Bridge bending and gluing

With the soundboard planed appropriately underneath, the time has come to glue the bridge down.

In the old days, bridge curvature on an Italian instrument was established at the moment of gluing, by simply forcing the bridge into the proper curve and clamping it. When the glue was dry, the curve would hold. I was tempted to try this, but in practice-bending my bridge I found quite a lot of force was required, even though the bridge did make the curve without breaking.

I decided instead to pre-bend the bridge more or less to the proper shape using dry heat. The bentside of the soundboard acted as a template of the appropriate curvature, which I traced in pencil onto a large sheet of plywood. Next I screwed little wooden blocks at intervals along this pencil mark.

By heating short stretches of the bridge with a 1500W heat gun, I was able to bend and clamp it to the blocks. The amount of heat required was considerable; my sources recommended heating just short of scorching. In fact, some parts of the bridge did get toasted a bit, but since walnut is naturally brown the scorch marks simply look like natural variations of colour.


Once the bridge cooled it held its bend, although wood bent through any means always "springs back" somewhat after it is unclamped due to its natural elasticity. So the resulting curvature was not perfect, but only minimal effort was now required for the bridge to assume its proper shape.

To glue the bridge down, I borrowed a trick from the kit-building world. First I placed the bridge on the pencilled curve on the soundboard and clamped it into place, bending it as necessary to match. Next, safely away from the future locations of the bridge pins, I marked places where I could drill holes right through the bridge. Each hole received a padded nail: a nail pushed through a couple of pieces of thick cardboard. These little pads keep the bridge from being crushed by the nail head when the nail is hammered down, and afterwards a set of pliers can grab the cardboard and pull out the nail.

Before gluing the bridge, the workshop humidity needs to be 45%. Later, when the ribs are glued, the humidity will need to be reduced further.

With nails in each hole, I used the nail points to make sure the bridge was precisely over the pencilled curve, and then I tapped the nails in partway:


I've left just enough room to get the narrow spout of my special glue bottle under the bridge. Note also that a portion of the soundboard rose has already been glued in place. A matching parchment ring will be glued on from above.

When I was satisfied with everything, the nails were driven home, through bridge and soundboard into the plywood sheet, until the nail heads started to compress the cardboard. I noticed that the outside edge of the bridge had a tendency to tip upwards, so to make absolutely sure it was flat against the soundboard, I slid some thin cedar wedges under the board, which forced it up against the bridge from underneath.


The bass hook—the straight part of the bridge in the extreme bass—was glued on next.

After the glue dried, the nails were pulled out and all the nail holes were masked with dark brown wax. Short brass tacks were nailed into the existing holes from underneath the soundboard to provide additional reinforcement for the bridge glue joint.

The bridge pins were located using the same register-holding jig used to mark the bridge pin locations on the soundboard. Pin pricks were made in the bridge to identify the pin locations for both sets of strings, then holes were drilled to a depth of about 10 mm. I used a #57 drill bit for these holes, which is about 0.1 mm narrower than my 1.2 mm bridge pins, ensuring that the pins fit tightly. All the holes were drilled leaning slightly towards the spine, so that in future the strings won't "climb up" the pin but stay pressed against the bridge. Then the pins were installed with a special pin-pushing tool that seated the pins to a uniform depth, leaving several millimetres projecting.

Tapering the soundboard

Now comes the part I've dreaded a little bit since the beginning of the project last summer: the soundboard has to be selectively hand-planed in spots on its underside. This will definitely have an impact on the tone quality and sustain of the completed instrument.

The process worries me because of my lack of experience: a professional would know how to adjust the planing to suit the specific pieces of wood he was working with at the moment. Experience and judgment would tell him whether to remove more or less material, given the peculiarities of the wood he had at hand: no two soundboards would be planed exactly the same way, even though the resulting instruments might very well turn out similarly.

There is no substitute for experience, so all I can do is follow the general principles and humbly accept the whole experience as a learning exercise.

Some of the general principles in thinning a soundboard are:

• Italian soundboard thinning takes place along the outside edges, and some sort of taper towards the edge is aimed for. Having said this, it is possible to find historical Italian instruments in which the soundboard doesn't seem to be tapered at all.
• Mass = inharmonicity. Inharmonicity is desirable in the bass, less so in the treble, so the treble areas will be thinned out more than the bass. This concept was already made use of back when the soundboard was assembled, by choosing "ringier" boards for the bass and "duller" boards for the treble.
• There is a correlation between soundboard mass and the resulting instrument's volume and sustain.

