Friday, December 13, 2024

Jouhikko project part 3: construction (temporary, no pics yet)

Constructing a lyre via hollowing isn’t a hard process on paper - there aren’t a great lot of steps, broadly speaking. And it can be as hard or easy of a process as you like, depending how intent you are on good results.

I am very intent on good results.

With the large design questions settled beforehand, I went over to my friend’s garage and got to work. The first step is cutting out the general profile on a bandsaw. Since making fine turns with a bandsaw wasn’t going to work, I radiused all of my inside corners with the drill press. On the three string, I left the wood inside the yoke in place for the time being - I wasn’t going to cut through the head of the instrument, and no jigsaw was available. For the four-string, I meant to do a tongue and groove joint anyway, so the bandsaw accomplished most of the material removal, save for some remnant at the shoulder.

After that, reducing the depth of the arms would give me handholds while doing everything else. As my boards were too wide for the bandsaw, I did this with a forstner bit in the drill press, by drilling from the backside. I couldn’t get exactly to size, because the bit has a spike in the center. That meant I had to stop just short of the spike leaving a hole in the arms when I was done, which left at least an eighth inch of extra material clinging on for the time being.

After this, I needed to define the slope on the back of the instrument, so I re-traced the angle and removed the relevant material with a belt grinder. Since hollowing was dependent on the drill press, having the exterior shape defined first meant that I could make the inside and outside parallel reliably.

The last and most time consuming step that I got done for both instruments that weekend was the rough hollowing itself. I started with a forstner bit, as with the back side of the arms. The table on the drill press was set to roughly 1/4” down, so as to avoid the aforementioned hole problem. The four-hole jouhikko was actually thicker than the maximum throw of the drill press, so my initial drilling was simply as far down as possible. The second drilling targeted roughly 3/16”, but this time, the forstner bit had its tip ground off. This caused a fair bit of wander until it found purchase, but it made a mostly flat bottom, save for a faint ring around the outside and a raised dot where the cutting surface was gone. Doing this, I averaged five gallons of wood shavings per instrument, as measured by overflowing the scientific precision garbage can.

On our second day together, we went over to a table saw and engaged in joinery cutting, where we cut the tongue and groove joint for the four-string and broke down of a spare maple board. Parts from that would be useful for pegs later.

At home, the longer process of refinement could begin. I took the jigsaw and cut out the last piece from the three-string, then looked to the joinery on the four-string. Unbeknownst to me, a lot of tension had been released from the wood, so the arms had moved together by roughly 1/4”. So, I trimmed the yoke to fit… and discovered my mistake. Fortunately I only had to buy a little extra wood to cut a new yoke, and my caliper and router work was precise enough that it fit perfectly. Almost makes me feel like a real woodworker.

More work on a bench sander squared up the outside of the bodies and removed the worst of the 60-grit texture. This handled the flat areas, but the bottoms of the arms had jagged excess from the forstner bit. This cleaned up with the router, as did the outside radius on the yoke. It also cleaned up the inside of the body beautifully up to an inch down. That’s as far as my 1/4” shank tool would go.

Special attention was required for the base and inside of the arms, and the tailpiece hooks. The arms took a half-round file to flatten out the imprecise cut at the base of the arms and keep the curvature at the base of each arm consistent in all dimensions. The file was also the best way to keep the insides of the arms straight.

I cut out the bottom (and on the four-string, the top) of the tail hook with a hand saw and did the rest of my contouring with hand files. The round file gave a nice profile for the actual hook portion with minimal extra cleanup.

To get the back face down to thickness, I put a straightedge across the face and measured down with a caliper. Both instruments were quite thicker than I figured necessary at this point, so cleaning up the last of the sides would go hand in hand with stock removal from the bottom. I set my cleanup depth in several locations with a small drill bit with tape around it - once the hole disappeared, I would be done sanding. Once again, the file sander was the best tool for the job, since anything that removed material faster didn’t fit in the space.

My original bracing was made from maple, until I read that it was best to match the wood that the soundboard would be made from. I ended up with extra cedar left over, so I carved bass bars and soundposts from the leftovers of the same boards on each instrument. The soundposts were hand-carved to rough circles about as thick as the soundboards were wide. The bass bars were whittled and sanded to a somewhat traditional shape, with a bump in the middle and a gradual taper to the ends, so the greatest stiffness would be right around the bridge.
 
My sound holes weren’t especially scientifically laid out, save that I had roughly the right clearance for where the bridge would go and space for the bracing. I went with a layout vaguely resembling the original on the Trossingen lyre, and did something a little different for the sake of difference on the Sutton Hoo - no surviving reference on that one, so I assumed that it would be roughly like its contemporaries.

Not all of my original maple bracing went to waste, as the maple jouhikko ended up a little thin in the back. When I found I could flex it slightly with my fingers, I was worried that it wouldn’t support the soundpost. A maple bass bar added to the back fixed that.

I had a particularly new experiment for this project in that I decided to use hide glue for assembly. Reports online said that regular wood glue dampened sound and hide glue didn’t, and if worst came to worst, I could open the instrument again. Rehydrating the glue to the correct consistency was an experiment, but overall I didn’t notice any of the smell that I had heard about, and the glue was easy to use.

I did cheat slightly and glued my soundpost to the top board after experiments trying to wiggle it into place through the holes went poorly. It also helped me set the length of the posts correctly. I found out afterwards from Reddit that my bass bar placement was technically backwards - it should be parallel to the strings. The way I installed it, it would be… on a violin, where the strings come together on the fingerboard instead of fanning out between the arms. That’s something to learn from on my next instrument, since I plan to do one more eventually. Meanwhile, both of my lyres repeated this mistake, since I glued them at functionally the same time.

Once the lid went on, I was able to file the edges flush and sand out all of my tool marks.

I copied my peg spacing from my existing instrument and drilled the pilot holes on the drill press. My laurel had a spare set of tapered peg tools, so I was able to ream the holes out with those. The taper lets the friction pegs work, setting their tension by pressing them manually into place. I carved the pegs from maple for both instruments. I tried mahogany at first for the 3-string, since interchanging the woods would be pretty, but the mahogany broke in half during the test fit, with no string attached. So that was a bust, and I had to start over.

After sanding the instruments to 400 grit, this is the point where the posts diverge, because I finished one instrument long before the other! Stay tuned for Finishing Out the Jouhikko.

No comments:

Post a Comment