Tuesday, September 26, 2017

Building an Archtop Guitar


I have been interested in guitar building for a long time and now have more time to try my hand at it. An account of my first attempt at building a steel string acoustic guitar is described in a previous blog:


Since then I have built six guitars, all variations on flattop steel string design. I have however had a notion that I would like to attempt an archtop guitar, another branch off the guitar evolution tree. Archtop guitars have a carved top and back more in the tradition of instruments in the violin family. They are the preferred instrument of jazz guitarists but other variations exist.

In May of 2017 I attended the La Connor Guitar Festival, just north of Seattle, where about 50 luthiers presented their wares. One of the builders was Wyatt Wilkie of Royston BC. A guitar on Wyatt's stand caught my attention. It was an archtop but not your typical large body jazz variant. This was a small bodied acoustic archtop with walnut back and sides and spruce top. I checked the guitar out in one of the sound booths and took a few photos:

I talked to Wyatt a few times over the duration of the show. He has a one man workshop in Royston on Vancouver Island in BC, where he builds custom archtop guitars and mandolins. His work on display in La Connor was impressive. The designs of his instruments and choices of wood and ornamentation were original and tasteful and the craftsmanship impeccable. Wyatt worked for Bob Benedetto for two years which is a nice addition to an archtop builder's resumé. I asked him if he had ever considered having someone come work in his shop for two weeks to build an instrument under his guidance. He said he had considered it but hadn't yet got around to doing so. That evening I sent him some photographs of my work to give him some impression of the types of guitars I had built and the quality of the work and expressed interest in a two week, one-on-one workshop or master class in Royston. After the La Connor show we exchanged a few e-mails and agreed to the last week in September, first week in October for the class. Since this was a first attempt at such an arrangement, we agreed that we would approach it in an exploratory manner rather than try to get too prescriptive.

Having set the dates I went ahead and arranged travel and accommodation. I also started to think about obtaining wood. One of the sources I looked into was Bow River Wood to Works who have an EBay storefront and from whom I had bought wood in the past. They have a shop in Chilliwack BC but in talking to them I learned that they also had a branch in Courtenay and could ship orders to that location at no cost. That was a stroke of luck. They offer a package comprising all the woods needed to build an archtop guitar. I went ahead and ordered one. Wyatt agreed to pick it up from the Courtenay location.

We didn't take a picture of the actual set all laid out, however this picture from Wood to Works website shows what the package looks like when received:

Wyatt picked up the wood in good time and joined the book-matched boards that would be used in the top and back of the guitar. He then ran them through his just acquired vintage copy carver (more on this later) to take the boards to an approximate archtop profile and save me a good deal of hand carving.

I travelled to Courtenay on Sunday the 24th of September, dropped by Wyatt's shop briefly to discuss arrangements for the next day and turned up on Monday morning to start work.

Day One

  • Cut sides to size and sand to thickness
  • Make end blocks
  • Bend sides and clamp in form
  • Make kerfed lining strips

Here's a picture of the two of us (me on the right) in one of Wyatt's two workshops, this one mostly for hand tools, the other for large power tools.

The first task was to cut the maple sides to size and thickness them to 0.090" in the sander.

Then the sides were bent on an old fashioned pipe and gas torch rig. Wyatt has a side bending machine and forms for the various shapes of guitars he builds but this simple set up is quick and convenient. I have limited space in my workshop back home so I have decided to try to get proficient in using a bending iron. This approach is very versatile in that with skill one can bend wood to whatever shape is required using a single piece of equipment. Alas it's not as easy as it looks.

The damp sides are clamped into the form and left to dry. During this time the neck and tail blocks are made from mahogany, the former with a curvature to match the inside of the end of the guitar. The neck block is square in the area where the neck joint will be cut.

I have recently started marking and cutting the sides of guitars I build to size then gluing them to the end blocks outside of the mold. Being able to see the end makes it easier to get a good tight fit, which leaves open the option of a simple, neat joint between sides at the base of the guitar. If this is not a presentable joint one can later insert a decorative strip or wedge. In the case of this guitar we're keeping the design simple to save time. Wyatt also glues the sides to the tail block outside the mold but glues the rim to the neck block inside the mold. This ensures that everything is aligned and flat and since the joint at that end of the guitar will be covered by the heel, it is less critical.

