[ Guitars by Cumpiano ] [ Classes by Cumpiano ] [ Newsletter archive ] [ Articles by Cumpiano ] [ Book by Cumpiano ]

William Cumpiano's String Newsletter #1

Go to newsletter number:

Saddle Adjustments and Verifying the Scale

1998 William R. Cumpiano All Rights Reserved
William Cumpiano's String Instrument Newsletter #2

In #1 we began a rather leisurely set of discussions surrounding the topics of when (and whether) to reset a guitar's neck. The benefits are dramatic and obvious: if the procedure is performed properly after the need is diagnosed accurately, the results are a usually a wonderfully responsive, quick and accurate action, and even, in many cases, improved pitch accuracy and even a bolder, more powerful sound. It seems that much of that which we cherish on a good guitar can be ours with the correct neck angle--and a careful set up. This is why, however modest the conception of any instrument may be, if the action is low and it plays in tune, you will find someone who will love it!

The discussions also revolved around the premise that because a neck reset is fairly advanced instrument repair technique, requiring stresses to be placed on the instrument to which it is not ordinarily subject -- procedures which always have the potential of putting the instrument at risk -- it is always prudent to scrutinize closely how appropriate the diagnosis of "neck reset" on an instrument is, always with an eye to see whether a simpler adjustment can suffice to restore the instrument to satisfactory function.

So whether the reader is a brave and handy person who wants to try to do their first neck reset (don't start on a 1925 0-45, please!), or simply a guitar owner who suspects that one is needed, but wants to be sure, this series will be of interest. The series focuses on steel-string acoustic guitars for now, but will deal in time with neck problems on classical/nylon string instruments.

In this issue we will cover:



What if, after you've verified that the rod adjustment and nut clearances are optimum, you find your string action is still unacceptably high? Your last remaining simple remedy is to cut down the saddle. This has a down side: as you lower the saddle you decrease the mechanical advantage of the string tension as it levers the soundboard. It's like shortening the arm on a crowbar. In effect, it's like turning the volume knob down on the guitar a little bit.

By this time, you will be familiar with how much saddle shows over the top of the guitar's bridge. If the there is ample saddle height, you're in good shape. If instead, only a thin sliver of saddle protrudes above the surface of the bridge, then (given the high action) you should have already been suspicious of a neck-angle problem on the guitar. Perhaps, though, the previous truss-rod and nut adjustments will have improved the action enough to delay the inevitable neck reset.

If you're fortunate enough to have a substantial amount of saddle showing above the bridge, you can proceed to lower the saddle by the following "scientifical" method.

Accurate action height adjustments are next to impossible without a good measuring tool. For the best results, get a high-quality (I recommend the Starett brand) machinist's 6" rule, with a non-glare "sintered" finish, and with accurate, legible 1/64" subdivisions. The lines should be engraved into the surface, not screen-painted on.

Stand the narrow edge of the ruler on top of the 11th and 12th frets, just beside the string, and read (on the 1/64th scale) the amount of air space between the crown of the 12th fret and the bottom of the string. Jot down the number of 64ths, noting measurements between the lines as the closest line plus .5 (for example if it reads a little over 4/64ths, jot down 4.5 . In this way document all the string heights in turn, and jotting them down in a sequence. For example the reading from bass to treble might be:

9.5 8 8 8 9 9

Immediately below these "real" measurements jot down your "ideal" measurements. Let's mark what I've found to be a reasonably comfortable "low acoustic" action:

9.5 8 8 8 9 9
7 6.5 6 6 5.5 5

Subtracting the two we get a list that represent the amount each string must drop, in 64ths, to reach the ideal.

9.5 8 8 8 9 9
7 6.5 6 6 5.5 5
2.5 1.5 2 2 3.5 4

Since the 12th fret is halfway between the nut and saddle (well, virtually) it is axiomatic that the amount of drop at the 12th fret can be achieved by twice that drop at the saddle. So, this last list of numbers (the differences in 64ths) when simply converted to 32nds, become the list of height drops for each corresponding string at the saddle.

Luckily, the saddle will invariably be marked by a tiny dimple where each string rested on it. Trace a pencil-line straight down from the crown at each dimple. Using your ruler, indicate with a pencil tick on the line each of the desired 32nd-drops for each string, corresponding to the list.

The sequence of ticks, when connected by a new line that is more or less continuous, denotes the new crown-line for the saddle--the one which is necessary to achieve the "ideal" action. Note that because we've averaged out all the "over 64th" readings to .5, and then doubled the error, the tick marks won't sit on a perfectly smooth line on the saddle. Now YOUR job, should you choose to accept it, is that when you trim the saddle down to the line you must create a smooth, continuous contour which averages out the slight inconsistencies in the pencil-tick marks. With the saddle held in your fingers or in the jaws of a little pliers, Trim the saddle top to the new line with a file or a small belt sander (of course, move the saddle over the sander, not the sander over the saddle!) Now, shape a new, longitudinal crown on the saddle edge. But wait! Here's you chance to tweak the instrument's intonation, maybe improving it with a little effort.

