Restoration of a Fischer Ampico grand for Jim Durfee
109) 
110) 
111) 
109) Rebushing the connecting rods. This is a
wedge-type pump. These can be very tricky. The spring loaded wedges are
intended to help keep the pump from developing knocking noises when the bushings start to
wear out. In fact, they don't really help at all. However, they can cause
additional knocks if they aren't restored and installed perfectly. First, each part
is removed and sanded smooth on a piece of glass to make them very smooth and very flat.
Then they get a thorough coating of powdered graphite mixed with alcohol.
Finally, the bushings have to be replaced. It is very important that the bushings be
so tight that you can barely get the arms back on to the spider at first. If they
are any less tight than that, the pump will develop a knock very quickly. Finally,
it is vital that all the parts be well marked so that you don't mix them up. Even
though they all look the same, they are definitely NOT interchangeable!!
110) This is the spill plate. It contains the spill valve and the amplifier
pneumatic. When this pneumatic is collapsed, it pulls more strongly on the spill
valve causing the internal pressure of the system to rise.
111) The underside of the spill plate, showing the new leather gasket and also
showing the fact that all the exposed wood got a fresh coat of shellac sealant. The
square item to the left is the spill valve itself. The original guide pins for the
valve were broken so I had to revamp the valve to work with an alternate type of guide.
112) 
113) 
114) 
112) The spill plate assembly and the manifold/pump cut out
assembly are totally restored. For the sake of originality, I reused the original
screens on the cut out (right.) I really feel that this is not a good design and may
(in the future) rework this. The metal screens make it very difficult for the gasket
to seal because they are buried under the gasket, keeping it from being totally flat.
In the future, I may replace these with thinner, more flexible nylon screen
installed in such a way that it fixes this design flaw. For this pump, the old
design was made to work properly but it was more work to get it right. Also note
that the large elbows on the sides of the manifold are new. The original pot metal
elbows were all shot and will be replaced.
113) Putting the pump back together. You can see that new bellows cloth has
been used to cover the airways in each bellows. This cloth gets a wood cover plate
as well. I don't know why Amphion didn't just use a gasketed cover plate so that the
inside flap valves could be partially accessed for cleaning but they didn't. This is
another design "flaw" that I'm tempted to start altering in the future.
114) The bearings have all been cleaned and repacked with grease. The metal
cover plate has also been repainted.
115) 
116) 
117) 
115) Installing the cover plate which contains the bearings
for the drive wheel.
116) Putting the spider back together.
117) Testing the pump to make sure it is drawing the correct amount of suction.
During this phase, the pump is allowed to run in for a while as well to make sure
that my bushings are, indeed, tight enough that they won't start to knock after a period
of time. These bushings are also very stiff at first so I use my own electric motor
to run them in so as not to over stress the motor that goes with the system.
118) 
119) 
120) 
118) The finished pump with the belly cloth stand off wires
also reinstalled.
119) A picture of the old flap valves showing just how many of them there are in
this pump. It's rather ridiculous, actually. The pump would work just as well
with half this many flaps. The primary purpose of the extra leather is to provide
some controlled leakage that helps keep the pump sounding smooth as it runs. The
leather to leather connections between the flap valves leaks on purpose.
120) Time to restore the pneumatic stack. This is actually one of the most
tedious designs of pneumatic stack there is to rebuild. There are others that are
worse but they are less common. This is probably the most difficult stack to rebuild
of all the designs that we find commonly.
121) 
122) 
123) 
121) Another view from the opposite side, before work began.
122) The three tiers have been taken apart. This shows one of the separate
tiers before it was individually disassembled.
123) The other two tiers before they were taken apart.
124) 
125) 
126) 
124) Removing the pneumatics from the trunks. After
the valves were removed, the trunk is marked to show the locations of the pneumatics.
Then the pneumatics are cut open and an iron placed over the fixed board where it
is glued down. As soon as the glue warms a bit, it is soft enough to pry the
pneumatic off with minimal damage to the trunk or pneumatic. Because these have been
done before, there was damage from the previous rebuild that came back to haunt me here as
well as some new damage which often occurs during this process. Taking the
pneumatics off the trunk is always a bit of a violent operation.