The wood supplied includes strips of aromatic cedar for the linings which are glued around the edges of the rim to add stiffness to the ribs and increase the gluing surface. The linings can be continuous strips of wood, bent to shape or can be kerfed or cut almost all the way through every 1/4" or so to make the strip flexible. We decided to make kerfed strips. This involved improvising a jig similar to one I have at home. The following photograph shows the completed kerfed strips.

Day Two

  • Glue linings to rim
  • Make neck blank
    • Rout groove for truss rod
    • Rout ledge for fretboard extension
The kerfing strips can be glued with the slots facing the inside of the guitar or with the slotted side against the rim. The latter results in a stiffer structure and that's what I generally do. Wyatt prefers the more traditional approach so that's what we did.

I typically glue a continuous strip of kerfed lining to the edges of the rim then glue short reinforcing strips at intervals to add stiffness to the sides and reduce the risk of splitting along the grain. Wyatt's approach is slightly different. It involves gluing short sections of lining, leaving room for reinforcing strips which extend all the way from top to bottom of the rim, then inserting short pieces of modified kerfed lining on top of the reinforcing strip. I hadn't come across this approach before but that's what we did in this case. Here are a couple of photographs showing what it looks like when finished:

The block of maple provided to make the neck of the guitar was big enough, with some careful bandsawing, to make two blanks.

I normally cut the slot for the truss rod on a table saw. Wyatt uses his pin router with a rather intimidating exposed bit. I left this task to him:

Day Three

  • Continue work on neck blank
    • Rout dovetail in neck
    • Rout ledge for fingerboard extension
    • Glue ears to headstock
    • Install truss rod and cover with wood gusset
Next the dovetail is routed where the neck will join the body. This is done with in a jig which tilts the neck at an angle of 5 degrees so that when mated to the body it will have the correct alignment between fretboard and bridge. The jig, like most of Wyatt's fixtures, is simple and fit for purpose. It looks like he did enough work to make it so and no more. There are some very fancy and expensive jigs available to do the same job. In fact, we frequently joked about how the luthier supplier companies go to great lengths to come up with new, fancy jigs and tools, when very often simpler home made solutions work fine.

Next a ledge is cut in the neck for the fingerboard extension block. Another simple jig does the job.

The ears are glued on and the truss rod installed and covered with a strip of wood which is then trimmed flush with the fretboard and headstock areas.

The neck can now be set aside for the time being.

Wyatt is primarily a custom hand builder and not a fan of the use of CNC equipment as part of the process. However he did recently acquire a rare and unusual copy carver. It is a substantial piece of gear with a story behind it. Only three of these machines were made. Robert Benedetto had one when Wyatt worked there.

They don't make them like this any more. Wyatt heard from a fellow luthier that one of these machines was available. He had it shipped to Royston and managed to find a local electrician who was interested enough to work on it to get it up and running. The principle is that of a pantograph. The stylus on the upper deck traverses a shape by means of X and Y stepper motors. A router bit with a spherical profile the same as that of the stylus tracks in unison with the stylus and machines a wood blank on the lower deck. The machine is massive.

Day Four

  • Initial carving of outside shape of top and back plates
Wyatt had preshaped the plates for the top and back of my guitar on this machine. This saved a lot of preliminary carving effort. The results are shown in the following photographs:

Maple Back Plate
Spruce Top Plate

Next a ledge is routed around the edge of the plates to a thickness of 1/4". This is the reference thickness to which both top and back will be carved to and eventually a recurve or concave groove will be carved just inside the edge of the plate.

Most of the carving is accomplished with small luthier's planes with convex soles. It's hard work, particularly in the case of the maple back plate. The goal here is to achieve a continuous, aesthetically pleasing shape. While the initial rough carving on the copy carver imparts a starting shape, there is still a good deal of wood to be removed so there is scope for variation in final shape. I asked Wyatt if he he had any templates to check progress or final dimensions. The answer was no! This stage of the carving is completed when the shape looks and feels right. It will then be used as a reference surface to drill depth holes on the back side of the plates so that that side can be carved to result in a desired initial thickness across the plates.

