Carving and smoothing the inside of the tops

Last post, I showed how I surfaced the inside of the backs to an even 3/16". Surfacing the inside of the tops is a little harder - because of the string tension, the tops need to be a little thicker than the backs under the bridge, yet because the tops are the primary sounding board of the instrument, they need to be thinner than the backs near the edges. Thickness at different spots on the tops ranges from 1/8" to 3/16, and needs to be graduated carefully for the best resonance of the sound board. Here's how I'm doing it:

I dug out about half the wood quickly, guessing at about 3/8 thick overall, because it would mean less jarring gouging work later. Then I mapped out 3 different depths to work to: 1/8" around the rim to about 3/16" under the bridge in the center. Then on to the drill press once again.

Using the same setup as for the backs, I drilled about a million holes with my hand-ground, no-spur 3/4" Forstner bit, angling the work a little as I went to make the bottoms of the holes vaguely parallel to the other sides.

Guess the number of holes and win a prize! Note how the center parts of the plates are a little "raised" - that's because the holes were 1/32 less deep than those nearby,

Next step was very pleasant - planing down to the bottoms of the holes with my hand-made double convex plane. This tool works really well, and is indispensable here. This is the job I made it for.

Planing getting done - about 1/2 hour a plate.

After planing, I used a hand-held curved scraper to reduce the thickness overall, checking repeatedly with the thickness caliper. Learning to sharpen and use this tool was another of my goals for this project. I must say, it works much more aggressively than I had thought. With a good cutting burr, it gets too hot to hold after about 5 minutes of use. WHich is OK, because my thumbs start to ache from holding it at the proper angle.  I'll use this tool again for scraping the recurve of the tops and backs once the body is assembled and the bindings are on.

After scarping the plate to the thicknesses I wanted, I gave the inside a quick smoothing with medium-grit sandpaper. I hold it up to angled sunlight like this to see where there might be irregularities - such as the waves in the upper rim and a slight bulge to the right of my thumb. These will be scraped and checked until each back has a fair curve to it, and then the plates will formed at last.

Rough carving the inside of the back plates

Shaping the outside faces of the tops and backs involved terrain mapping the wood blanks to a given shape, through the contour templates I made. Now that the outside faces are carved and smoothed, I can use these faces as a reference for carving the opposite (inside) sides.

The backs are supposed to be carved to a consistent 3/16." The tops, however, vary in thickness from 1/8 near the rims to about 3/16" in the center, under the bridge. To complicate this, the tops need a fair curve along the longitudinal axis for the parallel bracing.

To carve the backs to a consistent 3/16", I started by hogging out a lot of waste with this drill press jig. Basically, I made a soft (cork) topped 1" pin directly under a 3/4" Forstner bit which I reground to have no center point, and on which I extended the cutting blades into the middle. I set the distance between the cork and the bit at a hair over 1/4". and made as many holes as I could without overlap (which might have made the work slip). 

The result was a large number of reasonably flat bottomed holes, the bottoms of which were reasonably parallel to the opposite side of the back plate.

Then I made another pass at the drill press, this time with a larger bit, removing as much of the "honeycomb" between the 3/4" holes as I felt comfortable with. This would save me a lot of gouging in the next step.

Next job was to connect the bottoms of the holes with a sharp shallow gouge and that neat little double convex plane I made.  Took about an hour per piece, but was satisfying work.

After the backs were gouged, planed, and smoother a little more with rough (60 grit) paper and a hand block, I made "scraping maps" with my depth tool, to show me where hand scraping was needed. In the photo above, the center cross-hatched areas need about 1/16 off, while the other outline needs from 0 to 1/32. "  I went through this process of scraping and measuring about 4 times with each piece, until I was satisfied that it was "good enough" - by which I mean a fair, smooth-feeling curve on both sides, pretty close to the dictated thickness. 

Mapping thickness for scraping, using the depth caliper. The tool is on a high spot, which needs a lot of wood removed to get to 3/16. 

Rough carving the convex sides of tops and backs

Honestly, I have no idea how to carve wood. I can make really lovely rectangular objects, but once things get wavy, round and curvy, I'm lost. That's part of what I'm trying to learn building these guitars.

But I still want to succeed at this carving thing, so I developed a plan to start me out on the right path.

 As I showed in the first post, I made "contour map templates" of the top and back, based on the sections in Bob Benedetto's plans. Here's how I used these to rough carve the tops and backs.

