The Rosette

Next for the soundboard is reducing the thickness to close to final dimensions, then installing the decorative rosette around the soundhole.

The top plate started out around .150" thick, more than 1/8", which is much too thick. This is the part that must be vibrated by the action of the strings, creating the air movement inside the guitar that we hear as sound. No movement, no sound. I want it between .110 and .120. I want to balance structural integrity with movement. The structural ideal would be a top plate that would be stiff and strong. It would last forever--but would have little volume and the tone would suck. Great tone and volume would require a thin piece of wood, the thinner the better, but such a top would explode under string tension. So a delicate balance is required. There is no perfect figure, because each piece of wood is different. Two pieces from the same tree could vary widely. One piece would be very stiff, whereas another of the same thickness would be extremely flexible.

Sitka spruce is very popular for guitars because it is both strong and light, allowing it to be thinner than other species while maintaining strength. Other popular top woods, like cedar, are softer and so have to be left thicker to be structurally sound.

Using a plane, I start thinning the top, checking it frequently with a thickness gauge.

There's nothing more pleasurable to use than a sharp plane. You know it's going right when you get all these fluffy plane shavings.

After getting the top down to .125-.130", I switch to a cabinet scraper, more to remove the plane marks than to remove material. When the plane marks are gone, I hit it with some 100-grit sandpaper.

Before sanding down to final thickness I need to rout the channels for the rosette and then install the rosette. This is the decorative piece that goes around the soundhole. I plan to use a three-part rosette. The main rosette will be some paua (pronounced "pow-wah) abalone (short "a" followed by "baloney") shell strips which will be sandwiched between two thin pieces of purfling composed of thin lines of maple-ebony-maple. Inside and outside of this main rosette will be thin circles of black-white-black purfling. You'll see what I mean in a second.

I have a a plywood workboard into which is drilled a 3/16" hole. A metal pin goes into that hole, which is part of a circle-cutting jig to which my Dremel tool (essentially a small hand-held router) is attached. I lay out where on the top the soundhole will be located, then drill another 3/16" hole at the soundhole center.

Since the abalone shell is preshaped into a fixed diameter I have to locate everything off the inside and outside diameter of the shell. In the following picture the circle is the inside diameter of the shell. Then on the center line I have marked the top of the soundhole, and the inside and outside diameters of the two B/W/B purflings and the shell sandwich.

The purflings are 1/16" wide, and I happen to have a 1/16" bit, so I put that in the dremel. The areas to the left will be covered by the fretboard, so I test everything in this area before making the full cut. I make small cut in the neck area for the inside purfling and use a small piece of the purfling to test the fit. Perfect.
So I rout the entire circle and dry-fit the purfling:

Fits great, so I repeat the process for the outside purfling. The main rosette is trickier since I'm routing one wide channel but then fitting three separate pieces in. At first it was too tight, but the circle cutter allows for micro-adjustments, and I'm able to creep up on a perfect fit. After dry fitting everything, it's time to glue it in. I start with the two thin purflings, running the bottom edge through some wood glue then pressing them into their channels. They fit perfectly but just to make sure I use a small baking roller to make sure they are fully seated in their channel.
There's a neat trick for the main rosette. Since the shell is delicate and easily breaks, I don't want to try to work around forcing a sandwich into a tight channel while working against the clock (the glue gets tacky within five minutes). So I substitute a length of white Teflon that is the exact width of the shell. The glue won't stick to the Teflon, so after everything else is dry I can pull the Teflon out and easily place the shell between the two W/B/W purflings. This is what it looks like all glued-up.


The maple blends into the spruce so that it appear that there are two thin black lines surrounding the Teflon.

I put wax paper over all this, a plywood scrap on that for even pressure, and a three cam-clamps to make sure everything stays seated. I let this dry overnight.


The next day I pull the Teflon out to see how the shell fits--and it fits perfectly. (I'm saying this so often that it's getting scary.) I use some white glue (most of us know it as Elmer's Glue) because it drys clear, and fit the three pieces of shell into the channels.

Once again this gets an overnight dry, but mainly because I did it late at night. A couple of hours would have been sufficient.

The next morning I use a cabinet scraper to scrape everything flush with the surface of the top, followed by some sandpaper to get everything smooth.

Here's the finished product:
It looks even better in person. And under a finish the abalone really sparkles.