SOUNDBOARD BRACING - FAN AND FINGER BRACES

Initially, when the fan and finger braces were prepared, I had left them oversized and partly jointed with the gluing surface planed flat, some with all surfaces square while others still had sawmarks showing from the bandsaw, but all needed to be planed to their final size, 4x5mm for the three center fans and 4x4mm for the rest.

PLANING TO FINAL DIMENSION
I fabricated a little stop from some soundboard scrap, stuck it to the bench with double sided tape, butted up against two bench dogs. I laid all the braces on the side that was jointed to the bottom or gluing face, butted up against the stop and secured to the bench with double sided sticky tape, and proceeded to plane them all down to a 4mm width, constantly monitoring the progression by carefully measuring the stack around the periphery.

Once the fans were planed to width the whole procedure was repeated, this time with their gluing surface facing downwards. The whole group was planed down to 5mm thickness and the three middle fans were removed and the rest of the stock was planed down to 4mm.

Last, the fans braces were numbered in consecutive order from 1-7 and the fingers braces 8 and 9.

TO ARCH OR NOT TO ARCH...

...... That is the question. There seems to be a great variety in how builders prepare and glue their braces. Some arch the gluing surface to match the desired dome of the lower bout, while others leave them flat and force them into the dome when they are glued to the soundboard, arguing that it will build a certain amount of tension in the soundboard that will make it more lively and responsive.

Whether this is true of not, I don't know, but I decided to go with the latter. If nothing else it saved me from arching the bottom of seven thin pieces of wood.

FITTING AND NOTCHING THE BRACES
First each brace was cut to length with a small dovetail saw.

The next thing to do was to carve out little notches so the fans would fit over the flat bridge graft and the transverse aperture grafts. I carefully marked where the notch was needed on each fan and proceeded by using a little trick I learned from Bogdanovich' book. He basically kept a few pieces of scrap from the corresponding graft that he stuck on top of the vise jaws and used them as spacers for positioning the braces in the vise, leaving them to protrude the exact thickness of the graft.

I used my patternmakers vise for this job, prepped by planing the top of the clamped jaws flat and square. I positioned the braces by having them protrude a little over the shims with only light clamping pressure, enabling me to press them down flush with a piece of scrap spruce, then clamping them fully. Once in the vise making the notch was just a matter of cutting the ends of the notches and pare the waste down flush with the vise jaws.

The two outer fans as well as the smaller finger braces on each side of the soundhole had to be notched to fit the aperture grafts of the transverse braces. This was done in a similar fashion.

These last four braces also need to have to be mitered at one end. The outer fans to butt up against the soundhole doughnut and the finger braces to butt up against the outer fans. This I did it on the shooting board using a sliding bevel set at the corresponding angle as a fence. The braces were stuck to the bevel with, you guessed it, doublesided stickytape.

GLUING THE FAN AND FINGER BRACES

Like the flat grafts, I had made a handful of clamping cauls out of 1/8" maple to distribute the clamping pressure of the go-bars a little. In hindsight I might have made them a little thicker, 3/16' or so, to facilitate this but this is how they ended up this time. Each fan was mated with a caul using double sides sticky tape and a modest film of glue was applied to the bottom of the brace.

First the fan braces went on. I started with the middle fan and worked my way out alternating from side to side, followed by clean up of the squeezed out glue on the previous fan, until all seven fans were in place.

Each fan was initially held down by hand pressure alone until the glue started to grip. Then one go-bar rod was inserted in the middle of the fan holding the brace down into the domed shape. Then a rod was added to each end and last yet another rod on either side of the middle, between that and the end rods, making up a total of five rods for each fan.

CARVING THE FAN AND FINGER BRACES
I experimented a little by using different tools, a block plane, a little shoulder plane, different sized chisels and my little flat soled Ibex finger plane. I ended up using a 3/4" paring chisel for the 'peaking' or carving the cheeks.

In Cumpiano's book he suggests holding the chisel like a pool cue when carving the cheeks, however, I found 'pinching' the edge of the chisel against the fans gave me more control. This way the fingers holding the edge functioned like a fence preventing the chisel from accidentally slipping and marring the soundboard while the other hand controlled the force as well as the cutting angle. I found a wider chisel the worked better for the particular method.