It's a slow process, taking up most of the day. The work on the spruce top is a pleasant respite after the hard work of carving the maple. The planes are followed with scrapers of various shapes and sizes and then careful use of an orbital sander. During the latter stages of carving, the beginnings of a recurve contour are incorporated in the shape.

Day Five

  • Carve inside surfaces of top and back plates
    • Drill depth reference holes
    • Carve, plane and scrape to shape
    • Cut soundhole in top plate
Having carved the outsides of the two plates to near final shape, holes are drilled on the inside surfaces to be used as reference depths when carving to the desired shape and thickness. This is achieved by setting the drill press up so that when the bit hits the depth stop, it is the desired distance (thickness) from a wooden post located directly below the plate.

For the back, the drill press is set up to leave a thickness of 0.160 in. throughout the area to be carved. In the case of the top a flat area will be left in the vicinity of the neck block and fingerboard extension and an area along the long axis of the plate is drilled to a reference thickness of 0.240 in. The additional thickness along the axis of the plate ensures strength and stiffness where needed to resist the tension of the strings and the resulting compression of the bridge against the top. The photograph below shows the two regions marked out ready for drilling

Here's the back drilled and ready for carving. A flat gluing surface is left around the edge and in the vicinity of the neck and tail blocks.

The remainder of the day was spent carving. It's blister prone work. I had brought a pair of work gloves so I thought I'd try wearing them. Wyatt was not impressed, so I quickly gathered that blisters were considered an integral part of the process, or rather that he thought gloves would compromise one's ability to touch the plate frequently to check progress by feel as well as sight. As with carving the outside surfaces, the task was much easier for the softer spruce.

When the top is near final thickness, the oval soundhole was cut on the pin router.

Day Six (Saturday)

  • Carving/scraping to refine contours of both plates
I decided to spend a few extra hours on Saturday morning, to take my time to carve, scrape and thickness the plates to the point where I was comfortable with proceeding to the next stage, gluing them to the sides, after which the insides would not be accessible.
Blisters notwithstanding, I enjoy carving. The thickness across the plates is checked and marked using a caliper and various scrapers are used to fine tune the contours, including the recurve sections. By the end of this half-day session the two plates are in pretty good shape.

Week two

Day Seven

  • Prepare material for bindings and X-brace
  • Bend binding for guitar rim and sound hole
  • Mark and cut braces to conform to top plate contour
  • Sand braces to final fit then cut X-joint and glue in place
In the interests of time Wyatt and I had decided on very basic ornamentation for the guitar. This suits me fine, I like an elegant minimalist look. We go with a single strip of rosewood for binding the body and soundhole, no fretboard or headstock binding and a simple rosewood headplate.

The binding strips are cut and thickness sanded to 0.080 in., then bent to shape on the iron pipe. The soundhole binding is fitted, then taped and glued in place and trimmed to shape.

The top is to be braced with a single X-brace, positioned so that it runs directly under the posts and adjustable screws of the bridge. The quarter sawn spruce stock is cut and sanded to 8mm thickness (note mix of metric and imperial units). The curvature of the plate's surface is transcribed onto the brace using a drawing compass. The shape is then cut on the bandsaw and refined by working the brace back and forth in short movements over a piece of sandpaper stuck to the plate.

Once a satisfactory fit between braces and plate is achieved, the braces are marked, notched then glued together and clamped and glued to the plate.

Next the rim is prepared for assembly. The edges are marked with pencil then leveled with a sanding board to ensure a true surface for gluing to the top and back.

