Introduction: How to Make a Piano Keyboard

In part 4 of "How to Make a Pianoforte" W. H. Davies instructs the amateur piano maker "to provide {themself} with a set of 16in. keys to match the size of key-bottom; they may be bought at from 38s. to £3, according to quality". You could choose between more than ten keyboard makers just in London around the time he was writing, as well as purchase other components from many other kinds of specialists in work for piano makers. Today a new keyboard of this description won't be so readily available or affordable, and in America it's not likely you could even scavenge keys that would work from a discarded old one.

My keyboard is for a square piano and is more complicated than the one for Mr. Davies upright. The lengths are all different and the rails are at angles, the front and back ends of the keys are offset sideways, they aren't uniformly thick and besides there being fewer of them the keys as well as the action parts they set in motion are smaller. It works the same and has the same ordinary arrangement, though, with seven white and five black keys in each octave and no breaks in the back spacing so the steps are the same as for Davies' upright, and it even has the parts needed in part 5 of his instructions.

A keyboard closer to Davies' is described in "How to Make an Organ Keyboard" by "M. W." first published in Work Illustrated in 1884.

Step 1: Materials: Keyplank

Piano keys are wooden levers usually cut out in order like jigsaw puzzle pieces from a wide panel or keyplank.

Keyplanks are glued up to the width required out of a few boards with the grain going the long direction of the keys. The wood they're made of has to be dry, straight grained and free of knots, as well as stable with changes in humidity and relatively strong and easy to work. They're usually a light weight, soft wood with little difference between early and late wood, like soft white pine or basswood. English piano engineer Samuel Wolfenden considered pine a little too soft, and like the famous German piano manufacturer Julius Blüthner considered basswood a little too unstable, depending on where it grew. Different woods can be combined to make better advantage of their good qualities.

The keys shouldn't bend so they should be relatively thick, depending on the amount of space there is, but usually they aren't much more than 2.5cm or less than 1.5cm. Shorter keys can be made thinner, and thinner keys can also be made stiffer by increasing the height just around the fulcrum, using stiffer wood or by reducing the stuff they have to move.

Step 2: Materials: Keyframe

The keys are supported in the middle and their motion is also guided at the front and the back, and so they usually rest on a frame made with three wooden rails that are more or less perpendicular to the key levers.

The middle, or balance rail is tallest and works as the fulcrum for all the keys. It has metal pins driven into it that go through tapered holes in the keys that help keep the key in position. Its height depends on the height of the other supports, and its width depends on the lengths of the keys. It's usually bevelled starting at the lines of pins, and sometimes tapered a little so it's higher at the front row of pins to make clearance for the keys.

In modern pianos the front rail also has pins in it, and both it and the balance rail are made of clear and dry hard wood that resists splitting. In America cherry, and later, maple, are common, and in Europe I think beech and oak, but oak might make the metal corrode. The front rail should be tall enough so that when the front pins are driven in they don't stick out the other side, and wide enough to support two rows of pins about 5cm apart. Sometimes the rail is made in two thicknesses and thinner around the back row of pins so the adjustments for the raised keys don't interfere with the natural keys.

The back rail usually only touches the key when it's not moving and is usually covered with a piece of uniformly thick, relatively soft cloth to reduce noise and to make the working parts close to the right height. In uprights where the keyboard isn't shifted left and right, or pulled out like a drawer a lot it can be made of softer wood, like yellow poplar, or even left out entirely - otherwise it's the same as the other two.

In order to keep the rails in the proper relative positions they are connected by perpendicular pieces, one at either side and then one every 40cm or so, or at the breaks in the strike line, also often made of softer wood, yellow poplar or sometimes mahogany, unless they have to hold screw threads like in grands. These usually are no thicker than the shortest rail and about 5cm wide.

