Guitar Build
Or...Why not Just Buy a Guitar?
This is a question that my Uncle always asked me. Throughout my life, when I would see him and he’d ask what I was up to and I’d tell him I was building or fixing random thing X, he’d always laugh and ask, “Why don’t you just buy that?”
When I was younger it puzzled me but as I got older I realized he knew the answer. When I was 12 I built a banjo. I spent a lot of time on it, my grandfather helped, and it looked great. But it never sounded right. And back then it wasn’t very easy to troubleshoot that kind of thing, so it was relegated to the attic. And that cured me of building musical instruments for a while.
Fast forward several decades and I’m noticing that my arm is starting to hurt while playing the guitar, mild case of tendonitis, but it nags and returns. My son is convinced that my guitar is to blame. I have an old Gibson with the baseball bat neck and he kids me that it hurts his arm when he plays it. I have a feeling its more about my technique and age than the guitar but I lay off for a while and start to look at more modern guitars.
We seem to have a Bad connection
Several years earlier we had bought my son a guitar for his birthday, so I had been down this road a little bit. I’m not a great guitar player but have played long enough to have an opinion, and like a lot of people my ears and eyes outspend my talent by a wide margin. It would be nice to have an awesome guitar but I’m not that good and don’t play enough to justify it. The other challenge is that I can’t find anything I get really excited about. I’ve had my Gibson for a long time. Its not a great guitar, an L-30 which was a student model, but its from the ’30’s, sounds good and has tons of character. Instruments are kind of personal and I’m finding it tough to match the personality of my Gibson when I’m looking around the local guitar stores. I’m also having trouble figuring out the value equation; spending more doesn’t necessarily get you a better sounding guitar. It may, but it also may just get you something that’s flashier. And then you always run into guitars that aren’t setup correctly. Not a great vibe so far. The search continues.
At this point I’m pretty sure I want a solid wood guitar (vs. laminated back and sides – this is one of the delineations in guitar quality, although modern technology is blurring these lines) so one day as I’m googling this I bump into a guitar kit from Stewart MacDonald. Ironically enough, this is the same company I bought my banjo plans and supplies from back when dinosaurs roamed the earth. Now known as Stew-Mac, they offer several kits, and the reviews are generally very good. Which I don’t buy for a second. There are also people who believe the earth is flat. I chalk it up to cognitive bias and move on in my quest.
Stew-Mac OM Kit
I bang around some more guitar stores, online and brick & mortar. I’m reminded of the dairy cooler at the supermarket, the variety is endless. But I’m still not finding a connection. I pinball back to the kits, more reading reveals that they still get good reviews but they’re not inexpensive – basically this is for the experience, you’re not going to save a bunch of money building from a kit. Then I bump into a guitar kit building forum (who knew?). I read, join, and ask questions. There’s also a section for building from scratch, I ask questions there too. Very nice folks, lots of different backgrounds and skill levels, but I am assured that its very possible to build a guitar that sounds good. Maybe not great, but good. I’m still skeptical but it’s always comforting to find the wing of the asylum where your gang hangs out.
Hmmm.
Aedifico, Ergo Sum
I go back and forth between the stores and the forums. Further poking around in the kit & scratch building forums starts to give me a rough outline of the pros and cons of each approach. The advantages of the kits are that they give you a big head start which in turn means you’re more likely to finish. And you won’t need a lot of specialized, expensive tools along the way. The advantages of building from scratch are that you have more options and flexibility in terms of wood choices, design elements, construction techniques, etc., which I happen to find very appealing. I’ve built a bunch of stuff over the years, and no matter what it is, I always end up building from scratch. This started with balsa model airplanes: the kits are a great start but the balsa wood in the kits is usually pretty heavy and the tissue is crap – so if you want better performance you select your own, lighter wood, cut your own parts, use Japanese tissue… and you can see where this goes. So as I peel the guitar onion I’m starting to see the same opportunities, and by now I have enough ideas about what I want to do that I’m pretty sure the kit approach won’t get me there.
So although I have no idea what I’m doing, somehow my brain starts to crystalize a scenario in which I will select all of the best elements and assemble them into a Super Guitar on my first try. In the meantime I’ll get an inexpensive new guitar to play for now, and sell it after I build my masterpiece. Of course, brilliant! Why didn’t I think of this sooner? I’d be done by now.
I need a Plan
The next step is to decide on the particular design to build. I confess that as long as I’ve been playing the guitar I’d never paid much attention to the different body shapes, so now I’m learning that there are a lot of styles to choose from. Interesting and also confusing, again I’m at the dairy cooler. Most of the steel string acoustic guitar history in the U.S. goes through Gibson and the C. F. Martin Co. I’ve never owned a Martin but pre-war Martins are considered a high water mark in guitar history, so I start there. And as you can see below, there are a lot of Martin body styles to sift through. Fortunately the availability of plans narrows it down pretty much to the Dreadnaught, OM, and OOO. More reading, more glazing over, but ultimately I find myself gravitating towards the ‘small body’ OM & OOO models. I’m more accustomed to the body on my Gibson which is similar in size to a Martin OM, I’m more of a fingers-style player, and it turns out that I just like the shape of the OM. I have a brief fling with the OOO 12-fret slot head design, but ultimately come back to the OM as a good starting point. And because I’m building from scratch I’m going with a deep body (+1/2″) and short scale (24.9″). So I order the OM plan from Stew-Mac.
How hard could it be?
Castles in the Air
As I wait for the plans I read, ask questions, and start to build in my head. Guitars have been built for a few hundred years and there are a lot of different approaches. The basic outline for a steel string guitar is to bend the sides, attach the top and back, shape the neck, slot and fret the fret board, attach them both to the body and then string it up. But there are lots of ways to accomplish these things, ranging from traditional hand building methods to modern CNC engineering.
After more online research, three books and a lot of virtual building I opt for a method that uses an outer mold for the body and a bending machine to shape the sides. One of the things I notice early on is that there are very few right angles or flat surfaces in this project. Things that look flat to the inexperienced eye, like the top and back, are actually built with specific radii. And squaring things to curved surfaces can be a challenge. I also learn that the neck joint plays a critical role in the proceedings, so I like the idea of using an outer mold because the flat surfaces of the mold will provide reliable reference points, at least in the early going.
The plan arrives. In one sense it looks fairly straightforward. There’s an easy elegance, an air of confident simplicity to the lines on the paper; this is after all a classic design. But there is more to this than meets the eye. On paper it’s a woodworking project, a bunch of different shaped pieces that go together to form a guitar. But there is another dimension that doesn’t show up on the plans: the musical dimension. This is where the magic happens. It if its built too heavily it won’t resonate, too lightly and it will collapse. So I’m seeking a delicate balance where it sounds amazing but won’t implode. How to find that, I have no idea at the moment – but I’m confident that this will be revealed to the pure of heart when the time comes.
And so we begin our quest.
Into the Woods
I built furniture after college for several years but used primarily oak, ash, poplar and maple with some mahogany thrown in now and then. Very nice woods but also fairly boring in terms of colors and grain patterns. Now I have the world’s wood buffet spread out before me, and its very exciting: pale moon ebony, ziricote, purpleheart, leopardwood, all sorts of crazy stuff.
There’s a lot of importance attached to wood selection in guitar building, but again lots of opinions. One of the things that I’m enjoying about this process is that there’s very little agreement on any topic, and wood is no different. Fundamentally you want it to sound good and there are some tried and true combinations, but its also tempting to get sidetracked by the amazing aesthetics of some of the species, so this is fun research.
To the initiated, different woods impart different tonal characteristics, some warmer, some brighter, more sustain, better midrange, etc. The idea is that you can create the sound you want via the combination of body design and wood choices. The primary pieces are the soundboard (top) and the back/sides, but apparently everything plays a part: neck, fretboard, bridge and bracing material all contribute to the sound. So I listen to a bunch of videos and can occasionally notice differences but, like the finer points of wine tasting, most of it is lost on me.
And like wine, the finer points are more expensive. Brazilian rosewood is considered one of the best choices for the back & sides, but it is now protected and very pricey if you can find legal sets. Adirondack (Red) spruce was used for tops in early Martins until 1946 when the supply became tight and the more plentiful Sitka spruce was substituted. Old growth Red Spruce sets are available but also very expensive. So after several bouts of enthusiasm (and a few wayward purchases) I settle down and decide that its probably best to go with the popular combination of a Sitka spruce top, East Indian rosewood back and sides, mahogany neck, and ebony fret board and bridge. This should provide a good baseline from which to build my ear, and it’s an attractive combination.
