Varnishing Process for the 16-1/2″ five-string Viola:
All Smoothing and Varnish-prep is done:
When I last posted, the final woodwork had been completed. I had twice wetted down the wood, to raise the grain, and scraped and sanded away the rough raised grain. The wood was stable enough to commence the tanning process.
After the wood is smooth, there will be:
a tanning treatment,
a mineral ground treatment,
a sealer, to lock the mineral ground in place, and
finally, the varnish itself in a series of 6-12 coats, depending on color.
Tanning the Wood
People who live in very sunny regions (New Mexico, for instance) need no light booth: they simply hang their instrument out in the sun for a few hours and it takes on a deep yellow-tan color. I live in Oregon. Western Oregon, between Portland and the coast. We are more likely to achieve a patina of bird-droppings than a sun-tan, if we hang instruments outdoors. (Sigh…)
So, a number of years ago, I bought an old cabinet, about seven feet tall, lined it with aluminum foil as a reflector, wired it with a strong UV source (two 48″ fluorescent UV tubes in a shop-light fixture), and I hang my instruments in it overnight. To heighten the effect, I brush on a coat of very diluted Sodium nitrite and let it dry before I expose it to the UV. This works pretty well, and I have pretty much adopted it as a normal pre-varnish treatment.
Tanned front side of the 16-1/2″ five-string Viola.
Tanned back side of the same instrument.
Mineral Ground:
Years ago, an excellent luthier in Europe posted a detailed explanation of why and how he employs a mineral ground in his instruments, to improve projection. I tried it (because, “if it is good enough for Roger Hargraves…”) and immediately started getting better reviews on the sound of my instruments.
So…obviously, that became part of my process, as well. I use gypsum powder, suspended in coffee (gotta wake up the tone!) so as to achieve a little deeper color in the same move. I rub it in vigorously, trying to get the particles of gypsum to actually penetrate the pores of the wood, then rub off the excess with a rag, before it is fully dry. When it is dry, it obscures the grain, and turns a chalky white color.
Mineral ground drying on 16-1/2″ 5-string Viola. (5-string bass beginning in background.)
Mineral Ground is dry. I will sand off any excess mineral, and then apply the sealer.
Sealer
Sealer coat, front view.
Sealer coat, back view.
The sealer, in this case, is simply rosin, dissolved in turpentine and alcohol. The mixture soaks into the wood, causing the mineral ground to become transparent, then the solvents evaporate, leaving the rosin in the wood. The mineral ground will never again be visible.
Varnish Beginning
When the sealer is dry, I sand lightly, using 320 grit, to remove any lumps I may not have seen, and then I am ready to begin varnishing. I always begin with two base-coats of very yellow/gold varnish, so that the gold color will shine through the darker color coats.
Two coats of yellow varnish, side and front view.
Two coats of yellow varnish, back view.
I like the way the European Maple and Spruce are shining through the varnish. I think they will sound great, too. Tapping on the corpus, it sounds as though it will have a big, deep voice.
I will follow the completion of the varnish process in a later post.
Last “woodwork” tasks on the 16-1/2″ five-string Viola:
Last time, we finished up with the neck set, and the corpus closed, but all the edgework (and final shaping of the neck heel, etc.) left to be done.
Closed corpus, purfling weave sketched, heel/button need carving.
Carving the heel/button combination
The neck heel and the back button, together, make up the majority of the strength of the neck-joint. I once had a cello come in for repair, fully up to tension, but “something was loose.” Yeah, the ONLY glue still holding in the neck-joint was the glue between the neck heel and back button! I removed the neck, cleaned out the old glue, and re-glued the entire joint: but I never forgot that the heel/button connection alone had held the entire load of the string tension! So I make certain that this joint is perfect, and the two are carved as one piece after gluing.
There is also a specific measurement from the center of the neck-heel curve to the top edge of each side of the front plate where it joins the back: in violas, I shoot for exactly 27mm.
Heel and button carved: ready to begin purfling. I have laid out the purfling and incised it.
