Restoration of a Fischer Ampico grand for Jim Durfee
52) 
53) 
54) 
52) Here, I'm using a micrometer on the pneumatic cloth to
check if it is right. The previous rebuilder had recovered these pneumatics and the
cloth felt too stiff to me. Amphion used cloth that was .012" thick for these
pneumatics. The cloth that was on these was actually .014" thick. This
would have made the spring pneumatic too stiff. This would have effected the speed
and delicacy with which the device could operate.
53) Recovering the expression control pneumatics. It is critical to be sure
that the boards are mated and aligned just the way they were at the factory. If you
get them out of order or twisted, the cloth will not lay smooth and even when you screw
the pneumatics back onto the expression device. There are no pictures for this but
the pneumatics on the other expression device were cracked and brittle. Totally new
pneumatics had to be made to get an air tight seal. If these pneumatics leak or
their gaskets leak (even a little) the expression won't be right and the valves that
operate these pneumatics may not operate at all.
54) Recovering the spring and soft play pneumatics with .012" thick pneumatic
cloth. The hinge ends are not glued down until the glue has a chance to set a bit.
Then the pneumatics are closed before gluing down the hinge ends. The
prevents them from binding. Note that these pneumatics also received new hinges.
55) 
56) 
57) 
55) Replacing the original cork gaskets with leather.
Leather seals easier and with less screw tension.
56) Reassembling the expression. The pneumatics have been screwed onto the
trunk.
57) The expression arm is now screwed onto the trunk and onto the control
pneumatics.
58) 
59) 
60) 
58) The spring and soft pneumatics are ready to be screwed
back on the expression.
59) The control box which contains the actual control valves that all these
expression devices operate has been screwed back onto the trunk. The expression box
received new sealant inside and out. Amphion let these go unsealed in areas.
This meant that a little air seeped through the wood into the system. After taking
it totally apart, I fill these boxes with shellac and then drain out the excess.
This makes the boxes even more air tight than when they left the factory. I just
have to be careful to clean all shellac off the valve seats and other metal components
which don't need shellac on them.
60) The restored expression mechanism. The spring pneumatic is attached to the
same inner control valve that the three expression pneumatics (via the blonde colored
expression control arm) are attached to. They play a game of push and pull against
this valve. The three small pneumatics can be combined in seven different
combinations that effect the position of the valve. The large spring pneumatic tries
to open the valve while the three top pneumatics try to close it. By increasing the
suction level inside the spring pneumatic a crescendo can be effected because it will
become more and more able to overcome the upper pneumatics as the pressure inside the
spring pneumatic increases. This inner suction level for the spring pneumatic is
developed by a separate device known as a crescendo. There is one crescendo for each
half of the player action. These are the next devices to be restored.
61) 
62) 
63) 
61) Another view of the finished expression.
62) Both expressions are now rebuilt.
63) The two crescendos before work began on them.
64) 
65) 
66) 
64) A crescendo after much of it has been taken apart.
The valve box on the left was, again, gasketed with cork. This will be
replaced with leather for a better and more sustainable seal. It is vital that these
devices don't leak in any way. The pneumatics themselves have to be the most air
tight device in all of player piano restoration work. Any leak whatsoever will make
the speed of the crescendo/decrescendo incorrect and the music will not play properly.
Each stage of the restoration has to be tested carefully. After the pneumatic
is recovered, the air passages are covered and the pneumatic checked to be sure it is
positively air tight. Then the same test is performed after the restored valve
assembly is installed. In the case of these devices, the valves are covered on both
sides with leather. The outside metal seats that these valves seal against are also
covered in pouch leather. One of the basic laws of pneumatic systems is that leather
does not seal well against leather. To correct this flaw in design, I make the inner
leather on these valves slightly thicker and keep the outside face of the valve free of
leather. The wood of the valve is sanded on a piece of glass until totally flat.
Then the seat is releathered with new pouch leather. This makes for a much
better seal as the wood of the valve seals very well against the pouch leather on the
seat. The seat has to be the part that carries the leather rather than the valve
because the leather on the seat does double duty as a gasket.
65) The restored crescendo.
66) You'll have to forgive me; I forgot the camera the next day at the workshop.