(If you’re interested in trying to hear the difference between tonewoods, check out the video at the right from the Acoustic Letter)
Kid in a candy store
Tone woods are available from supply houses like Stew-Mac and Luthier’s Mercantile, but there are also smaller specialty businesses if you look around. For Sitka spruce, Alaska Specialty Woods seem to be the source, plus its a great story. All their material comes from old growth deadfall trees that they recover via a very controlled and environmentally friendly process. I visit the website and discover that in addition to beautiful quartersawn straight grain Sitka they offer quilted, curly, bearclaw, floating, denim, submerged, and ancient Sitka, all in several grades, and its pretty mind boggling. The full bearclaw is tempting but I reel it in and go with a slightly quilted top, along with a regular one for my son’s (eventual) Dreadnaught, a low grade set for practice and some bracing stock. Eventually I’m able to focus and place my order.
Back and side sets can be found in a number of places; the two that I’ve used are RC Tonewoods and the Exotic Wood Zone. Both great people, very easy to work with and patient with newbs. My first impulse purchase was a curly black walnut set from RC Tonewoods. My second was a bubinga set, third was mahogany…and then I settled down. Exotic Wood Zone specializes in East Indian rosewood so when I made that decision I combed through their website and ordered two sets.
The wood arrives, it looks good 🙂
Light Quilt Sitka, East Indian Rosewood
Fixtures
To build the body you need to bend the sides. Bending the sides is traditionally done by hand, around a heated pipe, until you achieve the desired shape…aaand that just doesn’t sound like a fun time. Open flame, hot steel, dry wood, no thanks. More recently a method has developed using a wooden form, a heating blanket and a press. I’ve done enough wood-bending to appreciate the advantages of using forms. The variances in grain and density can make a piece of wood bend and twist in unpredictable ways, usually at the worst possible moment, and all while the wood is cooling and losing its elasticity. The bending I had done previously was with steam; the heated blanket process is similar but more controlled, so I decided to try that approach.
Outer Body Mold
My first step is to build the bending form and body mold. People do this with CNC routers but perfection is overrated. I break out the carbon paper.
I started with the outer mold, which is built in two identical halves that clamp together. Tracing with my trusty carbon paper, I created a master on 1/4 plywood, rough cut the outline and then drum sanded it to NASA tolerances. Then I used that to create the plywood layers for each side of the mold with a flush cutting router bit. Then I glued them up. This is a little dicey since the stacked layers like to slide when you glue them, but this is not my first rodeo so I was able to corral them using my table saw fence, bar clamps and electric motors. But don’t try this at home, kids.
Side Bending Form
Next I tackled the bending form. This is about 3/32″ smaller than the outer mold, so the outline shrinks accordingly. Due to the glue sliding issue with the layers I tried a different method . On this one I created a master in 3/4″ plywood, attached it to the next (roughed out) layer, then used a flush cutting bit to rout it to match the master. Rinse and repeat until you achieve the desired thickness. This worked surprisingly well, hooray for continuous improvement. .
The Bend-O-Matic
After enough people charred their sides, burnt their hands, and set fire to their shops using a metal pipe heated with a blowtorch, a savior by the name of Fox came down from the mountain carrying the Fox Bending Machine. The only drawback to this idea is that it requires a bit of capital investment, but when you factor in wood costs, medical care and fire insurance you break even pretty fast. Now of course you could be like my Uncle and just buy one ready made, but we’re all about self-discovery around here and this one is easy: you need some plywood, a screw press, a silicone heating blanket, a themocouple, a router speed controller (this plugs into the thermocouple and controls your temperature), and some springs for the supports. Bam.
Back & Sides: Thicknessing
Now I’ve got to get the sides and back to proper thickness. The rosewood is rough and about 3/16″ thick; I need to get them to about 1/16″ (or around 095″ if we’re getting technical, which we are). Typical woodworking practice would be to run these through a planer, but that doesn’t seem to be what guitar people do. The preferred method is to use a thickness sander, and I’m about to find out why.
I have a small “lunchbox” planer. Its not bad for what it is, but it’s not for heavy duty applications. I run the sides through, seems okay. But then I run the backs through and uh oh, this isn’t going to work.
The grain on the backs is tearing out pretty badly, you can see the holes I’m getting on the left. If the tears are too deep I won’t be able to sand them out. Usually you can run the wood the other way but the grain is running all over the place, I’m trapped. I try a hand plane – nope. And now I know why people use thickness sanders! But no big deal, this can be done by hand. It will take a while, but scraping and sanding will eventually get it there. The tricky part is that I need to maintain a level and co-planer surface, down to about .095″. Card scrapers are great tools but this looks like a job for a scraper plane, which I’ve never used. A little research and it looks like the Stanley #80 is a good way to go.
So I head to ebay for the venerable a Stanley #80 Scraper. I find one at a reasonable price, its missing two of the thumbscrews but that’s what Menard’s is for. And after seeing what I’m up against I order a replacement blade from Hock, this wood is not messing around, so neither am I.
What I didn’t know was that Stanley made their own proprietary screw threads early in their history. So the missing thumbscrews are not available at Menards, or anywhere else, which is why the price was reasonable. You can’t make this stuff up.
Fortunately I find a Stanley aficionado who takes pity on me (and $10) and I’m back in business.
So I start working on the two back pieces. My strategy is to work one side of each piece until it’s smooth – I have to get past the tear out – and this will be the outside of the back. Then glue them together and work the inside of the joined piece down to final measurements. For this process I’m going to need some way to measure thickness across the piece, so we need another tool (yes!): we need The Luthier’s Caliper. But they’re kinda pricey, so…..
Now that I have my high-tech Luthier’s Caliper I can get to work. I sand, I scrape, I scrape, I sand. I measure. This goes on for a while, you might want to grab a snack.
Eventually I get both pieces down to a smooth level surface on one side.
The next step is gluing the back. So we need a really good joint and we need glue.
The Back Joint
First, the joint.
A jointer should give you a straight edge on a piece of wood, that’s what it’s for, hence the catchy name. So in a perfect world you make two passes on this baby and you have a perfect joint. Not so when you’re a slave to retro-fashion, as I am. Behold my beautiful, classic, 1950’s, but in need of slight adjustment, Delta 6″ jointer with the art-deco base. This beast gets them close, but not perfect – the outfeed table is a little droopy and the blades probably need to be sharpened, but this is supposed to be a hand operation anyway so here we go.
The way you do this is with a shooting board and a long hand plane. Effectively the same idea as the jointer but rotated 90 degrees and less accurate. The lack of accuracy allows you to massage things until they fit, at least that’s the idea.
Hand Plane Interlude
The “shooting board” is just an arrangement of boards that allows you to plane the edge of the workpiece at 90 degrees. But for this you need a half-decent plane, so I gifted myself a Stanley/Bailey #5 off eBay, a classic. This is a two-patent date plane with the curved frog rib, making it a Type 10 and dating it to 1907-1909. It was a good deal because it had a broken tote (rear handle), a common ailment we will soon remedy. And it needed some TLC, but once tuned up it’s a good tool for this job. Turns out Lee Valley Tools has a free PDF plan of the #5 tote, so I grab a piece of Walnut from my wife’s sculpture stash when she’s not looking and whack away.
I get a good vibe from this plane. The other #5 I have is a Millers Falls from my grandfather. It has no tote at all, but I really like that plane too. The balance of size and weight in the #5 seems to be a good combination.
I get the shooting board set up and start to plane the edges of the back pieces. I chase the joint a little but its not bad, and suddenly I get an invisible joint. You can tell – there’s a sound that it makes when they fit perfectly, they “clack” together, woo hoo! So now I just have to glue it together, for that we need glue.
"We shall not cease from exploration..."
Glue selection is an area where I could have simplified things, but it seems I bore easily. The basic Titebond type carpenter’s glue is great for guitar building, lots of people use it and swear by it. I used it when I built furniture and still use it, it’s great stuff. Hot hide glue (HHG) is traditional in instrument making and has been found as far back as Egyptian furniture, so it has a long and noble pedigree. The convenience of the modern Polyvinyl Acetate (PVA) glues like Titebond is undeniable, but if I have the opportunity to use to use a material that’s in Egyptian tombs and Stradavarius violins I’m like a moth to a flame.
There are pros and cons to each type of glue, but since I’ve never used hide glue I’m interested to at least learn more about it. Aliphatic resin glues are great but my experience with them has me concerned about a few things. Basically they can be gummy, difficult to remove and difficult to stain over. The HHG bond releases with heat and is very easy to sand and clean up, and it also takes stains and finishes well. But the big con for HHG is the open time: it’s used hot as the name implies, typically heated to around 130-140º F to work with. Once you take it from the heat it starts to cool and around 100 degrees it will start to gel, so you need to work fast to get joints together while the glue is still liquid. However, since HHG has been used for literally centuries, there are some workarounds for this issue, more on those later.
The other argument is around tonal qualities. Some people swear HHG glue sounds better, others say there’s no difference. My gut tells me that the hide glue is going to dry harder – basically its more brittle – and that this should help, if anything, in sound transfer.