125) After the pneumatics were removed, each trunk was completely taped off and then
the inner voids filled completely to the rim with shellac. Then the shellac was
allowed to drain out. This ensures that the trunk is totally air tight. I do
this now so that I can sand off any drips of shellac on the outide of the trunk at the
same time that I clean up the glue joints where the pneumatics were. This means less
sanding on the trunk and the less sanding we have to do the better. These wind
trunks are some of the most delicate structures in all the player piano world so you must
be careful to make sure they're air tight since the thin wood isn't enough by itself.
In fact, there are definite signs that these trunks are going to become a problem
for some rebuilder in the future. I suspect that some future restoration will need
to involve the total revamping of these trunks or, possibly, the making of new trunks
entirely. I had to revamp one of the trunks myself and another had to have all the
wood around the valve mounting rods plugged because it was stripped out. More on
that in a moment.
126) When you are servicing an Amphion, you must keep the valves in order because of
how the gaskets mate to the trunk. When rebuilding them, it isn't necessary to do so
because they are getting new gaskets and all the valves are identical in every other way.
This basket holds all the valves awaiting restoration.
127) 
128) 
129) 
127) This isn't strictly necessary. This photo shows
that I am sealing the insides of all the pneumatics with shellac (eventually both fixed
and movable boards got sealed on the inside.) The outsides are sealed at the
factory. However, I knew that the previous restoration had been so sloppy that the
outsides of the boards was going to require significant resurfacing in order to hold
strong and durable glue joints. To ensure that the sanding I did on the outsides
didn't leave the pneumatics leaking, I sealed the insides. In fact, it wasn't
considered critical at the factory that the pneumatics be all that air tight. Since
they can only leak while they are actually being activated and they are not activated for
that much time each, a small amount of seepage was considered acceptable. I prefer
them to be more air tight than that. Note that this sealing work was done before I
removed the old cloth. The cloth helped protect the edges of the pneumatics from the
shellac. The goal is to sand the pneumatics as little as possible. You MUST
have clean and true glue surfaces for the new cloth but you don't want to sand a great
deal of the wood away to accomplish this. Sadly, the previous restoration included
too much sanding and the result was so sloppy that I had to do even more sanding than I
would normally do if I were restoring an original unit that had never before been touched.
This means that these pneumatics have been worn away through sanding to the point
that they are much smaller and thinner than they were when new. It is likely that
the next time they are recovered or, perhaps, the time after that; the rebuilder will have
to make all new pneumatics for the system.
128) Time to rebuild the valves. Here, you see one of the valves has been
split apart to access the old pouch. The valves will get new valve leather, new
pouches, new gaskets and new sealant.
129) The new pouches have been glued into place. Several days of labor were
involved in getting from the previous picture to this one. Each valve must have the
original glue joints split apart with a very sharp knife. This is a most delicate
operation which can do great damage if not carried out perfectly. Then the original
glue joints must be sanded on a piece of glass to get them totally flat for when they are
glued back together. In addition, all traces of old glue squeeze out must be chipped
off the pouch wells; the old blotter paper gaskets on the wooden outside valve seats must
be scraped off the seat and the valve; the old leather must be scraped off the valve and
the valve body must be sanded on a piece of glass to true it up; all the inside valve
seats must be checked to be sure they are well seated and sealed; the outside valve seats
must be polished; the old pouches must be scraped off and the wood sanded clean; old glue
squeeze out must be cleaned off the outside valve seats; old pouch leather must be scraped
off the lifter discs; and all this is done before you can even begin to start putting new
leather into the valves. It is a very time consuming and tedious process.
130) 
131) 
132) 
130) Masking tape circles were placed in the center of each
pouch, then the pouches were sealed with rubber cement until they were totally air tight.
Then the stickiness of the rubber was eliminated with a dusting of talc.
131) Next the masking tape circles are carefully removed with tweezers to prevent
puncturing the new pouches and the lifter discs are glued onto the pouches. Some
rebuilders leave the old mess of glue on the lifter discs and glue them back in upside
down. I prefer to clean the old glue off and glue them back the way they were.
This avoids any chance of having the lifter discs too close to the valve stems
because of the blob of old glue sticking up. Also, this preserves the disc for
future reuse on another rebuilding cycle.
132) Now that the pouches are done, the valve bodies can be glued back together.
