Improving a Dark Years Martin

The Acquisition

A friend called to tell me of his recent acquisition, a 1972 Martin D-18. He thought the guitar really had some potential, but it was just disappointing in its current condition. The finish was damaged and the pickguard was curling. The soundboard was split, but the guitar sounded "okay". 

Most obvious to its new owner was the immediate need of a neck reset to even make the thing playable. And, speaking of necks, this one was definitely bowing forward.

I told my friend that, were it my own guitar, I would consider it wasted effort to simply put in just enough work to make it playable. Rather, I would want to correct its mediocre status to let it live up to its potential. With a few changes, this guitar could be really great!

If I remember correctly, his response was akin to, "I was really hoping you'd say that." He shipped the guitar to me the very next day. 

Perspective

So, what's so "wrong" with this guitar that I would dare to consider altering it?

The strongest response anyone has been able to elicit from this particular guitar is a softly uttered, "Meh". That should NOT be the case with this instrument! There are straightforward reasons behind its lackluster performance, and I know what to do to put that right.

As I understand it, Martin more than tripled their production between 1965 and 1971. Such rapid growth makes any manufacturer an easy target among critics, and with just enough "less-than-stellar" examples sneaking past Quality Control (let alone a few "lemons"), a company's reputation is easily tarnished.

We are told that many guitars of this era suffered intonation issues as a result of an improperly located saddle slot. But there were deliberate changes made to these guitars that impacted their tone, especially when directly compared with guitars that had been completed just a few years, prior.

Brazilian Rosewood backs and sides had been replaced with Indian Rosewood. That same Indian Rosewood also replaced the much smaller, thinner and lighter Maple bridgeplates that had been in use from the 1930s. The heavy metal square tube that was added to stiffen the neck certainly didn't help as much as had been hoped.

The latest celluloid pickguards were attached by melting them into the raw wood of the top using solvent, a practice that would prove to be a bad idea. The problem was not so much the method of attachment but, rather, attaching a material that would have such a propensity for shrinkage at a rate much faster than the wood it was glued to.

Perhaps most significantly, the bracing was more stout, leading to the conjecture that the company didn't want to be overrun with warranty returns. It makes perfect sense, especially when presented with two (2) options: A.) Build the most sensitive, fragile, returns-prone, but best sounding guitars ever, or B.) Increase production more than 300% to meet growing demand, while trying to find a happy medium between instruments that still sound good but are a bit more durable.

Martin wasn't alone in these changes; indeed, this was the consensus shared by most all of the guitar factories of the day. Many factory-built guitars in the 1970s would later be considered to have been "overbuilt". Public perception deemed that Martins built during this time were of lesser quality, thus of lesser value than what the Martin customer had grown used to. This period of time came to be known as the "Dark Years".

Vintage and Value

I want to address the topics of "vintage" and "value" head-on, as I intend to make alterations to this used instrument.

This guitar is most certainly not a  pre-war Martin. It is not some five- to six-figure investment that will be instantly devalued by collectors should it even be looked at in the wrong way, let alone modified.

Nor is it a post-war Martin that was un-earthed in original condition, unstrung, still in its case, in somebody's closet, in a climate controlled room for decades. Sorry folks, there is no $9,000 payday here.

It isn't even a 1970s Martin "case Queen" in really good condition, perhaps commanding as much as $4,000 from the unsuspecting buyer.

This guitar is NOT a collector's item. It is NOT one of the dozens and dozens of "Special Edition" or "Artist Edition" or "Anniversary" or "Signature" models. It is just a simple, unadorned 1972 production model D-18 that no one can make good use of in its existing condition.

Truly, it is precisely what its owner hoped it would be: a "project guitar" having no particular sentimental value, but having potential to be something so much more, something other than a wall decoration with a sign above it that reads, "Don't let this happen to you!"; another one of those guitars that any experienced player would glance at and say, "What a shame!"

This guitar deserves more! It demands a thorough investigation.

The Inspection

It is always good to know where we are in order to chart a more accurate course to where we need, or want, to go.

The headstock, headplate, and decal are intact. But for a few small dings, and an obvious forward bow, there is no sign of damage to the neck structure. The finish on back of the neck, however, is soft in spots. It is suffering from a chemical interaction and, left unaddressed, is a deteriorating condition.

The nut is in excellent shape, and the slots of the fretboard are cut properly. If it is original, this nut is Ivory.

The fretboard is in good shape, though a bit thin (I assume it has been leveled at least once in its 54 years). There are shallow fingernail divots at the first fret, diminishing to none by the fourth fret.

