Baseboard: Making the breadboard edge

After picking up a rabbeting router bit set at the Home Depot, I cut a half-inch wide tongue on the front edge of the bottom. Next, I ripped a strip of leftover red cedar to a 1" width and, after a bit of trial and error, routed into it a groove to match the width and thickness of the tongue. Here's a photo of the pieces, dry-fit together:


This strip is known as a breadboard edge and is used on large panels such as tabletops to hide the end grain. Here it provides a suitable surface for gluing on the case mouldings that will come later as a decorative embellishment.

Note that the bottom and breadboard are both still oversize at this point, so they need not be of identical length. The pencil tick on the light-coloured board (slightly below the X on the breadboard) marks the cutoff point that will establish the final width of the baseboard.

Baseboard: Trimming the front edge

After my two workers finished cleaning up the dried glue, I took the baseboard outside and sanded both faces with 100 followed by 150 grit paper. I used a random-orbit sander, which also helped to take out a few very minor irregularities in the joints between boards. Both my father and I agree that our work joining up the boards was very good: the sander needed to dwell on only a few spots to get everything level.

Next, we clamped a straightedge along the front, making sure it was perpendicular to the long side, and crosscut the front edge straight with a handheld circular saw. The process of biscuit joinery allows some slight sideways play in the boards when they're glued up, so the ends of the boards are usually a few millimetres away from being in a perfect line. As long as you cut the ends off, all is well and a straight edge is restored.

Additional personnel

This morning, as I was finishing my coffee, I heard activity downstairs and discovered my father, of his own volition, scraping dried glue off the joints in the bottom assembly. My mother was feeling for roughness and pointing out places he'd missed. So it seems I now have a supervisor (my mother) and an employee (my father). Thus far neither has said anything about salary...

Baseboard: Gluing together

All the boards making up the bottom were jointed cleanly on one edge, ripped along the other edge and jointed again. The final orientation of all the boards was decided upon and all were pencil-marked to make sure they didn't get flipped onto the wrong face or turned end-for-end by mistake.

The boards will be joined with glue and biscuits, which are little football-shaped splines made of beechwood that go into slots cut on a board's edge with a special power tool.


Biscuits help reinforce the joints and make it easy to align the edges during glue-up. Without biscuits the boards, slippery with glue, would slide around quite a bit while the clamps were being put on. An advantage of biscuit joinery is that minor bowing of the boards along their length can be taken out: the biscuit slots are all cut at the same position measured below the boards' upper faces, so the faces are all pulled into the same plane during assembly.

To start with, the four right-hand boards were glued up:


After this assembly sat in the clamps for a few hours, the remaining two long boards were glued on:

Wrestplank: Gluing together

The wrestplank (or pinblock) is also an important structural component in all harpsichords. Not only must it hold all the tuning pins tightly—although not so much that they can't be turned when the instrument is being tuned—but it must resist the pull of the strings and not get dragged backward towards the tail of the instrument.

The original Trasuntino has a walnut wrestplank 1.75" thick. Quarter-sawn wood would be helpful here, because its grain orientation allows the pinblock to expand and contract in the same direction as various other structural parts of the instrument.

Walnut doesn't seem to be available anywhere quarter-sawn, or perhaps only by special order, but that's not a problem. My friendly harpsichord maker told me to just saw up a bunch of slabsawn strips, stand them on edge and glue them up that way. Here are the raw materials:



The net result will be a block made of 2"-wide sections standing on edge. After planing to final thickness, the block will be covered with a veneer of soundboard material, as often is the case with many instruments, so the patchwork appearance of the multiple strips will be completely hidden.

I glued the strips up with polyurethane glue, which is quite strong and supposedly won't "creep" (i.e. shift slowly under prolonged stress) like conventional woodworking glues. This was my first time using it. It seems to work well, but has this habit of foaming dramatically as it cures. Not long after clamping everything, this beige sauce-like stuff starts oozing out of all the joints. It's supposed to scrape off easily when dry. I glued things up in two sets to start out:


Note that the leftmost set has strips of different lengths. This wrestplank is trapezoidal, not rectangular, so one end is wider than the other. It was more economical of material to not make all the strips full width. I used the same technique as I did for the baseboard in making these strips, by laying them down on the drawing right onto the outline of the wrestplank.

The final glue-up of both sets of strips:

By the way, the polyurethane glue does indeed scrape off when dry: a chisel seems to be the best tool to get it off, but it took a fair bit of work. You can see from the photo immediately above that I learned to apply it more sparingly the second time around. The squeeze-out I experienced the first time was a bit much. I have to say I'd be leery of using this glue on joints that would be easily visible in the finished product because the foam spreads out sideways much more than squeeze-out from regular glue.

After scraping the glue off, the wrestplank was planed on both sides to its final thickness, and the left edge was cut cleanly. I've left it long on the right side for now.



A straight taper will be cut along the back edge.

Baseboard: Selecting boards

In many harpsichords, the case sides are a major structural component made early on, and the bottom gets nailed to them much later. The casework of such instruments forms a rim that other components are attached to, so the boards it is made of must be strong and fairly thick. The bottom, or baseboard, of an Italianate instrument (this category includes some south German instruments made along the same lines) is, on the other hand, a major structural component. Various items are attached to the bottom to make a kind of inner frame, and thin case sides are glued all around this frame.

My instrument will have its bottom from that western red cedar I recently bought. A&M resawed it to slightly over 3/4" and our planing extravaganza got everything down to around 5/8" thick. This is thicker than some historical Italianate bottoms (only 1/2" on some instruments), but my research suggests that western red cedar isn't as stiff as other softwoods, so a bit extra can't hurt. It's pretty light material, and is said to be quite stable.