Step 3: Materials: Hardware

The playing portions at the front of the keys are covered with more durable and cleanable materials than the wood used making the keyplank. The natural keys have a thin, hard material on their fronts and tops, and the sharps are made taller by gluing thick pieces of hard wood or plastic with a color that contrasts with the naturals. Sheet materials like veneer or plastic can be glued on sooner and require less work compared with individual pieces for each natural key which you can get in plastic, bone or ivory, in three, two or even one piece, in the case of molded plastic keytops. I reused old two-piece ivory from an old piano, but this makes it more difficult because each pair has already been trimmed, levelled and shaped together for the exact key they were on originally, and like the ones I used they are often discolored, worn and damaged.

Modern pianos use two polished pins running in specially formed holes to hold each key in its position and to guide its motion so that no keys interfere with each other. You can read about some methods used in harpsichords here.

The balance rail pins are usually cylindrical and rounded or chamfered at either end, and long enough to be bedded into the rail at the lower end and protrude slightly above the key at the upper end.

The front rail pins are formed so that the upper end has an oval section, so that the flatter sides present wider bearing surfaces than if they had a circular section. The lower end that is driven into the rail is cylindrical and chamfered. A lot of writers warn against turning these pins to take up play as the holes in the keys wear larger, but this is exactly why they were invented. A couple of 19th century piano manufacturers makers did the opposite, so the hole itself could be adjusted by turning a screw.

The guide holes in the keys used to be closely fit in the wood, but eventually were lined with usually replacable bushings made of resilient cloth or leather, about 1mm thick and sold cut into long, narrow strips. The fitted parts of the holes may also be made in a different piece of wood, to make them more durable as well as to reduce work, usually balance pins work in glued on key buttons, which can be prepared separately or bought ready made.

The back rail is covered with a strip of about 3mm thick cloth cut about 4cm wide so that the keys don't knock or rattle as they return to their resting positions. There is padding on the balance rail and on the front rail as well, but now usually in the form of punched cloth rings, about 1mm thick and 12.5mm diameter at the balance rail, and 3mm thick and 25mm diameter in the front, with center holes a little tight on the pins. These are usually raised above the rail with different thicknesses of the same size paper and cardboard rings, so that the keys rest and travel uniformly.

The keys will also have some sort of device mounted at the back end. My keyboard, as well as Davies', has one of the main parts of the action glued onto the keys. Modern pianos have entirely separate actions, and use special screws driven into holes so that the play between them can be regulated, and many older pianos use more sophisticated mechanisms.

Weights are often used to change the resistance of the keys, usually they are made of lead with a slightly conical shape and are driven into holes made in the sides of the keys. I don't know if in Europe the lead weights are exempt from the Restriction of Hazardous Substances Directive.

Step 4: Keyframe

The keyframe can be rectangular, trapezoidal or even have no parallel sides depending on the spacing of keys and the strings and the angle of the hammers, and so the lengths of the rails and stretchers vary depending on the design of a specific piano.

My keyframe rails are made of cherry, and they are connected with mahogany stretchers. The front rail is 7.5cm wide and 14.5mm thick. The blank was longer than 60cm, a little wider than the finished keyplank.

The backrail is 8.5cm wide and 11.5mm thick, and the blank was more than 57cm long. It has a rabbet at the back edge, bevelled to 5mm at its thinnest, to fit similarly tapered blocks at the back of the keywell, but this is only useful in squares and grands.

The balance rail is 4cm wide, a little short of 2cm tall and at started more than 50cm long.

The side stretchers are 9mm mahogany planks, about 6cm wide. The treble stretcher is longest and is about 40cm long. They are quartersawn, but the grain was so interlocked they were more difficult to make than if they were flatsawn.

The stretchers are joined with the rails with open mortises and tenons with the tenons cut on the stretchers. I marked the mortises out by laying the rails on top of the stretchers on a pattern marked on my work table, cut nicks at their intersections with a knife and transferred them all around with a square and protractor. I marked the widths of the mortises and tenons with a marking gauge.