The next item to be restored is the primary valve chest. The best place to
see what it looked like before work began is in picture #2. Click
here to see that picture. It is the blonde colored device located at the bottom
of the photo. You will note black stripes running across the length of the valve box
in a couple of places. These were rubber cloth that was smeared with something
similar to silicone caulking. Since this device is always gasketed together and well
screwed together, this extra layer of sealant was hardly necessary. The last person
to work on this chest had only done two things. They had replaced one pouch inside
the box and had locked the chest together with this improper means of sealing. It
was virtually impossible to get this material off the box just to get it open. In
fact, it ruined one of my sanding belts for the large sanding station trying to get this
material off the wood. Mostly, it had to be scraped off by hand. In photo #66
you see that all the boards look very nice and new. Every outside surface (except
for those which the valves seated on) had been sanded clean and true and then given two
fresh coats of shellac. Then the boards had been left to dry over night. This
shows the repaired boards before I started work on the chest the next day. This
complete sanding and refinishing would not have been needed had the previous rebuilder
known what he/she was doing. This is a good example of why it is critical to only
allow competent rebuilders to do this sort of work. When amateurs and lazy
rebuilders attempt rebuilding these mechanisms, the results are often such that it takes
additional effort and, often, duplicate parts construction to repair/replace the damage
done by incompetent work.
67) 
68) 
69) 
67) The boards in the previous photo have been turned over.
You can see that I also resealed the pouch wells after having cleaned off all the
old pouches from the left most board. The center board which contains the valves
hasn't been touched yet. The right board shows some signs of shellac drippage.
This will have to be cleaned off and lightly sanded so that the new gasket will
stick well and seal well.
68) Gluing the new pouches into place. In the past, I've put these primary
pouches in without the use of tools; strictly by hand. However, I decided to make a
slight modification to my pouch setter so that I could put these pouches in with the tool.
To keep the tool working well for each pouch, it must by wiped clear of glue and
dusted with talc between each pouch. I also dust my fingers with talc as I go along
to keep them from becoming tacky with glue residue.
69) Forgive the overexposure on this photo. This is showing all the pouches
have been replaced, triple sealed and talced. The stack divider has also been
reinstalled. This is a metal arm which weaves its way between the pouches/valves.
It is held in place with a cork gasket in the usual installation. Because I
will be redoing this chest with leather gaskets, I used leather on the divider as well so
that everything will mate up correctly. The divider is glued to the pouch board and
held to the valve board by screw pressure. This enables us to have a two sided chest
that is totally air tight on both sides while still being very easy to take apart for
future service.
70) 
71) 
72) 
70) The primary valves have been pulled off the chest but
the leather hasn't been touched yet. They are temporarily stuck back together to
keep the halves of the valves paired up properly. Technically, you should be able to
mix and match these parts but I prefer to keep them together to prevent the chance of
loosing any of them. You also see one side of the valve board which shows that the
wood is a bit soiled.
71) The same as #70 from another angle.
72) The valve board is turned over in this shot showing the condition of the wood
and another angle showing the unrestored valves. Before putting the valves back into
this valve board, I will lap the wood on both sides with sandpaper backed with glass.
This will help the valves to seat better and will remove the last traces of damage
done to the outer surface by that mess of cloth and silicone stuck to the outside by the
previous rebuilder.
73) 
74) 
75) 
73) This shows one side of the valve board after it had been
carefully sanded to bring the surface to flat, clean perfection.
74) The valves are now getting new leather inside and out and are being reinstalled
into the valve board. The valves are traveled with the special gauge for the job as
they are installed into the board. When they are all in, a dot of glue on the tops
of the valve stems will lock them into the correct valve travel. I really hated to
see the inside valve facings go on this unit. There was really nothing wrong with
MOST of the inside valve leather. The only aspect of them that I objected to was
that Amphion had used a hide that varied a great deal in thickness for this application.
Some leathers were quite thin and others quite thick. It really doesn't
matter here so long as the nap is fine and the leather supple enough yet not too supple.
Most of the original leather on this was .035" thick and of a very fine nap
and stiff body. I have leather of a similar body but it is .045" to .050"
thick and the nap is more course. The leather I had with the same nap as the
original was .025" to .030" thick. The original leathers ranged from
.025" to .050" thick with an average thickness of .035". I had
to either go with leather which was thinner than the average original leather but of
similar nap or leather of courser nap and thicker body. I like using the thicker
body leather because it can last longer before it rots. However, I chose to use the
thinner leather because I felt it would be more likely to seal well in this particular
application. These valves are rather delicate and require an excellent seal to
operate well. I could easily have used the originals leather but I didn't want to
take a chance that the excellent condition it showed today would stay excellent for the
next twenty to thirty years. Besides, the inside leathers aren't even glued down so
replacing them is the easiest job imaginable. It really makes no sense to keep them
even though they were still good. You can see some of the new valve leather sitting
in a pile on the bench in this photo.
75) The primary chest now has all new leather on the valves.