So there are some differences with HHG but I find no deal-breakers. Just a traditional material with some very interesting qualities that has lost the convenience battle to PVA glues – which is understandable. But since it’s been around forever and has been used for pretty much every kind of adhesive or paint application, I’m very curious to try it. Again, there are lots of sources but I dig around and find “The Hide Glue Man”, Eugene Bjorn Thordahl and buy it from him at Bjorn Hide Glue.
Hide Glue is granular in the wild and comes in “gram strengths” which is another discussion. 192 & 251 are common in instrument building, Eugene recommends them, we’re good.
So now I need a way to heat the glue and keep it at a steady temperature. There are nice glue pots out there but they’re expensive. One of the hacks is to use a wax melter for cosmetics so I order one from Amazon. Unfortunately it’s pretty lightweight so I return it. My brother, a fellow mad-artist-scientist, points me towards surplus scientific hotplates on eBay and after a little patience I snag a basic Corning model that has seen very little sulfuric acid. I know its solid and will hold its temperature settings, seems like a good solution for now. Also gives me an excuse to buy a beaker, and this combination makes the whole operation look more scientific.
At this point I make a few sample batches of glue, being very careful about proportions, but soon learn that this is a pretty friendly and idiot-proof material. I do some tests on open time, and figure I should be getting things done in about 30 seconds to a minute. This is fine for my first glue-up so let’s proceed.
Gluing the back seems like it should be simple right? Well, believe it or not, its a little trickier to get two large, thin surfaces to mate up evenly than you might think. My first try is okay but not perfect. And this is where my appreciation for hide glue starts. Oops, joint isn’t great. Break out the hair dryer. In a few minutes the joint is apart and cleaned up, ready for another try.
For the second try I broke out more clamps and the VW parts, but still off a bit…what sorcery is this? “There is no try, just do or do not!” I’m doing not.
As it turns out this is a fairly common issue so there are a number of solutions, many quite involved using lots of clamps, wedges and rope. But with hot hide glue, I don’t have that kind of time. So I remember reading about a very simple method that takes advantage of the slight spring in masking tape, and decide to try that. In this method one board is angled up slightly and masking tape is attached across the gap, so that the tape is a little shorter than the level distance. When the plate is lowered, the tape is stretched causing the joint to pull tight. This is one of those things that’s so simple it seems like it can’t work, but after a few dry runs I get the idea and give it a shot.
This is a winner. Deceptively simple, great solution.
Now that the back is together I need to get it down to a uniform thickness. I’m shooting for around 2.5 mm/.098 “. One of the other things that you bump into in this world is that measurement can be in millimeters, inch fractions or inch decimals. And inch fractions, the one I’m used to, is the one that used the least. The digital calipers toggle between all three so are a huge help in translation.
The numbers in white pencil (left) show the thickness in millimeters, courtesy of my Luthier’s Caliper. I do this periodically as I’m taking it down, which helps to keep things as even as possible. I’m headed for around 2.5mm, finally I get there.
Good, my arms are tired.
Bending the Sides
Time to turn our attention to the sides. The first step is to get them to thickness and thankfully, the grain on these is more cooperative so the planer has been able to get them there with only a little final help from me. So they’re at 2mm and ready to go.
Next I need to figure out my bending rig. East Indian Rosewood is reputed to be fairly easy to bend, but this is a book-matched set and I don’t have any spares. I’ve made my share of firewood via my prior bending experiences, so when I bought some of my impulse sets from RC Tonewoods I also bought some orphan sides for bending practice. So I break out the Bend-O-Matic.
I do research on the process until I’m ready to try one of the practice sides, a piece of walnut. The wood is sprayed with water, wrapped in wet butcher paper, and placed between two metal slats, I used aluminum flashing, stainless would be better. Another piece of aluminum flashing goes down first, then the heating blanket followed by the the wood sandwich on top, everything held together with binder clips to keep the blanket in touch with the wood. I tuck the leads from the thermocouple into the sandwich, attach the spring loaded bending supports, and turn the press so that the center caul is putting enough pressure on the sandwich to keep it in one spot. Then I’m ready to to try the test piece. I turn up the heat, and wow, things start to get hot!
Heat ranges seem to vary by wood species but somewhere in the 200º-300º F range things start to happen. As you get to boiling temperature the water starts to give off steam and after a few minutes you can feel the wood getting pliable. I give it a few minutes and a few test bends and when it kind of feels ready I give it a shot – and hooray! I get the walnut side bent without any issues. So now we move on to the real thing.
It’s important to get the sides oriented to preserve the book matching, so I make sure that I know which is the left and which is the right, mark the outsides and front and back on each, and double check before I wrap up the first one. For rosewood, I’m shooting for around 300º F to start the bend (John Hall of Blues Creek Guitars has some very helpful videos walking you through this process). As the temperature comes up past 200º I start to get the steam and then gradually I feel the side start to soften. After a few minutes they feel like they’re ready and I start the bend.
I begin with the center caul, twisting the screw at a slow even speed until the waist is seated. When that’s down I move to the upper bout, going slowly because this is the tightest bend. It seems to go well so then I move to the lower bout. It all feels good but due to the flashing you can’t see the wood, so after letting is sit a few minutes I break protocol to check, and all is well! Then I put it back in the mold and re-clamp and re-heat it to help set the shape.
The other side is a repeat of the process and it goes well. Then I run around and show off my “guitar”.
Here’s the first side off of the bender, you can barely see the steam faded “O” in white pencil for “outside” on the lower bout.
Assembling the Sides
The next step is to get the sides into the mold so they can be trimmed to length at the neck and heel joints. These joints need to be a close but not a perfect fit, because the neck and the tail flash cover them. So at this point I make the spreaders for the mold from turnbuckles and laminated plywood cutoffs. These then snug the sides up against the mold. Once the sides are where they need to be I mark the ends and trim them. I also start to make reference marks on the sides and the mold so I can keep things level and parallel.
After the sides are trimmed and fitted into the mold, I start to work on the mahogany heel and neck blocks. These are cut down to shape (with a small ding in the heel block for wabi-sabi) and fittted. The neck block is flat at the glue joint, the heel block is slightly radiused to fit the curve of the body.
One of the tricks I learned from watching John Hall’s videos is that he uses a drywall screw to lock the blocks into the mold. This helps by keeping the assembly stationary in the mold and it also allows you to remove the sides from the mold and put them back in the same place when you need to. So during one of the dry runs I drill the pilot holes for these, and then glue them up.
Next I need to shape the top and bottom of the sides (the “rims”) to match the contour of the top and back. The top of this guitar is a 28 foot radius, and the back is a 15 foot radius, so I’m going to shape the rims using a “radius dish”. These are circular pieces of MDF that have been CNC routed to a specific concave radius and then covered with 80 grit sandpaper. I was able to snag a group of used supplies through one of the forum members, and this netted me a Bosch Colt trim router and two radius dishes with 28 & 15 foot radii.
At this stage I spend some time double checking measurements and being sure things are where they should be – and appreciating the trick of locking the blocks in with the drywall screws. With the radius and the taper from back to front, it’s good to have the sides locked into the mold via the drywall screws. Without those screws the only things holding the sides in place are the the spreaders, and with the amount of handling I’m doing at this stage I know the sides would be shifting in the mold if those screws weren’t in place.
I do some rough trimming with a hand plane and then use the radius dish to get the top rims to the proper radius. After the top is done I move to the 15′ dish and do the bottom edges.
Rims on the 15′ radius dish, showing the gap that needs to be closed for the back. You can see that the top edges have already been done.
Driving the bus: sanding the rims in a circular motion on the dish.
Kerfing
In order to reinforce and provide more gluing surface for the top and back, strips are added to the insides of the rims. These are called kerfing strips due to the multitude of slots cut into them (kerfing) that allows them to follow the bends of the sides. These come in a number of varieties, I am using mahogany “reverse” kerfing – this puts the slots towards the side, theoretically adding stiffness….but once the top or back is on its not going anywhere…so this is another one of those guitar building mysteries I will sidestep by deciding to err on the side of “it can’t hurt, might help”. And I also think it looks nicer to see the smooth side of the kerfing when you look inside the guitar.
This is the first glue-up where I doubt I’ve got the time to get it together before the hot hide glue starts to set. I do a few dry runs and it’s taking around a minute and a half to get one strip on and I know in real life it will be longer, so I need some help.
There are two ways to extend the open time of hide glue: one is heat and the other is by adding either salt or urea to the mix. In anticipation of this I’ve bought some urea and have done some experiments: adding 5% urea by weight is giving me over 2 minutes in open time. When I raise the ambient temperature to 80º with some halogen work lights I get even more. This should work.