To cut the boards to length, I simply placed the first board on the left edge of the drawing and looked to see how much of the drawing was left exposed. Each succeeding board was then measured from the front edge of the instrument to the bentside line on the drawing where it disappeared under the existing board, cut to that size plus about 4" extra for safety, and placed right next to the previous board.

The process continued until the outline of the instrument was completely concealed.

To make sure I would have enough material, I estimated the area of the baseboard by calculating it as a triangle sitting atop a rectangle. This area was divided by the width of the boards I could get and translated into a linear measurement, as if of one enormously long board. Then it was simply a question of getting enough individual boards to sum up to that one imaginary mega-board. The only thing to be careful of was to ensure that the leftmost boards were the full length of the instrument (about 92").

August 24: work begins

August 24: an auspicious day. After months of research and musing, the project transitioned from an intellectual exercise into actual hands-on work.

My father and I spent 2 hours in sweltering humidity planing all the cedar and walnut down. It was nice to discover how attractive that wood is once all the saw marks and outer roughness disappeared. We generated an incredible quantity of shavings: a thigh-high Shopvac canister was filled up 1.5 times over! No wonder serious woodworkers talk about a 35% waste factor when going from rough to dimensioned lumber.

The irony, of course, is that all this lovely wood will be almost totally hidden once the instrument is done. Only the underside of the bottom will be seen when the instrument is set on its side for transport.

Getting under way

To cut a long story short, I have done a great deal of research and hard thinking since last February, when the idea of actually making an instrument became appealing. I've lain wide awake from bedtime until 6 a.m. or later on a number of sleepless nights, my mind churning away on various thorny problems. My notes to self on the construction process occupy more than 18 pages so far (admittedly, some of them are reminders of basic woodworking things so that I don't do something foolish).

Fast-forward to August 16, when my father and I visited A&M Wood Specialty, an excellent supplier of all kinds of quality lumber. We came back with about 35 feet of 5-6" wide quarter-sawn western red cedar, two large walnut boards about 9" wide and a chunk of Douglas fir.

A few days later I picked up 8" and 6" wide poplar boards from the Home Depot. I take a dim view of the lumber they have on offer, but this poplar is purely structural and no one will ever see it. Plus it saved me having to get oversize poplar boards from A&M.

There's plenty of wood still left to buy, but this initial batch will help get things started.

Acknowledgements and caution

Before getting to the meat of the matter, credit is due to my principal sources of information:

• First and foremost, a certain harpsichord maker in New England whom I've known for more than a decade. He is of a retiring disposition and I don't know if he'd want me to put his name down here. Over the past six months he has been exceptionally generous and patient in answering a blizzard of emails I've sent. He also put me up when I came to visit him for a few days in August, answered still more questions, looked at my drawing and gave very helpful advice. At no time did he ever suggest that I was foolish or incapable of doing something this ambitious. Thank you, thank you.
• The book Cembalobau: Harpsichord construction by Martin Skowroneck. This legendary builder's reminiscences and observations about harpsichord building make up the book, which has the same text in both German and English. Available from amazon.de (the German-language site only; the book won't come up on the American site).
• The booklet Making a Spinet by Traditional Methods by John Barnes. As far as I am aware, this is out of print but the university library has it sitting forlornly on the shelf. I may be the only person that has ever taken an interest in it.
• The online harpsichord mailing list, HPSCHD-L. It's been going on for well over a decade, and although I haven't read it for a long time, I started doing so again back in the springtime. A forum for musicians, music lovers and harpsichord makers, it has searchable archives that have provided help, as well as a lot of conflicting opinions, on various matters. With judicious reading, it's possible to arrive at some kind of majority view on the questions that I've had during my research.

Finally, a caution: if you haven't realized this yet from the tone of the blog, I am not a professional instrument maker or woodworker, and my statements should not be taken as guaranteeing any success in this endeavour. I am documenting my thought process and work progress honestly, but I cannot say that I won't get into trouble as I build this instrument. I will rewrite the blog if I discover that something I said earlier led to difficulties later, but be warned nevertheless.

Welcome

Welcome to this online record of my project to build a harpsichord completely from scratch.

That means that my instrument will be built almost entirely from raw materials. It is not a kit such as those made by Zuckermann Harpsichords or Hubbard Harpsichords, by which is meant an instrument that is made of accurately cut parts sold largely disassembled to the buyer, who then puts everything together.

With few exceptions, all the wooden parts of the instrument will be made from rough boards that I will purchase and mill down to the appropriate dimensions. The exceptions will include items made from certain exotic woods and the critically important soundboard, which will be supplied as a series of boards about 1/4" thick and around 6" wide. I'll still need to joint and glue them up but at least I won't be resawing 8-foot long boards: we don't have the facilities for that here.

Speaking of facilities, I'll be using my father's woodwoorking equipment, which is spread out between the basement and the garage. In these two locations, there's pretty much everything one could need (except for major resawing, as noted) as far as large power tools and various smaller hand tools go.

Metal parts are a different story: I'm not a blacksmith, so I'll be ordering tuning pins, bridge pins, hitch pins and so forth from the appropriate sources. Most of the professional makers pretty much do that anyway.

This project is based on the 1531 harpsichord by Alessandro Trasuntino (see the photo in the sidebar), which is currently in the collection of instruments at the Royal College of Music in London, England. I obtained a full-size drawing of the Trasuntino which features views of the instrument from various angles. The internal details have been deduced from X-rays and by peering through cracks and such.

I intend to focus this blog solely on making this harpsichord. As such, I don't think I will add anything more to my profile, such as my real identity (although there are not that many harpsichordists in major Canadian cities, so someone could find me out). A few people will know who I really am, but apart from that I don't need the supposed glory of having my real name out there on the Internet. I probably won't even appear in any of the photos that will soon be forthcoming.