For old Woody, December is a tiresome month. WELCOME January, 2017 !
Now, everybody... Get back to work !
Note: If the pictures cover the text, just zoom in or out a bit.
Closing out 2016 with the completion of a mandolin started in 1982. 2017 in progress...So far, so good.
December, in addition to being the middle of the eating season, was a full plate in Woody's shop. Work continued toward the completion of the mandolin on three different fronts__
* Finish carving of the spruce top
* Joining the neck to the body with the dovetail joint
* Laying out and cutting the fret slots
Finish-Carving the Top
Last month the top was rough-carved to the approximate final contours and thicknesses. Finalizing the process primarily involves scraping, sanding, checking and re-checking the evolving contours for symmetry and thickness as shaping continues. A key element is the symmetrical distribution mass across the top.
One might think of the top of the mandolin as a trampoline. The center mass moves up and down due to the relative flexibility of its perimeter. Flexibility is controlled by thinning the outer perimeter of the top at the recurve (~20 mm from the edge of the mandolin). The center of the top is more than twice as thick as the recurve. The idea is to allow the center of the soundboard, as it is driven by vibrations through the bridge, to oscillate up and down smoothly and symmetrically. That way the energy from the vibrating strings is transferred evenly and completely across the top.
To many, the carving of the top is like watching glue dry in terms of excitement, both in the doing and in the describing. Controlled thicknessing is, however, VERY important to the instrument's tone and playability. The thickness of the soundboard should taper smoothly from thickest in the center under the bridge (~6.5 mm) to thinnest at the recurve (~3 mm). These suggested dimensions are 'advisory' only. What we are really trying to control is stiffness of the soundboard which is not always linear with thickness. That is, wood in the same soundboard may vary in density along any line across the soundboard width or length. Each soundboard is unique.
Pictured below is a carved soundboard back-lighted with a 100-watt spotlight. The dark stripes running vertically are from the tree's slow growth during a few unevenly spaced years, resulting in slightly tighter grain and varying stiffness across the board. This is not unusual, of course. It's just and example of the uniqueness of each piece of wood as carrier of vibrations.
You might notice in the back-light photo that there is a heavy dark spot just behind the bridge location (the rectangle). The back-light silhouette, and physical measurements, indicate that this area needs to have more wood removed... maybe a half millimeter. Removing this wood will enhance the volume and richness of the instrument's voice. How much is enough?... one stroke less than too much.
Bracing the A-Style
The structural plans that I use are from the 1920's Gibson A-style mandolin. By the 1920's Gibson structural design had evolved to incorporate X-bracing. The first mandolins that I built (back in the 80's) were based on a 1908 Gibson A-style that I owned at the time. The 1908 mandolin had only one brace and that was between the sound hole and the bridge, across the soundboard (transverse). Structurally, it proved to be inadequate and I had some trouble stabilizing the movement in the soundboards of a couple of mandolins. A few years later Gibson changed to X-bracing which is what I use now.
The need for bracing the A-Style mandolin comes from the location of the oval sound hole and the physical fact that a standard set of mandolin strings will generate 45-50 pounds of downward pressure on the belly of the soundboard. That's the equivalent of a small child standing on one foot placed in the center of the mandolin top. (Of course, a child would never do that.)
F-style mandolin soundboards are under the same string pressure, but F's are carved differently than A's. The bars that run the length of the F soundboard (roughly parallel to the gain) are for tuning the soundboard more than structural support. The 'tone bars', as they are called in violins and F-style instruments, impact the speed and amplitude of vibrations as they travel out from the bridge.
This being so, the bracing significantly impacts the tone of the instrument. Once glued to the underside of the soundboard, the braces will be carved until the target stiffness of the soundboard (resonance frequency) is achieved. More on this next month when we discuss "tuning" the top by modifying the height (stiffness) of the bracing. Right now I am thinking about structure, i.e., getting all of the various parts of the instrument joined together.
Above is the same nearly-finished soundboard as in the back-light picture. The soundboard is resting now after Woody has tinkered with it for about a week.
Dovetail and the Mandolin Neck
A dovetail joint holds the neck onto the body of the mandolin. The neck extends from the body at a 6-degree angle from soundboard plane. This means that the dovetail, a 14-degree mortise and tenon, must be cut at a 6-degree angle to the neck blank surface creating a compound angle dovetail. If this sounds tricky that's because it is.
The neck blank is a maple block, ~ 45mm x 400mm x 60mm, squared and surfaced on all 6 sides. By temporarily gluing the neck blank to a 6-degree plywood wedge (with contact cement), the dovetail tenon is bandsaw-cut at a 6-degree angle to the neck surface.
Be sure to layout the outline of the dovetail carefully. Pay particular attention to ensure that the centerline of the neck is a true extension of the centerline of the body. Using the band saw to both cut and rasp, the dovetail can be roughed out as shown below. Later, after the dovetail mortise has been cut into the headblock, the dovetail tenon must be worked by hand to get a tight fit.
The dovetail mortise is routed into the headblock with a standard 14-degree dovetail bit. I've built a fixture to hold the headblock and rim in place while routing. Use a straight bit to hog out the center of the mortise. This will create a rectangular slot to the required depth of the mortise (multiple passes, no pass deeper than half the width of the bit). After the slot is routed to depth, change to the dovetail bit for the last pass to make the 14-degree dovetail walls on the side of the slot. Pictured below are two blocks that I used for practice before I cut the finished joint on the mandolin headblock and neck.
On line, Roger Siminoff, and others, have good videos on how the dovetail is cut. These two videos will get you started.
Practice ... It's a good idea.
While Woody is waiting for glue to dry (or baseball season to start), I started the process of shaping and slotting the fretboard. The fretboard is left a bit oversize until the final carving of the neck is complete. That way, the fretboard can be carved and sanded flush with the edges of the neck. The fret slots are a different matter.
First, thickness the fretboard to approximately 5 mm. Following that sand a 12" radius into the top of the fretboard being sure to keep everything symmetrical to the centerline. Then, because my fretboard stock comes wider at one end than the other (trapazoidal), I have to glue an wedge onto one side of the fretboard stock to create an edge that is parallel to, and 24 mm from the centerline. This is necessary so that the fret slots can be easily cut perpendicular to the centerline in the miter box or CNC.
Using a 0.375 mm (0.025") router bit, the fret slots are routed on a CNC machine built for me by my son...he's very good at designing and building machines. For years, Woody laid out and cut fret slots by hand and eye in a customized miter box. Now, the CNC precisely routs all of the fret slots, position markers, and the slightly over-sized perimeter of the fretboard in one half-hour run.
That's where things stand right now. In January I hope to install the truss rod, shape the peghead, cut the sound hole and complete the soundboard bracing, and tuning. Hope to see you back next month.
JE Woody Strings