Note on hide glue strength: there is a concern/argument that when you add anything to hide glue the strength of the bond is degraded. The argument makes logical sense: given a certain volume of glue in a joint, if any of that volume is replaced by a non-binding agent you now have less glue in the joint – and less glue must mean a weaker joint: Q.E.D.
But since additives have been used for centuries and are still recommended, I looked for quantitative information on this topic, which led me to this very interesting Fine Woodworking article on glue strength.
The cuts are deep so the strips are relatively fragile, which you learn by breaking one! So the first step is to wet the strips and pre-bend them to the outsides
Then the drum roll for the glue-up of the first strip. This was a bit nerve-wracking because it went more slowly than I’d anticipated. The strip needs to be a bit proud of the side, like 1/32″, and I hadn’t accounted for the time it would take to make that adjustment. The longer open time was good and needed.
For the second one I made two changes. First, I ditched the small black spring clamps from Amazon for the traditional clothespins with rubber bands. The black clamps worked but some of the edges were sharp and marked the wood.
Second, I added ambient heat via a pair of halogen work lamps. I was able to get 80ºF on the worksurface which aided in slowing down the set of the glue.
While I’m here I do the kerfing for the top as well.
Back to the Back
Time to get the back braces on and shaped.
First I do a rough cut-out of the back on the bandsaw, not a big deal but every time I cut something on this project I’m nervous. Now the back looks more like a guitar, this feels like progress.
Next, I need a reinforcement strip for the back joint. This is a thin piece of cross-cut spruce, I cut it from my “practice top” wood and glue it down using my handy-dandy DIY Go-Bar deck. When its dry I shape it to a lower profile and we move to the braces.’
Next come the braces. Theses are quarter-sawn Sitka, hand-split to minimize runout.
I stayed with the traditional brace sizes on the plan in which the lower two are heavier than the upper two. Lots of people make them all the width of the top braces these days, but I reason that in 1931 Martin had about 50 more years of guitar building experience than I do now, so I stay with their design.
The braces are cut to width and rough height on the table saw. Sides are planed smooth to remove any saw marks.
Next is the layout of the braces and marking and cutting the reinforcement strip so the braces can run across the back. After that the braces are cut to rough length.
With everything marked we move to the 15′ radius dish. The braces are glued to the back in the dish in order to form the 15′ radius of the back, so the bottoms of the braces need to be sanded to the 15′ curve. I sanded them in their approximate layout to try to keep them perpendicular when glued.
Welcome to the
With the bottoms radiused, the ends of the braces are rough-cut to an initial taper. Once the braces are glued the real shaping takes place.
Gluing the back braces
So now we head to the Go-Bar deck. The back is placed in the 15′ dish and the braces are glued in.
Shaping the braces
In this step the braces are contoured and wood is removed to get to either a desired shape or a desired sound. This is another area where there is little agreement – whether or not to “tap-tune” the back and the top. Most factories do not do this and most hand-builders think its a huge benefit, so there seems to be ongoing debate about the value of this step. But like the hide glue, I’m here to learn so I’m going to listen as I go and see what happens.
After they are glued its time to shape them. Again I stay pretty close to the traditional shapes on the plan, and I am tapping the back as I go to listen to the changing tone. As material is removed from the braces, the general pitch goes down, but that’s about all I know at this point. It seems that there is much to learn here, but I am so new to this I’m just recording data, I have no database in my head yet.
This is the first pass, shaping the cross-section with a hand plane, being careful to avoid the reinforcing strip.
Next I work on the scallops, removing wood and trying to get some resonance into the back. I tried to record the sounds but it’s very difficult to get anything on a recording that translates. But what’s happening is that when the braces are at full size the sound is pretty stiff and toneless, kind of like a countertop or desk top. As the braces are shaved things are able to move and vibrate, so you start to get a little bit of tone and sustain.
I remove material until I’m getting some sustain and am also still within Martin parameters for size and shape. It’s tempting to keep going because the more material you remove, the better it sounds. But we can’t have the back collapsing and you can’t put wood back on, so I stop here and sand things up – so the back is pretty much finished.
To the Top
Now I move to the top. You can go ahead and attach the back here but I decided to do the top first and then attach both together.
The first step is actually to inspect the top for irregularities or pitch pockets. If there are issues you try to push them outside the pattern or into the sound hole. Fortunately things look good to my untrained eye. So the next step is to glue them together. The Sitka for the top comes sanded so we just need a good joint. I use the shooting board again and chase this one for a bit. But ulitmately I get a joint that looks good for gluing.
I use the tape method again but the Engineering Dept. has recommended replacing the VW cylinder head with fire bricks and hockey pucks, and this is much better. The Baldridge Award can’t be far away. Continua emendationem
Now things start to get interesting. The next step is to install the rosette, and this is one of those steps that makes you think wistfully about the kit approach.
One of the things I didn’t appreciate about building from scratch is the number of aesthetic decisions that you need to make along the way. The first big one I ran into was the design of the rosette. I looked at a lot of rosette designs and liked the design in the Kincaid book which is a center ring of Pau shell bordered by thin black and white veneer strips called “purfling”. It’s relatively simple to make and the materials do the talking.
So now I need the materials and a way to cut the channels.
I bought my shell from Australian Mother of Pearl Co. LTD. Not sure if they’re the source but their prices indicate that they’re closer to it than anyone around here. Since I don’t really know what I’m doing I order one mother of pearl (white) and two pau (mulitcolor) rings, plus an assortment of fret markers. Covid delays the shipment a bit but we stay in touch. When the package finally arrives its really exciting, they’re beautiful and the quality is fantastic.
Bling! (L to R) : 3mm Pau, 2mm MOP and 4mm Pau rosettes
Now we need to take a short detour into The Land of Cutting Circles with Routers. This is not uncharted territory but still requires a bit of navigation. I have the trim router that I picked up with the radius dishes, a Bosch Colt. I also have purchased a circle routing device. However, said jig does not mate to said router, so I need an adapter plate. I submit a request to Engineering and go have a cup of coffee.
Okay, now we can move forward so I need to familiarize myself with the circle cutting jig and make some practice cuts. I have purchased a set of 1/32″ and 1/16″ downcut router bits for this. What I am after is 3 channels around the soundhole. The outer and inner ones will have the black/white/black (bwb) purfling, the center channel will have the Pau shell bordered by bwb purfling on either side. And we need a good fit.
The practice cuts helps me to get a feel for the process and the tolerances between the bit, the material and the spruce. Its fiddly, but not too bad. This is one of those things that you just sneak up on, especially the first time through.
So I lay out the top and the sound hole placement and then drill in the center of the sound hole for the pin on which the circle jig centers.
The first cut is in the center of the middle ring. After I check the depth and then move gently to the inside and outside. Definitely nervous the first time through, because if anything slips or is cut too big, it’s tough to recover. I’m measuring with the caliper and the shell pieces and taking small cuts.
I was so wrapped up in this I don’t have pictures from all the pieces of the process so it’s a bit of a mix, but you’ll get the idea.
Next are the cuts for the BWB purfling for the center ring….
…and then the BWB outer rings. This shows the slight difference in height for the shell and the purfling in the center ring. Probably unnecessary but newbie nerves.
Some are more excited about the proceedings than others…
Once everything fits it gets glued up. Don’t be concerned about the gap at the left/top, that gets covered by the end of the fret board. But nice of you to ask.
After the rosette is in and sanded and I’m done marveling at it, I need to get the top down to thickness, so I flip it over. I’m going for 2.7-2.8 mm, and it’s a pleasure with a sharp plane.
The next steps are simple but fun. I cut the sound hole out with a few passes of the router from each side, and then rough trim the top. Exciting to see another guitar shape!
Bracing the Top
So here we are! This is where we get to re-create the legendary Martin bracing pattern from the early 30’s. In mid-1938 Martin moved the intersection of the X braces back from the sound hole about 1/2″ in order to stiffen the top due to problems with warping behind the bridge. This helped with the warping issues but reduced the volume (sound) of the guitars, so in spite of potential structural issues the pre-1938 “forward bracing” pattern is still preferred, and so this is where we’re headed. The idea is that the area below the sound hole is where the magic happens, so that real estate has to balance resonance with durability. Because Martin guaranteed their guitars and had to fix them when they came back with problems, the structural changes they made over the years were primarily to address these issues.
Here’s a tracing of a 1930 Martin OM bracing pattern – note the distance from the sound hole to the X-brace joint, the shape of the early “Pyramid” bridge, and also the single brace above the sound hole.
Here’s a tracing of a 1944 Martin OM bracing pattern. The X-brace joint has been moved back, the bridge has been widened to the more familiar “Belly” bridge, and braces have been added above the sound hole.