It was quite revealing how even slight changes in grain direction impacted the ease of the carving. With the grain it presented no problem, but as soon as it was against the grain, I could feel how the chisel wanted to dig in. Sometimes I ended up changing direction and started to carve from the other side, while other times it was purely a matter of taking less material of at each stroke to prevent tear out.

In the beginning I also tried to even out small irregularities from the chisel with a few strokes of the finger plane, but I eventually gave that up, partly because it was difficult to hold the plane off the soundboard and partly because the chiseling eventually got smoother as the work progressed and I started to get the hang of it.

The feathering of the end of the braces was done with a 1/4" paring chisel followed by a few strokes with the finger plane. I found the narrower chisel easier to control as the corners of the edge were less prone to dig into the soundboard once the ends started to thin out.

Once all the fans had received their preliminary carving the finger braces were glued on flat and carved in the same manner as the fans.

LAST THOUGHTS
Even though I tried to be very careful with my chisel I did mare the soundboard several places during the carving. I found the biggest problem was that curls generated by the chisel obscured my vision of the corner of the chisel closest to the soundboard and it was hard to follow when it was getting to close (or digging in). And it was virtually impossible for me to feel it when it happened as the chisel went through the soft spruce so easily.

Some builders advocate precarving the braces before gluing them as this minimizes the potential marring of the soundboard. This, I believe, is a very good idea, however, I also speculate that the clamping would have to be done with specially made cauls that would fit the pracarved profile or with something other than a go-bar deck.

FINAL PLANING OF FANS

MEASURING FAN WIDTH

MITERING OUTER FINGER BRACES AND OUTER FANS

MARKING THE NOTCHES ON THE FANS

CUTTING BRIDGE GRAFT NOTCHES

PARING BRIDGE GRAFT NOTCHES

GLUING THE FANS DOWN

PARING THE FAN 'PEAKS'

FEATHERING THE FANS

MARKING FINGER BRACES

CUTTING GRAFT NOTCHES

PARING GRAFT NOTCHES

GLUING THE FINGERS DOWN

PARING THE FINGER 'PEAKS'

FEATHERING THE FINGER ENDS

HUMIDITY CONTROL ON A SHOE STRING

It's been raining a bit recently here in Southern California. Needless to say that humidity have gone up, which is something that is close to lethal when you glue braces on your soundboard.

I have a two hygrometers, one that sits in a clock I have on the wall in the garage/shop, and a little digital cigarbox one I bought a while back for our piano. I while ago I took the little one out to the shop only to find that they were not displaying the same value at any given time. When the digital one showed 38% the analogue one in the clock read 58%. I couldn't believe that they are so far apart.

Two weeks ago I took my little digital hygrometer with me to my local wood and tool shop Rockler, who sells analogue hygrometers for clock makers. I wandered about in the shop until mine had settle and compared it to the ones they had on display. They all consistently read about 9% higher than mine. When I got home I adjust my shop hygrometer down about 10% and have used that as my reference, and now it is at least fairly consistent with the local weather forecast. I'm still not sure how precise it is and I might try and go to one of our local museum and compare it to one of their hygrometers. They should be quite accurate.

But to deal with the damp weather conditions, I choose to try out a trick from Bruce Hoadley's book "Understanding Wood". He suggests wrapping your wood in plastic until humidity is back to normal.

I'm in the middle of bracing the soundboard and hope to avoid to many unpleasant surprises. Ideally you want to get the soundboard braced and glued on to the body in a close to ideal humidity conditions as possible, otherwise you will be in for some nasty surprised later on.

A fellow builder from LuthierForum just posted several pictures of what looked like someone had stuck a hose to the soundhole and sucked out all the air from the guitar. It was not a pretty sight. Basically he had been building in too humid condition and when the guitar settled in the drier indoor winter air, the thin soundboard and back had contracted and caved in showing all the braces that were still trying to hold these plates in place.

Nasty business!!

HUMIDITY CONTROL ON A SHOE STRING

SOUNDBOARD BRACING - FLAT GRAFTS

INITIAL PREPARATION

Before any of the grafts could go on I traced the whole bracing lay-out onto the top very lightly with a soft pencil using the soundhole center and glue seam as reference points. Once that was done I cut the soundhole with my little Dremel router and a 1/16" carbide downcut spiral bit.

Before cutting the hole, I lined the bottom of the circlecutter attachment with cellotape to prevent the smudging I got when the rosette channels were done. The hole was cut it in two goes. First a little more than halfway down with the rosette facing up, then a second run with the soundboard flipped over with the rosette facing down. This was done to minimize potential tear out.