I cut the mortises with a fence on the bandsaw. I chopped out the waste with a narrow chisel and finished the sides and bottom with something like a plane-makers float which is a kind of thick hand saw with no set. I cut the tenons the same way and slightly too thick so I could plane each one to fit, and when they all fit I assembled it so I could mark and trimmed all the shoulders so they were flush. When the outside stretchers fit I marked for the middle stretcher, disassembled the frame and cut shallow mortises in each of the rails, and then cut tenons on the stretcher.

After fitting the middle stretcher I glued the keyframe together, making sure it was square and flat, and clamped each of the corner joints together to prevent the moisture from the glue from pulling the pieces apart. Once it was dry I planed the obvious high spots, and checked it on a flat table - it didn't take much to make it but it's not so important in a square or upright where it's screwed down. In a grand it might be better to have the sides higher.

Finally I marked for the balance rail and rabetted it on the bottom to slip over the stretchers. I only fastened it with countersunk screws so I could remove it to bevel it. This also helps regulating when it's finished.

Step 5: Preparing Wood for the Keyplank

The finished dimensions of my keyplank were a little less than 600mm wide at the front, about 37cm long and and 18mm at its thickest.

I used some pine pieces from the bellows of an old reed organ someone had cut apart with a saw - they flockpainted the case pink and squirted a lot of white glue on the bellows before discarding it. The trees were cut and dried about 120 years ago. The boards had some flaws which I had to arrange so they ended in places where they would be removed.

My octaves are 144.7mm wide, which is very small and almost the same size as 7/8 Donison-Steinbuhler standard keyboards you can buy for modern pianos. They are usually between 16 and 16.5 cm - Blüthner wrote that German keyboards had octaves about 161mm, and that English and French keyboards had octaves about 164mm, and as much as 166cm. 161mm works out to between 5 4/10 and 7 3/10 German inches, and the English and French from about 6 7/16 to maybe 6 15/32, to as much as 6 1/2 English inches at the greatest. In fact, like Steinbuhler, Blüthner, Wolfenden and piano engineer Siegfried Hansing all specify the overall width instead, so that the width of the octave is secondary. You can read an interesting paper about this here.

I was able to use all the bad pieces by cutting them in wedge shapes, which also made it so the grain followed the direction of the differently angled keys. I set them on top of the outline I drew on my work table, marked the joints and bandsawed off the waste.

Step 6: Gluing Up the Keyplank

I planed the edges as straight and flat as I could to prevent introducing any stresses the from gluing or clamping. My planks were near the final thickness so it was more important that the planks line up at the joints than making the panel sit absolutely flat, and the same way after it was glued I concentrated on making it smooth and uniformly thick instead of flattening it across the grain.

I marked out and trimmed the keyplank, and then routed out the bottom in the front 75mm about 4mm deep and glued in a bunch of pieces of cottonwood to make the bearings for the front pins, and routed out the back to make the keys the right height. After planing the cottonwood insert flush I jointed the front and rear ends of the keyplank. I glued on the molding for the front of the keys, a uniformly thin piece of maple wider than the thickness and longer than the final width of the keyplank. Once the glue was dry I trimmed it the right size, put a bead on the front using a scratch stock I made with some scrap and a broken hacksaw blade, and put a light coat of finish on it to keep from staining it with dirt or glue.

Step 7: Marking Up the Keyplank

I drilled and cut the keyboard manually so I had to mark different details on it.

I used the marking gauge to scribe lines on the top of the keyplank corresponding to the back of the wide portion of the white keys, usually about 4.5cm from the front, and another line about 3mm behind it corresponding to the front of the black keys. I scribed similar lines on the bottom but offset from the top one making about a 10° downward angle toward the front for the wide part of the natural keys which reduces some weight and makes clearance for the raised keys. Sometimes the raised keys are cut with a smaller angle toward the back.