I've taken a preliminary stab at the tapering process with a block plane, first tracing out the bridge curve on the underside in pencil and then planing outward from there with a block plane:

You can see a bit of a shadowy area along the bentside curve, which is the planed zone.

The grain has torn out here and there on one section, which probably means I inadvertently turned that board around at some point before it got glued into the completed soundboard. Luckily this is the underside of the board and a cosmetically even appearance is not important: you can see I didn't work hard at cleaning off dried glue spots either. In historical instruments the underside of the soundboard often shows plane marks and scratches. Many makers feel some irregularity to the board is actually a good thing: apparently this roughness keeps the board relatively neutral and prevents it from overly favouring any one particular vibrating frequency.

I might sand along the very edges of the board later, so that a smooth surface will be present when the soundboard is glued into the case.

Note that I also cut out a hole for the rose (a parchment or metal ornament glued to the underside of the soundboard). I have a charming parchment rose that consists of two layers: a disc glued underneath the soundboard and a ring around the hole at top. The hole was easily made with a drill and a 2.5" hole saw, which gives a hole slightly larger than the top ring. This way the hole edges won't be seen. I decided to locate the hole halfway between the spine and bridge, and twice this distance forward of the bellyrail; clearly there is one exact location that satisfies these criteria simultaneously, and that is where I drilled the hole.

Marking out: Hitch pin positions

With the bridge pin positions correctly located, the hitch pins were marked out. First I decided on the sidebearing: the angle at which the strings would come off the bridge pins. This is necessary in order to bring the strings into firm contact with the bridge. My friendly harpsichord maker pointed out that a shallow angle would allow the non-sounding string tails to resonate somewhat at the expense of part of the instrument's sustaining power, as vibrations would leak past the bridge. A sharper angle would lead to a drier sound. In historical instruments I have read about sidebearings of as much as 17 degrees. The Trasuntino has about 14; I followed my contact's advice and settled on 12. A little after-ring is a characteristic part of the harpsichord sound, which is, on the whole, not too dry.

To locate the hitch pins, I used the same wooden jig that marked the bridge pins but without the register. Since this is tilted at 8 degrees, I cut a plywood scrap off at 94 degrees, so that the extra 4 degrees would sum with the jig's 8 degrees for a total of 12. By lining the jig up with each bridge pin mark and pushing the scrap up to the mark, I could easily see a line going off to the right at 12 degrees. Where this line met the edge of the soundboard I put a little dot with a felt-tipped pen:


I allowed the leftmost 5 strings to go off to the left, instead of the right, in an effort to balance the sideways forces on the bass bridge hook.

Of course, all of these markings will be covered up by moldings that will be glued atop the soundboard against the interior of the case sides. So I will need to transfer them from the soundboard, once it is glued in, to the inside upper edge of the case.

Marking out: Bridge pin positions

In order to accurately locate the bridge on the soundboard's top surface and determine the positions of the hitch pins, I need to trace out the bridge curvature in pencil and locate the bridge pins.

The bridge curvature is a function of the Trasuntino's scaling parameters. In my version of the design, the majority of the instrument follows a true Pythagorean scale with c''=273 mm. This scale is accurate from the top down to c, an octave below middle c'. From that point downwards the strings shorten somewhat to keep the instrument from being inordinately long.

I made up a chart of Pythagorean string lengths derived from this scale, and, working from the treble downwards, I used a long wooden T-square with a tape measure to measure out each string length. I lined the square up successively with the tick marks made along the nut, and at the far end I put a little cross on the soundboard to mark the string length.

Back in September, when I was determining the curvature of the bentside, I mostly used Pythagorean C and F# string lengths to derive the curve and I did not make definitive conclusions about the bass string lengths. The time had come to be more precise, so, starting from c (below middle c'), I experimented with some foreshortened string lengths to see what the far end of the bridge curve would look like. Based on some foreshortening ratios from the original design, I found that the bridge would straighten out and also move leftwards well away from the bentside. I felt I didn't like this look so much. After looking through emails from my friendly harpsichord maker, I was reminded that Italian bridges were often made pretty much parallel to the bentside. In testing out this shape, the appearance was nicer and a side effect was that the bass strings were slightly lengthened, something which my contact had suggested to me could be beneficial to the overall design. The flip side was that the bass hook (the little straight section of bridge going off to the left) would now be longer and would carry about 10 strings instead of 4. This didn't seem to be a problem. I decided to put the hook as far in from the tail as the bridge was spaced from the bentside.