I also made clamping cauls out of 1/8" maple for all the braces and grafts.

GLUING THE UPPER FINGERBOARD GRAFT - FITTING THE SOUNDBOARD TO THE NECK
As I mentioned in the previous post, the final fitting of the soundboard to the rebate on the neck had to be done on the upper fingerboard graft, so that one went on first. In order to keep things aligned I clamped the adjacent transverse brace in place as a reference to but the graft up against. The bottom of the brace was covered with non stick cellotape to prevent the glue from the graft to adhere to it. Having the brace in place also helped preventing the graft sliding around in the wet glue. The squeezed out glue was cleaned up with a chisel.

When the glue had dried, the top of the soundboard was cut down to the template line with a saw and trimmed to fit with a plane on the shootingboard. As the graft was slightly wider than the neck, little notches were made at the end of the graft to allow for the sided. Likewise, about 3/16" of the lining on each side of the neck was pared down flush to the rebate.

When I did the clamping cauls for the braces, I also made a 'slope', a thin piece of maple tapering in thickness from 3.0mm to 1.5mm, that eventually will be stuck onto the solera to set the neck angle.

This slope was used as a guide for planing a slight taper into the fingerboard graft. I knew that the depth of the soundboard rebate was about 3.6mm, matching the desired thickness of the soundboard and graft at the 12th fret. Getting the right taper and thickness was then simply a matter of calculating where to place the slope on the soundboard and plane away until the correct thickness was achieved. However, I left it slightly proud to allow for some final sanding.

GLUING THE SOUNDHOLE DOUGHNUT
Next up was the soundhole doughnut. This was glued on in a similar fashion to the fingerboard grafts, thus with the two lower transverse braces as guides.

PREPARING THE TRANSVERSE BRACE APERTURE GRAFTS
One thing that I have chosen to modify from the original plan is adding little 0.5mm flat graft under the aperture of the three transverse braces. This is something that Jeffrey Elliott, who I loosely modeled the bracing pattern after, does. These were fabricated from soundboard scrap by planing an oversized piece down with my jack plane. They ended up a little thicker, about 0.58mm, but I figured that was close enough.

Once the stock was thicknessed, three individual graft strips were cut with an Exacto knife and trued on the shooting board until the strips were the right width, each matching their mating transverse brace. This was a fiddly and floppy affair and I had to clamp a little improvised fence on the shooting board to butt the graft strips up against while holding them down with one of the clamping cauls I made for the braces, to keep them straight in place.

Last, the grafts were cut a little proud of their intended length with a chisel. The final trimming will be done once the transverse braces are ready to be glued on.

GLUING THE TRANSVERSE BRACE APERTURE GRAFTS
For this, I wrapped three of the fan brace cauls in cellotape, marked their centers and stuck the grafts onto them with double sided sticky tape, using the already made transverse brace apertures as reference, keeping the grafts straight and flush to one side of the caul. I also placed a little, equally thin shim in the middle of the caul to support a go-bar rod.

When it came to the actual gluing, the upper fingerboard graft and rosette graft were used as guides to butt the cauls up against. First a go-bar rod was added to the center of the caul to hold it firmly in place. Two go-bars were then used added to each graft.

The four top aperture grafts were all glued on flat, while the two lower grafts were done together with the bridge graft on a the 25' radius disc.

GLUING THE BRIDGE GRAFT
This graft was originally done back in July together with the soundhole doughnut and the fingerboard grafts. However, when I revisited the soundboard thickness and thinned it further than I initially intended to, I also started to wonder if 2mm wasn't on the thick side for the bridge graft.

My Israeli luthier buddy, who got me onto the soundboard deflection testing, sent me an article about the Dutch born luthier living in Granada called Rene Baarslag. I the article Baarslag explains how he goes about the thinning his bridge graft, which is about 1mm thick and feathered to nothing at each end.

I have read about other notable luthiers, Romanillos being one of them, speculating about the need for such a graft in the first place. Torres didn't use them and neither has many other world renowned builders that followed in his footsteps.

Rightly or wrongly, I decided to thin my bridge graft to 1mm, flat across the entire length, and with that also making some equally thin spacers for when the fans will be notch. More about that when the fans are going on.

