Archive for September, 2015
Five String Progress Report #15: Varnishing
I did not take any photos of this process…it only consisted of rubbing a gypsum suspension into the wood, all over, then rubbing it off, as completely as possible. The goal was to fill the wood pores, so as to limit the penetration of varnish into the wood. Apparently the varnish tends to dampen vibrations, and deaden the violin sound just a little. I don’t know how much effect it really has, but I got the idea from Roger Hargrave, who said it made a difference in his work. If it is good enough for him, I am game to try it. My suspicion is that my ears are no longer good enough to hear the difference, due to a lifetime of work in heavy steel. But…I have had very good reviews on the few instruments on which I have used the ground, so…either it really helps, or I have just improved my building skills, lately.
My sealer is pretty simple: it is raw pine pitch dissolved in a combination of pure spirits of gum turpentine and alcohol. It penetrates pretty deeply, but the alcohol and turpentine (in that order) evaporate rapidly, and leave the pine resin in the wood. I rub off as much as I can of whatever is on the surface, and let it dry in the sun. Sometimes I have hung it up to dry indoors, but it does make the house smell of turpentine. My wife has very little sense of smell, and I don’t mind the smell, but others might, so I try to limit that sort of thing.
Once the violin was very warm, and quite dry to the touch, I rubbed it down with a paper towel, to remove any residue, then coated the back with a yellow varnish, and the belly with a brown varnish. I hope to even out the color somewhat and diminish the sharp contrast between the dark, curly Koa and the nearly white Sitka Spruce.
Anyway, that is about as far as I expect to go, today. I may get another coat or two of varnish on there, this evening, but it will not be an appreciable visual difference until it is done. I expect to use at least another six coats before calling it complete.
Thanks for looking.
Final Varnish Preparation:
An apology in advance: as I warned some time ago, I have a tendency to get out of “photography mode” and just pursue the tasks at hand, then suddenly realize that I was supposed to be doing “show and tell”. So, (sigh…) this time it will be more “tell” than “show.” But I do have some photographs of the tools involved.
Cutting the Purfling Channel
The first step in cutting the channel is to determine its boundaries. I usually use a compass to scribe a line 1.6 mm in from the raw edge of the plate, all the way around, including the corners and the ends, where the channel has to follow the purfling away from the edge. (This time I used a special tool, made by Jake Jelley, to do the same thing. I think it worked better.)
That marks the outer edge of the channel, as well as locating the crest of the finished edge. I extend that line at the same distance from the purfling at the ends, so, at the ends I have two lines: one forms the crest of the edge, the other the edge of the channel. The dark Koa wood did not easily show the pencil mark, so I had to scribe firmly, and usually several strokes.
I then cut the channel with a sharp gouge, trying to keep it shallow, but following the scribed line all the way around. I used a larger gouge for the upper and lower bouts: a smaller one for the C-bouts.
Scraping the channel and fairing the curves
Then I used scrapers to “fair in” the channel with the curve of the arching, and make sure there are no humps or hollows.
Finally, I use a tiny plane and a file to shape the outer edge, curving smoothly from the scribed “crest” to the outer edge, where, hopefully, it will smoothly join the curve from the inner edge.
Final Neck and Scroll-Work
I also double-check the scroll and neck shape and contours. They must be scraped to perfection before I can move on. I use a template (copied from Henry Strobel’s books) to check the upper and lower shape and size of the “handle” portion of the neck, and then try for smooth transitions between the two.
Final Varnish Preparation
At last I am ready for final varnish preparation. Everything has to be as perfect as I can get it, because every imperfection will definitely show up under the varnish. The tiniest blemish will show up like a neon sign once I begin the varnishing. Some people insist on only using scrapers, but at this point I feel fine about using very fine (400-grit) abrasive paper to remove the tiny blemishes.
I removed the fingerboard for varnishing…it was only temporarily glued in place, originally, to aid in the neck setting procedure. While varnish is drying, I will shape the underside of the fingerboard, and lighten it, to enhance tone and projection.
After the varnishing is completed, I will glue the fingerboard permanently in place. I will also install the saddle, nut and pegs after varnishing. At that point, the violin will be essentially complete, and set-up is all that remains.
So: I’m sorry I missed the “in progress” photo opportunities, but, from here on out, it will be just progress reports in finishing. The mineral ground is next, and then the sealer.