Day Eight

  • Glue back plate to rim
  • Shape braces on top plate
  • Coat inside surfaces with shellac
  • Glue top to rim
  • Slot fretboard
  • Rout plate edges flush with rim
  • Sand and scrape sides
  • Rout grove for binding
  • Install, tape and glue binding
This was a long day but with good progress as the component parts are assembled. The back plate is glued to the rim and while the glue is drying on that the X-brace is shaped

Inner surfaces are given a coat of shellac, staying clear of subsequent gluing area. The top plate is signed for posterity and then glued to the rims using Wyatt's robust press clamp and cauls. This is a good way of doing it as it allows full visibility of the gluing surface between the top and the rim. Any suspicion of a gap can be addressed by loosening the clamp, inserting a piece of card or folded sand paper and tightening things up again.

The fretboard is sanded to thickness (0.24 in) and slotted on a table saw with a 25 in scale template. The edge of the thin kerf blade can be seen peeping up above the jig. The fretboard is attached to the underside of the plexiglass template with double sided tape, run across the blade then lifted back to the starting position, moved so that the next slot in the template engages with a reference pin then run across the saw blade to cut another slot and so on.

The body assembly can now be removed from the press and the overhanging edges trimmed on the pin router

At this stage the body of the guitar looks neat and feels good. It has a pleasant resonance when tapped.

As mentioned above, the design has been kept simple, incorporating a single strip of rosewood binding around the edges. The groove for this is cut on a table router, cleaned up with a chisel, file, and sanding sticks, and then the pre-bent binding strips are glued and taped in place.

photo 1-1
photo 1-2

Day Nine

  • Remove binding tape
  • Wipe top and back with damp paper towel to relieve compression dimples
  • Scrape bindings and blend into recurves on plates
  • Cut taper dovetail mortise in body
  • Plane fretboard to 12 in. radius
The binding tape is carefully removed from the body and the results inspected carefully for gaps. The outcome was good. I did find a few small areas where I applied some rosewood dust and superglue to fill small gaps.

The process of drilling holes in the inside of the plates (see day five) results in small dimples in the plates where they were pressed against the wooden dowel below. I was preparing to sand the surfaces to the level of these indentations when Wyatt told me what would happen if I did. The compressed wood fibers would eventually recover resulting in clearly discernable bumps where the dimples had been. This would normally happen gradually and could give one a nasty surprise if the guitar had been finished and french polished or laquered in the meantime. Wyatt's remedy is to wipe the plates with a lightly dampened paper towel. The wood in the dimples expands and they disappear. Some sanding will be necessary later but most of the expansion has now occurred.

The bindings are sanded and scraped flush with the top, bottom and sides of the guitar and some more work done to give a smooth transition from binding through the recurve area.

The taper dovetail mortise is cut in the body/neckblock. This uses another of Wyatt's 'fit for purpose' jigs. Pipe clamps hold the body fast and a router template is clamped to the top then the mortice is routed with a dovetail bit. Bit depth is set to be slightly deeper than the corresponding tenon previously cut in the neck. This allows some room for final adjustment and fitting. The resultant small gap also facilitates the injection of steam through a hole drilled at the 14th fret position (after removing the fret), if the neck ever needs to be removed for resetting.

Next the fretboard is cut to final tapered form and dimensions. This is done on the table router with the appropriate template. The fretboard is then shaped to a 12 ft. radius curvature. I think Wyatt normally starts this on the belt sander then finishes it off with a radius block and sandpaper. At home I try to avoid generating wood dust when possible since I have only very basic air treatment equipment. I normally radius fretboards using a jig I made and a router but on this occasion I decided to do it using Wyatt's nice Lee Valley block plane. It was relatively easy to get to near final curvature, checking with the radius gauges, then finish the process with a radius block and 220 grit sandpaper.

The fretboard is glued to the neck. Small holes are drilled through the first and twelfth fret slots and small brad nails tapped through to use as location pins when gluing. The caul used has holes in these positions to allow for the pins. 

Day Ten

  • Thickness and taper back of headstock
  • Make tailpiece
  • Shape neck dovetail
  • First fit neck to body
  • Carve neck
The headstock is brought to final thickness with a gradual taper from nut to tip. I generally use small planes for this and I have used a Safe T Planer (bit of a misnomer) in a drill press. Wyatt took it to the pin router but ended up with some tearout of the grain so we decided to put a veneer on the back of the headstock after all. This will take a little more time but will also cover the ear joints and give me a chance to add a nice decorative motif where the veneers blend into the shape of the neck.