Bracing patterns courtesy of www.vintagemartin.com
Cutting the braces
Next I need to cut the braces to size. I want the grain running vertically (quartersawn) in the braces, this provide optimum stiffness from the wood. And although my stock is quartersawn and hand-split you have to check and sometimes make adjustments. The picture illustrates an example where the grain is running about 10º off vertical, so I’ve trimmed the piece parallel to the grain and will then cut it parallel to that face and then square up the top and bottom – after that we’re good to go and I get the braces roughed out.
Layout
Next, I transfer the bracing pattern to graph paper and then to the top in pencil.
The X-braces are rough cut to outline, the bottom faces are sanded to the 28′ radius in the dish and then I cut the joint and refine it to a good tight fit.
With the X-brace joint test fitted I’m ready to go. I get the Go-bar deck ready, glue it and….
Gluing
…one of the things you read about hide glue is that because it contains water you can sometimes get swelling in a joint or warping on a larger surface. And on my first try at the X joint this happens – the top piece binds with about 1/32″ to go, and won’t budge.
I start to panic but with hide glue this is pretty simple. I’m able to release the braces from the top with the hair dryer, and then the heat-softened glue cleans up very easily. To get the X joint apart I have to get pressure on opposite sides of the joint while supporting both pieces so nothing cracks. Again, I’m so wrapped up in this I wasn’t thinking about taking pictures, but I’m able to concoct a good contraption, apply the hair dryer, apply pressure and get it apart cleanly. Whew.
So back to the drawing board. I loosen up the joint a hair, scrape off all the old glue, and then glue them again. All good.
Next I add the two tone braces and the four shorter finger braces.
The next step is interesting. Now I add the braces above the sound hole, but instead of using the radius dish these are glued with the top flat because this is where the fret board is attached.
Then I add the rosewood bridge patch and the sound hole braces, and we’re done with gluing braces on.
Now we come to the shaping of the braces. I know enough to know that this is the belly of the beast in guitar making, and also that I have no idea what I’m doing. So I’m following the plans, going slowly, and listening as I go.
The steps here are to 1) taper the outside ends of the braces, 2) peak all of the braces so they are more of a pyramid shape in cross-section, and 3) carve out the scallops of the X-braces & tone bars in order to get the tone I’m looking for.
I’m tapping and listening as I go. It’s very interesting but with no frame of reference I have no idea what I’m doing. I even did some recordings with Audacity to see if I could learn anything but again, I’m just collecting data, I have nothing to compare it to. All that said, it’s a fun and fascinating process – but without more experience I stay fairly close to the available Martin measurements.
At this point I’ve got the ends of the finger braces flush and the X brace ends down to 3/16″, and am starting to peak the tone bars and finger braces.
At this point everything has been peaked and tapered. This is getting pretty close. But after a breather I take another pretty good whack at it.
The X-brace joint is traditionally reinforced with a cloth patch, so I decide to use linen. It’s trickier than it looks to get it to lay flat, but after a few tries I get a pretty good one.
After a final round of cleanup and sanding it’s as ready as I’m going to get it.
Closing the Box
In which we glue the top and the back upon the rims! Most of the braces extend beyond the sides so the sides need to be notched for the braces in order for the top and back so sit flush on the rims.
So to mark out the notches on the sides the basic method is to place the top or back on the rims, line it up and then mark the spots on the sides where the braces hit (as you can see below its a little tricky to mark these locations). Once things are marked you sneak up on the notches.
The back is a bit easier to do because the braces are run parallel to each other so the notches are all in alignment. It’s a little fiddly but not bad.
This shows the fitting process with the notches cut but not quite there yet. Need a little more trial and a little less error.
Back fitted, showing the fit from the inside.
Fitting the top is a bit trickier because the notches are not in alignment. The back can only move side to side in the slots, but the top can rotate, and we need to nip that in the bud. So I decide to use a series of blocks glued to the top of the mold to hold the top in the right alignment for notching and gluing. This turns out to be a great idea, but after my curtain rod experience I keep my mouth shut.
Clamping the top in the correct spot….
…and then gluing temporary blocks to the top of the mold to maintain the correct placement when the top is removed and replaced.
With the blocks in place its very easy to keep the top aligned which is a huge help in cutting the notches for the braces. The X-brace and transvers brace notches below.
One last thing! I have to remember to add the side reinforcements. Traditionally these are grosgrain ribbon. After looking at that option I decide to break with tradition and go with thin spruce pieces, I just like the look better.
So here we go with the back. I used heat and the 10% urea hide glue for longer open time. Had to work quickly but it went well.
Time to get the top on.
Trimming the back and the top flush with the router.
End Wedge
This is the piece of veneer that is used to create a better looking situation at the bottom joint than is likely to exist without it. It’s dealer’s choice, but this is where I have to start making decisions about the decorative scheme, because the the binding & purfling intersect down here.
Early on I liked the idea of ebony bindings. Most guitars seem to have light colored bindings so I liked the idea of doing something a little different, and I had purchased some ebony (or blackwood) bindings from RC Tonewoods a while back. As I played around with the ebony against the rosewood I realized that there is so much black in the rosewood that the joint would be hard to see – which might be cool but when I added a white line between them I liked the look. So to separate the ebony from the top and sides I use black/white purfling – the black is a cardstock material and the white is maple. This strip runs along the sides of the end wedge and miters into the strip that runs below the binding. This requires some planning which will all be clear when you look at the pics as we go.
The first step is to mark and cut the slot. This is easier to do prior to attaching the top and bottom but oops, I missed that one!
Testing a pattern (also shows the book matching of the sides).
Making these cuts gets your attention!
Clearing the slot – pretty good adhesion from the hide glue.
Trial fit with the purfling.
Trimming the ebony insert with the shooting board.
Glued up – after this it gets trimmed flush.
Binding
During my research, this is the step I was concerned about. In this process I have to cut channels in the edges of the top for the binding and purfling. These cuts are going with and against the grain, in soft and really tough wood. And they need to be pretty darn precise. So I send a request to R & D for a full analysis and wait for their recommendation in the break room while playing foosball.
There are many ways to do this, I will keep it short and say that I looked at all of them and decided to go with this approach because I liked the control it provided.
Binding Router Tower & Cradle
The approach I decided upon is one in which the router travels vertically in a tower and the guitar is held in a cradle. Once turned on, the router is placed on top of the guitar but free of the cutter, and then the guitar is moved into the cutting blade and around the perimeter. The depth of the cut is regulated by the router height adjustment, the width of the cut is determined by the bearing.
Again, this rig can be purchased but here at the E. Scrooge School of Luthiery we know the value of a buck. So I grab some nuts, bolts, and 12″ ball bearing drawer slides at Menards and bust out the plywood.
This is pretty straightforward. The cradle is self-explanatory; it’s job is to hold the guitar in a level position and to be able to slide around the router. My primary concern was that the drawer slides might have some play in them, but they were tight, yay Menards. So off we ride to slay the binding dragon.
I do the back first because its a single cut and it’s the back, so if anything goes wrong it will be less obvious back here (!). This cut is replicated on the top so once we survive the back we will flip the body over, level it again and then repeat this cut on the top, followed by the second cut for the purfling strip.
I do several test cuts to determine the correct bearing (width) and depth for the cut. And then after some deep calming breaths we go in.
Smoke is usually not a good sign when using woodworking tools. As it turns out, there is a right side up for the bearing and a wrong side – and if you see smoke, it’s probably a good idea to stop and flip it over. The green tape marks the divot that the non-spinning bearing burned into the side…
It’s a lot easier when the bearing is free-wheeling. I was honestly surprised at how smoothly these cuts went. Thumbs up on this approach.
This show the treatment at the end wedge. The full cut is stopped short and then then adjusted to just cut the binding channel over the end wedge.
Leaving the cut short allows you to create the mitered joint for the purfling (the finished product is shown below so you can see what needs to happen but we’re not there yet). This involves some chisel work but not too bad.
Flip it over, re-level and do the top
This is the start of the first cut for the binding.
The second cut is for the purfling strip that runs inside the binding – this is decorative so only happens on the top.
After a big sigh of relief I have to bend the binding and the purfling. It would make sense to do these together, but at the time I wasn’t sure what I was doing so I bent the bindings first.
When I decided to add the purfling strips to the sides it meant I had to bend them. The strips are so small that I sandwiched them between two strips of walnut from one of my test sides so they wouldn’t twist.
Following that step I then realized that my life would be much easier if they were attached to the binding strips before attaching the binding, so I glued them on. This phase of the operation was very tedious, my balsa model patience came in handy here, thanks Dad.
Bindings ready for the Bend-O-Matic.
Me realizing that this isn’t going to work.
Purfling strips sandwiched for the bend.
A much better approach.
As I started doing the trial fitting, I noticed a problem. I had gauged the thickness of the purfling off the ends of the strips, but the batch of purfling I received had a pretty good variation in the thickness, to the point where it was dipping below the surface of the top at some points as I fitted it around the edge.
