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

William Cumpiano's String Newsletter #3

Go to newsletter number:

Importance of the string's "back angle"

1998 William R. Cumpiano All Rights Reserved

In this issue:

Hello! again, to all you dedicated guitar lovers who want to know how the guitar does what it does and how it can be made to do it better. In this issue, you will find a continuation of my greater plan to address the whys and hows of neck resets.

If you haven't already noticed, you will at the end of this issue have discovered my devious master plan: by the time we finally get down to whether and how a neck reset should be done, you will all have been carried (without hardly noticing it) across the entire panorama of ideal vs. real-world guitar design, setup and construction.

In Newsletter #2 we had learned how to lower the saddle precisely, resulting in a desired action height. But we were dealing then in an ideal world of guitars with good neck angles. Alas, the real world rushes in: your Taylor, Guild, Martin--or Gurian-- may have little or no saddle left to cut down. And your action is still high and uncomfortable. Ah, tension, like rust, never sleeps.

What are, however, the consequences of lowering the saddle? Just that the action drops? No, like so many other things in life, there's a price to pay: you in fact reduce the instrument's power output and affect it's "feel" by a certain amount. If you have good auditory memory, by a noticeable amount. But why, you may ask?


When we raise or lower the saddle, we steepen or flatten the "back angle."-- the angle the strings drop from the crown of the saddle and into the string holes. We luthiers worry about those few degrees of back angle. Because when we change the saddle's height we consequentially affect the way the strings dump their tension onto the guitar.

We could say that the strings, when vibrating, are tapping a kind of sonic Morse Code on the saddle. If we did, we would be partially right: yes, it is indeed sublimely ordered information which emanates from the oscillating string. But the information is far more concentrated than dit-dit and dah. True, the string's movements result in a sequence of tiny tugging and releasing motions, what results as the string travels to the limit of its envelope (pulling oh-so-minutely on the saddle), and back to its at-rest location (lessening for an instant its pull on the saddle). Then, its own momentum carries it beyond that, to the opposite extreme of its envelope (pulling the saddle again). As it strains against the string's own elasticity, the string whips back in the opposite direction again, and so on. The saddle is receiving all this "code" and levers it onto the top of the guitar.

But all this is not just happening in just in an up-and-down or side-to-side plane: it is happening in all directions; towards the nut, away from it, up, down, sideways, and in a virtually infinite number of discrete rotations. This is how the string couples the sum of all its myriad harmonic motions to the guitar.

The sum total of the string's effect on the saddle, and because it is immobiled into the bridge, on the soundboard, is to induce it into rocking, swaying, jumping, tipping, heaving, rotating--like an airplane in a storm---restrained only by the soundboard's limiting resilience.

I like to picture in my mind what the surface of the soundboard must look like while you play the guitar, as if you were as big as a microbe. Undoubtably it is physically deflected in all planes, and it is further deflected due to motions and forces "feeding back" onto it -- energy being returned BACK to the bridge and then to the strings themselves. The soundboard feeds back the ordered kinetic energy imparted to it by the strings, returning it from the driven sides, back and the air mass partially trapped in the soundbox, all resonating back in turn. Add to the mix the myriad of motions wich the opposite end of the strings are feeding into the neck. Again, in turn, that energy eventually is returned to the box, but altered in its own peculiar way.

What we are dealing with is what the acousticians call a set of "linked oscillators:" springs attached to springs attached to springs attached to springs. And then the spring ends are attached to each other. The string is a spring, the thin soundboard--limited by its resilient braces, is a spring restrained by springs; the air partially trapped inside the soundbox is another spring. The neck shaft is whipping around like crazy, itself a big massive spring.

The largest, simplest motions, occurring at lower frequencies are the easiest to discern and to understand. Acousticians have identified and can even visually demonstrate their movements. But the rest--there are simply so many and they overlap so closely--are only partially fathomable by the application of statistical analysis.

My mental model of the guitar's soundboard while one is playing, is a flexible membranic surface which appears like the surface of the water in a pan heated to a rolling boil.

But I digress...

The amount and quality of the string's influence over the saddle and bridge is determined directly by the string's back angle. That is, again, the angle at which the string drops BEHIND the saddle and into the bridge. The back angle is changed by two things: the height of the saddle above the bridge, and the distance of the saddle to the string holes.

