STEINWAY MODEL “B” Restoration Blog:


This blog will follow the restoration of this 7′ Steinway grand. It will include a detailed description of the work as it is carried out.

Nov. 2 The workshop is beginning to take shape. An additional room was added to the back of the workshop to allow me to partition the woodworking tools. This will prevent wood dust from getting into the piano, which will be in a controlled area. The hand crafted workbench, now known as “The Captain” was made by Warren Atkins in San Mateo, Ca. This bench contains 8 drawers which are wide and deep, accommodating all manner of hand tools. The workbench and all of Warren’s machinery were purchased as he finally retired from his 60 years of craftsmanship. Nov. 4 A word about laying out a workshop: Rebuilding this piano requires a vast amount of space, which I have divided into three different areas. 1. The proper piano workshop is in two rooms with a six-foot opening between them at the back of the basement level of the house. We live on a hill, so the slope of the property puts the basement at ground level in the back. This area has to be shoe-box shaped, with enough room to have double the length of the piano. That way, the plate can be hoisted up with room to pull the piano out. The soundboard itself will need to be coming out soon, and I willl need to flip the piano upside down for that. All this activity needs to be accounted for in the planning of the shop. 2. The woodworking area is in an equally sized space which has been enclosed at the back of the house. The advantage of a separate room is the isolation of dust and wood shavings from the clean piano. This area will have most of the woodworking and finishing equipment, but I am still waiting for the contractor to finish the interior walls. Once that is done, the electrician will complete the lighting and wiring for the machines. 3. The garage will be the other area, housing equipment that is occasionally used. Parking cars has become an interest due to the first snow that arrived yesterday. Nov 12 The electrician will be installing lighting in the workshop today. This requires a bit of study these days, as lighting products are becomming quite different. T-8 florescent fixtures now have electronic ballists, so there is less flickering in cold temperatures; and speaking of temperature, the tubes have different light spectrums ( temperature ). The “tear-down” process has been started for the Steinway B. Case parts are stored, action removed, dampers removed and back action is out. Before lowering the string tension, I will take string height measurements. The keybed to the bottom of the strings is critical, and since I will be changing both the soundboard and the pinblock, this is something I can alter. The action geometry should allow the hammer to strike the string at 90 degrees. Strings that are too high or too low cause hammer to hit too late or too soon, with loss of power in both situations. This can be addressed by adjusting the height of the new pinblock as it is being installed. For now, all measurements go into the notebook.


Nov. 24

Warren's Bandsaw in action
Warren’s Bandsaw in action

The pinblock was removed by sawing and hammering along the edges. The remainder was cleaned up with a large chisel. This is not elegant work. The sawing is done by standing in the piano keybed. By the time you are done, there are wood chips everywhere. The curved edge of the pinblock has to be matched to the edge of the plate. First, the old block is traced onto the new block. The angle of the edge is never the same, so this is measured and the bandsaw table is angled to duplicate the angle. Warren’s Delta saw has been improved by him over the years. It cut straight and true, producing a very close duplication. Refining the fit, the plate is turned upside down, and that edge we need to match is coated with lumber crayon. When the new block is hammered against it, the high points are marked by crayon that transfers. Slowly, the high points are removed until there is a close fit. To increase the fit’s perfection, a final layer of a heavy duty epoxy is added to the edge and the block is firmly clamped in place. A layer of plastic wrap prevents it from gluing together.

The original pinblock.  Note cracks and black stains left by glycerin treatments,; pin tightener products that are distructive.
The original pinblock. Note cracks and black stains left by glycerin treatments,; pin tightener products that are distructive.
Plate support dowels cut off to lower plate height.  An attempt to increase down bearing.
Plate support dowels cut off to lower plate height. An attempt to increase down bearing.


With the new pinblock now fitted to the plate flange, it is now important that it also fit into the case. Up till now, we have purposely left it too long and too wide. Now, we refer to the two small index holes we had drilled through the plate. Measurements and templates from the original pinblock can now be used to draw cut lines on the new block so that it fits almost perfectly into the case. I purposely leave a little play so that when it is glued into place, it is the plate that orients the block and not the other way around. A heavy steel-filled epoxy is used for the glue for excellent adhesion and gap filling. The pinblock is also doweled into the case rim, for the maximum strength, fully fitted Steinway design.

The plate screws are into the block and is now part of the plate.  Using the plate as a clamp insures that the pinblock remains in this orientation while it is glues in place.
The plate screws are into the block and is now part of the plate. Using the plate as a clamp insures that the pinblock remains in this orientation while it is glues in place.

Dec 8

Now that the pinblock is glued into the case, the height of the strings above the action and the orientation of the iron frame are fixed and registered by the plate screws, it is now possible to proceed with the soundboard.