The soundboard (as well as the side around the neck heel, and even the pickguard) have several deep troughs of melted lacquer concentrated around the soundhole and pickguard. In the three seconds, or so, that I pondered the source of the scarring, the following ran through my head:

Intense heat, directly applied, such as when someone will lay a cigarette on a guitar, can certainly ruin a finish. But there were no telltale burn marks. Perhaps a soldering iron, or even a hair curling iron ...

... or maybe the guitar, in a stand, was repeatedly bumped up against a hot metal grate, such as the metal guard on a small work light, resting on the floor? Okay, overthinking has now been curbed. Look harder ...

There are several green fibers stuck to two of the twelve scars, and that green perfectly matches the color of the plush-lined case the guitar was transferred to prior to shipping it to me (not its original case). I recalled the owner specifically mentioning he did not want to send the guitar in the worn case he received it in.

I touched one of the scars with my fingernail and ... Yep, I was afraid of that! Soft and gooey. This soundboard has had an unfortunate encounter with a solvent at some point. A quick "sniff test" adds a perfumed oder to the mix.

I am no chemist, but I am thinking "solvent with additives". I have seen the same residual damage from accidental spills of fingernail polish / polish remover. This is really unfortunate, as this kind of finish damage doesn't improve or just "go away" over time.

The pickguard is curling away from the top, and has already successfully split the soundboard in front of the bridge. This all-too-common event is the result of directly attaching a celluloid pickguard to the softwood top. As the celluloid "dries out", it shrinks, exceeding the range of motion of the wood it is glued onto/into.

The saddle fits the slot perfectly, but is much too low, and the bridge appears to have been shaved, slightly. The string spacing at the bridge, specifically at the 3rd string, is off, being dictated by the (very small, but very consequential) slots cut into the front of the bridge pin holes.

Measurements

Strings #3 and #5 intonate "passably". The other four strings are all sharp, with #2 and #6 being really sharp.

A look inside the body reveals a surprisingly clean interior, and no apparent cracks or splits in the bracing. Checking the braces with a palette knife reveals some missing glue bonds, confirmed by rapping the soundboard and hearing the respondent "rattle".

The large Rosewood bridgeplate is home to six chewed-up bridgepin holes. The soundboard split (from the shrinking pickguard) goes clean through the top, and is split on either side of the X-brace glued to it.

The action is high, and is miserable to attempt to play. At ≈ 161 lbs of light gauge string tension, the action is right at 9/64 inches (3.3 mm, 0.130 in) at the 12th fret. This guitar shipped from the factory with medium gauge strings, not light gauge (like is on it, currently). Add that additional 25 lbs of string tension and this guitar, while sounding remarkably better for the way it is currently braced, would be rendered completely unplayable due to the increased action.

With light gauge strings installed, this neck has a pronounced forward bow under string tension. Contrasted with a straightedge laid across the surface of the frets, the gap measures 0.045 in. (about 3/64").

When string tension is removed (by loosening the strings), that gap only drops to 0.031 in. (about 1/32"), demonstrating a permanent bow.

Top deflection is a way of describing how far the soundboard will depress under a specific load. Though is is not necessarily an indicator all on its own, by using the same weight across multiple tests and/or multiple guitars, useful comparisons may be drawn.

I will measure deflection with the strings both tightened to pitch, as well as with the strings loosened. I first apply the weight, then "Zero out" the indicator.

By simply removing the weight, I can take note of the movement as the top springs back to its "before" position (as in, before I added 1 kg of weight to the soundboard, directly atop the bridge). Three (3) tests are sufficient for me to rely on the results. This soundboard is very "tight", in that it does not move much under a 1 kg load.

Top deflection (1 kg - strings under tension): 0.06 mm (0.002 in.)

Top deflection (1 kg - strings loosened): 0.06 mm (0.002 in.)

For reference, with the same weight applied, a high-end D-28 will typically measure ≈ 0.09 mm, while a more modern, boutique soundboard will measure in the 0.12 mm to 0.14 mm range.

I doubled the weight to 2 kg and repeated the tests. The results are more interesting, and a better example of how this typically works.

Top deflection (2 kg - strings under tension): 0.12 mm (0.005 in.)

Top deflection (2 kg - strings loosened): 0.17 mm (0.007 in.)

NOTE: Being one of the (increasingly smaller group of) players who purchased guitars, new from the factories, back in the '70s, I have some information that might be of interest to a younger generation. Guitars from this era were shipping with medium gauge strings. The extra tension was sufficient to "drive" the "over-built" soundboards. When you string one of this era's guitars with the calibre du jour, light gauge (or worse: extra-light gauge) strings, don't be surprised at its lackluster performance.

