From: dah22@cus.cam.ac.uk (D.A. Hubert) Subject: Re: BASIC GUITAR SET-UP 101 PART 1 Date: 27 Jan 1996 16:26:16 GMT Organization: University of Cambridge, England Looking forward to the rest. A couple small points. charles tauber (tauberc@candu.aecl.ca) wrote: : There are four, and only four, basic, universal adjustments that : affect the playability of a guitar. These are as follows: adjusting : the amount of relief (or "bow") in the neck using an adjustable : truss rod; adjusting the string height at the saddle; adjusting the : string height at the nut; and, adjusting the intonation. These four : adjustments are what I refer to as basic guitar "set-up". Probably jumping ahead here but is the order above the order in which you do your setups, I asume that the action at the nut is set before the saddle. : Practically, the way in which the tops of successive frets are made : given a greater initial displacement). Thus, there is no one : correct amount of neck relief; the amount depends upon the : preferences of the player, the type and tension of strings used and : the type of "attack" used by the player. As a general guideline, : approximately 1/64" or 1/32" of relief is typical. Are you familiar with the article in the Luthiers Mercantile catalogue by H.K. (can't spell the name) author of the classic guitar repair 1976 or something like that. What about his claim that the relief should not be evenly distributed but rather concentrated in certain sections. If you don't know the article I'll post a summary of his claims. Cheers -- Don Hubert, Social and Political Sciences University of Cambridge, CB2 3RQ, UK dah22@cus.cam.ac.uk ===================================== From: charles tauber Subject: Re: BASIC GUITAR SET-UP 101 PART 1 Date: 29 Jan 1996 17:02:41 GMT Organization: Atomic Energy of Canada Ltd. dah22@cus.cam.ac.uk (D.A. Hubert) wrote: > Probably jumping ahead here but is the order above the order in which > you do your setups, I asume that the action at the nut is set before > the saddle. No, I set the saddle height before the nut height. I'm not sure it makes that much difference which is first. Any thoughts? > Are you familiar with the article in the Luthiers Mercantile catalogue by > H.K. (can't spell the name) author of the classic guitar repair 1976 or > something like that. What about his claim that the relief should not be > evenly distributed but rather concentrated in certain sections. If you don't > know the article I'll post a summary of his claims. > Don Hubert, Social and Political Sciences Hi Don. Yes, I'm familiar with his article. I too can't spell his name without it in front of me. As a theoretical treatise, it contains several significant errors - the most noteable is that the amplitude of vibration is sinusodial rather than the parabolic shape he claims. That means that all of his calculations are based upon the wrong shape. I haven't crunched the numbers to determine the magnitude of the discrepancy because I find his method not very practical. As a practical quide to dressing fingerboards, I find his method of no value. I dress fingerboards using a carpenter's level to which I've glued a strip of 80 grit sandpaper. The natural action of the motion, and the tendancy of the sandpaper to dull faster towards the center, causes a slight backbow to be sanded into the fingerboard. This just happens to the the exact curve necessary to provide the desired relief when the instrument is fretted and strung. Go figure. Damn, and now I've given away my secret. I could have baffled you with "high-level" math and said that's how I do it, coupled with a complex jig. I bet I could even have sold a couple of jigs! Maybe next time. Charles Tauber, Luthier ======================================= From: dah22@cus.cam.ac.uk (D.A. Hubert) Subject: Re: BASIC GUITAR SET-UP 101 PART 1 Date: 30 Jan 1996 10:08:46 GMT Organization: University of Cambridge, England charles tauber (tauberc@candu.aecl.ca) wrote: : No, I set the saddle height before the nut height. I'm not sure it : makes that much difference which is first. Any thoughts? Depends on how close the nut is to correct action. I set the nut action first, independent of the saddle height, hold the string down at the 3rd fret and measure from string to top of first fret. Then knowing that the saddle height is exactly where I want it, I move to the saddle. Since my saddle measurements are not independent of the the nut, it seems to make sense to do the nut first. If for example on a new guitar, the nut were 1mm to high, my measurements for action at the 12th fret necessary for setting the correct height for the saddle would be off. But again, you probably have your own way around this problem that will come out in subsequent parts of your discussion of set-ups. : Yes, I'm familiar with his article. I too can't spell his name without : it in front of me. As a theoretical treatise, it contains several : significant errors - the most noteable is that the amplitude of One of his claims, and one that I have heard from other luthiers, is that relief is needed only up to the 12th fret. Thus the action test is first to fret the string at 1 and 18 or so, and check that the neck is straight at the 14th fret. Second hold at 14 and 1 and look for relief around 6 or so. Is there any merit to this approach, or should relief be greatest at the 12th? : I dress fingerboards using a carpenter's level to which I've glued : a strip of 80 grit sandpaper. The natural action of the motion, and : the tendancy of the sandpaper to dull faster towards the center, causes : a slight backbow to be