After finishing all this, I realized that I probably could just have set a compass to a fixed spacing and simply traced the whole bridge curve out without all these measurements. It really is quite consistently placed with respect to the bentside curve: no surprise, since the shape was derived from the same data. I have read that Italian bridges were often located by purely cosmetic means, with any ramifications to the scaling apparently secondary to a pleasing appearance. Live and learn, I guess.

I went over the space between each cross in pencil to make a solid curved line showing the bridge position. My plan was to then determine the hitch pin positions along the tail and bentside by going off the straight line of the string at a 12 degree angle and putting a cross right at the edge of the soundboard for the hitch pin.

But, in looking over my T-square, I grew concerned that perhaps it was flexing a bit, particularly at the far end when I was marking down string lengths of some 1800 mm. Any flexing might cause a lateral shift of the bridge pin positions, and my hitch pin positions would only be accurate if the bridge pins were well located. So I decided to double-check the bridge pin positions using the same register that had originally placed the tick marks along the front edge of the soundboard.

I made this jig to hold the register, with a stout oak board to keep the register from flexing and a guide to slide the whole thing along the edge of the spine. The register is clamped so that the edge of the first slot is 37 mm from the left edge, just as it was when the first markings were made.


All I had to do was slide the jig along and see if my original bridge pin marks were aligned with the edge of each slot, like this:


It turned out that most of pin positions were good; a few corrections were needed only in the bass region.

In the photo above you can see that I am working on the second register's pin positions. I only needed to clamp the register slightly to the right to align with the first right-facing string, and then I could mark off all the bridge pins for the second register, which would be spaced 3 mm to the left of the first register's strings (as seen from the nut).



Once the bridge is glued on it will obscure all my pencil marks, so I will actually have to mark out all the bridge pins a second time. I'll slide the same jig along the bridge, marking positions with a little awl, then drill each dimple with a small drill bit and push each bridge pin home.

Hitch pin positions are to follow.

Marking out: Nut and tuning pin positions

Some weeks ago I cut the soundboard to its final shape: this was as easy as clamping it to the frame of the harpsichord and routing all around it with a laminate trimmer.

With the final shape established and the board clamped in place again, the process of marking out the locations of the nut pins and tuning pins commenced. Once again, the registers that were made many months ago acted as the marking stick that determines the positions of key elements.

Before marking out, I had to calculate where to locate the leftmost string in order that the string band be centered on the instrument. This is determined by subtracting the width of the string band from the interior width of the instrument and dividing the answer in half. Back in the autumn, when determining the bentside curvature, I estimated this would be something like 35 mm; it turned out to be 37 mm, which is close enough.

The string band width comes directly from the distance between the first and last register slot, with one caution: a harpsichord with two registers has left-facing and right-facing jacks with strings on either side. Therefore the lowest string is left-facing but the highest string is right-facing. Marking out with the register gives the position of all strings belonging to one register only, not the other. The final right-facing string is located one more register slot to the right of the final left-facing string (a distance of 13.75 mm), minus the string spacing between narrow pairs (which I have decided will be 3 mm), giving a position of 10.75 mm to the right of the final left-facing string.

I clamped one of the registers in place along the back edge of the register gap, making sure the first slot was 37 mm from the left edge. With a pencil I made a little tick mark on the soundboard along the edge of each register slot:


I also put a tick mark for the final right-facing string based on the calculations above. Next, I drew a line on the wrestplank to mark the nut position, which has its centre at 46 mm from the front edge of the wrestplank. With a square, I transferred the first set of tick marks to this line. Then, with a compass set to a width of 3 mm, I marked the position of all the remaining right-facing strings:


Finally, I drew parallel lines 13 and 26 mm from the front of the wrestplank and, using a bevel gauge set to 10 degrees, I drew lines from each tick mark on the nut to intersect these two lines: each left-facing string the first line, each right-facing string the second. The intersection points mark the tuning pin positions, which are shifted 10 degrees to the right of each nut pin. I pricked these holes into the wrestplank surface, shifting pins that represented accidentals slightly forward to make visual identification of the pins easier:


The next step will be to use a large T-square and tape measure to mark out the string sounding lengths on the soundboard, which will locate the bridge, and to determine the hitch pin positions along the bentside.