Having such a "tight" soundboard on a '70s Martin is not at all unusual, in my experience. And it is good news, frankly, as it can be interpreted as an indicator of the potential of a given guitar. For more reference, I measured a Guild from the same period, with the strings tightened to pitch. Top deflection using a 1 kg weight: 0.08 mm. Top deflection using a 2 kg weight: 0.15 mm.

Soundboard rise is measured behind the bridge and is the difference between the height of the soundboard under tension and not under tension. This Martin is moving 0.27 mm (0.010 in.). This number is important to factor in when calculating how much material to remove from the neck heel during a neck reset.

The bridge rotates forward under string tension 0.8°, indicative of a fairly "overbuilt" (too rigid) X-braced soundboard.

The bridge is 8.3 mm (about 21/64") tall.

The saddle protrudes 1.6 mm (about 1/16") above the bridge.

Overall string height in front of the bridge measures 9.9 mm (3/8").

If the bridge height was closer to 9.5mm (3/8"), and the saddle projected another 3mm (1/8") above it, the extra torque applied from an overall string height of 12.5mm (1/2") would generate better performance with this (heavier) bracing. This is why lowering saddles, let alone shaving braces, (and don't forget slapping on uber-light-gauge strings) on these guitars, eventually mutes their output.

A straightedge, when laid along the fretboard, collides into the front of the (already shaved) bridge. At this stage, it is a good indicator of the need for a neck reset.

I will use a chart to display the actual body measurements of this guitar, as they differ based upon whether you are measuring the soundboard, or the back.

A few more comparative measurements are needed before I begin work on this guitar.

Body frequency (air resonance, or T(1,1)1: 93 Hz

Soundboard frequency (monopole, or T(1,1)2: 201 Hz

It is painful to play in its current state, but I need to capture a "Before" recording of this guitar.

Before and After

The Assessment

This Martin is plainly appointed, and the owner wishes it to remain that way.

The guitar sounds "ho-hum" and there are definite reasons for its lack of performance. It could and should sound fantastic! I don't want to "restore" the sound of this guitar, as I don't believe it was ever that great to begin with.

I am interested in taking the guitar where Martin could have taken it, had it been built just a few years earlier, or had they been able to spend a little more time with just the one instrument. I know that I would readily achieve that by replacing the soundboard, but I believe I can also accomplish my goal using the existing soundboard, with a few modifications made to the bracing. We'll see.

A few thoughts on the body measurements:

The Tail is about 1/32" shy of 5" deep. That qualifies these guitars as "deep bodies" by my standards. The depth at the Neck Block is about 1/16" shy of 4". The difference in depth between the two is just over 1 inch. For fun, you may want to measure your own guitars to see how they compare.

The body is nearly 1/8" shorter in its length on the soundboard side than on the back, and both the upper bout and waist are slightly wider. I have yet to measure such disparity as consistently on brand new instruments, but see it on decades-old instruments frequently. This lends credence to my understanding of body flexion, neck block rotation, and upper bout compression.

In addition to having disparate widths, the soundboard and back are not perfectly aligned, resulting in that wonky wobble when you try to stand the instrument on its side on a perfectly flat surface. Press the lower bout flat, and the upper bout wobbles. Press the upper bout flat, and the lower bout wobbles. One side may be more pronounced than the other.

In the image, below, the neck block end is facing you, the soundboard is facing to your left, the back to your right, and the (treble side) lower bout is pressed flat against the marble slab.

I have three paths in front of me, each representing an increasing amount of effort:

Beginning Disassembly

I removed the pickguard. This process always goes one of two ways, either with a clean removal or a wood fiber tearout. This one required just a little bit of heat and rewarded me with a super clean removal.

All of the parts and accessories get removed. That includes the strings, plastic pins, saddle, nut, tuners, and the strap button. 

The only item here of any real "value" as far as I am concerned, is that Ivory nut.

The split in the soundboard is more clearly visible, now:

This type of soundboard split is not unique to Martins. Below is a Guild that is six or seven years older exhibiting the same type of split.

Here are photos of the headstock, front and back, with the Grover Rotomatics removed:

Bridge Removal

Focus now moves to removing the bridge. I use a small heat blanket connected to a power source to warm the glue holding the bridge on. A thermocouple lets me control the temperature to keep things safe and secure.

There are a few ways to accomplish this, and they all usually result in a delicate balancing act with some serious heat involved. I built a simple jig that mounts to the neck and holds a swappable caul. This leaves two hands free to manage the heat.