I started with glued-up slabs of African Mahogany for the backs (and sides) and Western Red Cedar for the tops.

I used my Wagner Safe-T-Planer (no longer made, alas, although you can find some here and there) in my drill press. I use brass measuring bars to set it at diminishing heights, starting with the first cut from the center of each blank.

Here's my "terracing" setup. I made 1/8" cuts from the top down. The green hose is a directed fan clearing chips.

After each cut, I marked the next level with the countour templates. I was able to use the same depth setups for the tops and backs, so I did them all in one shot - about 2 hours at the drill press. Not bad.

After all the terraces were cut, I used a round-bottom spokeshave to remove the waste from line to line.

With all the terraces faired together, I took a final pass with the Safe-T-Router to 3/16" along the area where the "recurve" will be. The recurve is a shallow negative curve along the perimeter of the tops and backs, which lets the tops and backs resonate more. It's also beautiful, and will be a challenge to scrape and carve. Here is an incurve on another archtop (not my work)

A final rough-carved top. A little sanding to fair out the plane gullies, and it's ready for glueup.

Making a double convex palm plane

I'm so cheap. I knew I'd need a special tool to carve the insides (concave sides) of the tops and backs, and that tool was a double-convex palm plane. Lee Valley / Veritas makes a nice one, but it's $80 delivered, and besides, I like making things. So...

My son gave me a nice cast-iron palm plane this last Christmas, and I decided to use it as the base for the plane. All of the parts were a little rough. I started by truing the sole flat, and regrinding the cap flat and thinner at the base. I had a nice scrap of bubinga for the falsie sole.

I glued the newly-flattened plane sole to the bubinga with JB Weld, which glues anything to anything, although not beautifully. 

On my stationary belt sander, I first ground a longitudinal radius of maybe 5 or 6 inches, then a radius of maybe 2 inches or so. This was all by eye and feel. When it felt fair in my hand, I sanded it up to about 250 and polished it with a 220 sanding sponge. Then I spent about 3 hours chiseling and filing out the slot for the blade. 

I ground and honed the blade to a slightly wider radius than the sole. If you don't do that, the edges don't extend as far below the sole as the center when you set it up.  I did the rough grinding on my stationary disc sander, and I have a set of ceramic Shapton stones I like for sharpening. The blade was not very flat to start with, but I'm lucky it was a little concave, as you can see, rather than convex, which would have taken forever to flatten. 

Here's the assembled, working plane. Not very pretty, but it works really well. On first tests, I found I had to make the front part of the throat more vertical for chip clearance, and I filed the cap really thin and sharp near he blade. It makes consistent long curls in scraps of both cedar and mahogany, but I haven't dug really deep with it yet.  I may or may not decide to clean up the exterior appearance. If it works as well as I hope it will, maybe it'll be nice to it.

First, some new tools and jigs

I'm a pretty good cabinetmaker, with a decent shop. But almost all my tools are dedicated to making rectangles of one sort or another; I'm good at plumb, flat, square and level. So I needed to make some special tools and equipment. Some of these won't be used until midway in the game, but I had some time while I was getting basic materials together.

This is a mold for holding the steam-bent sides in position (with the pressure cauls) and for holding the sides vertical when assembling the bodies of the guitar(s).

These are 1/8" interval "contour map" templates of the outside (convex side) of the tops and backs. Getting these contours took all my skills in logic, tracing, and page layout software. I traced horizontal sections of the top and back from the plans, scanned them and imported them into Indesign, and applied a 1/2" grid to each image. Where the outline crossed a gridline, I made a mark, then transferred all these "altitude marks" to a full-sized tracing of the front and back. I then spray-mounted these tracings to 1/8" masonite, connected each dot of the same height with a curve I felt appropriate, and cut all the resulting lines on a jig saw. These templates will be used in rough carving the tops and backs, in a way I feel will be safest for a first-time archtop carver.

This is a carving clamp for the tops and bottoms. It's made to fit one of the vices on my bench. I can use it for both sides. A decent sheet of 3/4" AC plywood was on sale at HD, and I grabbed it for this. A couple of coats of water poly, and simple hand knobs for the clamps.

I made this thickness gauge based on Bob Benedetto's design, and improved it a little with the easier to read indicator. It's really neat - you put the part of the top or back you're carving in the jaws, and pull the trigger. The string on the roller pulls the springy top down so the two dowels touch top and bottom, giving you a readout. All made from scraps laying around, including a nice chunk of 3/4 Hondouras from about 30 years ago.