Purfling
I used to struggle with cutting the purfling slot (I still do, but for different reasons) because I was trying to cut the full depth in a single pass, or maybe two. One of my teachers corrected me, saying that the first pass around, with the knife, is just to “darken the lines” left by the marker. Then it is relatively easy for the blade to follow the groove for subsequent fast passes, each making the slice a little deeper. Finally, I use a special tool to pick out the waste wood from between the lines.
Purfling-pick with front plate.
One problem I faced with the back plate that I had not noticed so much, on the front plate, even though it had the same issue: This purfling is a little wider than what I usually use, so, in spite of the fact that I marked out the correct width, my pick tools (all of them) are made for the narrower purfling, and they do not readily make the slot the correct width. That meant a lot of going back and widening things just a little bit (0.5 mm, usually.) The European spruce of the front plate is soft, and quite forgiving. The harder European maple back plate does not give at all, so if the slot is too narrow, the strip is not going in, at all.
Another issue is that the purfling weave is on top of a fairly thin portion of the back plate, so I could not cut my slots as deeply as I wanted to. Thus, there was very little wood-support for the purfling, and the pieces were difficult to fit, whereas, around the perimeter, I could cut a slot for the full depth of the purfling strips and achieve full support. Ah, well…that’s life. But there were some joints I am not so happy with.
Anyway, this is how the purfling went:
I cut the center bout slots, first, along with the corners.
The goal is to complete the whole slot before inserting any purfling. (Notice the shallow purfling-weave slots.)
Installing the purfling
As I did on the front plate, I installed the center-bout strips first, dry, and then the rest of the perimeter. I glued the perimeter in completely, before beginning the purfling weaves, themselves.
In the case of the purfling weaves, since the slots were so shallow, I glued each piece as I installed it, then worked on the other end of the instrument while the glue from that piece set up and began to hold.
Outer perimeter complete…working on the weaves.
Purfling weave is nearly complete. Notice that some joints are not as clean as others.
Completed purfling weave: still needs to be planed flush.
Completed purfling weave…”warts and all.” I may elect to go back and improve things a little. (Probably not.)The other weave turned out a little better.
The Channel
Once the purfling is all in place, and planed flush, it is time to carve the “channel.” This is a slight “ditch” that runs all the way around the perimeter: the bottom of the “ditch” is usually at the purfling, while the outer edge of the ditch ends exactly at a line called the “crest,” which is about 40% of the distance in, from the outer edge of the plate to the outer edge of the purfling. The inner edge of the “ditch” will be planed and scraped back to “fair” into the surface of the arching, without any lumps or hollows.
Notice the pencil-line “crest”, between the purfling and the plate-edge.
Edgework
Finally, after all the surface of the plate is correct, I plane, scrape and sand the edges themselves, so that the outer curve of the plate edge perfectly meets the inner curve of the channel, all the way around the plate.
In this case, I did not take the picture until after I had completed the next step, which was to wet the whole structure down with water, in order to deliberately raise the grain, so that any imperfections, or compressed areas, will rise up and be seen…and subsequently, be scraped and sanded flush again. All this to say, please understand the “rough” surface of all the wood.
Edgework complete, but still rough with raised grain.
So…that means the whole instrument is now complete, minus the varnish prep-work, and the actual varnish and set-up!
Front ready for varnish.
Side ready for varnish.
Back ready for varnish.
Varnish Sequence
I will post the varnish sequence as it occurs, but, for now, know that the sequence will include at least two “wet-it-down, let-it-dry, and scrape/sand-it-smooth” iterations. The idea is to produce a surface that will no longer respond to moisture by raising the grain. This is particularly important on the handle portion of the neck, where the moisture from players’ hands will certainly be in contact with the wood, every time the instrument is played. But, under the varnish, the slightest discontinuity will become glaringly obvious, so that is important as well.
After the wood is smooth, there will be:
a tanning treatment,
a mineral ground treatment,
a sealer, to lock the mineral ground in place, and
finally, the varnish itself in a series of 6-12 coats, depending on color.
Almost done with the Wood Work parts on the 16-1/2″ five-string Viola!
Completing Arching of the back plate.
When I last posted, I was nearly done arching the back plate, and nearly done shaping the neck:
Arching nearly complete; Neck nearly complete.