This requires a great number of clamps and a skilled hand. Hide glue MUST be
used for this. You have to get the glue onto the valve and the clamps in place
before the glue starts to gel. In addition, great care must be taken to ensure that
the parts don't slide and become misaligned. They will still work even if they are
slightly out of alignment but you may have trouble getting the gaskets to seal to the
trunk if they are too far off. You can sand off any slight misalignment but this
will only shorten the number of additional times the valve can be restored in the future
if you sand off even more wood to correct alignment problems. Any valve that isn't
absolutely perfect in alignment after clamping, I leave alone. I prefer to have the
glue joint very slightly rough than to shorten the life of the valve. I make up for
any slight discrepancy by how I glue the gaskets down. Of course, if I were to have
created a serious misalignment, the valve would either have to be sanded or broken apart
again to correct it. Therefore, this gluing up process must not only happen very
quickly but with exceeding care and dexterity. To make things even more complicated,
I glue the outside valve seats into place at the same time I glue the pouch boards back on
so the work is doubly difficult. Part of the process that happened before this that
I forgot to photograph: The valves had new leather put on them and then they were
placed inside the valve body and checked for how far they moved between the seats. I
make my own paper gaskets that are much thinner than are commercially available for
adjusting these. I vary the thickness of the leather I glue on each valve and I
adjust how many paper gaskets I place under each valve seat in order to get the valve
travel just perfect. Too much valve motion and the valve will spit, waste air and
will repeat more slowly. Too little valve motion and the pneumatics will operate
sluggishly. Each valve is carefully traveled in this way as the leather is being
glued onto the valve bodies because of the fact that each wood valve seat is quite a bit
different from the others. Later in the production of the Amphion design, they
switched to a metal seat that was far superior. These wooden seats will sometimes
leave a great deal of space for the valve to move and sometimes will leave very little.
Sometimes the discrepancies are so huge that the thickness of the leather used on
the valve body must be altered in order to make it possible to glue the valve seat back
in. If the same leather was used on every single valve body, some of the outside
valve seats would require so much gasket material that it would be nearly impossible to
get them to stay in place. On others, without changing the leather thickness, the
seat is made so that it would be impossible to get the valve to stop spitting and repeat
quickly. In short, I hate these old style wooden seats. I vastly prefer the
metal seats they used in the later design. In fact, this piano represents an unusual
transition time in the making of these valves. The old design used the wooden
outside valve seats like this piano has but used a different type of gasket design and
placed the valves upside down on the stack. Later, they went to the metal valve
seats, changed the gasket and placed the valves right side up. This piano uses the
old style wooden valve seats but uses the new gasket and the right side up configuration.
For this reason, this piano is of some historical interest as an example of the
evolution of the Amphion design.
133) 
134) 
135) 
133) The valves are all glued back together. At this
stage, however, they still leak quite badly. They must now be recoated with a thick
layer of burnt orange shellac. The wood structures that make up these valves are
very thin and delicate. Therefore, they are incapable of being air tight without
this thick layer of shellac all around the outside of the valve block.
134) Here, you see the valves recoated with shellac and drying on a tray.
After the shellac is dry, they will get their new cork gaskets. Finally, each valve
will be tested individually to make sure it it totally air tight and functioning
properly. Any faulty valves are taken apart and adjusted before I move onto the next
step.
135) Now its time to recover the pneumatics. To the left, you see some
pneumatics which I have cleaned up and flattened. To the right are those which are
still in their original condition except that the old cloth has been pulled off.
Since the cloth was not all that old, it came off somewhat easier. I was able to get
almost all the cloth off before I started sanding. This would have enabled me to do
much less sanding to the wood normally. However, when I got to the sanding station
with the pneumatics, I discovered that the last time they were sanded, they were not
sanded in such a way as to make each pair of boards matched in size and shape. Some
additional sanding had to be done to true them up the way they would have been at the
factory. Because of all the sanding that was done, any additional sanding in the
future might start to compromise the function of the pneumatics. Things aren't that
bad on them yet (they will work just as well as when they were new this time around) but
any additional sanding in the future will probably be more than they can take. They
will end up too small and too thin to operate properly. Therefore, the next time
they are restored, they should probably be replaced. In addition, the wooden lifter
fingers that are glued to the movable boards have to be split off the boards during
restoration. The last time this was done, they were glued back on very
sloppily. This made it necessary for me to sand the boards even more as well as the
lifter fingers in order to produce good glue joints. If I hadn't done this, there
could have been trouble with the lifter fingers breaking off the pneumatics during use in
the future. Also, the job looked really sloppy before. I always want my
finished pieces to look as close to factory original as possible. Because of how the
previous "rebuilder" treated these pneumatics, my hands were somewhat tied in
terms of making a really good cosmetic job, but I did my best to leave them looking neat
and clean as well as solid and strong.