Making the bridge and nut

The bridge and nut define the sounding lengths of the strings in a stringed instrument. On a harpsichord, both usually have a similar if not identical appearance, and their position on the instrument determines the name: the bridge is glued to the soundboard, while the nut is glued to the wrestplank's top surface. A bridge, therefore, conducts string vibrations into the soundboard while a nut, together with the heavy pinblock, reflects vibrations back towards the bridge.

There is occasionally an exception to this: some harpsichords have their nut placed on a small piece of free soundboard, meaning that the wrestplank ends in front of the nut instead of continuing all the way to the register gap. This is called a "hollow wrestplank", but is not a concern in the Trasuntino design, which places the nut on the wrestplank.

The shape of the bridge reflects the scaling choices made by a harpsichord's designer, and the exact curvature is a function of how closely the design adheres to or departs from a theoretically just scale. I have already indicated that the Trasuntino follows a just scale through much of its range, so all but the lowest octave of the bridge follows a Pythagorean scaling curve. The shape of the nut cannot be too radical as it must fit within the confines of the wrestplank. Nuts range from perfectly straight to slightly curved; typically they are closer to the gap in the treble. The Trasuntino's nut is unusual in being parallel to the front of the wrestplank.

Unlike most other stringed instruments, the harpsichord's strings do not touch the wooden bridge and nut surfaces without first bending around a metal bridge/nut pin. If a harpsichord string were to touch wood first, the tone would be feeble and choked. Therefore the cross-section of the bridge and nut must have a little slope or hollow on the edge facing the sounding portion of the string. This provides a little clearance for the string and prevents it from touching a wooden surface before it reaches the pin.

Hardwoods are the material of choice for bridges and nuts, and walnut is appropriate for Italian and certain other harpsichords (although, interestingly enough, the original Trasuntino actually has a cypress bridge). I started by obtaining a walnut offcut a bit longer than 6 feet. Since this is nearly the length of the finished bridge and I still need a nut of about half that length, I made sure my walnut was of a width that allowed me to work on both its edges, thereby producing two identical pieces. All operations were carried out at the router table.

The first step was to establish the height and width of the bridge: 13 x 8 mm, respectively. I did this by cutting out a centred groove with a straight-cutting bit, making the groove 13 mm deep and adjusting its position until the remaining material was 8 mm thick:


Next I used a portion of a router bit designed for decorating small items like jewellery boxes to cut a profile that included a cove and groove:


The cove will provide the string clearance I discussed earlier, while the groove will receive the brass pins. I deepened this groove slightly with a 1.6 mm straight-cutting bit, and then chamfered the back edge with a 25-degree chamfering bit:


This chamfer makes it easier for the strings to slope downwards on their way to the tuning pins and hitch pins.

All that remains is to cut these parts free at the table saw.

Soundboard in the rough

I've been busy yet again, this time as part of the continuo team in a recording of Handel's Israel in Egypt, but work has still been taking place on and off since my last post.

The soundboard was completed "in the rough" earlier this month, after all the intermediate sections were glued together into the final assembly of 8 boards. After scraping all the joints and examining the trueness of the surface with a metal straightedge, I discovered, not with any great surprise, that the act of scraping had left little valleys at the joints between boards with higher areas in between. The correct way to address this would be with a large handplane, but we don't have a good one, and I'm too cowardly to try this for fear of doing something irreparable. So my father made a sanding block about 10" wide, and, starting with 80-grit sandpaper, he sanded both the front and back face of the board. Using a wide block ensured that the hills would be flattened without making the valleys any worse. Though time-consuming, this approach proved quite effective, and the resulting flat surface ensures that the bridge and ribs can be glued on securely when the time comes.

Here's a view of the soundboard in the rough, with the large sanding block visible in front:


With the off-cuts that came from my soundboard planks, I made a small sheet to cover the wrestplank:


Gluing this onto the wrestplank not only hides the fact that the plank is made up of a bunch of strips of walnut but also constrains the plank's cross-grain movements to a degree. A second sheet has to go on the underside of the plank, since veneering technique requires both faces of a board to be covered, otherwise uneven expansion and contraction would occur.