FRETS.COM Repair Blog, April - May, 2010
Please click on the small images to see the large views.

Some comprehensive maintenance on
1937 Martin D-28
© Frank Ford, 2010; Photos by FF

The patient is a fine 1937 Martin D-28, an instrument that has had 73 years of experience being a guitar and holding the load of tension brought on by six steel strings tuned to pitch. At this stage of its life, the time has come for some serious regular maintenance. I'll be resetting the neck, leveling the fingerboard and refretting, replacing a broken bridge, replacing a cut-down nut, regluing the pickguard, and repairing some damage to the peghead, along with a few other small operations.

I thought it would be interesting to present this project in "blog style," taking the various steps as they occur. If this proves to be fun, I'll probably do more this way. The project days will be sequential but not contiguous, because I won't be able to work on this job every day.Here's a listing of the days, so you can jump to them quickly:


Day 1: Inspection Day 2: Disassembly Day 3: Reglue Pickguard
Day4: Pickguard Cleanup Day 5: Fill Tuner Holes Day 6: Fill Tuner Screw Holes
Day 7: Drill out Veneer Damage Day 8: Tool Making Day 9: Veneer Inlay
Day 10: Drilling Tuner Holes Day 11: Finish Touchup Day 12: Another Small Tool
Day 13:The Bridge Plate Day 14: More Bridge Plate Day 15: Top Crack Repair
Day 16: The New Bridge Day 17: Resetting the Neck Day 18: Refretting
Day 19: Setup Day 20: Setup & Final Touches  

Day 1: Inspection

Unfortunately, I did forget to take a comprehensive series of "before" photos because I hadn't thought of doing the blog until I'd started taking things apart. While this guitar is in fine condition for its age, there were a few features I did shoot for the record.

First the bridge, showing its obvious crack between the pin holes, and the not-so-obvious finish damage behind from previous regluing:

Additionally, the bridge had been cut lower than standard height to lower the action at some time in the past. So, there are a couple of good reasons for replacement.

Inside, everything looks good, althugh there are a few stains here and there:

The bridge plate has enlarged slots that trap the string balls, allowing them to pull through the top and bear against the lower surface of the bridge, so those will need attention:

There's a small shrinkage crack in the top at the edge of the fingerboard:

I'll put some reinforcement there because this crack can pose a serious structural danger if it is allowed to spread. More on fingerboard top cracks.

Sometime in the 1970s, the original Grover G-98 tuners were removed and replaced by modern (and superior) Grover Rotomatics. That operation required drilling the peghead tuner holes from 1/4" to 3/8" diameter. The luthier who did the job clearly had a difficulty on one hole, the low E. I guess the drill bit had caught and ripped a big hunk out of the pehead veneer, which was reglued at the time, without finish touch up:

Before you get to "How could they have DONE that?" or "Why replace the ORIGINAL tuners?" take a moment to remember that in those days the ethic was entirely different, as was the cash value of vintage Martin guitars.

Day 2: Disassembly

Today, I'll be taking off the neck and bridge. I need to remove the neck to get full access to the dovetail joint so I can recut it and change the angle of the neck as I reglue it to the body. Preparatory to getting out the steam generator and actually attacking the dovetail itself, I first need to unglue the end of the fingerboard that sits on the top of the guitar.

Using an electric heat blanket that reaches a temperature of about 305 degrees Fahrenheit placed directly on the frets, I can keep the heat localized to the fingerboard itself:

From time to time I reach into the guitar and feel the temperature rise under the top in that area. When the top starts to feel as though it might be around 120 degrees, I remove the heat blanket and start to pry the fingerboard loose from the top, using a very smooth flexible putty knife with a blunt rounded edge:

This is one of those operations that's largely a matter of "feel," where experience contributes a good deal to success, as I try to sense whether the glue or the wood is giving way as I press the knife under the fingerboard.

Knowing that the fingerboard was glued to the top with hide glue, I listen for the telltale crackling sound as the heat-weakened glue line gives up under pressure from the knife. Had this been a modern guitar, assembled with aliphatic resin glue, I'd expect to hear a squeaking sound and feel the gummy resistance of the softened glue.

