Friday, February 29, 2008

BRACING THE BACK

CUTTING THE BACK PLATE
The back plate was first marked using the template, first one half then the other. In order to draw a saw line, a second marking was done with the pencil stuck inside a 1/4" wide brass washer, rolled around up against the template.

The back was sawed out on my bandsaw using a 1/4" 6TPI raker blade. The marking was done on the 'true bookmatched' side that eventually will be on the outside, in order to eliminated potential tearout from the saw on that face.

GLUING THE BACK CENTER REINFORCEMENT GRAFT
In Cumpiano's book he does this by lining up the back graft against a metal ruler and so I went and did the same and it worked very well. Though I think I was a little frivolous when I added the glue as the ruler ended up being heavily soiled in all of the squeeze out. But, after a couple of minutes, once the glue had set, I was able to remove the ruler and clean it up without the back graft moving and all was fine and dandy. I waited about half an hour to clean up the glue on the back itself.

For the gluing I made a 20" long caul out of some scrap walnut that I stuck to the back graft with some double sided sticky tape before gluing. The caul was about 11/16" wide, just a hair narrower than the graft and about 1/2" thick.

The graft was glued in the go-bar deck, using a 15' radius disc and a number of go-bar rods - 13 I believe. But, also of importance, the go-bars were stuck in there slight off center along the graft gently pushing the graft towards the metal ruler and thereby preventing it from slipping out of place.

CUTTING THE GRAFT PROFILE INTO A GENTLE CURVE
I have seen many different tools being used for this depending on the preferred profile and look - sanding blocks, chisels, gouges and planes.

I started off by using a block plane to cut the initial facets, but I kept having difficulties controlling it as the plane wanted to rub up against the back plate itself as the facet started to form. Soon I switched to my little bullnose plane instead. As it is a rabbet plane, the blade sits flush to the sides and I felt it gave a lot better control for faceting the edges, where the block plane kept rubbing up against the back plate. One caveat with the bullnose plane though, it had to be kept off the back plate all together in order not to mare the back plate.

Like when I did the heel carving, I started of planing a primary facet of both sides of the graft and finished the curve off with to secondary facets on either side of the primary one. This produced a gentle curved over the graft consisting of seven facets in total.

I think using facets as a great way to monitor the progress and evenness of the contour or curve you are trying to get. Using a light positioned at a low level produces a clear visual reference of the consistency and accuracy of the facets.

The graft was finished off by sanding with 320 grit paper, smoothening out the edges of the facets until one single smooth curve was present.

CUTTING NOTCHES IN THE GRAFT FOR BACK BRACES
The Torres plan I scaled down for this guitar calls for a certain positioning of the back braces. However, when I laid them out on the back I noticed that the spacing was smaller between the middle and lower brace than the spacing between the middle and the upper brace. I couldn't help but think that it looked a little odd and out of balance.

Being the vain person I am, I took the liberty to move the middle brace closer to the upper bout so these two braces were at an equal distance from the center of the soundhole. I don't know if this is going to be detrimental to anything, but, from a little bit of research I found that the spacing between the back braces to be seemingly rather random; Some keep the spacing even while others moves the middle brace closer to the upper bout, yet others again keep it closer to the lower bout. In Cumpiano's book he suggest using four braces instead of three.

I started by marking the position of the braces on the graft with a pencil. Then each brace was held in position on the marks and squared off with a engineers square held against the graft. Once the braces were in place, a mark was scored with a scalpel on either side of that brace.

With a small razor saw, cuts were done on the inside of the marks for each brace and the waste was pared down with a 1/8" paring chisel. As when the Transverse braces were trimmed to length on the soundboard, I had to be careful not the mare the back plate with the saw. As the cut progressed I gradually moved the sawing to the tip of the saw, sawing with small strokes only taking care not to keep the heel of the saw off the back, stopping just before the cut was done. The rest was taken down with a 3/4" chisel while clearing out the waste.

Finally a last fitting was done where small adjustments were performed to widen the notches here and there for a snug fit.

GLUING THE BACK BRACES

With the notches done, one last mark was made before I was ready for gluing. With a plastic drafting square a mark was made with white pencil at the template line. This was done to have a reference mark for keeping the braces square to the graft when they got glued on.