Trying to show the variation in thickness on the purfling strips, but doesn’t really do it justice. They were supposed to be 3/32″ wide and that’s how I cut the channel, but in some places got close to 1/32″. So I had to add spacers where it was thin to be sure it didn’t drop below the top – you can see them in the background.
Another look at the small pieces of veneer glued onto the purfling ledge to raise it up. Had to do this around the waist in a few spots too.
Now that everything looks good on the dry runs, its time to glue. This can be a challenging process. You have to clamp the pieces in and down at the same time so you need three hands and clamps that only exist in another dimension. There are a few approaches, my plan is masking tape with cloth webbing backup. I’m also using Fish glue, which is recommended for its better tacking qualities, if that’s the right term – basically it’s stickier which helps hold things where you put them. It’s also liquid at room temperature so you’ve got a more normal open time.
You need lots of tape and you need it fast.
You start at the bottom joint: the binding is trimmed to this length and the purfling is mitered to fit the miter on the end wedge. Should have taken a better pic of this.
No comment.
The back binding is a repeat of the process.
After my first flush of victory passed I inspected the joints more closely and found a few issues. And again, these are not a major problems with animal glue. The fish glue didn’t seem to release as easily as the hide glue, but after some persuasion with the heat gun it did loosen and I was able to close the joints.
Gap
Heat applied, gap closed
After fixing the gaps, the bindings are scraped flush and the body is done for now.
Neck Joint
Now we tackle the neck joint. I had read a lot about this part of the deal because it’s critical to the playability of the guitar. It needs to be straight and angled down about 1º, but the actual measurements are fine tuned to the specific guitar.
This joint is traditionally a sliding compound dovetail. More recently it’s done as a bolt-on mortise and tenon, which is much easier to do and service. But since I value tradition and hard work for hard work’s sake, you know which way I’m headed.
The sliding dovetail lends itself very well to fixturing but I’ve made so many jigs at this point I’m jigged out. And I’ve also cut enough dovetails that I ain’t a-skeered, and as we all know, pride goeth before a fall.
The first thing that needs to happen is to get the area of the body where the next is going to be dead flat…and hopefully perpendicular to the centerline of the body. Mine was actually off very slightly, correctable but it created a bit of confusion later. This is a project in which you have to measure what you’re measuring and also be sure to measure how it relates to everything else in order to check yourself before you wreck yourself.
Then I start the layout of the dovetail mortise. The inside lines are the dovetail and the outer curved lines are the outside of the heel.
Cutting the dovetail mortise
Cleaning out the mortise.
Layout and first cuts of the dovetail tenon.
It was at this stage that I realized I needed a jig, whether I liked it or not. The cheeks on the neck, being on opposite sides of the tenon and cut by hand, were not quite in the same plane. So I whipped up a quick fixture for facing the cheeks with the router. It did the job. A full-on Neck Angle Jig is in my future.
Cutting back the cheeks and starting to refine the fit.
Getting closer. Using carbon paper to mark the high spots on the tenon.
Very close in this sequence. I got it to a little less than 1/64″ proud with a perfect fit so planed it off the rest of the way.
It was a little hairy at times but I got a good fit and was really glad to do the sliding dovetail by hand, once at least :-).
I’ve been so distracted by the neck joint and alignment that I almost forget about the truss rod. The truss rod allows the neck to be adjusted, or “bowed”, slightly up or down. The adjustment access is traditionally via small plate on the peghead but I go modern and do it from the sound hole. I grab the one my friend Diane recommends off Amazon and start the process.
I set up my “router table” to cut the 1/4″ slot, and end up with a good fit.
Fretboard
This is fun. I mean, the whole thing is fun but some parts have so much concentration involved that the fun comes after. This is fun the whole way through.
I bought the fretboard stock from Exotic Wood Zone, and it’s a nice piece of rough-sawn Ebony. Now that I have my thickness sander I put it to good use bring it down to thickness, I’m shooting for about 1/4″.
Ebony is really interesting to work with. Very dense and heavy, it has short grain and it’s almost like working plastic. One of the things I decided to do is to bind the fretboard – this means putting two pieces of wood on the edges of the fretboard in order to hide the ends of the frets and to add a decorative element. This is usually done with a contrasting wood, but being no slave to convention I decided to do it with Ebony. So I guess you’d call it a “self-bound” fretboard. In any event, after getting it to thickness I sliced off two strips about 1/16″ thick to use as the binding.
Bringing the ebony down to thickness. Imagine everything outside the frame covered in black sawdust.
Binding pieces sliced off. Also experimenting with various white pencils.
The next step is to slot the fretboard. I always suspected that this was the problem with my banjo never playing in tune (although now I know it was more likely that I didn’t know where to position the bridge), so there’s no way I’m laying this out by hand if there’s a better way, and thankfully there is. The idea is that you have a precise template that is notched at the correct intervals for the frets. This template is then attached to the fretboard stock. The cuts are made using either a miter box or table saw setup. I elected to go the table saw route so built a sled with a pin that engages the notches in the template. You engage the pin in the first slot and cut the first fret. Move to the second slot and cut the second fret, and so on. Hard to describe, but brilliant idea.
I buy a CNC’d 24.9″ scale template from LMII, along with a .023 thin kerf saw blade (get the blade stabilizers from Amazon, much cheaper). I build the sled, attach the template to the fretboard with double sided tape, and we have a contraption that looks like this:
Here it is all slotted.
There are still all sorts of ways I can still screw this up, but at least the frets will have the right spacing.
Yay! But back to it. Now I need to trim the sides so have to layout the fretboard, taking the binding strips into account. I create the centerline, measure a bunch, rough it out on the bandsaw and then bring it down with the jointer and a hand plane.
Masking tape does the trick for the glue-up of the binding pieces.
Now we move on to attaching the neck, fretboard, and carving the neck. This can be done a few different ways. After surveying the field I decide to sneak up on it. I will dry fit the neck and fretboard, then finish the heel area of the neck and rough carve the rest of the neck. After that I will glue the neck while checking alignment, and then glue the fretboard while checking alignment again. And then when everything is okay, carve the neck. Basically I want to be able to check things along the way. Sometimes everything is fine until you add glue to the joint and clamp it, and a small change at the body joint could mean a larger change at the peghead, so I want to have a little room to maneuver if I need it.
Now I need to start refining the neck.
Carving the Neck
I start by thicknessing the peghead. I begin with the old saw and chisel method and then remember something else that I got from my GuitarFriend Diane, a tool called a drill press planer, or “Saf-T-Planer”. This is an ingenious device for removing material on small pieces or in tight places where a hand plane or planer won’t work. This is exactly the kind of thing its designed for so I decide to give it a go.
Sawing and chiselling, zzzzzzzzzz…..
Drill press planer in the background, results in the foreground. This is an older version, it burns a little but works and makes me want the new version from Stew-Mac.
Needs to be around 5/8″ thick, off to a good start.
After this I need locate the fretboard so I can start trimming it on the other dimensions.
Marking the centerline…imagine lots of measuring, clamping, un-clamping, and micro-adjustments.
With the fretboard lined up, 1/16″ holes are drilled through the 1st and 12th fret slots for locating pins. These are removed after glue-up and the holes are covered by the frets.
Trimming the sides.
Roughing out the profile of the heel cheeks.
Roughing out the shape of the heel cap.
Refining the heel.
Roughing out the diamond.
Trimming the diamond to outside dimensions.
Round 1.
Heel Cap & Headplate
At this point we interrupt the neck carving to figure out the heel cap – I want it on there before I do the final shaping of the heel. My thought was to go with an ebony heel plate – simple, matches the bindings, easy, I can keep moving. The idea of adding a white line to match the purfling on the body briefly surfaces, but I have no maple veneer so I go ahead and glue up the ebony by itself. It’s not thrilling. The heel cap is definitely not a focal point but I thought it could at least be more interesting. No immediate solutions present themselves so I decided to give it a rest and turned my attention to the peghead and the headplate.
The headplate is the veneer that goes on the peghead, and is another design element that I’ve been gumming to death during this entire process. Factory headplates are not usually flashy but on the custom side you see some very artistic treatments, so again I’m inspired to do something interesting if I can. The shape of the peghead also plays a role, so I have a few things to nail down.
In my early days of enthusiastic wood purchases I bought a bookmatched ziricote headplate from RC Tonewoods that was just such a crazy piece of wood that I bought it with the idea that I would figure it out later. Part of me wanted to use it and part of me was concerned that it would be too much, since the rest of the guitar is fairly low key. My lack of experience with the visual details had me chasing my tail a bit and again thinking of the advantages of the kit approach, where all these decisions are made for you.