I set these types of blankets to ≈ 190° F (88° C) for bridge, fretboard extension, fretboard, and headplate removal. I never leave a heated blanket unattended (a developed habit). Upon occasion, a drama may occur, involving some combination of heavy runout, or a surprise adhesive, and/or not allowing that extra minute or two of adequate heating to soften the glue prior to sliding in a palette knife. But "patience" is the operative word.

This bridge came off so cleanly after only about 10 minutes of heating, it made me curious. I took a little harder look at the two mating surfaces. It is not completely unusual to find bridge wood fibers on the soundboard, along with soundboard wood fibers stuck to the bridge. But it is unusual when they don't match.

There are just enough sections of missing Spruce on the soundboard that are not on this bridge (and should have been present if they were just torn off during this removal) to make me think this bridge may have been removed, before.

Regardless, the bridge is off without drama.

Fretboard Removal

Whenever possible, I prefer to take the fretboard off of the guitar leaving all the frets in place (regardless of their condition) and leaving the neck still attached to the body.

If you remove the frets, first, you exponentially increase the risk of splitting the fretboard during its removal. If you remove the neck, first (as you would for a neck reset), and then attempt to remove the fretboard, you have to come up with a clever means of holding the neck, either by the heel or the headstock, while you both apply heat and work with palette knives. It can get ugly.

I carefully scrape the lacquer at the edge of the fretboard, along the line formed where the fretboard contacts the neck, in an effort to mitigate lacquer chipping. I used to heat a blade and scribe the line, but that line is not always where you think it is, and can be especially hard to locate on these heavily-tinted necks.

I lay a long, trapezoidal heat blanket atop the fretboard, lay a strip of cork insulation on top of that, then rest a metal radiused sanding beam on top of the pile. A couple of clamps hold the assembly tight against the fretboard. After bringing the blanket up to temp, I let it warm up the board for about 15 minutes.

I set the large blanket aside and switch to a short rectangular blanket, of the same type that I use exclusively for removing the fretboard extension (after all, that is what I am about to start with). Stay with me, as this should make sense, soon.

I lay a short, wooden radiused sanding block atop that blanket, and set a weight on it to maintain good contact. I quickly bring the small blanket up to temp and begin testing the extension for glue release.

I use a long palette knife to carefully separate the fretboard free and away from the soundboard.

But I'm not stopping after releasing the fretboard extension from the soundboard. I slide the small heat blanket assembly a short distance up the board and work the palette knife (or knives, as needed) to separate the fretboard from the Mahogany neck. Cast your gaze upon this beautiful site:

Lather, Rinse, Repeat. I work my way up the fretboard, letting the heat and the knife's resistance to the glue guide me. It is so important to maintain sufficient temperature transfer, and let the heat do the heavy lifting. If my setup is too cool, I risk damaging the wood. If it is too hot, the wood will discolor / burn.

As you can see in the photo, above, I reversed the direction of my short blanket once I got up to the first fret. Using my little assembly of blanket, caul, and weight, his let me get better contact with the very end of the board.

The results speak for themselves. A very clean fretboard removal.

You get a clear view of the infamous 3/8" square tube. Take note of the dark stripe running down the length of the steel. The tube is marked on one side, the side intended to face "UP" during installation. It is the side that will be in contact with the fretboard. I will reference this stripe again, a bit later.

The underside of the fretboard is very clean, as well, but for a section of the fretboard extension. There is a reason for that, as we shall see, shortly.

Forensic Moment

You may have already caught site of it, but here is a close-up showing evidence of a previous neck reset. Two holes were drilled through the board at the 15th fret, spaced so as to avoid hitting the steel square tube, all to gain access to the space between the end of the tenon and the back of the mortise in the neck block. That almost worked.

When the neck was re-glued during the previous reset, it was shimmed with those thin, light-colored strips of wood (Maple), on both sides of the dovetail.

Here is a better view of the area of the soundboard beneath the fretboard extension. The Spruce is in remarkably good shape, especially considering that this is the second (presumably) time the fretboard has been removed from that area.

Here is a corresponding view of the underside of the fretboard. Neck removal #1 tore its share of fibers loose from the Spruce top. Gone was the 220-grit smooth sanded surface that was present when the guitar was built.

A generous amount of adhesive was (re-)applied after that first neck removal, adhesive that had an opportunity to find lots of new surfaces to bond with, making future removal(s) exponentially more difficult.