Thanks for looking.
Five-String Fiddle Purfling Weave
Over and Under Illusion
Some makers, especially those making violas da gamba, Lutes, etc., make much more complex purfling weaves. Some of the Celtic designs employ the technique in very sophisticated ways. The point is to make an illusion of 3-D “over and under” weave in the purfling. As far as I know it has zero effect on tone; just appearance.
When I left off, last post, the purfling groove was nearly complete, but not quite: I finished picking out the last bits of wood in the “fleur de lis” areas, then went all the way around checking for depth and width.
My purfling bending “iron” is an old-fashioned solder-iron affixed to a brass cylinder with various diameters. I don’t know who made it…I got it from my friend Jake Jelley.
Starting with the completed purfling groove, I first cut and bent the center-bout purfling strips, and inserted them into the grooves, making sure the mitered ends were all the way into the corners of the “bee-stings”, as the sharp miter-ends are commonly called.
Then I cut and bent the long upper and lower bout strips, and fitted them carefully into place, jamming them tightly into the miters at the corners, and trimming them to fit exactly at the other ends.
Then I began gluing the long strips in place, so that they would stay put while I installed the short ones. I tipped the center strips up and inserted hide glue in the groove, then pushed them back in place, and forced them to the bottom of the groove, so that the glue was squeezed out all the way along each strip. Then I repeated that procedure on all the upper and lower-bout purfling strips. Afterward, I could begin work on the “fleur-de-lis” designs.
You can see that I had to decide, initially, which strip goes “over” and which goes “under”: In reality, of course, they are all at the same level, but, choosing which gets cut off (thus looking as though it goes “under”) and which goes on through an intersection (thus appearing to go “over”) determines which way the “weave” seems to go. Once I pick a direction, I need to pay close attention to see that it continues with the “over and under” look, to make the “weave” illusion appear correctly. I also try to make both ends the same way (starting “left over right”, for instance).
All that is left on the back, now, is the channel, the edgework, and final scraping. We are officially “on the home stretch!”
Thanks for looking.
Five String Fiddle Back Plate Purfling
I use the same purfling marker to begin the layout of the back purfling as I did on the front purfling, except that, as I have a habit of using a “signature” fleur-de-lis on the upper and lower ends of my five-string fiddles, I have to stop short of the corners and ends, and sketch those areas in by hand. However, I had noticed that, since I have literally been sketching them in by hand, no two were alike, and they were pretty time consuming. So, today I made a small template out of tag-board…a junk-mail offer for some thing or another, that just happened to arrive at the time I needed such a thing. (Serendipitous, that….)
I used the purfling marker to lay out everything except the corners and ends, then used the template by poking through it with a needle, to lay out the ends, and sketched in the corners with a pencil and knife.
And, now I am ready to cut all my purfling grooves, pick them out, and begin installing purfling.
Cutting the groove and picking out the waste wood.
This part is hard on the hands. Some very good luthiers, today, now do this part using a dremel tool, but I tried it a couple of times and had some rather nasty accidents. I reverted to cutting the grooves by hand. It is hard on my hands, but I end up doing better work. I just have to take breaks now and then.
Something I had to bear in mind on this fiddle, is that the Koa grain is so curly and wild that I could have no confidence that the purfling pick would not chip out a larger piece than I intended. So, I had to move carefully, and take small “bites.”
Also, inlaying the “purfling-weave” (the fleurs-de-lis) was risky, as the graduation was already complete, so I did not have lots of extra wood to work with. I had to make sure I did not cut too deeply. I worked carefully, and took my time, and got through the challenge without mishap. Aggravated my arthritis somewhat, but that is OK, too; I will just take a break for a day and do some other things.
The chimney needed to be cleaned, and commercial cleaners refuse to do it, as they say our roof is so steep and high, that it is too dangerous. (sigh…) So we bought a set of chimney brushes, and, every year, we do it ourselves. That took a few hours Saturday morning. We heat with wood, and it is important to clean that chimney every year.
But in the afternoon and evening, I went back and got back to work on the fiddle. Section by section I sliced along those marks and cut the grooves as deep as I thought I needed them, then began picking out the wood from between the cuts.
Anyway: I think that is about as far as I am going to get, this weekend. I will try to finish the purfling by Wednesday, get the heel and neck and scroll at the absolutely finished level, then start doing all the final edgework, and prepping for varnish.