I adjusted the gauge to about the middle between the front of the keyplank and first bottom line, and scribed this across the top, and then scribed a second line about 1cm behind the front of the black keys. These lines correspond to the two rows of holes for the front pins. It can also be helpful to make lines on the bottom about 5mm to either side of them to show the limit of the mortises, especially if the mortises are going to be bushed so that the wood between them can be planed cleanly down by about the thickness of the cloth.

I trimmed the keyplank to its final width, with table saw for the parallel sides at the front and bandsaw for the angled sides at the back as well as the front, and then marked out the divisions for the individual keys. First I marked the heads, or the wide front parts of the natural keys, making even spaces up to the first line on the top of the key, and up to about the back of the raised keys between every E and F, and B and C keys.

To find the divisions for the back of the natural keys and for the raised keys Blüthner, Wolfenden and writer on pianos Lawrence Nalder all described dividing the space occupied by C, D, and E, into five equal spaces, and F, G, A, and B, into seven, but Wolfenden points out this can make it hard to play between F# and G#, and G# and A#. The two different spacings make the rear portions of the C and F keys wider than the neighboring B and E keys, though this can be remedied by sawing to the side of the line instead of in the middle, and requires two sizes of sharps, which can be adressed in the same way. Architect and writer on organs George Audsley suggested to subtract the width required for equal sized sharps before dividing into three and four spaces respectively, which results in an even greater difference between the neighboring tails, and Wolfenden suggested alternatively that the Cs, Ds, and Es be made slightly wider than the Fs, Gs, As and Bs, with uniformly narrow sharps.

Each writer advised making a wooden pattern from the divided keyboard to save the effort of dividing it again. Blüthner suggested drilling holes through a wooden pattern so their positions can be transferred with a punch, and Audsley described that in the key making department at the Steinway factory they used patterns with embedded points so that all the marks would be made at once.

I already had a pattern to work from that I used to mark the divisions in the front. I glued the pattern for the divisions at the back of the keys to a piece of thin plywood and punched holes for both sides as well as the middle of each key. The balance pins fall in two rows so the leverage of all the keys is the same, and I scribed lines on the top of the keys set back a little from where they set in the rail because having the pins angled back makes the keys easier to put on and take off.

I lightly pencilled lines connecting the marks for the sides of the keys at the front and back, and punched marks for the balance pins centered between the pencil marks, along the corresponding scribed line. Technically their most efficient position falls in line between the where the key is pressed and where it works the action, but it's also important that they are centered on each key and aren't surrounded with short grain. When there's a large sideways offset between these it's common to make multiple angles so that the portion around the balance pin is parallel with the front and the back parts, in a shape sort of like a dog's back leg, to have more wood and stiffer straight grain at the fulcrum.

Step 8: Drilling the Keyplank

In order to drill them the marked up keyplank and keys have to be kept aligned together the same way that the keys would be on the keyframe. Blüthner suggested nailing them using wire nails.

Most drill presses will reach the front pins, but the sides and back of the keyplank and keyframe will stick out a lot and there won't be much support. You can fasten them to a stationary table drill by hand or use a small drill press placed on top of the keyplank with the head lowered almost to the base. I used a homemade drill press that has a relatively long reach and clamped the keys and keyframe together to a table that can slide in front of it. It makes it much easier drilling in lines and at compound angles.

I drilled each hole from above, and in three passes - first a tight driving fit through the keyplank into the keyframe rails with a bit about 90% of the pin diameter, then a running fit just through the keys and then a hole large enough for clearance as the key rocks. A special step drill bit like Audsley described would be more efficient.

Since the rails aren't parallel I drilled it in two settings, starting with the balance rail first so I could bevel it up to the centers of the pin holes before driving in a pin in the top and bottom holes to re-index the keyplank on the keyframe for drilling the front pins. I put a shim under the front so the holes through the keyplank would be straight and clean. Bluthner suggested drilling out the waste behind the natural heads, too, but I did this after sawing, and only a hole large enough to turn a fret saw blade. I also used the drill to make the angled holes for the backchecks, which required repositioning several more times to get the compound angles right.