136) 
137) 
138) 
136) Gluing new hinges onto the pneumatics. The
previous rebuilder had reused the original hinges. I never do this because
attempting to save the original hinges usually means that you can't clean the edges of the
boards as well, leaving a good chance that the pneumatic will leak or even come apart in
the future.
137) Gluing felt circles inside the pneumatics. This was not done at the
factory nor did the previous rebuilder do so. Having this felt glued inside serves
one purpose. It keeps me from pinching the cloth when I clamp the pneumatics back
onto the trunks. If a hard crease is created in the cloth during rebuilding or even
during original manufacture, the cloth will develop pin holes at the creases much earlier
in its life span. This felt circle prevents that crease from forming and allows the
cloth to stay air tight for years longer.
138) Cutting the strips of rubber cloth, needed to recover the pneumatics, from the
bolt.
139) 
140) 
141) 
139) The first three sides are glued onto the pneumatics
first. Then the glue is left to dry. After it dries, the excess is trimmed off
and the pneumatics neatly closed. Only then does the hinge end get glued into place.
In this way, the hinges remain as free and flexible as possible. If the hinge
ends were glued down with the pneumatics left in the open position, the pneumatic would
not be able to close readily. This is called being "hinge bound."
140) On the left of this photo is one of the trunks boards. The thin wood had
failed over time. The suction inside the board had literally sucked the wood inward
and ruptured it. The last rebuilder had glued a layer of pneumatic cloth over the
tier to stop the leakage. This worked to stop the leaks but didn't address the
deeper issue of weakness in the wood. To the right you see thin strips of wood that
I planed and cut to fit over this trunk. The trunk was sanded down as far as I dared
to get the wood as flat and true as possible. Then these strips of new wood were
glued over the weak area.
141) Here you see the new wood being clamped into place. After the glue dried,
the wood was sealed with shellac. This trunk will no longer have any problem with
leakage or weakness in this area. In the fairly distant future, the other two tiers
will probably need a similar repair.
142) 
143) 
144) 
142) Restoring the lifter fingers and poppets. The
poppets are the lifter wires and graphited dowels which contact the backs of the keys on
the piano's action. The wood dowels have been relubricated. The fingers that
glue to the pneumatics have had all the old glue removed and the wood sanded clean and
true. The previous rebuilder simply snapped these fingers off the pneumatics and
then glued them back on again with some extra glue without doing any cleaning or
preparation. This left the fingers very difficult to remove on this restoration,
causing a great deal of unnecessary damage during disassembly; not to mention the fact
that it looked terrible. When I started on these, I felt that this was probably the
last time these fingers would be able to be reused; that the next restoration would
require new fingers. After I finished restoring them and gluing them back onto the
pneumatics, I felt that they had been revived well enough that they may survive even
another rebuilding cycle provided the next rebuilder knows what he/she is doing.
143) Gluing the pneumatics back onto the trunks. Originally, a layer of cheese
cloth was glued to the trunks and the pneumatics glued over that. The idea was to
make it easier to get the pneumatics off in the future. The idea is a good one but
it doesn't really work as well as Amphion though it would. The cheese cloth really
doesn't make it that much easier to get the pneumatics off; therefore, damage is incurred
during disassembly despite the cheese cloth. I gasketed the pneumatics with a layer
of construction paper before gluing them down. Experience has shown me that this
will hold very well throughout the life of this restoration cycle. Then, when the
next rebuilder has to remove the pneumatics once again, he/she will find that the
pneumatics come off exceedingly easily with little or no damage to the pneumatics or the
trunks. My main goal here is to try to preserve the trunks. I think the
pneumatics will need to be replaced but making new trunks is very costly and difficult so
we must attempt to preserve them. This includes helping the next rebuilder even now.
144) The top tier is positioned to show the lifter fingers have been glued back into
place with the poppets removed. I will leave the poppets off until I'm ready to
reinstall the stack into the piano. This will keep me from damaging the poppets by
accident. The lower tier shows that the valves have been reinstalled. Knotted
tubing is placed over the signal nipples and each tier tested to make sure none of the
valves or trunks are leaking in any way.
145) 
146) 
147) 
145) The reassembled stack.
146) Another angle.
147) The drawer before restoration began.