The braces were glued to the back in the go-bar deck using a 15' radius hollow form. I used 5 rods for the upper and middle braces, whereas the lower brace was glued on using 7 rods.

A note: The braces had previously been carved when I made them a while back. In retrospect I don't know if I would do that again. I think it made them more difficult to clamp in the go-bar deck. The tips of the rods had to be positioned more carefully in order not to slip off and using a caul was not really possible either. I can see why some people prefer carving them before gluing, as it may avoid problems like damaging the plate should the chisel accidentally slip and that sort of things. But, I think I prefer using cauls and I don't really mind carving the braces once they are on. Maybe another clamping method would be more suitable for the precarving approach.

SCALLOPING THE BRACE ENDS
I made a little template out of Lexan for this and used it both the mark the braces and as a reference while carving and it work quite well. I had drawn some reference lines parallel to the base that I used keep the template held aligned to the top of the brace and thus not following the radiused bottom.

The scalloped was carved with a 1/2"chisel with the belly down. On the top and middle brace they started 2" inside of the template and 2½" on the bottom brace. All of the scallops tapered down to a 1/4" at the ends.

I was quite surprised to see how much the grain in the mahogany was interlocked. Often one half of the brace, along the length, would cut fine with the grain while the other half would tear severely cutting against the it. It was quite difficult to control, but with a freshly sharpened chisel and thin slivers I was able to keep the tearout to a minimum.

Each scalloped was finished off with 320 grit sandpaper rolled onto a piece of 1" dowel.

THE LABEL
Since this guitar is a 3/4 classical size it lends itself to be played by a child, and I though what more appropriate than to have one of my sons do a drawing for it. Albert, my 4 year old, came up with a drawing of a person, me I think, playing a guitar that I thought was quite delightful (as parents do).

Kay put it all together in Adobe Illustrator and voila! the label was done. I printed it on some Strathmore Series 400 Drawing Paper, an acid free paper that is slightly cream colored and had a little bit of tooth to it.

Our Inkjet printer did the printing and that left something to be desired. The problem was that the inks bled hugely as they are water based and the glue itself contains water and when the paper absorbs it, the moisture makes it expand and buckle up. One way to remedy this is to add water to both sides and, Haha, the colors bleed.

A fellow builder had suggested using fixative to lock the inks before gluing/wetting the label. I did some test with this and found that it mostly solved the problem, but I had to literally soak the label in fixative to prevent the bleeding, and even that didn't control it 100%. Some bleeding still occured but I found that dabbing the label with tissue paper restrained it.

The label was glued in using normal yellow wood glue thinned with water. The glue was applied to the paper with my finger and once the whole surface was covered, I gently wetted the printed face with water too. This kept the label fairly flat. I held it in position for a few second and then dabbed the excess moisture off.

Despite all the care I took, the inks still ran a little and left the final appearance a little less crisp than I would have liked. But, no big deal. I wonder how it would have turned out had it been printed on a color laser printer instead. Next time........


MARKING THE BACK WITH THE TEMPLATE

CUTTING THE BACK

GLUING THE BACK GRAFT

GLUE CLEAN UP

CONTOURING BACK GRAFT IN FACETS

SANDING FACETS SMOOTH

MARKING BACK BRACE PLACEMENT

SQUARING OFF BACK BRACES

MARKING BACK GRAFT FOR BRACE NOTCHES

CUTTING BACK GRAFT NOTCHES

PARING WAIST

ADDING SQUARE REFERENCE MARKS FOR BRACES

GLUING BACK BRACES

MARKING BRACE END SCALLOPS

SCALLOPING BACK BRACES

LABELS

ADDING THE LABEL

THE LABEL ON THE BACK

THE FINISHED BACK

THE FINISHED BACK

Sunday, February 10, 2008

FITTING THE SOUNDBOARD

We have had quite a bit of rain here in Southern California lately, a lot more than average. This had bumped the humidity up to unacceptable levels for this crucial part of the building. The two hygrometers in my garage seems less helpful than I would like them to be, displaying erratic measurements that are in sync sometimes and more than 25% apart other times. Confused, well I am.

