this orifice is critical
to getting the tempos correct for the rolls. Also, this plate must be in perfect
condition and totally flat so that the slide valve that runs across it can seal against it
perfectly. Without this perfect seal, the wind motor would tend to continue to creep
along slowly even when you turn the tempo down to zero. In the case of the celluloid
tempo plate in this governor, the plate had failed over time. It was crumbling.
It is possible to buy new brass tempo plates. However, since so many other
components which are usually reusable have been found to be useless on this job, I took
out some extra time to save the customer some money. I made a new tempo plate from
plastic key top material. It is totally flat and I have carefully cut a duplicate
orifice for the control of the tempo. I cleaned out the old tempo plate (which came
out in small crumbled bits) and bedded the new plate down into a pool of PVC-E glue.
This glue sticks well to this type of plastic and will serve to keep it air tight
as well.
164) This photo shows the large number of separate components that the wind motor
governor breaks down into. In this photo, some of those components have received
restoration and others are still waiting their turn.
165) The completed governor, tested to be perfect and ready for reinstallation.
166) 
167) 
168) 
166) From left to right in this photo you see the roll tracking mechanism which has
the play/rewind valve assembly screwed to it; next is the player rewind pneumatics, and to
the right is the automatic shut off pneumatic and valve.
167) Here are those three components after being broken down into their various
parts. In this photo, some restoration work had been done but much remained to
complete.
168) This is a control valve assembly that operates various function for play and
rewind.
169) 
170) 
171) 
169) Here, that valve has been broken down to its individual parts.
170) Now all the parts have been restored and are ready for reassembly.
171) The finished valve assembly.
172) 
173) 
174) 
172) The motor governor, play/rewind, tracker, signal cut-out block and auto shut
off have all been restored and are ready to be reinstalled. I was impressed as I
worked on these with how the previous rebuilder had clearly neglected to remove all the
old pneumatic cloth before attempting to glue his new cloth into place. It's no
wonder these things were leaking. They probably leaked badly from the moment he
finished with them. I never let a component pass onto reinstallation until I'm
satisfied it works perfectly and is absolutely air tight. To prove this point, I
rebuilt the play/rewind pneumatic three times before I was happy with the result.
173) This is the very sad wind motor. It had been poorly rebuilt in the past
so that it probably never did work very well. In addition, the rats that had lived
in and around it had ruined every bushing and caused every metal component to begin
rusting. When I started in on this, I feared it may not be restorable.
However, as I went along, I realized I could save it. However, rebuilding this
component usually only takes half a day. In the case of this unit, it took two days
to rebuild it in order to correct everything that was less than perfect. I even had
to correct a mistake made by Amphion. They did not seal the insides of the
pneumatics like they had the outsides. This had eventually caused the pneumatics to
cup because moisture could travel out of the wood and into the wood through the naked
interiors easier than the finished surfaces on the outside.
174) A photo of the back of the motor before I began. Some of the rat's nest
is still clinging to parts of the motor on this side.
175) 
176) 
177) 
175) This photo shows the wind motor broken down into its
various components. I'm always amazed at the complexity of these devices.
176) Some of the work has been completed. The pneumatics have been resealed
inside and out. They have been sanded to accept the fresh glue joints. They
have been rehinged with felt dots inside to prevent putting a hard crease into the
pneumatic cloth during the gluing up process. Because of the damage this motor had
been subjected to, I had to split the pneumatics off of their trunk blocks. Usually,
they are recovered while the blocks remain in place. However, I felt that the glue
joints holding them together had been badly weakened and were really only staying together
because of the nails used to reinforce the joints. I pulled the nails and split the
pneumatics off (with considerable ease, I might add.) All the metal has been
polished. You will note that the crank shaft at the bottom of the picture polished
up very well after a great deal of elbow grease went into it. All the slide valves
have been lapped for a perfectly flat surface, they have been sealed to ensure an air
tight valve and have received new seal cloth to ensure longevity. Usually, I have to
rebush the parts of the motor which spin on the crank shaft itself. That is
considered normal. However, all the bushings that were associated with the light
connecting arms that ran down to the slide valves had also all gone bad. This
quadrupled the number of bushings that had to be remade.
177) While waiting for shellac and glue to dry, I turned my attention to the
transmission. This is a part that I seldom mess with because they are usually fine.
Even though the frames are made of pot metal, they are usually intact. This
transmission, however, showed signs of problems because the rewind shaft had a stiff spot
in its motion. When I pulled the transmission out to search out the problem, I found
that the cause was a warped frame. When I started to work with it, the frame simply
broke as well. It is just as well that it did break since the warpage would have
prevented the roll from moving smoothly in any event. In this photo, you see the
transmission along with a new cast aluminum frame which I purchased for replacement.