Setting the neck
I forgot to take photos during this process: sorry. I get wrapped up in the work and forget all about taking photos.
Setting a neck follows this course of action:
Prepare the neck heel– angles and dimensions all correct. Heel absolutely flat, all mating surfaces absolutely smooth.
Lay out and cut out the mortise in the neck block on the corpus. I try to give myself some room for adjustment. It is always a mistake to try to cut exactly to the layout lines in the first attempt.
Check all measurements with every single change:
how does the centerline of the neck fit, relative to the centerline of the corpus?
how does the transverse level of the neck match the level of the corpus?
how does the pitch (front-to-back) angle compare to what is correct?
how does the distance between the nut line and the top edge of the front plate compare with the correct distance?
how does the height of the lower edge of the fingerboard above the edge of the front plate match the correct height?
When the final fit is perfect, checking ALL measurements, remove the neck and slather in the hot hide glue, and immediately ram the neck into place one last time, making a fast re-check of all measurements, to make sure nothing moved out of place.
Let it dry!
Neck set complete!
You will notice that I set my neck before closing the corpus. I find it easier and faster, and it allows me to achieve a perfect fit against the back button.
Graduating the Back Plate
The next step was to carve out the interior of the back plate, to achieve ideal thicknesses all over. This is called “Graduating the plate” or, simply “Graduation.”
Beginning graduation of the back plate.
Calibration of “dots’ begun: circled dots are already correct. Others need carving out.
After I established the current thicknesses for all the above locations, I carved each dot to the thickness I actually desired, leaving a series of pits all over. The following photo was actually of the front plate, but it is the same idea:
Dots (front plate) completed and ready to be connected, using a plane.
After connecting the dots, the graduation is very nearly complete. I scraped, and checked thicknesses, and scraped some more, until it looked like this:
The shadow line shows the curve of completed graduation.
Preparing to close the Corpus
With the back plate graduation complete, I was ready to attach the plate to the corpus, except that:
the corpus was still attached to the mold, and
the blocks and linings had yet to be shaped.
So I sketched in the planned shapes of the six blocks, trimmed the neck-heel flush, removed the mold, and then set about shaping the inside of the blocks and linings.
I used a knife and a chisel and a gouge to shape all of them, finishing with a scraper. The very first step was to sketch in the block shapes.
Sketching in the block shapes.
Then I cut the neck-heel off, flush with the neck block. I used a flush-cut saw for this step. It works well, but you have to watch carefully to make sure it is not wandering off the line.
Neck-heel cut off flush with the neck block.
Removing the Mold
Then I leveled the back of the garland, using a sanding board, and I was pretty much ready to remove the mold. I popped the glue-lines loose from where the blocks were attached to the mold and lifted the mold out. It is a collapsible mold, so I removed the bolts and took the mold out in three pieces. That left me with a completed corpus, and a neck already set, but no mold. At this point the structure is extremely fragile: I have to be verycareful until the back plate is glued on, making the structure rigid and strong again.
Mold is out! Preparing to shape the blocks.
No mold means no support! Be very careful!
The next photo is from an earlier instrument, because I forgot to take a picture of the shaped blocks in this instrument:
The general shape of the finished blocks (earlier instrument.)
Closing the Corpus
Then I installed the label, and dry-clamped the back plate to the corpus, using spool-clamps, checking carefully to see that my position was correct, and that the overhang was equal all the way around.
I then removed a few clamps at a time, and slipped hot hide glue into the joint, using a thin palette knife, and quickly replaced the clamps. When I had spool-clamps all around the perimeter, I added one last clamp, a padded C-clamp, to ensure that the joint between the neck heel and the back button was fully closed and tight. I used a glove to pad the neck, and clamped it securely.
Closed Corpus, with bass in the background. (Gotta get going on that double bass again!)
The next thing will be to trim the neck heel and back button together, as they are the key to the strength of the joint, and must hereafter function as one.
Neck-heel and back button awaiting final shaping. Planned purfling-weave sketched in.