Thanks for looking.
Five String Progress #11: Back Plate completed and installed.
The inside arching took a lot of time and energy, but it is just part of the job. Once it begins to get closer to completion it is a lot more encouraging, but initially, it is just a lot of work.
Bit by bit, however, the project begins to take shape:
Here, you can see the curvature of the plate, but you can also see why I have to stop using the planes, relatively early, and revert to scrapers: the curly wood tears out badly under the plane.
I also have to measure the thickness over and over, using a graduation caliper, so as not to cut too deeply. This process is called “graduation”, I suppose, because the thicknesses have to gradually change from area to area. They are not entirely symmetrical, but there is a general plan and some practical limitations.
Now we work with scrapers:
Final inside preparation for gluing:
I almost forgot to take pictures! At this point the plate is complete, except for purfling. In the past, I have always installed the purfling before attaching the plates to the corpus, but on this instrument I decided to try purfling after plate installation. I think I like it. I have better control of my edge over-hang in terms of both size and shape.
So: the inside edgework has been done, the label is installed, and the plate is ready to be glued in place.
What I do, nowadays, is to carefully dry-fit the plate to the garland so that it is exactly the way I want it, clamping securely over all the blocks (about eight clamps in all). Then I remove a couple of clamps at a time and slip hot hide glue into the joint, wipe it down with a rag and hot water, and re-clamp that area, adding as many clamps as will fit. I work my way all the way around the plate, and never have to hurry, or suffer any fear that something will get out of alignment while I am working.
For the first four of five instruments I made, I would apply glue all the way around the garland then engage in a panicky race to get the plate aligned and all the clamps in place before the glue gelled. Not good. Usually, about that time, the phone would ring, too… (sigh…).
This way is very peaceful, by comparison, but I have learned to be less compulsive about answering the phone while gluing, too.
And here is the completed corpus, with the back plate glued in place and secured with spool clamps:
Once the glue is dry, I will be ready to begin purfling the back plate.
Thanks for looking.
Five String Progress #10: Arching
Completing the back plate arching
When I left off, last time, I was too tired to continue carving, so I took a break and completed other responsibilities for a few days. Saturday, I came back and spent some time carving and scraping:
As you can see, the basic shape is complete. Scraping will be the method of moving wood from here on, on the outside… the inside is still a flat, rough plank. But I continued scraping for a while on the outside before beginning the inside.
Once the plate is essentially the exact shape I want it, (checking with low-angle lights, etc.) I move to finer scrapers– sharpened at 90 degrees, and used gently, flexing the blade to match the curvature of the plate.
Beginning the inside arching and graduation
Now I can flip the plate over and begin carving out the inside of the back. Here is the cradle without the plate. Notice that the plywood cutout matches the shape of the plate fairly closely, while the thick pine board simply supports the plate while I am carving. The plywood is what holds it still, laterally. The spring clamps prevent the plate from flipping out of the cradle.
The back plate has already been marked for inside arching. I will have to monitor thickness constantly, but here it is, ready to carve:
And, the “fun” begins again. This Koa wood is by far the most difficult wood I have ever used on a back…but it has to be done, so, chip by gouged-out chip, here we go:
I will post again when I am ready to install purfling.
Thanks for looking.
Five String Fiddle Progress: Back Plate
Flattening the back of the garland
After the glue was completely dry, I removed all the clamps from the linings and cleaned up the interior: that is to say, I tapered the back linings, so that they faired smoothly into the ribs, and I shaved and scraped the blocks to their final shape.
Then I rubbed the back surface of the garland (rib and block structure) on a “sanding-board” to level the back of the garland, and to ensure that the back of the neck heel was completely level with the back of the garland. That way, I can fit the back plate absolutely flat against the garland, and trace the shape.
I didn’t take a picture of this procedure, this time. Here are some taken when I was building a cello:
I make pencil-marks on the edges of the linings, ribs and blocks, all around, and scrub until the marks disappear. When all the marks are gone, the garland is flat. It gets pretty vigorous and physical, but it is quite effective.
Here is the flattened garland, ready to trace the back plate:
As you can see, the blocks are quite smooth, now, and the linings taper gently into the ribs. After this point, there will be no more changes to the interior of the corpus, except as it directly affects the back plate.