Back in the day when those tuners were replaced, it would have been fairly common for a luthier to saw the fingerboard all the way through at the fourteenth fret, and remove the end of the fingerboard as a separate piece to get visual access to the dovetail joint. As we've reset so many necks in the intervening decades, that practice has been abandoned in favor of not compromising any structural integrity. So, we work pretty much in reverse, taking the guitar apart in the same order it was originally assembled. That way, when the neck goes back on, we can say with confidence that it's good as new.

Since I still had my heat blanket controller on the bench, I switched over to the one that fits the bridge:

Notice that just as before, I have small lead weights to keep the blanket in place. I also have the blanket wrapped in aluminum foil, but I can't tell you for certain if that has any real benefit to the process. It does keep things a bit cleaner, and I think there's possibility that it may improve heat transfer, but that's about it.

Same as before, I try to keep alert for the feel and sound as I lift the bridge from the spruce top:

As the knife penetrates the glue line, I try to be aware of the direction of any grain runout in the spruce. I don't wnat my knife to be carried downward into the top as it lifts a splinter of spruce. While I have a firm hand on the knife, I have to avoid slipping for fear of scarring the finish around the bridge, and each time I pull the knife out to reposition it, I check the underside for any traces of glue that might adhere there. Even the smallest bit of glue can dig into the finish when I'm pressing hard against the knife!

Now, this is the way the textbook whould show the job - bridge off neatly and cleanly, with only a few spruce fibers stuck to the bottom:

Doesn't get much better than this. The top is perfectly intact where I'll be gluing my new bridge, and I can proceed with confidence that my replacement bridge will be able to have the same footprint as the original without the need to patch either spruce or lacquer.

Now, I'd like to introduce a seriously cool new addition to Gryphon's repair shop - the steam generator. This unit is part of a professional steam ironing station, and the flat area on the stainless steel tank is where the iron would ordinarily rest:

I had not known of these rigs before I started looking through Internet sources for an upgrade to our old steam generator - the venerable Krups home expresso machine. This particular steamer is sold as a jeweler's steam cleaner, and comes without the iron. Instead, it has a simple hose coupling. In the picture above, I'm holding the end of the steam hose to which I've fitted my own shop-made steam needle assembly and handle.

That black item in the foreground is a foot switch, and, of course, is normally on the floor. You can see two lighted switches - one for the boiler, and the other to lockout the steam valve that's operated by the foot switch. Behind the handle, there's a pressure gauge mounted on the same pipe as the electric steam switch and a separate manual valve that controls the volume of flow. The boiler is set to maintain 50 PSI steam as long as the power is on, and it does so without any attention other than making sure there's water in the tank.

So, once it's up to pressure, I step on the valve, and out comes the steam - pretty neat, eh? And, because it's running at 50 PSI, the steam is HOT - 221 degrees - so less condensation comes through the needle.

Now, steam is pretty nasty stuff if it contacts sensitive surfaces, say, like my hand. Or, possibly even more important, like a guitar's finish, which doesn't heal from a second degree burn. Recently I decided to upgrade my neck removal fixture, and have come up with this one, I'll call Mark III:

It's a heavy plywood base, with T-slot tracks at the edges and tall threaded rods for clamping the guitar down. Additionally, there's a polyethylene manifold and a tangle of "Loc-Line" flexible coolant hoses to direct streams of cold air wherever I think steam might excape to damage the finish.

Here's how it goes together. First, I set a pair of supports (you can't see 'em in the photo below) under the guitar body, and adjust this little built-in jack against the heel of the neck, to aid in pressing it upward as the glue is softened in the joint:

OK, now you can see those support blocks:

And, the padded bars I use to clamp the guitar body firmly in place. Also, at the front edge, you can see two orange clamps I use to hold the whole thing solid on the bench.

Once everything is clamped up securely, I can adjust the position of all those air squirters so they play cold air on all the areas where steam might escape:

And, finally, the air supply itself - the outflow hose from the shop vac:

Not a lot of PSI here, but lots of CFM, and that's what I'm after.

Time to get to work, yes? First order of operations is to remove the fifteenth fret, which happens to be located directly above the end of the dovetail in the neck block:

At that position, there's about a 1/8" gap, and I need to hit that gap as I drill a pair of 1/16" holes through the fingerboard:

One hole is for the steam needle that will poke way down into the joint to get steam all the way to the bottom. The other is a steam escape hole - yet another safeguard against damaging finish from escaping steam.