To achieve a nice geometric effect, a gentle concave curve is sanded into the neck where it joins the headstock.

Veneers are glued and clamped to conform to the shape.

When it comes time to carve the neck, the intersection between the curved veneers and the curve of the neck will hopefully result in a nice effect.

Next I start work on the tailpiece. I decide to adopt the design Wyatt had used on his small walnut archtop guitar which was on display at La Connor. In this case the wood will be rosewood and possibly maple rather than ebony and walnut (see photograph at start of blog).

The process begins with some work on the pin router to cut the slots for the string recesses and the tailpiece cable fastener. Then holes are drilled from the two ends of the tailpiece to intersect these slots. The tailpiece is then trimmed to shape on the bandsaw.

Next comes work on the neck to get it ready to join to the body. The tenon which was routed in the jig is now shaped into a tapered dovetail to match the one routed in the body.

As the dovetail starts to take shape the neck can be test fitted to the body and adjustments made. As it settles into position, the contour of the top plate is transcribed to the fretboard extension, using a drawing compass, and it is carved to shape also.

Once the joint is an approximate fit, the rest of the neck is carved prior to final jointing and gluing to the body.

I normally carve necks with spokeshaves. I like doing it and as mentioned above I have a preference for carving rather than sanding where possible. Unfortunately the only spokeshave Wyatt had was a very small one (he normally does the initial shaping of the the neck on a belt sander). It was pretty heavy going and sore on the thumbs (my iPhone didn't recognize my thumbprint for several days afterwards). It wasn't good for Wyatt's small spokeshave either.

As the neck is carved, the intersection with the headstock veneers starts to look attractive. By the end of the day the neck is carved close to final shape and the dovetail is a good approximate fit to the body.

Day Eleven

  • Scrape sand body and binding
  • Refine neck/body fit
  • Install frets
  • Drill holes for tuners
  • Glue neck to body
Final day. The body is cleaned up and brought to near final contours with an orbital sander with 220 grit paper and scrapers.

Final fitting the neck to the body is a crucial task. A good fit and precise alignment is fundamental to the guitars action and playability. The task takes me a long time. Chalk is applied to the dovetail and after test fitting the joint, the obvious high points are trimmed with a chisel and sandpaper. As the neck dovetail settles down into the body, the fretboard extension starts to contact the contoured surface beneath it. Carbon paper is placed between the components and carefully withdrawn while some pressure is applied to the joint.

I made good progress at first but slowed down during the late stages. There was a lot going on. adjustments to the dovetail can also affect alignment of the neck to the centerline of the guitar so this must be checked and factored into subsequent adjustments. It takes some skill to determine exactly where resistance to fit is located - is it in the dovetail or under the fretboard extension. The goal is to have a seamless joint with everything precisely lined up. I got to a point of diminishing returns so in the interests of time, Wyatt assisted with the final fitting. He was more confident and assertive with the chisel at this stage than I was. Eventually we are satisfied with the fit and move on to the installation of the frets.

These are cut to approximate size, set aside in a holder with numbered holes for each fret slot, then tapped partially into the fret slots. An arbor and caul with the same curvature as the fretboard (12 ft radius) is used to press them home.

Holes are drilled for the tuners in the front of the headstock and then larger diameter holes drilled part way through from the back to accommodate the wider bushings on that side of the tuners.

Now it's time to glue and clamp the neck to the body. It is left clamped for about an hour. This is the stage at which I had to call it a day. I was about to head south to Victoria to stay overnight and catch a plane early the following morning.

The guitar is not quite finished. There's still some work to do on the tailpiece and a bridge to be made before it can be strung up and set up. This will have to wait until I'm back in Houston.

As mentioned at the outset, this session was a trial run for a master class which Wyatt is planning to offer. The goal was to test the process and get some idea of how long the various tasks would take in this situation. Along the way we both identified opportunities to save time. For example we made all the components from sawn wood. Time could be saved by buying pre-made kerfing, binding and perhaps a pre-slotted and radiused fretboard. Prospective students will have to have some prior building experience and have probably little to gain from doing tasks the long way if they have done them before. Time spent on these parts of the schedule could be better used if focused on the more critical tasks specific to archtop building.