Nevertheless I felt it would be best to stay cautious and wait until the rain has settled, and so the soundboard was put to rest in it's protective and comfy sleeping bag - a black bin liner, while I was monitoring our local weather forecast closely on weather.com, that even show an hourly humidity forecast. I waited until weather.com and my own two hygrometers were more or less in line around 50% RH before proceeding.

MARKING THE LININGS
This didn't go as well as I hoped. First I laid the top on the assembly, only holding it down with one hand while trying to mark it with the other. But I struggled with this as I had a hard time keeping the top in place. It kept sliding around a little, making the whole procedure obsolete.

I decided to try and clamp the top to the end block and have that as a reference point. This made everything a lot easier to control. I placed a small engineering square on the neck and used that to butt the soundboard up against, flush with the rebate on the neck and right along the middle axis.

However, I later realized that this too was a flawed approach. Clamping the soundboard to the end block made it bend a little and pulling it away from the neck, 1mm or so. I only realized this when the notches had already been cut in the linings and a test fit was done. I had to widen all the notches to accommodate for this inaccuracy. Not a big deal, but a better way to do this would probably have been to clamp the soundboard to the heel block with a 15mm shim, the same height as the transverse braces, keeping the soundboard evenly suspended above the rim.

The linings were marked with a scalpel, first on the top using the transverse braces as guides, then on the sides with the aid of a small square.

CUTTING THE NOTCHES
I cut the linings right on the inside of the markings with a small razor saw and the waste was cleared a 1/4" paring chisel. Once the major part of the waste had gone, I had to widen the notches a little more to fit the transverse braces. This was quite delicate work, especially in the upper bout where the notches were quite skewed. Thin slivers had to be taken off little by little in order to prevent the linings from tearing. The correction for the inaccurate marking mentioned above was also done this way.

The ends of the transverse braces also had to be fitted a little as they were sitting a little to tight up against the sides. This too was done in thin slivers with a chisel.

PREPPING THE BRACES FOR THE FEET
Quite often brace feet are just a little piece of wood glued to the sides right on top of the braces. But, with a rectangular cross section this means that they don't fit the braces - the corners will 'stick' out. For all practical purposes this doesn't really matter and no one will probably ever see it, but as a kind of an exercise I wanted to try and see if I could cut them in a diamond cross section and fit them in little notches on the brace ends, so their sides would run flush with the braces while the ends would follow the tangent of the sides. This meant the brace end notches and the corresponding foot cross section had to be cut and fitted individually.

With the assembly mounted in place on top of the soundboard, I marked the braces approximately at the width of the linings. On the two upper braces I used a little piece of spruce scrap that was held against the sides. Holding it with its middle over the braces, I figured that it was close enough to being at the right angle of the sides tangent. The lower brace was marked using a sliding bevel set by eye.

The notches were cut with a razor saw. The saw was fitted with a little spruce stop clamped to the saw blade with two spring clamps and set to a 2mm deep cut. The waste was pared away with a 1/4" chisel.

At this point I also roughed out the feet from scrap spruce planed into a rectangular cross section, the same thickness as their corresponding braces. Then, one side was planed into bevel. Once it matched the notches on it corresponding brace, it was stuck to the bench with double sided sticky tape, with the bevel facing down, and the opposite bevel was planed, leaving them with a slightly oversized diamond shaped cross section.

The final making and fitting of the feet was left until after the soundboard was glued onto the rim.

THE SOLERA, THE CORK SHIM AND SLOPING NECK SHIM
In order to make room for the soundboard dome, I made a cork shim out of 1/8" cork glued to some thin cardboard. I cut the cork so there was an overhang of about 2mm on either side of the rim including linings.

A larger pad was glued to the upper part to support the clamping pressure from the soundhole clamp
on the two upper transverse braces. Likewise, a smaller pad was added below the soundhole to support the pressure on the lower transverse brace. As this brace is arched, the cork was thinned to about 2mm, or about 5/64".

As the body will be suspended by the cork shim, I had to make a shim for the neck as well. This was done in maple tapering from 3.0mm to 1.5mm over a length of 29cm. I did however, cut it slightly shorter to give way for the spool clamps near the neck, that will be needed for when the back is glued to the assembly. As the cork was stuck to thick paper I had to do the same with the slope. I suppose that I could have just made it that little bit thicker, but I didn't. Oh, well.