New cracks appear alongside repairs due to aging of wood.  It shrinks across the grain.
New cracks appear alongside repairs due to aging of wood. It shrinks across the grain.


imageThe decision to replace this soundboard was confirmed by measuring the crown with no string tension. A scant .5mm is present, but that will disappear or go to a negative figure under string tension. It is also evedent from the reappearance of cracks, that there is no future in repairing this soundboard again. In removing the soundboard, another problem was discovered. The hide glue holding the soundboard in the frame was deteriorated and broke away very easily. Hide glue is organic and like all organic substances, is effected by age. Hide Glue crystalizes, looses it’s strength in the process.

Previous restoration shims in soundboard
Previous restoration shims in soundboard

Soundboard removal….piano placed upside down.
Dec. 9

Ribs are copied from the original board, except that they are given a crown shape. The thickenss of the rib is the same at the point it is contacting the bridge ( not the middle), and tapered in thickness as it gets closer to the rim. This not only helps in forming the soundboard crown, it also contributes to the tapering of the whole soundboard to make it “diaphramatic”.


Ribs, pre-crowning
Ribs, pre-crowning
Ribs being fitted.
Ribs being fitted.

Jan 1, 2014. Happy New Year!

Now, for a quick review of soundboard “crown”. The “body of knowledge” on this subject is fairly difficult to get your hands on. Articles in journals and encyclopedias tend to settle on a crown that has a 60-ft. diameter. Traditionally, this is achieved by gluing straight ribs onto the bottom of the flat soundboard panel, but using a dished out table as a clamping surface. Clamping force in the middle would be impossible to achieve with normal clamps. The traditional solution was to use go-bars, long slats of wood ( usually Ash), slightly flexed to fit between the clamping surface and a sturdy ceiling built above. Each of these can excerpt 30-50 lbs of force, and a soundboard might use 50 go-bars at once. Modern methods use the same dish shape clamping table, but replace the go-bars by using pneumatic pressure.

Prior to gluing on the ribs, the soundboard panel is dried in a special “hot-box”. The drop in % moisture content means that the dimensions of the panel will shrink, mostly across the grain. When ribs are glued across the grain in this shrunken state, the bottom plane of the soundboard is restricted from expanding when normal moisture returns. This combination of the clamping shape and the planned distortion from increased moisture content, forces the soundboard into what is called ” compression crowning”.

This is my worksheet for figuring the crown of ribs, and the curvature of the jigs and clamps used in the process.
This is my worksheet for figuring the crown of ribs, and the curvature of the jigs and clamps used in the process.

The problem with compression crowning is that the soundboard panel suffers damage to the wood cells that are under this long-term state of compression.
It can lead to compression ridges appearing, and eventual cracks. To reduce this, many rebuilders and manufacturers are cutting the shape of the crown directly into the ribs. This allows the crowning to be done with less of the moisture removed before gluing the ribs. ( 6% EMC, instead of below 5%).
This rebuilder uses a combination of “radius crowning” and “compression crowning”. The ribs are cut to the original height, then tapered so that the thickest part is under the bridge ( NOT the middle of the rib). This also contributes to the DIAPHRAMATIC action of the soundboard, which will also be tapered in thickness.

Jan 20, 2014

The soundboard now has the ribs glued to the underside and the crown is already present. Attention has been directed to the bridge which is being recapped. The top layer of wood on the bridge is being replaced with rock maple. The boards being used are 1/4 sawn so the grain is straight and the wood is dimensionally stable. The four sections are being arranged like puzzle parts. The bridge shape is curved, so the sections of straight grained wood are spliced together at obtuse angles to keep the grain direction following the bridge. Sound travels along the grain and we want that sound energy staying on the bridge as much as possible. Once these angles are cut and the boards are indexed to the bridge, the shape of the bridge will be traced on the underside of the cap and cut out on the bandsaw so that a slight overhang remains. Gluing and clamping are a critical part of the restoration. Once this is done, the cap will be trimmed flush with the bridge with a router. ( Danger!)

Jan 23, 2014
The bass bridge is connected to the long bridge with a curved section of Maple. The purpose is to conduct vibrations from the low tenor over to the bass section of the soundboard. While the two actual bridges are vertically laminated,making them quite ridged; this section is somewhat flexible while it is not glued down to the soundboard, and subject to changes in humidity. In other words,
The bass bridge changes it’s location in relation to the long bridge on it’s own! This next step will locate in a very positive way the exact location of the ends and middle of both bridges. Any wondering will be corrected.