Last the bridge graft and the lower transverse brace aperture grafts went on. This was all done on the 25' radius dish. The bridge graft was the only graft that went on 'freehand' without any reference points to hold it up against. Needless to say that I was a little nervous about it. I applied a modest film of glue on pressed the graft down on the soundboard with my hands. A little sliding did happen initially but eventually the glue started to grip and go-bars were added one at the time starting from the center of the graft and working my way outwards in both direction, slowly letting the glue set for a little before adding another. Once the glue had dried a little the squeeze out was carefully cleaned up with a chisel.

A COUPLE OF NOTES
This is the first time I use a go-bar deck as a clamping device and I must confess that I love it. I don't know how it would have been in comparison to use regular clamps, but the go-bar was a breeze to use - efficient and precise.

Prior to gluing I used a little plumb to mark the underside of the go-bar clamp to roughly know where to put the go-bar rods. This made for a nice, straight downward clamping pressure that also minimized the chance for sliding before the glue was gripping.

TRACING THE BRACING PATTERN

CUTTING THE SOUNDHOLE

GLUING THE FINGERBOARD GRAFT

TRIMMING SOUNDBOARD TOP

TRUING THE SOUNDBOARD TOP

PARING THE LININGS FLUSH TO THE REBATE

SLOPING THE FINGER BOARD GRAFT

GLUING THE SOUNDBOARD DOUGHNUT

GLUE CLEAN-UP

PLANING TRANSVERSE BRACE APERTURE GRAFTS

MEASURING GRAFT APERTURE GRAFT STOCK THICKNESS

ROUGH CUTTING TRANSVERSE BRACE APERTURE GRAFTS

TRUING THE TRANSVERSE BRACE APERTURE GRAFTS TO WIDTH

CUTTING TRANSVERSE BRACE APERTURE GRAFTS TO LENGTH

GLUING TRANSVERSE BRACES APERTURE GRAFTS

PLANING BRIDGE GRAFT AND NOTCH SPACERS

GLUING THE BRIDGE GRAFT

BEVELLING THE SIDES EDGES AND MAKING THE SOUNDBOARD REBATE

This has been a huge hurtle for me to conquer. It appears that most builders just leave the edge of the sides, to which the soundboard will be glued, flat. And, perhaps for good reason. But, while some argue that it is preferable to build in a little bit of tension into the soundboard, and mating a domed top to a flat gluing surface would definitely do that, I was determined to try and fit this the best I could and thus bevel the edge.

INITIAL CONSIDERATIONS
I started by drawing an elaborated version of Rik Middleton's string height diagram featured in his book. I found it a very helpful guide to visualize how different parameters, such top doming and neck angle, would impact the final setup of the guitar.

Initially, it was my intent to have a 25' radius dome in the soundboard that tapered off linearly above the lower transverse brace, forming a kind of tear drop shape, but from Middleton's diagram it became obvious that this wouldn't as easy to do as just that. I started to experiment with alternative scenarios such as having the lower transverse brace arched more to create a more elliptical shaped dome, another version had all the transverse braces flat, and so on and so forth.

Through a lengthy correspondence with two of my luthier buddies, Udi Vachternman and William Londree, I ended up deciding on a more traditional solution where basically everything below the soundhole will be domed to a 25' radius and everything above it will be flat.

With all this, I also settled on a neck angle where the nut is raised 1.5mm. This will give a string height of 3.4mm with a 3.0mm saddle sticking out of the bridge. These numbers are of course theoretical as some distortion is likely to take place once the strings are on and tuned up to pitch, but it should hopefully be in the right ballpark.

I had a friend from work help me out with calculating the right angle for the bevel in various places on the rim. Measured with the center of the dome at the saddleI, it came up somewhere in between 1.05º and 1.26º, changing with the undulating shape of the lower bout. As it was not practical for me to work closer than a 0.5º proximity, I decided to settle on 1.0º for all of the lower bout and tapering to flat between the lower transverse brace and the next one up.

THE JIG
To do this, I constructed a little fence jig to hold a spokeshave. It's a simple little thing that consists of a base that functions as a fence and a little tilting shelf that holds the tool. The shelf has a rare earth magnet that anchors the tool down and at the end a bolt replaces one of the tool handles and locks it in place. In future I might alter it a little by adding a slice of veneer over the magnet to protect the blade and rework the shelf part so the jig can be used both left and right - at the moment it only goes one way.