You have to have a machinists' yardstick to do this. If you are going to do any kind of bridge work or intonation diagnosis, you just have to have one. If you don't and if you don't want to spend the $50 - $75 it'll cost for a good one, skip this section. It is a metal ruler, 30 - 36 inches long, finely engraved with markings in English Standard and something called "Ford" measurements (actually very few people know that it's called that). Ford rule measurements (yes, first conceived by ol' Henry Ford) are inches divided into tenths. And each tenth is divided up in ten divisions, yielding 100ths (and a headache trying to read them), or (easier for my fifty-year-old eyes) in five divisions, yielding or 50ths (that is, with four engraved ticks that are read .02", .04", .06" and .08" . Of course the measurements in between the ticks are .01", .03", .05" and .09"). The advantage of the Ford rule is that you can measure off, to .01" accuracy, the results of a mathematical division involving inches. It like the next best thing to metric. I know, I know, we should all be using metric, but hey! Give me a break: life is changing too fast as it is....

The advantage of this decimals-of-an-inch ruler is that we can create our own list of fret intervals--and actually use them. And check to see if bridges are in the right place. For example, take your guitar, measure from the front face of the nut to the center of the crown of the twelfth fret with the yardstick. Say it reads 12.75." (we know by rote that it must be a 25.5 scale, i.e., twice 12.75). Now look down the centerline to where the saddle slot is on the bridge. The center point of the saddle should be slightly OVER twice the nut-to-12th fret measurement. Often it isn't. That's bad.

Sometimes it is. Is that good? Well if it is a tenth of an inch, .01" (that is, the total nut-to-saddle ruler measurement is 25.6 ), that's good - on both nylon and steel string guitars by the way. But wait, (ON STEEL STRINGS ONLY) if its .15 over the scale length -- that's even better! Yup, trust me: this is 27 years of obsession talking here. I may be crazy, but I'm not stupid! The extra .5 actually makes everything tune up a little better than not having it. It also insures that when you tune your low E string down to D, for "dropped D tuning" the D string will intonate nicely, not go haywire, as it does on most .10" center-compensated scales.

Does this "extra" compensation always work? Well, not always. On all modern full sized guitars it does, and on most smaller guitars too. The .15" compensation seems to work just dandy on scales from 20" to 25.7" in length. For nylon strings, .10" appears to work for all quinto, requinto and full-scale nylon-string scales. When DON'T these centerline compensation figures work? They work only as long as the straight front edge-line of the bridge is indeed parallel to the frets. And we find that more reliably on the most carefully made guitars.

Check it this way: Choose a fret - - any fret, but one preferably closer towards the end of the fretboard. Say the 14th fret. Measure from one end of the fret to the forward-most corner of the corresponding end of the bridge. Say its 11.2". Now measure from the OTHER end of that same fret to the forward-most corner of the OTHER end of the bridge. Is it 11.2" ? Yes. The maker had his/her act together. Is it 11.3"? The maker had a bad day. The bridge is crooked (even though the centerline compensation may be spot-on) and the intonation is a goner. If all the above measurements are off the chances are you've been struggling with tuning the guitar for years, blaming yourself, blaming the weather, blaming the strings, and so on. You've probably had the poor repairman tearing his hair out when he checked the centerline intonation to be fine -- but didn't check the bridge's parallelism (parallelity?)--and you kept coming back again and again with your sick puppy that the repair man wished would just die.

If the bridge is only slightly off, you may be able to crown the saddle selectively to rotate the actual bearing line of the strings to approach the above ideal more closely. The saddle's thickness, .09" gives you a surprising effective margin to work with, either by moving the crown fully forward to the front edge, the back edge or leaving it in the middle--or having it travel diagonally across the length of the saddle.

Well, guys, I've gotten off the planned track. The best laid plans of mice and men often come to naught. This is close to three pages and I promised myself that I would keep these short and often. I don't know if you're like me, but really, it's not my favorite activity to read a lot of text off a computer screen. I'd rather be lounging on a hammock under a palm tree with a lemony-sweet tequila drink in one hand and a lurid paperback in the other. That's fun reading. I'll catch up to the promised list of topics next week.

Ciao. Oh...by the way, I will close with Epi Stathopolous' Warranty Card as I promised above. Epi Stathopolous was a Greek (...no!) guitarmaker who made some of the greatest American arch-top guitars in history. His label was Epiphone. After he died, someone kept selling that name over and over and over, squeezing every last bit of respectability left out of it, until all it is now is a cheap illusion. BUT there was a time when there was a man behind the name. And what a man he was. His whole personality could be summed up in the following text found on a little card that was included with all his guitars. <>

"Now that your are the proud possessor of this instrument, learn to love it. Remember it is one of your most intimate possessions--closer to you perhaps than anything else you may own. For it is the voice of the music within you, singing or sighing with your mood, and forever faithful to your innermost whims. So protect it from careless hands. Keep it clean and dry. It is well made and rightly designed to stand constant use. Have faith in it to respond to your skill and magnify your best efforts. Give to it the best you have and the best will come back to you."

Epi A. Stathopoulo

Text of a warranty card included with a 1941 Ephiphone Broadway
(courtesy Dr. Peter Weitzman)