Since I couldn’t make up my mind about the ziricote I got some other headplates and did some mockups: ebony, rosewood, pale moon ebony, and figured mahogany.
They all have their merits, but I’m still drawn to the ziricote. I start to play around with the orientation and like what happens when I put the outside edges together. It’s more subtle but still striking, and the center strip will line up with the line of the tuners. It seems a shame to not use the center portion of the piece, but I start to like this idea.
As I do with a lot of things on this project, I circle it for a while, ask other opinions, do other things, but each time I come back to it I still like it.
I glue it up.
As I’m looking at how it lays out I realize I might be able to use the cutoffs to mirror the headplate on the heel cap…yes, I’m liking this idea.
I trim the headplate and start to play around with the cutoffs. I really like the part where the lighter wood transitions to the dark. So I glue that up and trim it for the heel cap.
So here’s what I end up with, a nice detail that I like much better than the plain ebony.
Now, back to the neck.
From here it’s just a lot of carving and creeping up on the final neck dimensions.
As I get closer I need to get the headplate on so I can get the peghead trimmed to shape. This will be easier with the neck off the body so I do it now, but will trim it later with the fretboard on. I decided to use a black and a white (maple) veneer sheet under it to coordinate with the purfling. Martha Stewart on line one.
Now I can glue the neck on. After all of the time spent fitting it, gluing it is very simple.
The top has to be cut to allow the truss rod to sit down. Fairly straightforward chisel work but doing it with a router is an option.
This shows the channel that I did not put in the popsicle brace that would allow access to the allen screw on the truss rod. Oops. Fixing this was interesting, won’t miss that again!
Fretboard & Frets
At this stage we trim and affix the fretboard, sand in the radius, inlay the fret markers, and then put the frets in.
Decided to put a slight radius on the end of the fretboard before gluing.
The locating pins make gluing the fretboard a non-event.
And now a slight detour to trim the peghead. My order of operations is a bit odd but since this is my first time through all this I was more cautious than I needed to be. I wanted the fretboard in place before I finalized the peghead but with the pins nothing moved, so I could have done this earlier. I’m going to do this using the router table and a flush trim bit, much easier to keep things square and straight.
Have to do it face down so the template goes on the back with double-sided tape.
Need a spacer on the front to clear the fretboard.
No disasters.
The next step is to radius and flatten the fretboard. For this I broke down and bought a 15′ radius sanding bar and a leveling sanding bar. The accuracy of the fretboard is pretty important for a good setup so I wanted a fighting chance. There are lower cost ways to do this but I wanted to remove any variables I could at this stage.
First passes, sanding bar in the background.
Using chalk to keep track of progress.
All done except the piece over the body – this is tapered down about 1/64″.
With the radius done it’s time for some bling! This is where we inlay the fret markers. When I bought the inlay for the rosette I bought an assortment of mother of pearl fretboard inlays, so now is the time to work out what I’m doing. I have circles, squares and diamonds to play with, and I go through just about every iteration before deciding on a pretty classic pattern.
Inlay practice!
Final pattern.
Layout.
First one in.
12th fret – saved the diamonds for last, they’re a little tricky due to the acute angles.
All in!
Glued and filled (with ebony dust, available on any surface in the workshop).
Sanded.
Detail (with locating pin holes at 12th fret).
Now it’s time for the side markers, these are 1/16″ MOP dots.
I took it as a challenge to do the ones over the body – a long 1/16″ drill bit did the trick.
If you look closely you may see a rookie mistake – I made the 12th fret marker one dot instead of two. Fortunately its very easy to hide things on ebony 🙂
Frets
With a regular fretboard you cut the frets a little long and trim them after you put them in. But with a bound fretboard the tang has to be cut flush at the ends so the fret can continue over the binding to the edge of the fretboard. This is a little fiddly so I made an attachment for the Dremel that allowed my to grind the ends of the tang down flush. It’s a little tedious but 44 grinds later you’re ready for install.
These have been marked (black sharpie) for trimming, I’m down to the last ones.
To install them you position the fret in the slot, tap it to get it started and then hammer them in, with something firm but softer than fret wire! I bought a rubber headed mallet for this but ended up using an old rawhide mallet, liked it better.
Some people use CA glue in the slots as insurance, some don’t use glue at all, I used some hide glue. They seemed to fit well so not sure if it was necessary.
After the frets are in, the next steps are beveling the edges, leveling the tops and then re-crowning the frets.
To bevel the edges I made a file beveling thingy out of a piece of 2×4 and a file, careful to keep the file from hitting the soundboard.
This shows the fret ends beveled to the edge of the fretboard. I will probably go a little further to take the edge off the fretboard but this is good for now.
Some people use CA glue in the slots as insurance, some don’t use glue at all, I used some hide glue. They seemed to fit well so not sure if it was necessary.
After this the frets need to be leveled, crowned and polished. This is difficult to photograph, but the basic process is this:
1) Level the frets with a sanding bar. First you go over the frets with a black Sharpie so you can see what’s going on, then you sand with fine paper until you get level across the frets.
2) After this, you re-crown the flat areas with a fret-crowning file.
I performed these steps and then left the final polishing for later after the finish is on.
Bridge
We need a bridge, now is a good a time as any.
At this point there’s a little explanation needed for the next pics. The bottom of the bridge has to conform to the top radius, so I created a sanding form by sanding a piece of mdf in the dish to get the 28′ radius. Then I put 80 grit sandpaper on it and sanded the bottom of the bridge until I got a match.
The wings of the bridge are done on the drum sander, it works perfectly for this step. You bring it up once before radiusing the top of the bridge and then hit it again after to get back to the line.
After the bridge I headed back to the neck and trimmed the peghead veneer at the nut slot. This is tempting to do freehand but if you get off line with the saw it gets ugly fast, so I cut a piece of plywood at a 15 degree angle as a stop and this worked well.
Next is the peghead layout. This was more challenging than I anticipated. The issue is that you don’t want the strings touching any of the other tuners, and it gets tight. I drew lots of lines on lots of graph paper, this would be a good thing to put into a CAD program. We’ll see how it comes out.
Once I got the holes drilled I played with the tuners. I tried to go with 18:1 open back Grovers (traditional) but they were unavailable so found Gotohs which are very similar. 15:1 but the quality is really nice.
So we’re getting close. From here we need to get the finish on, then glue the bridge. So prior to finishing I need to mask off the bridge location so I can glue to bare wood after the finish is on. I’ll spare you the calculations but the objective is to get the saddle in the correct location based on the scale length.
Finishing
So now we are at the finishing stage….not to be confused with completing the guitar! The steps here are are sanding until you can’t sand any more, pore filling, and then finishing. I’m going to be using shellac and a technique called “French Polishing”, hopefully. Shellac is a natural and very friendly finishing material, with the drawbacks of not being totally heat- and/or moisture-proof. However, since my gigging days are over (before they started), that’s not really a factor in my decision.
The sanding part is pretty straightforward, basically just getting everything down to 220. For the pore filling I need sawdust. So between sanding the guitar and creating the sawdust for pore filling, I created a lot of sawdust that Saturday. And my face and neck turned bright red. And then I learned an interesting fact: many exotic woods, including east indian rosewood, contain irritants that can cause contact dermatitis. Which explained the mysterious intermittent rash I had been dealing with for months!
Before learning about wood irritants.
After learning about wood irritants.
After this we do the pore filling. This is done on “open pored” woods such as EIR and mahogany. Legend has it that you don’t need to fill the top because the Spruce doesn’t need it. We will test that theory later.
Pore filling is one of those things that has a million different approaches and products. There is a traditional approach using pumice and shellac; the pumice abrades the wood creating sawdust that is then forced into the pores and sealed in with the shellac….theoretically. I decide to use the sawdust and shellac method the Robbie O’Brien likes. This removes the pumice and substitutes pre-made sawdust in it’s place. You sprinkle the sawdust on and go over it lightly with a pad loaded with shellac. Sand, rinse and repeat until pores are full.
I did three passes and called it a day. I could tell I was making progress but that they weren’t completely filled. I thought I would get some fill from the shellac and decided to proceed.
French Polishing
This is definitely a woodworking merit badge. French Polishing was developed in the early 19th century and was considered a top finishing method until the 1930’s when spray equipment and nitrocellulose lacquers arrived. Like a lot of older techniques it has an air of mystery about it but it was actually just a very practical and economical, if labor intensive, way to use the material. Shellac is expensive and difficult to brush smoothly because it dries so quickly (being alcohol based), and until the early 20th century fine sandpaper was not readily available, so it was difficult to smooth a rough surface after the fact. French Polishing takes advantage of the quick drying characteristic by using a thinner mixture, a pad to apply it, and oil to rub it out (oil?!? I know, I still don’t get it). Because it pretty much dries as you apply it, you can build the finish quickly. The applications are thin and rubbed out in each application, so when you finish, you have a smooth, glossy surface. There is no need for sanding or buffing.