I don't have any hard numbers on what percentage of necks are reset with a "Not my problem" approach regarding future repairs. From my own encounters, I am convinced it is a compounding issue.

No soundboard can withstand this, at least, not for very long. Whether, like me, you consider this to be a damage condition, or not, you have an issue with your guitar that time will exacerbate.

Removing the Square Tube

For the curious: the steel square tube does not just sit, loosely, in a slot, as you would expect for a 2-way truss rod. It is glued in place. I laid the fretboard heat blanket on the surface of the neck and warmed the square tube for ≈ 10 minutes. After setting the blanket aside, I switched to my trusty heat gun. I worked the body end of the neck first, focusing heat on ≈ 6 inches of metal at a time.

I switched to the headstock end and repeated the heating. Using a slot screwdriver as a lever and a sacrificial stick as a fulcrum, I worked a little on both ends of the tube, pausing my prying to re-heat the metal. First on one end, then on the other, back and forth.

The heat softens the glue sufficiently to permit a gentle prying to release the bond. Again, patience is key - let the heat do the work, not the prying.

No drama. The square tube lifted out cleanly. Total elapsed time: < 10 minutes.

NOTE: Neither Martin's steel square tube nor its predecessor, the T-bar, resided in channels cut all the way up to the nut. As you can clearly see in the photos, the groove stops 3/4" shy of the nut. The wood from the far end of the headstock, under the nut, and extending 3/4" more to the steel tube, is not reinforced. Whether you think that is a big or a little thing, it is most definitely a thing.

The question of neck material depth comes up from time to time, so this is a good opportunity to document it. Using a depth gauge, I can measure the thickness of the neck remaining between the bottom of the groove and the back of the neck.

One revolution around this particular dial indicator is 10 mm. That means there is 6.7 mm (0.26 in) of neck material remaining beneath the 1st fret. This neck increases in thickness as I move toward the heel, though the groove remains the same depth.

Stiffness is all about height. For example, imagine incorporating an "I" beam into a guitar neck. For optimal results, one plate would be affixed down the center of the back of the neck, and would be tied to a second horizontal plate running along the surface of the fretboard.

While it might make for the world's stiffest neck, such a design is impractical for (hopefully) obvious reasons. Additionally, an I-beam would not have the torsional benefit of a square tube, meaning an I-beam would provide less resistance to twist. But a square tube cannot achieve the same height (depth) in the neck.

The square tube measures precisely 0.375 in (9.53mm).

Martin would have selected a 3/8" square tube as a stiffener because it was as deep as they could reliably and consistently inset a SQUARE into a neck that is ROUNDED on its back without the edges of the square breaking through. 

I placed the side of the square tube having the dark stripe running down its length up against a dead flat straightedge, I could just barely get a 0.030" feeler gauge between the two. It is most definitely "bent" forward, bowed by the same amount that I measured in the neck with the strings removed.

Here is a photo with the square tube rubber-banded to the straightedge and held up to the light. You can clearly see the bend in the steel. 

Neck Removal

Prior to switching to "foam cutters", I have reset many a cantankerous (heavily slogged with glue) Guild neck using this HeatStick, though it does require drilling two LARGE holes through the fretboard  that are NOT covered by replacing a fret. With the fretboard safely out of the picture, and with the massive gap between the neck tenon and the mortise in the neck block seemingly begging for it, I decided to try this yet one more time (I should know better, by now).

There is no "hole" into which to insert the hot metal wand. Instead, there is a cavernous slot, filled with air, into which I drip a small amount of distilled water in order to generate steam. The goal is to soften the glue. This scenario always works in my head, but it is not as successful in practice.

If you are unfamiliar with the whole disassembly process, you might assume that removing a neck where the fretboard has already taken off would be a breeze. Well, that isn't necessarily the case. An attached fretboard works much like a lid on a pot, holding the warmth in the dovetail joint. My pot has no lid, so the glue and surrounding wood is cooling rapidly.

As with the 6 times I've tried using a HeatStick like this, previously, I switch over to steam. (If I ever suggest I am going to attempt this method yet one more time, somebody, please, smack me.)

A tried and true method for neck removal uses a steam wand. With the fretboard removed, it does take longer than when the fretboard is still attached, due to the open area being subject to more rapid cooling. But it works, and works well.

The neck is off, cleanly. The previous neck setter did a decent job.

You may recall, I had measured the permanent forward bow in the neck at 0.031" before I removed the fretboard (see the "Measurements" section, above).

With the neck off, the fretboard removed AND the (bent) steel square tube removed, the Mahogany neck has a permanent forward bow formed into the wood, measuring 0.030". I'll fix that with the carbon fiber. Stay tuned.

