Friday, February 8, 2013

Building an Early Lacôte Guitar - Part 2

In my last post I had assembled the guitar body and finished the neck and peghead. Before I glued the neck onto the body I checked the strength of the neck. Wedging the heel against a notch in my workbench I tried to bow the neck -- mimicing the pull of the strings -- by pressing against the back of the peghead with my thumbs and down on the face of the peghead with my other fingers. The neck bowed a little less than a millimeter from level. I was satisfied as this is what I expected. If it had bowed more than that I would have routed a recess in the top surface and installed a long rectangular ebony strengthening rod. I glued the neck to the body using a screw through the front block into the heel by maneuvering the screw driver through the soundhole. At this stage the neck is plain wood. Eventually I will lacquer it black like the original that I am copying.

1823 Lacôte

The purfling detail on the original guitar is composed of five bands of ebony and five of ivory that wrap around the perimeter of the guitar body in one piece. Each band is about 0.7mm thick. I substituted holly for ivory. I make my own banding by band sawing lengths of appropriate woods with widths of 30 or 40 millimeters. After smoothing and thinning to the required thickness I stack them one on the other, as many as 10 high, and glue them together. After they have dried I square the two rough edges and saw off several pieces and smoothed with a small plane. Each piece is then composed of multiple bands that only need to be soaked in warm water to dissolve the glue and then dried or bent while still damp.

Even though I carefully prepared the banding I found discrepancies in the thickness of a number of pieces. If I were using a decorative edging of 3 to 5 pieces of banding I wouldn't worry about a tenth of a millimeter here or there. But multiplied over 10 pieces a small error can cause problems. I use a home-made device to check and re-size each piece of banding. The tool is composed of a block of wood with a flat bottomed ramp cut into the block at a 45 degree angle. A scraping blade is clamped in a position over the exit of the ramp leaving a gap through which the thin strip of banding can be passed. High spots are scraped off by the blade. It is a primitive tool but with patience it will turn out perfect work.

The width of the purfling is a little wider than 8 mm composed of ten pieces of ebony/holly banding. A final band of ebony that is a little thicker and, of course, deeper is added to the edge.

I use a small router fitted with a simple guide to cut the rabbet. Setting the guide to the proper depth I make a series of passes gradually widening the recess. I test fit the 10 ebony/holly bands and stop when the wider ebony piece fits comfortably.

The purfling is intended to wrap around the soundboard tongue with four mitred corners. This is a tricky procedure that displays the luthier's skill. If it is well done it looks pretty classy.
In this photo I masked the corner of the  rabbet with cellophane tape and glued the last 30 - 40 mm of the pre-bent banding together. I wedged it temporarily into the recess so it would dry in the proper alignment.

I lifted the banding from the recess, peeled off the cellophane mask and cut the end of the assembled banding to the correct angle. Once that was done I could proceed to glue the entire assembly in place. It is necessary to work slowly and methodically making sure that each piece of banding is covered with a layer of thin glue and that each piece in turn is properly seated in the recess.

Fitting the mitred banding perfectly around the soundboard tongue was complicated by discrepancies in the thickness of the ends of each piece of banding.  I found it necessary to fit them individually -- each mitred piece matching its mate.

Once all of the banding was in place and well dried I scraped it flush with the soundboard.

The rosette is composed in the same style of multiple bands of ebony/holly except that they are narrower and present greater difficulty in making and assembling. There are 20 bands, plus the wider ebony band that edges the hole. It measures 13.5 wide.

I assembled the rosette before I finished thicknessing the soundboard. This allowed me a little lee-way with the depth of the recess and the process of levelling the face once I was finished or if something were to go wrong. I routed the recess to the maximum  outside diameter and took off a little extra toward the inside as well. This allowed for the expansion that the thickness of the glue might cause. Lacôte assembled his rosette rings so that the ends, which are faintly visible, were spaced randomly around the circumference in order to further obscure them and I did the same. Using pre-bent bands I started at outside  and worked in, one at a time, cutting the pieces to length with opposing bevels so as to avoid making a band either too long or too short. It was not easy going. I quickly developed sticky fingers which made it difficult to do everything including pressing the most recent ring in place against its neighbour. When I had all of the bands in place there was a little space left that I filled with thin strips of styrene until the glue dried. I wasn't happy with the result. Every imperfection showed up. A kink or two here and there and several gaps between bands as well. I routed it out and started over.

Next time, I'll fret the neck including the embedded soundboard frets. I'll also describe staining and varnishing the body and lacquering the neck.

No comments:

Post a Comment