I took a whole bunch of shots as the steaming was progressing along, hoping to get one that showed the wisps of condensation coming up out of the escape hole, but none of them showed it. So use your Photoshop imagination, and pretend you see some, OK?

Here's the steam coming out of the hole:

The air blowers are operating full blast, and as steam escapes from under the fingerboard it is immediately converted to cold water, so there's no finish damage - whew!

After a little while, the heat and moisture from steam have melted the glue sufficiently, and the neck rises up from the body a bit:

Once the neck moves upward about 1/8" it is completely released from the joint, and I can lift it free with ease.

Well, the steam pot is still up to pressure, so I'll just take this opportunity to steam clean the glue off the end of the fingerboard to save scraping or more distructive processes later when it will have dried:

Day 3: Regluing the Pickguard

Well, here's another process I invented myself. However, like so many other ones I've developed, a bunch of other guys probably invented it for themselves as well.

I recall phoning my pal and veteran guitar repair guru, Dan Erlewine, about 15 years ago to tell him about some new techniques I was working on. I said, "You're really going to flip when you see some of the new things I'm up to right now. It's some pretty exciting stuff." Without missing a beat, he replied, "Oh, so you're getting into magnets, are you? I've talked to some other guys who are, too." Dang - I had not given him clue one, and he had it already. This business is like that. Seems lots of us come up with similar ideas at similar times. In the case of magnets, no doubt it was the recent availability of the super strong neodymium ones that pretty much captivated us all. Thus started a brief rivalry between me an Dan as to who could come up with the wildest and most creative uses. I'm still working on them, of course, but I think the magnetic spin sander will remain my best entry in the contest.

But, I digress. . .

With age, and accelerated by heat, celluloid pickguards shrink, but the actual amount of shrinkage varies from batch to batch of the plastic. So, it's not predictable whether a pickguard will shrink so much that it needs attention in the first few years of the guitar's life, or whether it will take a couple of generations for problems to arise, as it did with this one.

When the guitar was made, the pickguard was attached to the spruce top by "welding" the celluloid directly to the unfinished wood. The back of the pickguard was coated with solvent, making the celluloid tacky, and the pickguard was pressed down onto the spruce. A lead weight was placed on top, and the solvent was allowed to evaporate overnight, firmly bonding the celluloid to the spruce. Next day, the weight was removed, and the guitar went through the finishing process, with lacquer sprayed evenly over the spruce and the pickguard.

Today we refer to this style of pickguard as "under the finish," although it was the standard method of attachment until the development of the "self-adhesive" pickguards most makers use today. These have a very sticky adhesive - similar to the stuff on the back of packing tape - and a piece of protective releasing paper on the backside of the pickguards. The paper is removed, and the pickguard stuck down the way you'd apply a bumper sticker to your car. These new pickguards don't cause the same kind of trouble as the older style, and don't require tricky work to reglue them. If they fail, they can be replaced easily.

Well, I did it again. Let's get back to the business at hand.

As the pickguard shrinks with age, it either breaks loose from the wood, or it takes the wood with it. Most of the time, the early pickguards are so firmly bonded to the wood, they actually compress the top underneath. Sometimes that compression can be very serious indeed, causing severe cupping of the top and cracking of the top at the edges of the pickguard. Check out this example.

In previous decades, the "approved" repair was to laminate cross-grain spruce under the entire area of the pickguard to stabilize the top and prevent the spread of cracks. Unfortunately, the pickguard continued to shrink, and, while cracks didn't expand, the cupping of the top became even more extreme.

I prefer to release the tension on the top by carefully prying off the pickguard. Working slowly with a very thin palette knive, I can separate the spruce from the celluloid with minimal damage to the spruce:

I say "damage" because inevitably there are thousands of spruce fibers actually imbedded in the celluloid, and they tear free from the guitar top as I lift the pickguard. Once the pickguard is free from the top, it can then be reglued using regular wood glue, which takes advantage of the imbedded spruce fibers to reattach the pickguard firmly in place. Water soluble wood glue would never stick to celluloid; it would not be possible to reglue the pickguard using the original solvent technique because the solvent would destroy any finish it touches.