DRYFITTING
Finally a dry fit was performed.
First I had to take half of the solera support feet off in order to get the oversized soundboard in place. Once it was in, the feet was put back on again using the template to place them accurately. The soundboard was held in place by the soundhole clamp and the side/neck assembly was placed on top. It was an almost a religious experience to see it all come together.

Eight spool clamps were
inserted into the holes around the periphery of the solera at an even spacing, ready to receive and clamp the rim down.

GLUING IT ALL TOGETHER
Once everything was lined up, the side/neck assembly was taken off and a modest film of glued was added to the edge of the rim, head and end block and placed back on top of the soundboard again. I used the eight spool clamps already inserted to clamp it down along the sides and big deep reach Bessey clamps and cauls for the heel and end block. The neck was held in alignment with a small bar clamp.

I had a little bit of
squeezed out glue to clean up, particularly around the heel and end block. This was done with a spruce chisel made for this purpose. It was the first time I used wood to clean up squeezed out glue, but it worked really well, and I didn't have to be so careful not to mare the surfaces. Spruce chisels can also be made in all shapes and sizes in no time, making them very suitable for this particular job where accessibility was an issue.

MAKING AND FITTING THE BRACES FEET
Initially each roughed out brace foot piece was cut into two 2" long feet, one for each side of the individual braces, but as I started to dry fit them, I shortened the feet for the two upper transverse braces, about 1/8" and 1/4" respectively, as the width of the guitar rim decreased steadily from the waist towards the neck, and I wanted to leave an equal gap between the top of the feet and the lining above them.

A bevel was planed lengthwise, tapering the feet towards the top and the gluing surface was sanded to fit the curvature of the rim.

Most of this final fitting was done by eye. I had to redo some of the feet from scratch mostly because they got planed or sanded too much and weren't fitting snugly in the notches anymore, while I accidentally tapered the wrong face of others. But, eventually they all fitted fairly well where they were supposed to.

The feet were glued in and clamped to the sides with spring clamps.
I tried to angle the clamps so any sliding would go downwards locking the feet in place. A trick from Cumpiano's book. I also fabricated small cauls for the feet, mostly to protect the feet from being marred by the clamps.

TRIMMING SOUNDBOARD OVERHANG
I clamped the guitar neck in the patternmakers vise and went at it with a 1/2" chisel, first in rather broad strokes, then in finer ones as I worked my way down towards the sides. It went relatively quickly.

A spokeshave was used for some of the finishing work. I set the blade slightly skewed, fully recessed at the side that was on top of the rim, whereas the other side was set to cut. This way the danger of the spokeshave digging into the rim was minimized, while the depth of the cut could be adjusted from a more aggressive cut to a very fine one, by holding the shave further over the rim or not.

LAST NOTE
I am was really pleased with the fit between the rim and soundboard. Both the beveled linings and endblock, as well as the slightly sloped neck rebate seemed to fit perfectly, with no visible gaps to be seen at all, inside and outside the rim
, both during dry fitting and gluing. I don't know how it would have looked, had I not gone through the trouble of bevling the linings around the lower bout, but this appeared to work very well.


MARKING LINING NOTCHES

MARKING LINING NOTCHES

CUTTING LINING NOTCHES

PARING LINING NOTCH WASTE

MARKING NOTCHES FOR TRANSVERSE BRACE FEET

CUTTING NOTCHES FOR TRANSVERSE BRACE FEET

PARING NOTCH WASTE FOR TRANSVERSE BRACE FEET

SETTING BEVEL FOR BRACE FEET

PLANING BEVEL IN BRACE FEET STOCK

PLANING BEVEL ON THE OTHER SIDE OF FEET STOCK

CHECKING BEVEL ON FEET STOCK

SOLERA WITH SHIMS

GLUING SOUNDBOARD TO ASSEMBLY

CUTTING BRACE FEET TO LENGHT

SANDING CURVED IN BRACE FEET GLUING SURFACE

PLANING THE TAPER


DRY FITTING ASSEMBLY

GLUING BRACE FEET

CUTTING OVERHANG

FINISHED ASSEMBLY

FINISHED ASSEMBLY

FINISHED ASSEMBLY

FINISHED ASSEMBLY