sanded into the fingerboard. This just happens : to the the exact curve necessary to provide the desired relief when : the instrument is fretted and strung. Go figure. Damn, and now I've : given away my secret. I could have baffled you with "high-level" math : and said that's how I do it, coupled with a complex jig. I bet I could even : have sold a couple of jigs! Maybe next time. This appears to suggest that you anticipate the neck moving significantly due to the string tension, moving from "back-bow" to correct relief. In my experience the amount of movement here cannot really be estimated in the abstract, some move lots others little. Or are you referring to your own instruments here. Again, thanks for the initial post, its great to have some comprehensive post to respond to rather than a series of smaller questions that happen over months or years. Cheers, -- Don Hubert, Social and Political Sciences University of Cambridge, CB2 3RQ, UK dah22@cus.cam.ac.uk ===================================== From: charles tauber Subject: Re: BASIC GUITAR SET-UP 101 PART 1 Date: 30 Jan 1996 15:17:37 GMT dah22@cus.cam.ac.uk (D.A. Hubert) wrote: > > charles tauber (tauberc@candu.aecl.ca) wrote: > > > : No, I set the saddle height before the nut height. I'm not sure it > : makes that much difference which is first. Any thoughts? > > > Depends on how close the nut is to correct action. I set the nut action first, > independent of the saddle height, hold the string down at the 3rd fret and > measure from string to top of first fret. Then knowing that the saddle height > is exactly where I want it, I move to the saddle. Since my saddle measurements > are not independent of the the nut, it seems to make sense to do the nut first. > If for example on a new guitar, the nut were 1mm to high, my measurements for > action at the 12th fret necessary for setting the correct height for the saddle > would be off. But again, you probably have your own way around this problem > that will come out in subsequent parts of your discussion of set-ups. That seems to make sense to me. I'll try that, thanks. > One of his claims, and one that I have heard from other luthiers, is that > relief is needed only up to the 12th fret. Thus the action test is first to > fret the string at 1 and 18 or so, and check that the neck is straight at > the 14th fret. Second hold at 14 and 1 and look for relief around 6 or so. Is > there any merit to this approach, or should relief be greatest at the 12th? > Interesting. When writing the article, I originally had written the 14th fret. After some thought, I decided to change the 14th to 18th, even though in practice I usually use the 14th or 15th fret. I'll rethink the whole thing. I practice, relief probably isn't necessary beyond the 12th or 14th fret. Why? The neck is cantilevered from the body. The truss rod moves, essentially, the head end of the neck up and down, causing the bow in the neck (or vice versa, actually). Thus, practically, you can't effect a bow in the neck above the body joint using the truss rod adjustment anyway. In theory, if you fret a note a the 12th fret, the maximum amplitude of the string will be half way between the saddle and the 12th fret. This maximum is probably about over the sound hole, and off of the fingerboard. Fretting higher notes shifts the maximum amplitude even closer to the bridge and away from the fingerboard. It also seems to me, now that I think about it, that the magnitude of the amplitude (say that 10 times quickly) is smaller for notes fretted higher up on the fingerboard. This is an empirical observation. I'll have to think about why this is true. Anyway, if this is indeed true, the net result would be that less and less neck relief is required the higher up the neck you fret. In that case, the maximum amplitude would then be for the open string, occuring at the 12th fret. Hmm. > : I dress fingerboards using a carpenter's level to which I've glued > : a strip of 80 grit sandpaper. The natural action of the motion, and > : the tendancy of the sandpaper to dull faster towards the center, causes > : a slight backbow to be sanded into the fingerboard. This just happens > : to the the exact curve necessary to provide the desired relief when > : the instrument is fretted and strung. Go figure. Damn, and now I've > : given away my secret. I could have baffled you with "high-level" math > : and said that's how I do it, coupled with a complex jig. I bet I could even > : have sold a couple of jigs! Maybe next time. > > This appears to suggest that you anticipate the neck moving significantly due > to the string tension, moving from "back-bow" to correct relief. In my > experience the amount of movement here cannot really be estimated in the > abstract, some move lots others little. Or are you referring to your own > instruments here. The amount of back-bow is small - perhaps a mm or two. I find this works for my own instruments as well as the repairs that I do - Martins Gibsons, Taylors, Larrivees, etc. It doesn't work on Rammirez guitars or those with similarly non-true fingerboards; these have to be treated specially. > > Again, thanks for the initial post, its great to have some comprehensive post > to respond to rather than a series of smaller questions that happen over months > or years. > > Cheers, > > -- > Don Hubert, Social and Political Sciences > University of Cambridge, CB2 3RQ, UK > dah22@cus.cam.ac.uk Your welcome. Thanks for your comments. Charles ===================================== From: dah22@cus.cam.ac.uk (D.A. Hubert) Newsgroups: rec.music.makers.guitar.acoustic