I've had a few guitars with slight neck twists: some were 'dogs' and other set up surprisingly well. I have a theory about this and I'd be interested to hear other peoples' thoughts on the matter.

I like a lowish action and try to set-up my guitars accordingly.

I have found that is possible to get a good indication of a likely neck twist by looking along the guitar from the headstock toward the body. The eye needs to be about a metre from the headstock to see this. Idealy both headstock and body should lie in the same plane, if they don't they may be a slight twist (unless of course the headstock is twisted).

This can be confirmed by lieing the guitar down on a level surface such as a table and looking in the other direction: from behind the rear straplock, along the face of the guitar and up the neck toward the headstock, any twist, compounded by neck-relief will cause one side of the fretboard at the nut to appear higher than the other.

The combined effect of the twist AND any neck-relief means that, effectively, there will be greater relief on one side of the neck than the other.

I've heard it said that the thicker base strings need more room to vibrate than the thinner treble strings, therefore they benefit from greater relief or neck-bow. If the neck is twisted in one direction this will happen, but if in the other the situation would be reversed.

Thus it could be beneficial if there were more string relief on the low 'E' side than the high 'E', as would be the case if the plane of the headstock appears rotated slightly anticlockwise relative to the body when viewed looking down the neck from the headstock end (i.e. the low 'E' side of the nut is higher than the high 'E', when viewed on a table from the other end: test '2' above).

This is born-out by several guitars I've seen: those with headstocks twisted anticlockwise set-up better than those where the headstock has rotated clockwise.

No guitar tech book I've read mentions this, including Dan Erlewine (appologies Dan if I've spelt your Surname wrong!) I'd be interested to know if this idea is backed-up by anyone elses experience?

I have found that is possible to get a good indication of a likely neck twist by looking along the guitar from the headstock toward the body. The eye needs to be about a metre from the headstock to see this. Idealy both headstock and body should lie in the same plane, if they don't they may be a slight twist (unless of course the headstock is twisted).

I have an example of what you describe. But on mine, it's not due to the neck twisting, but a slant in the neck pocket. I know these pockets are cut with computer guided lathes or routers, but using a bubble level on the base of the pocket showed that it was clearly off horizontal by several degrees. Water damage maybe, I dunno?

The way I discovered the twist was by putting a 6" ruler on the headstock and sighting it down the neck to the body. (Boy was I p*ssed.) In order to adjust for the twist, I simply folded some aluminum foil to build up the low side of the pocket. Worked fine.

"those with headstocks twisted anticlockwise set-up better than those where the headstock has rotated clockwise."

I've had two guitars with mild anticlockwise twists and they set up very well -- even a little better than non-twisted necks. I came to the same conclusion as you -- that type of twist gives a little more relief on the wound strings than the plain strings.

So far, I haven't had a guitar with the opposite twist. I'd expect those to require extra relief and higher string height, but since the only guitars I work on are my own, I hope I never get a chance to confirm that part of the theory.

But sighting down the neck can be deceiving. The top of the nut is usually shaped/carved so that it's lower on the plain string side and that can make a straight neck look twisted. If you have a really good eye that might not throw you off, but many people will see a twist that isn't there.

A better way is to just do the normal relief check (fret/capo a string at the 1st and 17th fret and measure the gap between the bottom of the string and the top of the 8th fret), but do it on both the low E string and the high E string. If the neck isn't twisted the measurements will be the same, but with a twist one will have more relief than the other.

That's quite true, and I had actually taken that into account: I have a fairly new CP60s which takes an insane action without complaint! The neck twist is anticlockwise by a few degrees, BUT the neck pocket is actually slanted in the opposite direction! I can tell that by 'eyeballing' the proportion of maple visible on the neck, beneath the rosewood fretboard, at either side of the top-end of the neck in the neck pocket. Good point though!

I've had two guitars with clockwise twists, an 25 yr old MIJ '72 reissue; and also a American YJM - which went back to the shop! From what I've read on this site twists are relatively common on YJMs, must be the very slim neck and scallops? Both these guitars did not set-up well.

Oh, and just in case someone mentions it: I'm quite aware that one of the reasons the CP60s set's-up so well is the 12" radius board, this is especially true went notes are bent.

[quote="strayedstraterBut sighting down the neck can be deceiving. The top of the nut is usually shaped/carved so that it's lower on the plain string side and that can make a straight neck look twisted. If you have a really good eye that might not throw you off, but many people will see a twist that isn't there.

A better way is to just do the normal relief check (fret/capo a string at the 1st and 17th fret and measure the gap between the bottom of the string and the top of the 8th fret), but do it on both the low E string and the high E string. If the neck isn't twisted the measurements will be the same, but with a twist one will have more relief than the other.[/quote]

Yup, that is quite true, which is why I suggested siting along the guitar whilst place flat on a table, from the rear straplock up across the body, to the top-end of the fretboard where it meets the nut! Most my my guitars have rosewood fretboards so it is quite easy to see this without the any curve on the top surface of the nut confusing the issue.

I also take your point about measuring relief under both 'E' strings, that's how I discovered a significant closkwise twist in a YJM I briefly owned! I posted on here all about it at the time (searching 'YJM' would probably bring it up), although is does occur to me that after the frets have significantly worn it may give a false indication. I tend to play (well, try to!) lots of blues in A with an over relience on bending and vibrato perhaps, so this is particularly true of any guitar I've owned for a few months! However with a new, or refreted guitar measuring the relief would be a very good indication. After a fret-dress however (depending on how it's done), a neck twist may still be present on the fretBOARD, but actually dressed-out along the fretTOPS, especially if a long radius beam is used.

Maybe I didn't word it very clearly; what I was trying to say was to look at where the fretboard joins the base of the nut, not so obvious with a maple board though I guess!

Anyway thanks for responding on this Guys! Sometimes I wonder if I've got too much time on my hands!