The front mortises have to be made oblong with perfectly parallel sides, and Bluthner suggested using a broach but didn't describe it. I made one that works like a couple of stuck together knives I learned was based on an old one discovered in France, but I used one that is shaped like a double sided metalworking broach. I chucked it in a benchtop drill press and kept the mortises aligned by using a fence and locking the spindle. It leaves the narrow sides kind of rough depending on the type of wood, but it looks like the mortises in old piano keys.

I routed a shallow groove around the row of natural front holes and glued a strip of scrap from the keyplank into it so the holes don't show through the keytops. It had to be really thin so the tops of the front pins don't hit it when the keys are pressed down.

Step 9: Gluing the Natural Keytops

Audsley and M. W., who gave instructions for gluing on the natural keytops, suggested attaching them at this stage. Audsley recommended using single pieces of ivory for each key if the keyboard will be played often, adding that their greater cost will be offset somewhat by their more easy application, and he described the more difficult process using separate pieces for the wide and narrow parts of the naturals. Since they're hard they work sort of like guides for sawing. Sheets of plastic or veneer can be stuck on at this stage as well.

I used some two piece ivories I removed from a set of keys someone gave me from an old German upright they took apart. They were all chipped and discolored, many were worn thin and a bunch had been deeply scratched. I ended up with just enough usable ones for 29 keys. I think a lot of people would have thrown them away.

I bleached them and jointed all four edges square and to the right size, which also removed the chips and marks from fitting, but I wanted to keep the existing surfaces so I had to sort them by color and thickness so the matches aren't perfect and it looks like an old keyboard instead of a new one. M. W. suggested reserving the lightest colored ones for the treble but didn't say why. After bleaching the ivory it was kind of translucent so I bleached the front part of the key plank. The freshly mixed hide glue I used dries clear (and doesn't smell bad) so I added whiting to it, too.

I clamped an aluminum straight edge to butt the back ends of the front parts of the naturals because it won't stick to the keyplank as a wooden one. Audsley instructed then to apply the glue and then put the ivory heads in place and butted against the straight edge, and to clamp them down with a large heated metal bar covered as a clamping caul with a piece of cloth to protect the ivory, but my used ivories were all different thicknesses so I used individual cauls. These are tapered and made with a lip so that they draw the heads against the straightedge, and they let me see how each one lined up so I can make adjustments to their positions while the glue isn't set. I wiped off as much of the glue underneath on the front as I could while it was gummy.

Once the glue dried I used a fine square file to join the back edges because I thought the suggested skew plane might make them chip.

The tail pieces have to be carefully jointed before they're attached, and ensure a good joint Audsley described drilling small pilot holes into the key at their far ends so they can each be tacked in place with a tapered peg before clamping them the same way as the heads but mine were too small and short for this to work reliably. I glued them using individual cauls as well, and I used a gang of cams strung on a rod to clamp them because I only have a few clamps with a very long reach.

Most of the excess glue was easy to remove with a damp cloth, but I had to soak a few and a couple got too wet so the glue at the edges re-gelled and the ivory cupped so I had to reheat and clamp them again later, but since I did the pairs it compensated for the small height differences at the joints.

I jointed and rounded the fronts and radiussed the corners slightly with a fine three-square file.

If the keys weren't so varied in length I would have drawn a dark angled line across the width of the keyplank or numbered each key so I couldn't jumble them later.

Step 10: Sawing the Keys

George Dodd showed cutting the the keyplank into individual keys with a bowsaw with a wide blade, and Audsley recommended using a bandsaw with a fine blade and a sliding table. I don't have the right combination of these so I used different saws to cut the straight parts in the front and the angled parts in the back.

I cut the fronts with an old Craftsman scroll saw, the kind with a spring and with about 2.5cm blade travel. I made a new blade guide to work with a broken pullsaw blade I ground to fit. It has a 0.6mm kerf, cuts straight, leaves a decent surface and survives cutting through glue and ivory without heating up or dulling.