The spokeshave was set up to cut as fine shaving as possible, the right handle was screwed off and the spokeshave was fastened to the jig. I set my sliding bevel to 1º and transferred that angle to the jig and was ready to go.

CUTTING THE BEVEL
First the guitar was fastened to the bench by resting the body on the 15' dish the back rim was fitted to. A support for the neck was built up with some MDF and other scrap and the hole thing was secured with two clamps, one at the top of the neck and one at the end block.

Cutting the bevel was quite straight forward. The only thing I really had to concentrate on was to keep the fence flat up against the sides while working my way through the job. Once the kerfing got down close to the actual side, I lowered my work light so it was virtually level with the rim. This revealed all the areas where the kerfing wasn't cut down far enough by casting a little shadow on to the edge of the sides and allowed me to monitor quite closely where more shaving was needed and where it wasn't.

When I got to the endblock, I repositioned the clamp slightly off to one side, worked the endblock area down, moved the clamp back again and continued on the other half.

Once the whole rim was done this way, I reset the jig to 0.5º and worked the area from the lower transverse brace and up, trying to gradually work the shallower angle in. The same was done with jig set to a flat angle, though this taper was gradually added, starting halfway between the upper and lower transverse brace.

The last few inches near the heel I couldn't get to with the spokeshave was pared away with a chisel.

MAKING THE SOUNDBOARD REBATE ON THE HEEL
Since the neck will be set at an angle to the soundboard, I've also been debating whether to cut the rebate to compensate for this or just do it parallel to the neck. My first instinct was to compensate for the neck angle. However, I decided to compensate by angeling the fingerboard graft on the soundboard instead, once the soundboard is ready to go on.

I started by marking the heel 3.5mm down from the edge with a scalpel all the way around. Ready to pare away. Cumpiano's method of making the rebate is similar to the way he cleans out waste inside the head slots - by first paring the waste down to the marking while leaving a peak in the middle and then slowly taking the peak down to a flat, a method I really like when I did my head slots.

But, this proved to be a real chore as the already attached sides got in the way big time. Instead I had to work my way through the task in a quite random manner, using both sides of the chisel, paring sideways from the center and out and from the end and in towards the side joint. It was a real pain in the neck to say the least. But down it went.

As I got closer to the marking, I started sanding the rebate a little allowing the scratches to show the peaks and then carefully chiseled these away and repeated the cycle until the sanding left a somewhat evenly scratched surface on the entire rebate. I finished the rebate off by scraping it lightly with the chisel, leaving it a little shallow on purpose to do the final fitting for when the soundboard goes on, being satisfied just to have a it fairly flat. Oh, well - next time.....

IN CONCLUSION
Since this is my first build and I have gone a little astray from tradition, I have to ask myself if this was the right thing to do, and to be honest I don't really know. I was surprised to see how little an angle was needed to fit the 25' soundboard dome and since I ended up rounding the correct bevel up to the nearest 0.5º it wasn't accurate anyway.

Another caveat was that it was difficult to monitor whether the rim ended up on a perfectly flat plane - that some parts of the sides ended up being cut deeper than others. For this I mainly had to rely on my hope that the side were flat already and all I had to do was to cut the kerfing down to match it. I don't know if this is easier to achieve by sanding the whole side on a flat sanding board prior to installing it, as it is normally done. To me, it seems that it might be just as difficult to sand a loose and floppy side evenly along it's length, but I don't know.

So the question remains whether this effort had any value at all or if I was just being another newbie trying to reinvent the wheel. One thing is for sure though, it made me go places and read up on things I would have if I had just followed instructions and as always it was fun to do the research and think it all through.

That and cut the heel rebate before the sides go on!!!

DOMING, NECK ANGLE, STRING HEIGHT DIAGRAM

EDGE BEVEL DIAGRAM

SPOKESHAVE JIG DISSEMBLED

SPOKESHAVE JIG ASSEMBLED WITH TOOL

SETTING THE BEVEL

SETTING THE SPOKESHAVE JIG

BEVELING EDGE OF THE SIDES

PARING KERFING NEAR HEEL

SPOKESHAVE SHAVINGS

THE LOWER BOUT SIDE BEVEL

MARKING THE HEEL

PARING THE REBATE ON THE HEEL

THE FINISHED REBATE

THE FINISHED RIM