Tracing the plate
I clamped the plate to the corpus, carefully centering the glue-seam of the back plate on the centerline of the neck and the end block.
Then I traced around the garland using a ball-point pen and a flat washer whose flange is exactly 2.5 mm wide, so that my line will be 3 mm from the ribs, all around. I watch carefully to make sure the washer stays flat on the plate…they have a tendency to flip up and follow the pen. I work along the perimeter, giving several strokes to every part, so that the line will be visible against the dark wood. Notice that this procedure makes “round” corners. I will modify them before cutting them out, so that they are the correct shape.
And, there is the plate, almost ready to be cut out:
One thing I do, that I forgot to photograph, is that I re-shape the corners. I use a straight-edge to connect a line from the end of the rounded corner where the pen circumscribed the corner of the rib to the center glue line at the location where the purfling will cross the far end of the plate. (Or, you can just use a 30/60/90 triangle to lay out a 30-degree angle off the center glue line on the end of each corner. But I do it with the straight-edge.) Then I continue the curves of the inner bouts to connect with the straight lines I just scribed in, and the corners are complete. I use a sharp scraper to remove any ink lines that are not part of the perimeter outline, and then I really am ready to cut out the plate.
Cutting out the plate
I use a band-saw to cut within a millimeter of the line, and then use an oscillating spindle sander to perfect the edge, right up to the lines. These and my drill press are pretty much the only power tools I use, though I have occasionally used an angle-grinder with a coarse sanding disc to remove rough excess wood, on larger instruments.
I run my fingers around the edges looking for lumps, and work those out as well, using a file as needed. Here is the completed plate blank:
At this point I also sketch in the interior graduation plan–just the outline of the inside boundaries of the plate, so that I know where I am going to carve. There is no reason it has to be done at this point, but it helps me remember which side was outside and which inside…and that really does matter, as the plates are vitually never exactly bilaterally symmetrical. If I forgot and arched the wrong side, the completed plate would never fit the garland (ask me how I know…).
You can see that there are numerous bark inclusions in the wood, which I will fill with matching wood…but I will wait until the arching is complete before I attempt to fill the holes, so that I do not plane away my plugs. It is unsusual to use wood with holes in it like this, but it is also unusual to use anything other than maple for a violin back. Five-string bluegrass fiddles are not burdened by the same 300+ years of tradition as violins, though, so exotic woods can be used. This wood has amazing flame and figure in it, and the plugged holes will not detract from the beauty of the wood.
Before I begin arching, I establish the edge-thickness all the way around, using a “wheel-style marking gauge“, set to 4 mm. It has a tiny sharp wheel that does the marking, and it makes very cleanly scribed lines.
I used a Japanese-style pull-saw to cut the approximate height of the plate thickness on each of the corners, so as to avoid unnecessary stress on the corners while carving the arching. It looks a little odd until the arching is complete, but it avoids the possibility of breaking off a corner. I use a large, sharp gouge to reduce the edges and begin the arching; then a toothed finger plane to continue the curves until I am very close to the desired shape. at that point I will switch to a smooth, curved-sole plane and bring the arching to nearly exactly the desired shape. From that point on, a variety of scrapers will be my only shaping tools, for fear of tearing out wood along the figuring .
Here are some photos of the progress. This is very hard, difficult wood to carve, and my hands are giving out, so this is about as far as I am going tonight:
So…that is it for today. Pretty fancy wood, isn’t it? The customer bought it in Hawaii 33 years ago, or so, and has dragged it around all these years, until he decided that he wanted a five-string fiddle. Probably the only opportunity I will ever have to make a fiddle out of curly Koa, but it is really going to be a beautiful instrument.
My hands are getting too tired to work effectively. I will try to get more done later this week.
Thanks for looking.
Five string Fiddle Progress: Neck-set and more
Completing the Scroll
Last week, I had left the scroll nearly finished, but still lacking the outer fluting under the neck, and still pretty rough. (Honestly, I continue “fine-tuning” scrolls right up ’til I begin varnishing.) So, the first thing was to get the scroll and neck completed well enough that I could set the neck.
I am never fully satisfied with my work, but I have to decide at an appropriate point that it is OK to move to the next step. However, scroll and neck carving is much more difficult after the neck is set, so I want to have them pretty close to how I want the finished product to look, before I begin setting the neck.