I cannot simply sand the neck flat, as I would if this was a pre-carved neck blank. Sanding this neck immediately reduces it's width, which must not be permitted. I'll fix the bow with the carbon fiber. Stay tuned.

The D-Tube

GOAL: I want an inherently stiff acoustic guitar neck that is lightweight.

The steel square tube from the 70s fails to meet either criteria. The steel weighs 115.5 grams and is contributing plenty of weight, already. Replacing it with a larger piece of steel (for added stiffness) takes me in the wrong direction, weight-wise, not to mention, it just won't fit. There isn't sufficient depth remaining to increase the size of a square groove without risking blowing out the back of the neck at the corners of the metal.

In this Martin neck, the steel has bent. The photographic evidence is indisputable. Furthermore, the bent steel has de-formed the wooden neck.

The 3/8" square tube is insufficient to prevent forward bow.

• Strung to pitch with light gauge strings, the neck exhibits 0.045" (3/64ths", 1.14mm) of "relief" (forward bow).
• Unstrung, the neck bow is reduced by 1/3, dropping to 0.031" (1/32nd", 0.79mm) of permanent "relief" (forward bow).
• Fretboard removed and bent steel square tube removed, the wooden neck exhibits permanent forward bow of 0.030" (1/32nd", 0.76mm).
• The steel square tube, when tested against a dead flat slab of marble, as well as against a dead flat steel straightedge, exhibits permanent forward bow of 0.030" (1/32nd", 0.76mm).

Re-read the previous sections, above, if you missed all of that.

Let me repeat: The 3/8" square tube is insufficient to prevent forward bow.

You don't have to take my word for it. Let history do the teaching ...

If my assertion was not true, there would be no need for compression fretting to "fix" (counter, correct, whatever) bowed necks fitted with steel square tubes. There would be no need for epoxying carbon fiber inside the square tube. There would be no need for re-gluing the steel tube, believing it somehow "slipped" out of place.

In my experience, the 3/8" square tube obviously works better on some necks than others. Environmental factors such as temperature and moisture, along with choice of string gauge and the number of years under tension, all certainly play a role in determining just how long a given neck will resist curling before succumbing. I know (but can't prove) that the individual, inherent stiffness of each wooden neck makes a contribution, as well.

To me, the important thing is that, while others had jumped ship long ago to embrace the mechanical solution of the adjustable truss rod, Martin worked diligently toward maintaining a symbiotic relationship between an inherently stiff neck and string tension. They got the combination really close using both the T-bar and the 3/8" tube.

Timing is everything. I have long speculated that, if Martin had access to a carbon fiber product such as the D-Tube, or Matt McPherson's necks, they may have never embraced the adjustable truss rod. Who knows?

What we do know is that stiffness needs height, not width. A beam includes two horizontal planes having a fixed vertical distance. The further apart those planes, the greater the moment, the more stiffness. Thus, to increase stiffness in my neck beam, I need to increase its height (or depth, in this case).

In the photo, below, I am holding a section of the carbon fiber D-Tube next to the steel square tube. Note the the dimensional differences. Height matters!

I can't achieve sufficient height with a square profile.

I am not constructing a high rise, a cross-section of whose floors reveal flat profiles. I am stiffening a funny shaped architectural wonder whose cross-section is flat-ish on one side and really rounded over on the other.

Gluing a square beam inside such an oddball shape, I must be conscious of the stress at the transitions (the bottom corners of the square). The closer I get to the back of the neck, the more likely I am to witness cracks/splits at those corners. I know because I have tried.

Enter the D-Tube. Being 33% taller than the steel tube, this accounts for its increased stiffness. It is twice the width of the steel tube, providing me with unrivaled torsional stability. At 46.5 grams, the D-Tube is 60% lighter than the steel tube. Mission accomplished.

With D-Tube, I achieve my goal. Mission accomplished.

The New Neck Groove

I am modifying an existing 3/8" x 3/8" x 14" groove, widening it to 3/4", deepening it to 1/2" and lengthening it to 14-5/8".

• Before: 9.5mm x 9.5mm x 355mm
• After: 19mm x 12.7mm x 371

You may recall me measuring the remaining wood beneath the 3/8" groove. There is roughly 1/4" wood remaining, of which I will be removing approximately 1/8". It would be foolish to remove that much material if I was intending to loosely lay a truss rod in the the new slot.

I will be permanently bonding the wood with the stiffener, so there is plenty of wood remaining.