I rented Robbie O’Brien’s French Polishing video and followed his instructions pretty closely. I started to create test panels but felt confident and antsy enough to just plow ahead.
Robbie uses a 1 lb. cut that he thins further to about a 3/4 lb. cut. The first passes are the filling stage – and this is when you start to see things you couldn’t see before! I quickly notice two low spots on the back that needed to be addressed. So I sanded the back down until they disappeared, and then started again.
Getting started with the fill coats.
Hmmm…what’s that…?
A few low spots. Back probably needed to be thinner anyway…
Take two.
Back to fill coats.
Hmmm, this might work after all
Okay, so I kind of got the hang of it on the back. Honestly, it’s not that hard and I really like the look – it has a very organic feeling that’s really appealing, and it’s very thin. My pore filling leaves something to be desired but I’m okay with it, kind of like it actually. I’m not sure I want a mirror finish, it starts to feel too far from the wood if that makes any sense.
Fill coats on the sides.
Starting to polish.
Mon dieu!
This gives some idea of the process and the buildup of the coats. By the time you get to the end of one coat its pretty much dry enough to start again. It’s a fun and very satisfying process, but it does take some patience.
Now we move to the top. Conventional wisdom says that spruce doesn’t need to be filled prior to finishing, but the road to hell is paved with conventional wisdom, or something like that. Straight grained spruce is one thing, spruce with figure is another. And I have some slight quilting in this top. It took me a while to figure out but the wave in the quilt creates spots of end grain. This causes the top to fill unevenly (end grain is more absorbent than flat grain). But once you get past that it looks amazing and provides the perfect opportunity to use the word “chatoyance”.
Fill coats.
Starting the polishing coats.
Pretty close.
Good progress, will probably leave it there for now. One of the things I’m after is a thin finish, and although I’m getting a nice gloss, in the video to the right you can still see the wood grain telegraphing through the spruce. For contrast take a look at the mirror finish on most factory guitars.
So at this point I was getting pretty cocky about the French Polishing and then I hit the neck and peghead. This is different because the area is smaller so you have to wait longer between coats for it to dry, which takes some getting used to. You also are short on good places to hold on that aren’t newly coated – so you might get a few fingerprints in places you didn’t want fingerprints. And, the neck joint is tricky because you can’t really get the pad into the corner and if you do, you hit the side which is already a finished surface. So I chased my tail a bit here but eventually got something serviceable.
At some point you have to move on. I’m happy with the finish, and there are definitely lessons learned and I’m pretty sure it will be better next time. I will take filling more seriously and also do more filling and sanding with shellac prior to the french polishing. I also want to experiment with adding Manila Copal to the shellac which is supposed to provide a higher shine, and also with Royal-Lac which is a much more durable version of shellac. But my appreciation for shellac continues, its great to work with and has an unmistakable warm look and feel.
Time to glue the bridge.
Gluing the Bridge
Measuring….ahh-gen….
Squaring up.
Taping off the outline of the position.
Cutting the tape!
Peeling it back, carefully…
Woop woop!
When I looked at gluing the bridge I thought, “this is a job for three C-clamps”. My friend Diane said “it won’t work, I have the C-clamps on my wall to prove it”, but I had to find out for myself. It doesn’t work, because the braces are in the way on the wings. So, yes, another jig. This one is easy.
I actually ordered a bridge clamp from Amazon but it was kinda janky so ….
(….background music…….)
T-nuts, always fun!
Ta-da! 5-minute bridge clamp.
Marking the holes for the bolts.
Holes drilled, taping the bolts in place.
Glued up.
Drying (…more background music…..).
Mission accomplished!
Deez Nuts
Now to the nut. Another guitar rabbit hole discussion regarding material, tone, sustain, etc. My friend Diane says camel bone is the most dense so I’m using camel bone.
Camel bone nut & saddle blanks
To thickness the nut blank I used double sided tape to attach the nut to a piece of maple so I could trim it on the table saw.
After some additional sanding we have a fit.
Trimmed to length.
Rough profiling of the top.
I’ll be using light strings so the gauges run from .012′ to .053″.
Nut slotting files are expensive so we will attempt another hack: feeler gauges notched with a file.
Can you see the little teeth? They do work, sorta kinda…
You start the cuts with a razor saw and then widen the slots with the appropriate size “file”. But this isn’t feeling very precise…
Time to drill the rest of the bridge pin holes and then ream them to fit the bridge pins. Broke down and bought a nice 3º taper reamer.
First string on. I’m using bone pins so ended up numbering them due to slight variances in diameter.
First sounds!
Intonation
If you’ve been following along you might assume that the distance from the nut to the bridge would equal the scale length (24.9″), right? And you know what happens we we assume, right? You’d never be in tune, that’s what happens.
Well, technically you’d never be in tune on any fretted note. Because when you fret a string you are stretching it slightly, and stretching it causes it to go up in pitch, or become “sharp”. And since we plan to use the frets this needs to be addressed. The easy way to explain this is that the fret board repeats itself at the 12th fret. So the note you have on any open string repeats at the 12th fret, one octave higher. So we need the open string note and the 12th fret note to be the same, and that’s where this step comes in.
To start with, about 2mm is added to the scale length at the high E string, and about 6.5mm at the low E. This gets you in the ballpark. Then you create a way to simulate the saddle break on the strings, and, using your trusty tuner, you maneuver the break point back and forth until you get the same note on the open string and the 12th fret.
These measurements are fairly well known at this point so the bridge can be slotted off the guitar and then glued in the correct spot, with fine tuning occurring on the saddle itself. Due to my inexperience I wanted to do it on the guitar in case the bridge moved during the glue up. More complicated but hopefully more accurate at my level of experience.
First attempt: music wire and walnut scraps. This worked but the walnut was covering the front of the bridge which is the measuring point, and the single wire limited the adjustability.
Second try: 1/32″ plywood and individual pieces of music wire. Better, but the plywood was too soft and the wires dug in so they were difficult to move.
3rd one: I found a strip of brass and in addition to adding some class to the proceedings, it was hard enough to slide the adjusters on.
So the break on the high E is right at 4mm back from the edge of the bridge.
And the correct break on the low E seems to be right at 7mm. Note the natural bone pins with pau inlays…
The individual strings fall pretty much in line, with slight adjustments made later on the saddle if needed.
The Final Jig
So now we have to slot the bridge, which is a bit of a nail-biter and of course requires another jig. And as you can imagine, an expensive store-boughten option exists, but the E. Scrooge School of Luthiery happens to have a stash of plexiglass cutoffs and a bored engineering department, so I rap on the glass and wake them up.
This is the beauty we’re after,courtesy of StewMac.
From humble beginnings…
“If my calculations are correct…”
Starting the router baseplate.
Gluing the piece to the baseplate that rides in the slot.
Testing the felt bumpers.
Cutting the slot for the router guide to NASA tolerances.
Checking layout and hole locations.
Ready for a test run…
Getting serious – clamps on, end stops in position, ready to make the first cut.
The slot is 3/32″ so I’m using a 1/32″ bit, trying to hit the middle on the first pass.
Lots of test fitting…
All done! Whew.
Cutting to length…
Presto.
Setup
The strings go on, the nut is slotted, the saddle is shaped, intonated, then given it’s final shape.
Broke down and bought the StewMac nut files…
…which made this job much easier.
Final intonation on the top of the saddle.
After this you’re supposed to go back through everything and check the neck relief, string heights, etc., but I confess that as I was tuning and strumming and picking and grinning, my patience finally ran out 🙂 Everything was pretty close to where it should be, and I couldn’t stand it anymore. They say that there’s a stabilization period, during which the instrument kind of settles in to the new string tension, so you should revisit everything in a month or so and do another setup then. So until then, time to have some fun.
Drum Roll Please...
And now…I have a guitar. It’s a bit of a shock actually! It’s been such a long process (about three years) that I’ve become used to having it to work on or think about. The problem solving aspect has been fantastic and of course it’s quite a learning experience, but the best thing is that it sounds good. Hard to describe. It feels heavy to me but not in a bad way, maybe ‘substantial’ is a better word. It vibrates. And has really nice sustain. This is where some of the talk about various things adding to the sound comes back to me: Robbie O’Brien feels that it may not be one big thing but a lot of little things that make a difference, which is honestly a little confusing in the beginning when you are focused more on the big things like design and wood selection. But now its easier to understand the idea that its a journey of a thousand steps and that each step plays a part in the outcome.
It will be fun to see how it evolves. One of the things you read is that it take a while for them to settle in and that they sound much different in 6 month’s time.
Haha. Let us tune.
Ta-da!
It still needs a few little things but it’s done.
And it sounds good. 🙂
Sound demo by the good guitar player in the family 🙂