But I will leave those steps for a later post…this one is already too long. (sorry…)
When I last posted, I had traced the shape of the garland onto the plates and was ready to cut out the plates. I decided to wait on the back plate, but the front plate was ready to go, so I cut it out, using my band saw, and smoothed the edges, using the spindle sander and files. (I have built precisely one instrument without power tools of any sort: One of my early teachers required it, so I complied, but it convinced me that, at my age, I need to save my joints for the things that I have to do by hand, rather than beating them to death just on principle. Besides, I am convinced that if the old masters had possessed power tools, they would have used them without question. They were very practical people.)
So, with the front plate cut to shape, I first marked the edge at a thickness of 4.5 mm. I used a wheel-style marking gauge, with a sharp disc, to mark the thickness and scribe it into the edge of the plate, all the way around. Then I began cutting away waste wood to achieve the desired arching shape. I checked a poster (Published by The Strad) of the “Conte Vitale” 1676 viola by Andrea Guarneri . It is one of the most frequently copied violas in the world, as it is a large viola that works very well, and copies of it frequently work very well, too. I am modifying the pattern a little for superior playability, but I have made this model before, so it is not “guesswork.”
I forgot to take pictures, initially, but here are a couple, belatedly:
Remaining scribe-line for edge-thickness. I plane down to the line all around, eventually.
Line beginning to disappear.
Arching complete.
Arching complete and f-holes laid out.
F-hole Layout and Incision
Once the arching was complete, right down to scraping, I laid out the f-holes, and incised them deeply into the European Spruce of the front plate. Incising the f-hole outline allows me to turn the plate and sight over the edge of the plate at the profile. I want the main stem of the f-hole to be essentially parallel with the plane of the ribs, when seen from the side. I use this as a final correction for the arching, and without exception, it has required me to correct the shape of the arching before moving on.
F-holes laid out and incised deeply.
Graduation
Once the arching is truly completed, and I am satisfied with the f-holes, I begin graduating the inside of the plate. This means that I am carving the inside of the plate to “match the outside,” in that it will be an appropriate thickness all over. I usually want the center area between the f-holes one thickness, the band running up the center to each end slightly thinner, and the wing areas outside that area quite a bit thinner. There is no “set” thickness, and each luthier has to make choices in order to achieve what he or she wants from an instrument. Getting what you hope for depends on those choices you make, and the choices were (hopefully) made intelligently, based on the type and density of the chosen wood, the shape of the arching, and so forth. Getting the arching and graduations right is a lot of carving on a large instrument, but it pays off in quality of sound.
Beginning to carve the interior for thickness graduation.
Carving away all this waste wood produces a lot of shavings. Fortunately, we heat our home with wood, and the shavings are great for starting the morning fire. 🙂
Once I am getting close, I carve “dots,” calibrated to specified thicknesses.
“Connecting the dots,” using a finger plane.
Frequently I can see the traces of the incised f-holes from the inside by the time I am finished with the graduation of the front plate. And, believe it or not, I always can easily see light through the spruce plate, in the thinner areas, if I hold it up to a lamp.
Cutting out the F-holes
I use a special tool to cut the upper and lower eyes of the two f-holes, then use a small knife to finish cutting them out.
I use an f-hole cutter to open the upper and lower “eyes” of the f-holes.
Ready to cut out the stems of the f-holes. (Earlier instrument…I forgot to take pictures…)
Bass Bar
Once the f-holes are cut out and refined, the next thing is the bass bar. This is the only brace attached to the inside of a violin, viola or cello: it supports the bass foot of the bridge, and provides for clarity and strength to the bass notes. An instrument with a weak bass bar will not sound good.
Bass bar fitted, glued and clamped.
With no point of reference, it is hard to realize the size of that plate: so here is a standard violin-sized plate for comparison: A 16-1/2″ viola is pretty big.
Size comparison with violin plate and bass bar.
Bass bar glued, and ready for trimming to shape.
Proposed shape of the completed bass bar.
I cut the bass bar to the desired shape, using gouges, planes and scrapers.
Completed bass bar.
Another size comparison. Same violin plate.
Inner Edgework
I know it will be difficult to accomplish the inner edgework after the plate is installed, so I always do that first. I also trim and shape the linings, so that they taper smoothly into the ribs.