My process is to treat the pickguard just as though it were a wood veneer being laminated to the spruce guitar top.

Oh, yes, and I should mention that while I release the pickguard, I don't take it off completely. I leave it attached to a portion of the soundhole inlay, which, also being celluloid, is welded solid with the pickguard. That little bit will hold the pickguard in place so it won't shift when I squidge the slippery glue underneath and clamp it flat.

Speaking of clamping it flat - that's what I do, and with a vengence. Reglular wood glue can handle 100 PSI of clamping pressure, but the thin guitar top could not handle much at all without serious support above and below. I'll need heavy, flat clamping blocks, or "cauls" - one for the surface of the pickguard, and one to fit among the braces inside. Martin dreadnoughts all have about the same size pickguard, so I already have a regular acrylic-faced outside caul I use every few days on more modern guitars, but I don't have an inside caul for a 1937 D-38 - at least not yet.

First, I'll make a drawing of the brace locations in the pickguard area:

I use little magnets, one on top and one inside, to mark the outer edges of each brace, and connect the dots to transfer the brace position to the paper on which I've traced the pickguard outline.

Then, after cutting and building up a very thick hunk of plywood to the general pickguard shape, I cut out sections to clear the braces:

Here's the finished caul, after flat sanding the clamping surface:

The notches are probably twice as deep as they need to be, but I do that because in coming years, the caul is likely to get resurfaced any number of times to keep it flat and clean.

Enter the "Wizard of Ooze." I use a liberal amount of slightly thinned aliphatic resin (you know, regular yellow carpenter wood glue) and smear it all around under the pickguard. It gets pretty much everywhere, and most people chuckle a bit when they see me slather it on all over:

The glue wipes off easily enough, and does no damage to any wood or finished surfaces, so I press down and push out as much as I can, wiping the area clean with a damp rag or paper towel:

I lay down a piece of wax paper to resist the glue (I wouldn't want to create a 3/4" thick pickguard, you know.) and clamp it all up. I can usually fit in a heavy deep throat iron clamp and three of those orange "Pony" adjustable ones:

And, I'm not bashful about the clamping - I give it all a good hard squeeze before setting it aside to dry overnight.

Day 4: Cleaning up the Pickguard


Not what you'd call a big day today. I did get the clamps off from yesterday's operation, and after cleaning up the residual glue , the pickguard looks just about as is it did when the guitar was new:

Day 5: Filling Tuner Holes

Ordinarily, I don't cotton to using epoxy on vintage instruments, but there are times when I believe it really is the best solution to a difficult problem. Today's problem is the need to fill roughly drilled holes to fit the 1/4" tuner shafts of the Waverly tuners I'll be installing. They are direct copies of the original Grover "G-98" tuners that were original equipment on the 1937 D-28, although they are made with superior materials and preision than the originals.

Just as the originals, these Waverly gears have a decorative trim bushing that presses into the hole in the peghead and is held in place by the friction of a tight fit. While I could save some time and simply glue dowels into the holes, the bushings would then be held by vertical grain, and much more likely to slip out than if they were pressed into the flat grain of an undamaged peghead. The forward thrust of the tuner bushing is generated by the string tension, and ideally, that thrust would be held by end grain that's far less compressible than side grain. My job, then, is to fill the holes with plugs that have the grain oriented the same as the surrounding wood.

Since the holes were drilled roughly to 3/8" and I don't want to enlarge them for plugging, I make up some mahogany plugs using standard plug cutter:

Here's the epoxy part. Once the plugs are in place and drilled for the new tuners, there won't be a lot of solid mounting inside the walls of the hole between the original wood and the plug. So, I chose to use epoxy because it has tremendous cohesive strength and it fills gaps solidly. In addition, it works well to fill and hold the end grain between the plug and surrounding hole.

Because epoxy is there to hold the plug in place and fill the voids in the hole, I really don't have to clamp the plug in place. I do have a bit of a mess on my hands because I can't afford to have any of the epoxy setting up on that nice lacquer finish. For the next hour or so I watch for the epoxy to start to set, wiping off any excess with paper towels. No solvent that would dissolve epoxy can be used on lacquer, you know. Time to set it aside for the epoxy to cure tonight.

More, on day 6 . . .