Subject: Re: BASIC GUITAR SET-UP 101 PART 1 Date: 31 Jan 1996 09:37:13 GMT charles tauber (tauberc@candu.aecl.ca) wrote: : Interesting. When writing the article, I originally had written the : 14th fret. After some thought, I decided to change the 14th to 18th, : even though in practice I usually use the 14th or 15th fret. I'll : rethink the whole thing. I practice, relief probably isn't necessary : beyond the 12th or 14th fret. Why? The neck is cantilevered from the : body. The truss rod moves, essentially, the head end of the neck up and : down, causing the bow in the neck (or vice versa, actually). Thus, practically, : you can't effect a bow in the neck above the body joint using the truss : rod adjustment anyway. In theory, if you fret a note a the 12th fret, : the maximum amplitude of the string will be half way between the saddle : and the 12th fret. This maximum is probably about over the sound hole, : and off of the fingerboard. Fretting higher notes shifts the maximum : amplitude even closer to the bridge and away from the fingerboard. : It also seems to me, now that I think about it, that the magnitude of : the amplitude (say that 10 times quickly) is smaller for notes fretted : higher up on the fingerboard. This is an empirical observation. I'll : have to think about why this is true. Anyway, if this is indeed true, : the net result would be that less and less neck relief is required the : higher up the neck you fret. In that case, the maximum amplitude would : then be for the open string, occuring at the 12th fret. Hmm. Having considered your comments and reread the original post, I wonder about the logic of maximum relief at the 12th fret. This would be true if the suspension points at both ends (nut and saddle) were the same height above the top of the frets, they obviously are not. More importantly, relief must allow for fretting at various points along the fingerboard as well. In these cases the difference between the height of the suspension points is at its greatest. This would suggest to me that one of the most crucial areas of relief will be in the first few frets byond the one that is fretted. This seems to be born out in experience, low action guitars tend to buzz on the next fret or two, rarely in the middle (assuming that the frets are at least straight. The higher up the fret board you go the greater the angle from the low suspension point to the high, and the less relief needed both at the mid- point and in the first couple of frets. The conclusion, relief must offer sufficient clearance for the shape of the string when vibrating (your starting point) given the difference in height of suspension point (and therefore the angle of the string) there is no need for compensation beyond 12 as the angle of the string away from the fretboard more than compensates for the necessary relief this need for relief is greatest when fretted at the first fret (the nut creating a slightly higer suspension point relative to the first fret than the first fret to the second) and the further up the finger board the more the angle of the string accomidates the necessary relief relief must take into account the frets immediately following the fretted one as well as the mid-point relief therefore must increase steadily from the first fret. With each fret the amount of increase will lessen as the change in string angle offers more clearance both for the next fret and for the mid-point, until relief is no longer required, (in fact, following this logic better playing action on the upper frets could be had by a slight upward bow) It will take me more thought than I can spend at the momement to work this out further. If this logic is correct, at what stage in practice does the angle overcome the need for relief. cheers, -- Don Hubert, Social and Political Sciences University of Cambridge, CB2 3RQ, UK dah22@cus.cam.ac.uk ========================================= From: David Means Subject: Re: BASIC GUITAR SET-UP Part 4-I Date: 26 Jul 1996 08:24:46 -0700 In article <31F79C0A.756E@candu.aecl.ca>, Charles says... > >While I'm not entirely happy with Part 4, it is the best so far after >several attempts over as many months. It is not as simple a presentation >as I would have liked, and is a bit too textbook-like. In my attempts to >make the information accessible, I have purposely ommited things and >glossed over others. However, I hope that it is reasonably easy to read >and presented in such a way as to be understandable. > Charles, No need to apologize in any way for this document. It is far and away the most concise and accurate treatment of the arcane topic of temperament and its practical application to guitar intonation that I've ever seen. Good job! Yours is the first treatise I've encountered that has logically connected the theory to the practical application. Speaking of practical matters, one technique that I have been using for some time when making a compensated saddle is to make sure that I use the smallest diameter string segment "V" that will provide a clean break for the string where it crosses the unbeveled saddle blank, and then to mark the saddle on the side of the "V" toward the nut with a very sharp pencil. I think the very slight offset thus created in the marked position compensates for the fact that the "experimental" string height at the saddle is slightly greater than the final string height. Sort of compensating the compensation. Maybe it's overkill, but it's always seemed to work for me. Thanks for all the effort and the great results! Dave Means dmeans@fcc.gov =======================================