Installing the Front Plate
I dry-clamped the plate to the Garland, and then, using a thin palette knife, slipped hot hide glue (on the thin side, for easy removal if needed) into the unclamped areas and immediately applied padded spool clamps totightly hold the plate until the glue could dry. Then I removed the first few clamps and inserted glue there, and reclamped. My wife thinks the instrument looks as though it is wearing hair-curlers at this point. 🙂
Front plate installed, glued and clamped.
And that is where the instrument rests for tonight.
This is the first large 5-string viola I have made. So far, most players have been quite firm about wanting the same scale length as a violin…and an instrument that will fit in their fiddle case. So that is what I have mostly made. But lately, there seems to have been an increased interest in five-string violas. Some were interested specifically in a smaller viola (hence the current, nearly completed 5-string 15″ Viola,) but more recently, there were two players who were really interested in a large 5-string viola. A 16-1/2″ Viola with a high E string.
So… here we go! I already had European Maple and Spruce billets set aside from earlier “shopping trips,” and everything else that I needed to build a big viola. All I had to do was to finish the instruments already on the bench! They are now out of the way, except for completing the varnish and final set-up of the 15″ 5-string viola… so I am good to go. Here are most of the materials, with the plates already bookmatched.
Materials for new 5-string Viola.
Making, Bending and Installing Ribs
The blocks were already in the mold, and shaped. It was time to start working on ribs. I thinned the ribs to the proper thickness, using a fixture I made for my oscillating spindle sander (one of my few power-tools.) Then I bent each of them to the proper shape, using a hot bending iron.
When I had all six ribs shaped, I installed the two center-bout (often called “C-bout”) ribs, and secured them with hot hide glue and clamps. You can see that clamping to a curved surface is not easy. I held the mold in a small vise, then secured each joint using hot hide glue, cylindrical wooden cauls, and f-style clamps.
Center ribs installed on 16-1/2″ five-string Viola.
When the glue holding the center-bout ribs was dry, I trimmed the ends of the ribs, using the spindle sander again, and installed the lower bout ribs. The joint at the center of the instrument, between the lower ribs, has to be pretty close to perfect, as it will always be visible and any discrepancies will be glaringly obvious under the varnish.
Center ribs trimmed to match the curvature of the blocks.
Lower ribs installed.
Finally, I installed the upper ribs. There is no joint between the upper ribs: in fact, they don’t even have to touch. The neck mortise will remove the middle section regardless of how good my joinery is, so I leave a gap there to allow for easy installation of the ribs. (Meaning, I only have to concern myself with how the ribs fit the corner blocks and that they cleanly follow the mold up to the neck block.)
Upper ribs installed.
Making, Bending and Installing Linings
While I was bending ribs, and still had the iron hot, I went ahead and cut and bent a supply of linings. The linings, like the blocks, are made of willow, because I like the way it works. The linings serve to triple the gluing surface of the edges of the ribs, where they contact the plates, as well as strengthening the rib garland.
Linings bent and ready to install.
I made a small mortise at the juncture between each rib and each block (24 of them,) and then installed the linings dry, to make certain they fit correctly. Then, one-by-one, I removed each lining, applied hot hide glue to both the rib and the lining, and quickly reinstalled the lining and secured it with a series of small spring-clamps.
Linings with hot hide glue and spring clamps.
Tracing the Shape of the Plates
When the glue holding the linings was dry, I removed the clamps and used the spindle sander to trim the ends of the rib corners. I also leveled the front and back of the garland, so that I would be able to trace the shape of the plates. The European Maple back plate is on the left, and the European Spruce front plate is on the right. I used a small washer as a spacer, to establish the edge overhang, and a ball-point pen to trace the shapes. You can see that I have begun work on the neck, as well, which is also made of European Maple.
Completed garland, traced plate-shapes, and partially carved neck.
The garland is temporarily out of the focus of the work, now, so I hung it up, out of harm’s way, until I am ready to begin installing plates.
Rib garland completed and set aside for safe-keeping.
The next step is to actually cut out the plates and begin shaping them into the voice of a Viola. I will let that wait until a later post.
.JPG)
.JPG)
