Can it be done? I've tried but it's been hard to be precise and this step has daunted me because it needs to be precise.
I borrowed my friend's miter box but it's not the latest model so not only do I lack a template for the 25.4" scale but it looks like there's no way to lock it in anyway. I'm going by the figures in my plans for the OLF medium jumbo as well as those from the Kinkead and Natelson/Cumpiano texts. Just to be sure, i also have a couple of guitars on hand to directly compare to.
Here's where I'm hitting a snag: in order to get a precise measurement I'm using a 6" digital caliper to measure the distance between the frets but I think my math is off. My resources all list the distances in relation to the nut but obviously my digital caliper won't reach that far so I've had to break out the calculator and crunch the numbers in order to get a fret to fret distance. The first few frets went well but as I went higher up the fingerboard I've been coming up short.
What am I doing wrong? Do I need to account for the width of the slot (which is the width of the blade of the Japanese pull saw) too?
BTW, the width of the nut slot also seems to be ambiguous. Am I correct in assuming that it really doesn't matter if I go with a blank nut?
I do this quite often. The nut slot width has nothing to do with it as all measurements are taken from the face of the nut. It is best to adjust your measurements so you are measuring to the edge of the slots and not the centers. So if your slots are .023" wide either add or subtract.012" and you are now marking the edge of the cut. I make those marks with an exacto knife. Just line up your saw toi the appropriate side of the mark and make your slot. I think your problems are coming from tolerance stacking in your layout. the first 4 frets seem to be well dimensioned compared to the other FB, which would be the first 6" or so, the problems start after that due to the fact you are exceeding the length of your calipers. So at that point if you use your calipers in conjunction with a 6" machinist's rule you can lay out another 6" of frets. After that a 12" rule should get you to the end of about any FB. I lay out by the overall dimension from the nut and use the fret to fret dimension as a check. I check the entire layout at least twice both ways before cutting any slots. I want to be under a tolerance of .005" with my layout. If you do not have a nice set of machinist's rules, you can make yourself some spacer bars using your caliper to measure them.
You never know what you are capable of until you actually try....
Brain is dead on. You have to take base measurements from a given point. If you measure off each fret as you go , you magnify error so by the time you get to the lower frets , they may be off in one direction or another. Remember than getting a fretted instrument to intonate on all frets with all strings is a mathematical improbability.
At one point you will have to make a datum point from the nut ( also known as zero fret position ) unless you have a long enough caliper. Be as accurate as you can. I think if you are off .010 you may be ok . The best way is to make a block that you can measure as close as possible. May be 2 of them. If you can make them 6 inch on the dot , you can stack them as you go up the neck and out of range from the 0 position. You can also make a stack template, A series of blocks that are the size of the spacing. Remove a block slide the board to the next one and cut.
Keep the pictures coming, we like to see other people work and designs .
John Hall
Blues Creek Guitars Inc
Authorized CF Martin Repair Center
president of Association of Stringed Instrument Artisans http://www.bluescreekguitars.com
For any one else using this fret box our templates will fit.
We make our templates from florescent clear green 3/16" acrylic with a slight round notch. This notch fits the SM box or it will also fit a number 4 finish nail. We put a center line on the board so your fret board doesn't have to be un-cut to begin. Simply use double sided tape to hold in place. Here is our instruction on using the fret scale.
How to use fret board construction aid:
You begin by finding fret board material of appropriate rough size that is fairly well dressed to finished dimensions.......... draw a center line down the material or at least to either end. Now attach the fret board to the appropriate scale using double sided tape or small amounts of hot glue using the center line on the template aid ....align with the center lines drawn on the fret board ( it will make a difference as to the saw you are using....... as to where the fret board will be cutting either on the bottom side of the template or the top). This aid can be used with hand miter, table saw, or radial arm saw provided the area with which the aid will be used has a minimum of 3" plus room to clear the notch that is approximately 1/16". If you are using a hand miter you would typically begin at the nut notch ( which will be cut off flush eventually ) a notch guide should be attahed to the floor of the miter if hand miter or on the sled if table saw or on the table if radial arm saw at an appropriate and safe distance from the saw blade....... then simply begin....... cut first slot/fret then move aid and fret board to next notch and proceed to finish..... in the case of say a guitar 19 to 22 frets or this many times plus the nut would be the required movement over the blade or the blade moved through the material. If you are using a table saw a table saw sled is recommended same procedure but this time the fret board should be on the bottom of the acrylic template and moved back and fourth over the blade with the aid of the sled and the notch controlls the spacing and the slot cut. Typical saw blade width will be approximately .022" If using a radial arm saw or a sliding miter saw then the fret board will be on the top side of the template and same style procedure will be used........ Keep in mind that all tools can be dangerous if used with-out proper safety apparel, proper knowledge, and skills....... use at your own risk. But enjoy our craft.