I drilled small holes at the ends of the short cuts in front of the sharps, and then bandsawed from the back side to the ends of the long cuts to divide the keyplank into smaller, more managable sections. I took the big blade and guide off the scrollsaw, and cut to either side of the sharps threading in an ordinary blade and then starting it with a footswitch. Afterwards I cut the sections into individual keys with the bandsaw.

Step 11: Fitting the Keys

The bandsaw blade left wider spaces between the keys and reduced the amount of fitting I had to do, and I was able to remove the saw marks with a coarse mill file. I made the angled undercuts on the naturals with a small backsaw and removed the waste and relieved under their covered parts with a chisel so that they're about 4mm wider at the front and on top than on the bottom.

I had to deepen the wide part of the balance pin holes a little, and some of the wood had sprung a little to one side, and because they were used ones most of the key pins had been bent so the spacing was uneven at the front and at the back. I was able to fix this mostly by twisting and bending the pins as prescribed by Hansing, using pliers with protective wooden sleeves so I wouldn't nick them, and only had to plane a couple of keys that looked kind of too close together.

The ivories often are rounded with a file on the sides neighboring other ivories, and the heads chamfered at the back edge in front of the sharps. This makes them look funny when someone reglues them on the wrong keys.

Step 12: Gluing the Buttons and Sharps

I made short moldings for the buttons out of cottonwood, and drilled and mortised them like the fronts of the keys. I sawed them out for each key a little oversize, and first glued on the buttons for the naturals, and then the sharps, using an aluminum straight edge clamped to the assembled keyboard and keyframe. I heated the buttons in in sequence on an upside down travel iron so I had a little time to adjust the keys so the back spacing worked out before the glue set.

I planed down the keys that were wider than the ebony sharps, and made the spaces on both sides even, and then glued the sharps on the same way as the buttons. They're just far enough back from the ivories so they don't interfere when they're pushed down

I left my keys with the original surface, but this is probably a good stage to surface the natural keys with progressively finer sandpaper and a block, for example by clamping in place on the keyframe with the sharps removed.

Step 13: Installing the Hoppers

Modern piano keys have special capstan-like screws stuck in round holes to push the action parts in motion, but my keys have hoppers glued into slots like the ones for Davies' sticker action upright. I used a Woodruff keyseat cutter chucked in my drill press to make the slots. I didn't have one the right size so I set a smaller one the right height above the table to cut the upper side of the slot and then stuck the right size shim to bring the key so it cut the other side, and removed the waste with a narrow mortising chisel.

I glued each hopper in with the key off the keyframe, and then replaced them so I could check the alignment with a straightedge and redo any that were out of line while the glue was still wet.

Step 14: Weighting the Keys

The keys often don't fall the right way by themselves, either for mechanical reasons or for how the action is supposed to feel when it's played. Usually this is adjusted by adding lead weights to increase the weight on one side of the balance rail, but more flexible systems have been designed using springs, and more recently magnets.

In grand pianos skilledfactory workers carefully position key weights by hand with the action all assembled so it takes a specified minimum weight to make each key move. This would have been impossible to do with my keys anyways, but recently this subject has been explored a little further, which you can read about here and here. My keys were all front heavy but in this kind of action they need to be back heavy by as much as 15 grams and return to rest by themselves so that they repeat and have the right amount of resistance, and one of the only places to put weights would have been inaccessible since it's directly below one of the action rails.

Ordinarily pre-made conical weights are pressed into holes drilled in the sides of the keys, but I didn't have any ones small enough. Davies suggested pouring molten lead into the holes and swedging them, but instead I made some old ones narrower by pressing them flat with an arbor press. I used a plate of harder metal to stop the ram at the right thickness.