The original groove is a rectangular slot. The new groove is radiused on the bottom, forming a semi-circle. To cut it, I will use a core box bit at the router table.

When slotting a new neck, the neck blank has parallel edges to reference against a fixed fence, the purpose being to center the slot. An existing neck, however, has been shaped into a trapezoid that is narrower at the nut. I can't just push it along a fence to cut or modify a groove down its center. I have a sled for the task.

The two (2) buttons in the lower-right corner of the photo, above, center the slot in the neck with the slot in the sled.

The neck is held in place, securely, and the buttons are removed (from the underside) prior to commencing routing.

I control the depth of the cut (height of the bit above the table), mindful to "sneak up" on my target depth. I control the length of the groove by using a stop block on the table, to prevent cutting through the headplate.

NOTE: For new builds, I prefer to slot all the way through the length of the neck, out onto the headstock, reinforcing the entire length. Gluing a D-Tube into a neck slotted this way makes it nigh unto impossible to snap the headstock (due to a fall, or severe mishandling). This dramatically reduces the potential for forward bow at that section at the base of the headstock, beneath the nut, right before the end of the truss rod slot. For an existing neck that does not feature a truss rod cover (that would hide the incursion), to slot the neck all the way through out onto the headstock would require removing the headplate. That is major surgery, having severe consequences when things go wrong. I will forego that level of modification, here, but extend the length of the groove as far as I dare.

Several passes later, raising the router bit incrementally each pass, I produce a cleanly routed neck, ready to receive the D-Tube.

For a close fit, I round the end of the carbon fiber nearest the nut.

The D-Tube can be glued in place with a variety of adhesives, including hot hide glue.

TIP: Hide glue has a sonically vitreous quality about it. By extreme contrast, rubber cement has a sound damping, or deadening quality about it. Carbon fiber can add a bright, sometimes too bright, sometimes downright plastic-y sound to a guitar. I say "can add", as it depends on several factors.

I think hide glue is too bright to add to the already bright character of the D-Tube. If I was building a Walnut and Cedar guitar, for example, hide glue can work. For this Mahogany guitar, I won't even use FGCI SuperBond, my go-to epoxy for guitar construction.

I want an epoxy with more rubber in it, both for increased peel strength and for less sibilance in the neck. My choice is 3M's Scotch-Weld 2216. It is a bit more fiddly to mix, but it is just the ticket, soundwise.

TIP: Save yourself some grief by first applying some tape across the flat surface of the neck. Include the D-Tube and tape over the whole thing. The purpose of the tape is to isolate any squeeze-out onto the surface of the tape, not the surface of the neck (or the D-Tube).

Trace the outline of the D-Tube using a sharp knife. Remove it to apply adhesive.

The first neck I ever fitted with a D-Tube was a 5-piece Mahogany lamination blank. I labored over the groove, cutting the world's most precise fit {snicker}. Resting in its cradle, the D-Tube was perfectly flush with the surface.

I did not account for the added thickness of the layer of glue. By the time the epoxy had set, the D-Tube sat just slightly proud of the neck. I was forced to sand away some of the carbon fiber so the fretboard would lay flat without rocking. Not ideal. Don't do that.

The secret for success with a D-Tube is to rout the groove a little deeper, so the D-Tube sits slightly below the surface. Apply epoxy, leaning toward "sparingly" as opposed to "generously". 

Press the D-Tube into place and, if there is any squeeze-out, clean it up. Unlike typical wood glues, epoxy should not be clamped tightly. Rotate the neck face-down onto a flat, protected surface. I apply a single clamp in the center of the neck to hold the neck down perfectly flat while the epoxy sets.

I am not trying to pinch the D-Tube up tightly against the bottom of the groove. I am providing a level surface for the D-Tube to rest against while the epoxy layer between the base of the D-Tube and the bottom of the groove cures.

The tape does its job and protects the wood surface from any squeezeout (of which there should be little to none, depending on how much adhesive you applied).

The end result is a Martin neck retrofitted with a DragonPlate D-Tube.

The surface of the retrofitted neck is now dead flat, forming the ideal plane to re-attach the original fretboard onto.

Back Removal

I am taking the back plate off to gain access to the insides. Back removal is typically less harrowing than removing the soundboard, unless I am not intending to keep the soundboard intact.

I rout off the bulk of the Boltaron binding. This is tough stuff, a durable thermoplastic that resists dents and shrinkage, as well as flame. What more can a guitar player ask for in a binding?

I will be re-binding the guitar after I re-attach the back. At that time I will cut the binding channels to full depth. This approach (not routing a full-depth channel, yet) leaves me some "wiggle room". I carefully saw around the perimeter ...