Our overall assessment of the two week exercise is that with some revision of the schedule and assuming a reasonable level of competence and experience on the part of future participants, it should be possible to build an archtop and have it set up and in a playable state (in the white) by the end of the course.

I thoroughly enjoyed my two weeks in Royston and thereabouts. It's a beautiful area, not that I had much time to explore. At this time of year most of the tourists have gone so it was easy to find accommodation. I rented an apartment in nearby Courtenay. Here's the view across Comox Bay to the mainland from the shore close to Wyatt's place.

I managed to get across to the other side of the bay (to the left on the photograph) one evening to check out Wyatt's recommendation for fish and chips. Good fish and chips are hard to find in the US. I tried the halibut and chips at Troller's and it was excellent - the view from the outdoor tables is looking back across the bay towards Wyatt's place.

The work was challenging at times. Wyatt was always available when needed and always forthcoming with advice and help. I had the rare opportunity to work one-on-one with a master archtop builder. I could probably have built an archtop guitar myself using resources such as books, DVDs and online information. However, it would have taken me a long time and I would have undoubtedly made mistakes along the way. This was a much better and faster way to learn and a great way to gain insight into what it takes to make a living as a luthier. Wyatt is an ardent fan of early jazz and has an extensive collection of CD's and original 78's. I am also fond of jazz from that era. With CD players in both work shops vintage jazz was an integral part of the ambiance.

Thanks also to Wyatt's wife Emily. She had to put up with having a visitor around the place for two weeks.

Jake Booth

Since Then:
After getting back to Houston, I was able to step back and think through how I would like to finish the guitar off. After some additional sanding and scraping, I finished making the tailpiece then made a bridge and nut. The tailpiece is based upon Wyatt's original design on the guitar that initially caught my attention in La Connor and led to this endeavor, however, I did not add wings. I did inlay a mother of pearl celtic triquetra in the tailpiece and one in the headstock. I've done so on all my guitars to date so it has become a logo of sorts.

Once all the parts were available I was able to string the guitar up 'in the white' to check alignment and see how everything looked together. I'll leave it like this for a few days to let everything settle under string tension. Here are some photographs of the guitar at this stage.

There's still a lot to do by way of setup and then preparation for and French polishing. Even at this stage, however, the guitar sounds good.


I have been trying my hand at French polishing for my last few instruments. This is an ancient technique involving shellac dissolved in alcohol. The shellac is derived from a resinous secretion from Lac beetles which is deposited on tree branches to protect larvae as they develop. It is harvested, cleaned and sold in flake form. It dissolves in alcohol and can then be applied to wood, gradually building up a beautiful lustrous finish. Apart from being a traditional and fascinating technique, it appeals to me because it is a natural product with low toxicity whereas spraying nitrocellulose lacquer, which is the most common finish on commercial guitars, involves working with rather nasty VOC's.

To finish the guitar I gave it a good detail sanding to provide a smooth surface for polishing, which tends to reveal unnoticed imperfections. I then applied a seal coat made up from blonde shellac flake dissolved in denatured alcohol. This was followed by build coats of a product called Royal Lac, which is shellac with the addition of synthetic resins to increase its durability. After a few coats I sanded carefully with 1000 grit paper and applied additional coats using a small pad consisting of cheese cloth inside a piece of old cotton t-shirt. As the layers build, a drop of olive oil can be applied to the pad to make it less inclined to stick.

Additional coats can be applied and burnished to as high a shine as one wants. In this case I stopped short of a really high shine. I prefer this look and one of the advantages of French polishing is that it is easy to repair or to apply additional coats at a later time should one want to.

Below are some photographs of the finished guitar. It sounds good, quite different tonally from the flattop guitars I've built to date and better than the two archtop guitars which I already own (a Hofner Jazzica and Godin 5th Avenue). I am pleased with the outcome.