I checked the resulting weights at the front with a couple uniformly spaced sample weights for four different pairs of neighboring natural and sharp keys and just drew lines connecting the positions I had marked out. I made a little jig to drill the outside holes for the leads, with a cut off nail to hold the jig in place, and opened the wood between them with a sabre saw chucked in the jigsaw and squared the sides with a chisel. I pressed the weights in with a drill press and snipped the excess off with some flush cutting shears. I used a third weight placed much closer to the balance rail to bring the keys to the final weight, but positioned them individually to make up for the differences from the widths of the keys.

Step 15: Bench Regulating: Keys

All the different movements of the piano action have to be adjusted to take up variations in the parts and positions so that they all work evenly. This important stage is called regulating and happens to be missing from the online copy of Davies' instructions. Different kinds of actions have different things to adjust but first step usually involves adjusting the keys.

I glued the front edge of the backrail cloth onto the rail and stuck a cloth punching on each balance and front pin, checked to make the natural keys so the fronts and tops were square, and then used a small fine triangular needle file to enlarge the key mortises in order to make the keys all move freely, taking only a couple passes at a time before checking the key on the keyframe. I made the front mortises a little looser than the balance mortises.

In order for the regulation to be durable, and to reduce the amount of noise the keyframe needs to be fit, or bedded to the piano in the exact position it will occupy. In uprights and squares it is usually screwed down in a bunch of places so it follows the contour of the keybed. In grand pianos it usually has to move sideways when it's installed, and besides carefully lowering high spots by hand there are some mechanical methods of making it fit.

It's more accurate to regulate the action inside the piano, but in grands and squares it's cramped and difficult. I rough levelled the keys by clamping the high spots of the keyframe to a flat table and propped a straight edge on a pair of equal sized parallel blocks to either side of the front of the keys. I made the blocks so the straight edge was above the level of the keys and so when they were raised to that height they would go down a little more than necessary. The gaps between the straight edge and tops of the keys are bigger than the thicknesses of the paper punchings used to close them, so it's easy to check if they're much too thick.

I levelled the naturals, and then readjusted the blocks and levelled the sharps and each set took about three passes. You have to make sure the keys are all the way down on the pins, and for a final regulation the paper punchings are supposed to be placed underneath the cloth ones. There's a special forked tweezer made to lift all the punchings together. William Braid White pointed out that it's more rational to make large increases in height to groups of keys by shimming under the balance rail.

Step 16: Bench Regulating: Action

I fitted the action brackets to the keyframe so they were square and located the holes for the mounting screws for using transfer punches and drilled them by hand. I had to file a little at the front of the ledges on the brackets so that the underhammers and hammers were in the proper mechanical arrangement with the hoppers and keys. The hammer rail is also supported in the middle by a stud that passes between two keys and is fixed to the middle stretcher, I used transfer punches to locate the necessary holes as well.

Then I rough regulated the action. Piano tuning teacher Oliver Faust gave instructions for regulating the "old English square" action and illustrated one made in Boston in the 1830s, but I didn't end up following them very closely. I set it with about 1mm lost motion between the hoppers and underhammers, enough that the keys could return to rest even moving slowly, by moving the underhammer rest rail, which is the unmarked narrow piece screwed to the underhammer rail in Faust's picture. I adjusted the heights of the hammers above the underhammer rail by a little more than the thickness of the cloth strip that would be glued between them by either filing or gluing shims on the bottom of the wedges depending if they were too high or too low. I adjusted the hoppers so they flipped out from the underhammers when the hammers were a little short of the height of the strings above the key bed. When the hammers reach this height my keys go down a little over 8mm.

I removed the rails, and relieved and clothed the ends of the keys for the damper levers which meet them at angles. I pressed in the backcheck wires and glued on their coverings, and then set the key dip to 8.5mm, both by adding paper punchings on the front rail and by gluing a cloth strip the right thickness to the bottom of the underhammer rail. A gauge block made as thick as the required key dip and with a lip kind of like the cauls for gluing ivory is helpful selecting the right punchings.