... and I get that first look inside the box. Opening up a guitar that has been sealed shut since it left the factory never gets old. It always feels a little like stepping back in time.

Note the water stains on the sides, on either side of the tailblock. It is not uncommon to find water stains inside a guitar whose neck has been steamed off, depending on who did the work and how careful they were to keep water out of the body. But these stains don't have the typical "rivulet" path.

They more closely resemble water staining from side bending. There is more evidence of stains that have been sanded out near the waist and again in the upper bout. Curious. Regardless, it is merely an aesthetic thing, having no impact on the performance of the guitar. They will get cleaned up.

Bridgeplate Removal

Heat, a palette knife, and patience. A clean removal is most rewarding.

Popsicle Brace Removal

The Upper Transverse Graft, or "Popsicle Brace", is a bit more challenging to remove with heat. It can be done, but I choose to use a small block plane and a bullnose plane for the task. Again, a clean removal is most rewarding.

Note: The pencil lines are "original".

Transverse Brace Resized

I am reducing the thickness ("width") of the Upper Transverse Brace from 1/2" to 5/16".

Scalloped Bracing

I have scalloped all of the braces, excluding the Upper Transverse Brace and the forearms of the X-Brace.

I have also taken the time to reduce the width of the entire X-Brace, from 0.3335" to 0.3225".

OF INTEREST: I had measured the top deflection (how far the soundboard could be depressed under a 1Kg load) prior to beginning disassembly. Here is a pertinent snippet of that information:

With the bridge removed and the interior braces scalloped, the top is much more flexible.

Top deflection (1 kg - no bridge, braces scalloped): 0.18 mm (0.0075 in.)

(The bridge is a substantial brace, being the sole exterior brace. Attaching the bridge will tighten the soundboard up again.)

The purpose of scalloping the interior braces is to increase movement in the top. The risk of scalloping the interior braces lies in relinquishing structural support to the tension of the strings. There is a (relative) point of failure that one must not cross.

Having access to an incredible amount of measurement and performance data, today's guitars can be consistently manufactured lighter and more "responsive" (less string energy is required to generate a pleasing sound), all without crossing that threshold of "Oops!".

When compared with a "pre-war" D-18; specifically, a guitar built from 1935 through 1937, there are a few obvious differences. I assume that an early '70s Martin wasn't being hand-built with either the same selection of materials or the attention to detail.

In those first dreadnoughts, the entire bracing pattern sat 7/16" closer to the soundhole. The two tone bars were positioned closer together and were angled more toward the waist than the tail. The braces were narrow and scalloped and the bridgeplate was sized to the bridge.

If I was replacing the soundboard, I would pattern the bracing after those guitars. 

I'm obviously not starting from scratch, where I have complete control over every aspect of each component. The braces are glued where they are glued. Armed with a few hand tools and an already-built guitar, let's see if I can get anywhere close to success without the aid and benefit of computer models and CNCs.

Kerfed Lining

I removed the back by cutting through the kerfed lining ("kerfing") at the back edge of the guitar sides. I used a thin kerf saw (in lieu of a router) to sever the back from the sides.

I could simply re-attach the back, as is. Few would question it, as this is a very common technique.

The consequence of this approach is a reduced body depth. Granted, that saw blade was only 0.020" (0.5mm) wide, leaving a kerf about twice that amount. That amounts to a rounding error in most guitar shops I am familiar with.

To be fair, by the time the sides are prepared for reassembly (smoothed, planed, radius sanded, etc), and the back is cleaned up, that figure increases and the body depth decreases.

Been there, done that. I will remove the kerfed lining and replace it, allowing me to re-attach the back and maintain full (original) body depth.

Upon occasion, I will encounter sections of components that were (are) not fully bonded, as in, there is no adhesive bond between wooden pieces in certain areas.

On this guitar, I have already counted four sections of the kerfed linings that were not glued (see the photo, below, for one of these). This can also apply to sections of bracing, as well as to the neck and tail blocks. I have not identified a pattern, and am left to speculate.

The sides are extremely thin and extremely brittle. They will readily crack, a condition that is exacerbated as kerfed lining is removed and the sides have even less support.

I find that having curved bracing chisels, tools designed for the task, is a must. Even more important is their sharpness. A dull chisel encourages the craftsman to push harder, and that is the source of the majority of mistakes.

On the other hand, one slip with a razor-sharp chisel can be a career-ending experience, or worse. Not to mention, you can ruin a workpiece in a moment.

More to come ...