Before removing the soundboard, jigs were constructed. They are pieces on wood with short pins sticking out of the edge. The pins fit into corresponding holes in the inside of the rim. Each jig has a finger of wood glued to it that reaches out to touch a corner of the bridge. With these installed, it creates a sort of dotted line. By placing the bridge on the soundboard and moving it around until fingers of wood touch the bridge, we can drill index holes through the bridge and through the soundboard. Now, with the soundboard removed, a metal pin will put through each corresponding hole to begin the process of gluing the bridge to the top of the soundboard.



Our Pneumatic Soundboard Press
Our Pneumatic Soundboard Press



Feb 4

The soundboard is nearly ready for glue up. Naturally, all of the finish coatings must be applies to the bottom side of the board, and that was done yesterday. The important part about soundboard finishes is a layer of shellac. Not the hardware store product. This is made fresh in my workshop.
Shellac starts as a substance excreted by the lac beetle, which is collected and refined into button shaped discs, or further refined into flakes. Once mixed with alcohol, it becomes a penetrating resin which when applied to raw wood, makes the most effective vapor barrier, slowing the transfer of water vapor from the air into the wood cells. No finish coatings will stop this transfer, but shellac comes the closest and that is why this is the traditional method. There is a top coat of varnish to add gloss, and the choices for varnishes have become limited. EPA regulations have limited the solvents that can be used, so many of the products have become very hard to obtain, or they no longer handle in the same way. We opted to use Minwax Polyurethane because it behaves well when applied over shellac and hardens well.

All other preparations involve fitting the corners of the soundboard. They will be difficult to trim once the board is glued in. Also, it’s time to collect all the clamps and make clamping blocks called cauls. These are padded with thick cloth and are beveled 1.5 degrees to match the bevel in the inner rim. This is critical to establishing good crown.

Feb. 12

Soundboard in hot box, dried to6% EMC
Soundboard glue-up.

The soundboard was glued in today! It was dried for a few days in the hot box until. EMC was stable at 6%, then glued into the rim before it could expand. There are about 50 clamps holding it tightly in place. The glue is a specialty modified PVA formulated by the soundboard manufacturer. It has good handling characteristics, and dries glass hard.





DSC_0217 DSC_0219 DSC_0221 DSC_0223

March 1:

The bridge cap was planed to height, marked out for bridge pins, drilled for bridge pins.

With the soundboard glued in, the down-bearing is the next challenge. Crown is established in the board, but as the strings are deflected from being in a straight line by the bridge, they apply force the soundboard, known as down-bearing. The soundboard yeilds to this force, but at some point, the resistance of the soundboard is is equal to the downward force of the strings and a truce is declared. We never know where exactly that is going to fall, but it is expected that at least 1/2 of the crown will disappear. The height of the bridge cap is how we will regulate the down-bearing, but we will have to allow for this loss of crown….it will be left thicker in the middle to account for the motion, and less in the treble and bass. Every soundboard will be different.

Methods of establishing the down bearing vary. Some calculate the angle of the string as it exits the bridge on the way to the hitch pin. Some load the soundboard with force using go-bars to flex the board downward. I choose to follow a combination of methods. I don’t want to overload any part of the piano. .008 inches of deflection in the highest treble. .021″ in the very middle. The deflection between the middle and the edges is divided up evenly along the length of the bridge. The bass bridge needs even less, starting at .008 and going down to .006 for the lowest notes.

Once the height is established and the bridge cap is planes to height, the location of the bridge pin follows. This is where the errors in the original bridge may become obvious. Strings should travel in a straight line between the hitch pin and the tuning pins. If the hitch pins are wrong, the factory will always hide the flaw by placing the bridge pins so that they look correct. The rebuilder does not have to be a slave to the symmetrical looking strings, and can choose to lay them out correctly, even if it looks wrong! By pulling a thread stretched. between these two ends, the bridge pins can be accurately located; then checked using the templet of the original. When the two disagree, an investigation will be interesting!

I drill the bridge pins with a high speed drill… a dye grinder with a 1/4″ pin vise. It is light weight and easy to control. Once this is done, the bridge is notched so that each of the three strings of a unison will have the same speaking length.

Finishing the top of the soundboard involves several coatings of shellac to seal the wood as a barrier to moisture. No finish will absolutely seal the wood, but shellac is the most effective. This was over-coated with modern waterborne lacquer for a very durable finish coating. This piano also got a soundboard decal.

The cast iron frame was re guilted with finish provided by Steinway. The agraffes were removed, polished and reamed, then returned to their original location following the finishing. All the raised lettering had to be repainted by hand. This one included all the patent dates, making the job more difficult.

The wire and tuning pins were German manufacture, but the bass strings were made in the USA by Richard Aldridge. Working from direct measurements, the strings were made for this piano…..not just any Steinway B.

All Steinway parts, but the hammers were from Hamburg.

Ken Eschete – Bentside Arts