At this stage in a factory the keyboard and action might be reassembled and delivered to another department for trimming and finishing. I haven't reached that step yet.

Step 17: Fitting

In order to work properly the keys have to line up with the hammers and strings, as well as fit without interference in the space provided for them in the piano case which means they have to be located back and forth and up and down as well as side to side.

The keyboard's position is much less critical in uprights than it is in squares and grands where it affects the sound and function of the piano. Their actions are attached to the keyframe instead of the piano case and are much more difficult to adjust for the best mechanical combination when they're assembled than uprights, though the individual side to side positions of the hammers are easier to adjust, at least in smaller instruments, by turning the hammer flanges using a long handled tool that has a socket that fits the sides of the flanges.

My keyboard will be positioned using small blocks that are tacked in place and glued on. Two of the ones at the back that determine how far back the keyframe can be pushed are stepped and angled to match the rabbet at the back of the keyframe - these are positioned to get the best sound from where the hammers hit the strings, and long flat ones will meet the sides of the action brackets to get it in the same position sideways.

In most pianos the keywell is noticeably wider than the keyboard, and the extra space is filled with decorative blocks a little taller than the keys held in place with concealed screws. In grands these are what positions and guides the keyboard, and sometimes the fallboard hinges to them. In squares like mine the blocks are really the sides of the keywell which were planed to size and glued in place after the keyboard was positioned.

The keywell is also deeper than the keyboard, and the extra space in the front is filled with a long and narrow wood strip that matches the case. In uprights and grands the fallboard usually rests on it when it's closed to cover the keys, and the slip is thick enough for mounting a small lock and is held in place with several large screws through the keybed. In squares like mine it's much thinner and glued to the front of the keyframe so that it helps with alignment, in other ones it's pressed into a slot in the keybed. It's tall enough to disguise the bottom edges of the keys but not so tall that when the key dip is adjusted to the proper amount they hit it or go below its top edge.

The nameboard or part of the fallboard usually is padded on the bottom edge and works like the organ keyboard thumper shown by Audsley and M. B., and grands have an additional rail screwed down in a few places so the keys don't get jumbled when the piano is put on its side for moving. My nameboard is curved so once the keyboard is fitted I have to mark the outline of the nameboard on top of the outside sharps at both ends and cut them shorter to make clearance.

Step 18: Finishing

Once the keyboard is fitted it only has to be regulated more precisely and voiced to make it ready to play. The rest of the piano has to be completed, the dampers installed and regulated, and the strings stable and freshly tuned.

The action has to be adjusted closer to the piano than it was on the bench, but it involves the same things. The keys may have to be re-levelled in the piano, and since mine doesn't have keyblocks I'll have to shim the outside keys like Braid White describes, and if the height of the keys changes the lost motion may have to be reset, too. The hammers will have to be spaced perfectly with the strings, and possibly squared up, which is done by heating the shank briefly with a lighter and twisting the hammer so it's straight. Let-off can be adjusted accurately gauged against the strings, but then the backchecks have to be adjusted with the action removed.

When the action is ready the hammers can be voiced to make up for any note to note differences in springiness, hardness or weight so that they don't sound different from their neighbors. Skillful voicing can give a piano maker a lasting reputation, but Wolfenden warned that it can be more disappointing than any other part of the trade. Braid White described that ever since felt was used for hammer coverings voicing has been done using needles to manipulate its compression and tension, but my hammers are made with different layers and will have an outside layer of new leather, so I might not use all the needling techniques Nalder described. It may be easier to harden them without the deterioration they describe from ironing (or else by lacquering), though, by using different thicknesses and tensions in the leather.

I think at least in old factories finishers were the most experienced and skilled workers, and sometimes they signed or stamped their name on top of one of the keys, but in small letters and somewhere not visible when the piano is all assembed. Piano manufacturers aren't usually so restrained though usually the player only has to see the name in a couple predictable places.