Setting the Neck
Setting the neck begins with careful layout of the heel “footprint” on the neck block of the violin body (often called “corpus”). I already had the centerline of each laid out, so it was a matter of transferring lines accurately, and then cutting along those lines so that when the wood was carved from between the saw-cuts, the heel of the violin neck should fit, snug and straight, into the prepared neck-mortise. I always leave a little extra, so that the mortise is too small to begin with: it is much easier to take off a little more, than to replace wood.
I did not take pictures of this procedure– simply wasn’t thinking about photographs, and I forgot. It went very smoothly, this time, though, and I think I had it ready to fit in around 30 minutes, or a little more. (Usually it takes me longer.)
Here’s the neck mortise, ready to receive the neck:
And here is the neck, dry-fitted into the mortise:
Once I knew that everything fit the way it ought, and that it was going to be straight, tight, and at all the correct angles, I was ready to glue. The things I have to check are five points of measurement:
- The distance from the upper end of the fingerboard (where the nut will be) to the upper edge of the top plate (both sides) has to be 130 mm.
- The neck has to be measurably straight, so that the centerline of the neck and scroll are a continuation of the centerline of the corpus.
- The neck can’t be twisted (rolled side to side)…it should be level with the plane of the ribs, side to side.
- The height of the upper edge of the heel of the neck (underside of the fingerboard) should be 6 mm above the top plate.
- The height of the end of the fingerboard above the top plate should be between 19 mm and 23 mm, with 21 mm being optimum.
I realize that there are different ways of approaching virtually everything in lutherie, and there are sure to be experts reading this, who are shaking their heads, but: this is the way I was taught, and it has worked well for me.
I prepared the hot hide glue, and brushed it into the mortise, and onto the bottom of the neck-heel, and along the edges of the heel. Then I quickly rammed the neck heel home in the mortise, and checked to make sure all my measurements had held (primarily the height of the fingerboard above the plate.) All was in order, so I set it aside to dry. Here is the completed neck joint, with the glue squeezing out around the joint:
After the glue dried, I planed the neck heel flat, then removed the mold by breaking the glue-joint at each block (six places) and simply lifting out the mold. Then I trimmed the blocks and was ready for the back linings.
Installing the Back Linings
The first thing I do to prepare for installing linings is to cut mortises in both sides of each block, into which to insert the linings. I use a thin knife and a very small chisel to cut the mortises.
I make the linings by first sawing the chosen wood to about 2 mm thick, in 2-3″-wide “planks”, about 18 inches long, and then using a wheel-style marking gauge to cut off strips 7 mm wide. I wet each strip and bend them around a hot bending iron, until I can fit them into the corpus. I want them to fit tightly. The center bout linings are bent in such a way that without clamps they would tend to buckle away from the ribs, so I use small spring clamps to dry-fit them. The upper and lower bouts will stay put on their own.
Then, one by one, I remove each lining and brush hot hide glue along the portion of the rib that will receive it, as well as on the lining itself, making sure the ends are liberally coated, as well as the full length and width of the gluing surface. I quickly re-insert the rib, and clamp it in place with as many small spring clamps as I can fit along its length.
That is it for today. I’m tired, and 3:45 AM will come all too early (back to work tomorrow).
The next step will be to level the back surface of the ribs and neck heel, and then trace the back plate from that pattern, so that it will fit perfectly. (The neck heel is still not in the correct finished shape, but the back plate button and the neck heel will be shaped as one piece, after the back plate is glued in place.)
Next time we’ll start carving the back.
Thanks for looking.
Five-string Scroll Carving
Step by step progress
I finally got back to the five-string. Things have been hectic. The boys are in New Jersey, this week, and the daughter is back in Switzerland, so I spent today trying to carve this scroll. This is the hardest wood I have ever carved for a fiddle scroll– not in the least like Maple, which is tough, but not so difficult to carve.
I had left the scroll partially carved, and the neck roughly shaped. Pilot holes are in place. rough-cuts made to outline the scroll.
So I completed most of the neck shape and temporarily glued the fingerboard to the neck.
Then I went to work, shaping the neck and fingerboard as a unit, and continuing to shape the scroll:
You can see that I have begun undercutting the turns of the scroll, and chamfering the edges, but there is a long way left to go. Monday is a holiday, and I intend to spend it working on this instrument, so perhaps I will have more to show next week.
Thanks for looking.