We have a large number of scales....... Let us know if we can help.
When one measures each distance from the last fret slot rather then from a common data point(nut location)one ends up with what is called cumulative error. The further you go the further you will be off the mark.
Tom
" A person who has never made a mistake has never made anything "
nkwak wrote:Can it be done? I've tried but it's been hard to be precise and this step has daunted me because it needs to be precise.
I borrowed my friend's miter box but it's not the latest model so not only do I lack a template for the 25.4" scale but it looks like there's no way to lock it in anyway. I'm going by the figures in my plans for the OLF medium jumbo as well as those from the Kinkead and Natelson/Cumpiano texts. Just to be sure, i also have a couple of guitars on hand to directly compare to.
Here's where I'm hitting a snag: in order to get a precise measurement I'm using a 6" digital caliper to measure the distance between the frets but I think my math is off. My resources all list the distances in relation to the nut but obviously my digital caliper won't reach that far so I've had to break out the calculator and crunch the numbers in order to get a fret to fret distance. The first few frets went well but as I went higher up the fingerboard I've been coming up short.
What am I doing wrong? Do I need to account for the width of the slot (which is the width of the blade of the Japanese pull saw) too?
BTW, the width of the nut slot also seems to be ambiguous. Am I correct in assuming that it really doesn't matter if I go with a blank nut?
OK, I admit it: I can be stupid at times. I blame my lineage, which is reputed to be full of people with hunched shoulders and flat foreheads. In the last picture I compared my work against a finished guitar but did not realize that my Larrivee has a 25.5" scale. I'm working off plans for a guitar to have a 25.4" scale length. So I pulled out my Martin D-16GT which has a 25.4" scale length and things are pretty good up until the 12th fret.
So things aren't as bad as I thought, but they're still not quite there either. I can work with that.
Can it be done? Well, I did it once, with satisfactory results, and with tools more crude than
the ones I see in these photos.
The reason I was so bold as to deface a nearly perfect piece of ebony this way (back in the Fall
of 1973,) was that I had a guitar that needed a fingerboard, an 18-inch stainless-steel
straightedge that had its first six inches ruled in 64ths of an inch, a small but pretty good rock
maple miter box, a hardware store dovetail saw that cut approximately the right kerf, pencils,
and a Casio calculator that was accurate to fourteen decimal places. These were all the tools I
could bring to the job; I was dirt poor at the time; I had bought a very good top, a billet of
brace wood, and this fingerboard from Michael Lewis, up in Vancouver, B.C. (I say "up"
because I live in the State of Washington.
There are two reasons I'm telling you all this: first, to assure you that it can be done, no matter
what length of scale you wish to use for your guitar, and second, to inform you of a particular
pitfall to which one can succumb if he isn't informed of it by more experienced predecessors.
So I would first embolden you with the same line of reasoning I used on myself in order to get
the courage to attack the ebony. (Attack the enemy?)
It goes this way: LOTS of guitars have been built with tools that were no better than these,
and with the wrong value for the divisor you actually need (twelfth root of two.) In much the
same way that the math constant "pi" is approximated for quick estimation by the improper
common fraction 22/7, the twelfth root of two is approximated by the improper common
fraction 18/17. The error is about 0.0039, but this is a divisor, so it gets compounded, just like
interest on a loan, over again every time you place another fret.
Back in the days of Torres, 18/17 was as close as they could get to the true value of the twelfth
root of two, which is 1.0594630943592952645618252949463, correct to 30 decimal places.
Thirty-one are shown, but the last one specifies an approximate value of the error in the
thirtieth one. For comparison, 18/17, in decimal terminology, is 1.0588235. . . .it gets pretty
meaningless to talk of hundred-thousandths of an inch when we just can't eyeball that
measurement, but it is NOT meaningless to leave all those decimal places trailing in the
calculator and round each consecutive value to the nearest ten-thousandth when we're
constructing our table of fret locations.
And fortunately, the error in "consecutive eighteenths" was in the direction required for
saddle compensation, and close to the right amount, so just setting the saddle at the
"theoretical" location and not lengthening the scale a bit for tension compensation worked
pretty well, in addition to which it was possible to locate the 5th, 7th, 12th, 17th and 19th frets
in accordance with halves, thirds, and quarters of the scale length and work up by consecutive
eighteenths a maximum of four frets from each of those places, then use a trivial saddle
compensation derived by experimentation. Then, too, these legendary luthiers were building
for gut strings, not steel.
