When talking about a twist in the neck, we usually mean that the neck has moved either diagonally or across the width. When it happens along the length of the neck, most people will use the term warped instead of twisted. A warp can be adjusted out with the truss rod, a twist can’t (not an absolute truth, but generally speaking).
Semantics aside, let’s see what this neck did - EXACTLY.
The player had some complaints about the action. He tried fixing it himself, even replacing the saddle a few times, no luck. None of the saddles were the right hight or thickness (or material in my opinion), so that had to be fixed as well.
The edges of the frets were sticking out and very sharp.
As a reference, this wasn’t an old guitar. It was purchased new in 2015.
I started out by strapping the guitar in the neck jig, and checking out what the truss rod would let me do to get the fretboard as flat as possible, in as many spots as possible. As it turns out, not much was possible because the rod was close to maxed out.
Another issue to fix.
In the following video’s I’m checking the amount of relief using feeler gauges; not shown because I haven’t got 3 hands.
You can see the light shining through in different amounts and different places.
Between E and A
Between D and G
Between B and E
Using a bunch of feeler gauges, I started writing down the amount of relief in thousands of an inch on each fret, in these 3 different places. For example: at the 4th fret we had .006" of relief at the high E, .005" between D and G, and no relief at the low E string.
Eventually I ended up with the little schematic on the bottom, showing me were the high spots were.
Why no measurements beyond the 14th fret? Because it's not relevant. I was planning on sanding some more fallaway in the fretboard extension anyway.
Not much noteworthy was happening there, and the truss rod wouldn't affect it anyway.
The top numbers on the paper are the string gauges the customer wants to use, and the action at the 12th fret when it came into the shop, just for reference. And yes, it's pretty high for most players.
Time to jig the neck under simulated string tension.
With the same strings we plan on using, tuned to pitch, tilting the neck jig to 85° (playing position), I zero the dials.
Just turning the jig back to 0° (working position), makes the neck move under gravity.
The dials read out in thousands of an inch.
Time to take off the strings and the nut. I like to take the nut out diagonally to the slot. Taking it out sideways or straight from the top can sometimes cause little chips in the lacquer.
Even if you're very cautious and scribed the lines beforehand.
This is a pretty big guitar, with a deep body. The neck had to move quite a bit to get the dials back to the zero position, and the jig couldn't reach. So I quickly made a little extension block out of some scrap ebony, covered the top in leather, and double stick taped it to the jig.
To show you how sharp the fret ends are, I ran a block of scrap wood along the fretboard edge. If it rips up wood, it will rip up your fingers for sure.
Without any strings, in the working position, the neck is straightened out.
A deviation of half a thousand of an inch on the higher side is negligible, as the support rods will take up the difference.
Using some wax crayons, I marked out the schematic you saw earlier. Using yellow for the high spots. Why cover the entire board in white too? Because in the end I want to see a flat surface along the full length of the board.
On to the sanding!
I'm using a 600mm fret leveling beam, so that it will cover more than the length of the fretboard.
Using a leveling beam that's shorter (or just long enough) will cause you to sand away more in the center of the neck, and less at both ends. Using this longer one while sanding, the ends of the beam never cross the ends of the board, and sandpaper is in constant contact with the entire surface.
Just a few strokes proves my theory holds water.
After quite a bit of time, we got to this point. Notice the sanded area on the fretboard extension.
Almost there. I alternated between sanding and scraping for the pearl inlays. They are much harder to sand, and will end up giving you a false reading with the straight edge against the board. You can see the white streaks of pearl dust.
We have a flat fretboard!
But that's not enough. This board is flat without any frets (that could cause some compression) or actual string tension. The simulated tension can only do so much.
I sanded in some fallaway, to make life easier and to get a better setup in the end.
With the truss rod maxed out, and no tension in any way, the neck had a backbow of .018".
I wanted to increase that, so we would gain control of the rod again.
It would have to stay maxed out forever if we didn't fix this. Taking seasonal changes into account, that's not a good idea.
Some more localised sanding and scraping.
Slowly getting there.
I'll try to explain the theory behing sanding in drop-off and fallaway to increase truss rod capability. With the truss rod tight (and maxed out) the fretboard was perfectly straight.
By turning it a bit looser it becomes bowed. By sanding it again, only taking material of at the first and last frets, we have a straight fretboard once more (pencil line on the drawing).
But now the truss rod isn't maxed out. By tightening it again, we can create some backbow, gaining control of the rod.
I reinstalled the old nut, restringed the guitar and turned it to the playing position.
(This happened multiple times during this process, checking every step of the way)
Why aren't the dials reading 0? Because this is the amount of compression the frets were causing. I now have a reference when fretting to get these back to their 0 position.
Cleaning and deepening the fretslots, and prepping the board with sandpaper and steel wool to get a nice sheen. I skipped a step using a radius sanding block to make sure that didn't change to much. Forgot to take pictures, but nothing much changed.
The customer wanted to use Jescar Evo Gold wire, its hardness is in-between nickel-silver and stainless steel. It's hard on your tools and your hands. Prepping each fret with the most care possible.
Going in nicely, using hot hide glue and a handheld fret press.
For the upper frets, I switched to cold hide glue to have a little more working time before the glue would start to gell. I hammered them in, because the fret press won't work here, and then clamped them with a 12" radius block.
The Evo Gold wire is magnetic, so I put a large magnet on my flush cutters, so the pesky little off-cuts won't end up where I don't want them. They simply stick to the cutters now.
Filing them flush with the edge of the board, and establishing the correct angle.
Using an old StewMac concave fret file, I start rounding over the edges of the frets. The rest will be done by hand filing and sanding.
I'm not gonna go over the entire fretting process, as I feel that the main part of this job was rescuing this badly twisted neck. But these shiny gold frets sure do look nice!
Made a bone nut and compensated saddle, and she's out the door.
Thanks for reading!