That 31-place value for the twelfth root of two, incidentally, is from Windows calculator. You
can take the common Martin long scale length, 25.4, and divide it by that monster, divide the
result by that monster, divide the next result by that monster , , , out for twenty-four
consecutive frets, and at that point, Calculator says "6.35," (no trailing decimals,) which is
EXACTLY 1/4 of 25.4. So it's a VERY accurate value, and serial error is not a problem.
Windows calculator does the job. But then, so did my old 14-place Casio. The series of
divisions yields the remaining string length, which you then have to subtract from your scale
length to get the "nut-to fret distance," about which I will have more to say in just a bit.
I absolutely agonized, for several days, about my inability to measure lengths in excess of six
inches to the nearest thousandth, before it dawned on me that measuring to the nearest
hundredth was close enough. What I'm driving at is that our eyes, fingers, and the width of the
finest lines we can draw are such that as a practical matter, no matter how careful we are, we
won't get closer than a hundredth, and it doesn't matter, when we consider the probable
accuracy of a symphony violinist's finger placements. I've known dobro players who couldn't
get to within a quarter-inch of a C#, but that's another matter. I've also known many dobro
players who NEVER sound out of tune.
So I only needed to get to the nearest hundredth, and I had a measuring tool that could get me
to the nearest 1/128th. The 1/64th marks and the spaces between them were nearly of equal
width. I then converted all my decimal distances to 128ths of an inch, rounded each fret
placement to the nearest 128th, and laid them all out on a fine pencil line I had drawn down the
center of the board, and then used a square to draw lines (again, fine pencil lines,) across the
board from side to side at each fret location, all this was VERY painstaking work. I
acknowledge a debt to Irving Sloan's "Classic Guitar Construction" for much of the
knowledge which enabled me to re-top this guitar and put a new fingerboard on it.
I put the board in the miter box, and went to sawing. Depth stops were added to the miter box
for the saw's back to engage, and I eyeballed the saw's placement on the lines so that the edge
of the kerf on the side toward the nut just sat on the line. I realized that it was important to do
them all the same way. What I did NOT realize at the time resulted in an error, which brings
me to the "nut to fret distance" PITFALL that I mentioned earlier.
So here we have it, the error: all my measurements were made from the precise nut end of the
fingerboard to a pencil line which in fact represented the MIDDLE of the fret, as closely as I
could mark it. I then proceeded to saw a fret slot about .023 to .025 wide on the side toward
the saddle of each of those marks. I then trimmed the fingerboard to its taper plus a sanding
allowance, glued it to the guitar, smoothed its edges, put finish on the sides, pounded in and
dressed the frets, finished the top, put on a nut, strung her up, and what did I have?
I had a guitar I could not tune. In particular, I could not get the open 6th string E and the 3rd
fret G on the 6th string to BOTH be in tune. Other than that, I didn't notice much in the way
of intonation problems, and I have a fine ear for pitch. I didn't notice the rest because this
worst problem was so bad. It took me a LONG time to figure out what was wrong, and as I had
a Gibson L7 to play, it didn't seem urgent.
Two years later, I ran into a discussion of exactly the problem I noticed, and that discussion mentioned that some old Martin guitars also exhibited the "can't have both" problem between the open 6th string, and the 3rd fret note on the same string, the reason being that the fingerboard was laid out the way I had done it.So the solution was to cut half a kerf off the end of the fingerboard. What I did was take off the nut, construct a maple dummy the same height as the fingerboard, glue it in place against the nut end of the fingerboard, sand it flush with the fingerboard, cut a new fret slot that was 50% ebony and 50% maple, remove my dummy, and put the nut back on, half a kerf closer to the saddle, and voila! a guitar I could tune! So let that be a lesson to you.
What you really want to do when you're measuring for your fret placement lines is to add half a
kerf to the end of your fingerboard and measure from there. Get a piece of veneer that's about
.012 and temporarily glue it to the nut end of the fringerboard before you do your marking.
Something like that. Doing a zero fret slot will also work, but you don't want to saw the board
off at that slot, you'd want to actually use a zero fret. The interface between the end of the
board and the edge of the nut is, theoretically, your computed distance from the utmost crown
of the frets, but you can't center a saw on the finest scribed line, you're sawing to one side of
the mark, so be warned.
That I had this problem also illustrates that you DO need to be to the nearest hundredth, so
the best you can do with tools like these is just barely good enough.