So yes, we reduce the back angle when we cut down the saddle and, correspondingly, the string's effect on the soundboard. We could analogize the system to somebody with a crowbar. The string's motive force is the guy yanking on the crowbar. The saddle/bridge height over the soundboard is the length of the long arm of the crowbar. As we drop the saddle, we shorten the arm of the long arm of the crowbar, reducing its mechanical advantage over the soundboard.

And the back angle? I can wring the last bit of juice from this crowbar analogy by likening the back angle to two factors on the crowbar: as we lower the saddle, we loosen the guy's grip on the crowbar--from a two-fisted grip when the saddle is at full height over the bridge, to a simple thumb-and-forefinger grip when the saddle is a merely a protruding sliver. As the entry-holes in the bridge get closer to the saddle, we shorten the SHORT arm of the crowbar, similarly increasing its power, just like lengthening the long arm did.

I promise, this is the last stretch of this tired analogy. What happens if the saddle is very, very tall, and the strings holes are socked up real close to the towering saddle? Well, in this extreme case (and believe me, I've seen more than a few student guitars like this) the exposed part of the saddle starts to bend. The front lip of the bridge threatens to tear off and, as well, the bottom belly of the bridge wants to start peeling off the soundboard (but if it's stuck on there real good, the soundboard and all the braces will want to start bending or peeling off, instead). It's not pretty. But boy, it's going to be a LOUD, raucous-sounding guitar while it holds together!

If you've had to drop the saddle radically when setting the action to a more comfortable level, chances are you will have reduced the string's back angle so much that there now will be insufficient pressure on the saddle. The result is often an annoying whine when playing all the notes on that string, or at least a distinctive wimpy, wowing sound.


One remedy to this problem that occurs when you cut down the saddle excessively, (recommendable only if you're cheap, or the guitar is cheap, or if the guitar is badly beat up anyway, or if you're not fussy) is to "hot rod" the bridge by "ramping," or extending the little string-notches in the string holes, so that the string now enters the bridge top sooner. This can restore some of the "lost" back angle, increase the string's down-pressure on the saddle and eliminate the wimpy whine.

Alternatively, judicious shaving of material from the bridge top also can restore the back-angle. Indeed, you may actually improve the function of the bridge by contouring its surface to better match the contour of the fretboard. The payoff is that more of the saddle is exposed, and the back angle can thus be restored. That's what the builder should have done in the first place. However, shaving the bridge on an expensive, name guitar is usually bad practice -- unless the bridge is too thick to begin with. They do come through that way sometimes, you know. I know Martin use to make bridges available of varying thickness for the bridger to chose from, to best suit the neck angle of the guitar at hand. So you can expect to see original bridges on Martin guitars of the same model and vintage varying from about -inch to a full 3/8-inch. Any bridge over 5/16-inch in thickness is a potential candidate for re-countouring.

One thing: have you noticed that even though the steel-string saddle is angled, the common line through the centers of the bridge holes isn't? The result (which I believe is inadvertent and unplanned) is that the treble strings, whose bridge-holes are further from the saddle than with the bass strings, will have a flatter back angle and the bass strings will have a steeper angle. This can't be a good thing. It can only exacerbate the weakening of the treble response when the saddle must (invariably) be cut down later. For another thing it can only hobble the instrument's treble-string response under all circumstances.

I, for one, now determine the bridge-hole centerline to be strictly parallel to the slanted saddle. Visually, nobody has even noticed it or even commented. But it can't help but give the treble response more snap than on the traditional design. And, as the curmudgeonly builder and textbook author Arthur Overholtzer once said, "I don't mind tradition, as long as it doesn't interfere with some of my ideas."


If you haven't guessed by now, I'm subtly telling you all the ways to avoid or put off the inevitable neck reset. But is there some kind of measurement that can tell you, at a glance, just how "off" the neck is? Well, certainly, if your action is still high after you've done all the above, it will only come down with a neck reset. But here is another test:


To check whether a guitar's fretboard has correct inclination (and by extension, the neck angle is correct) first remove the instrument's strings. Adjust the truss rod to allow the fretboard to read the neck as straight as possible against a straightedge place along the centerline of the neck (keep it well aligned, now. If it crosses the centerline at a slight slant, it will inevitably read a hump in the middle). The straightedge may rock over a single high or loose fret. (Tap or file down the offending fret, or repair it with a dot of cyanoacrylate).

Notice if the straightedge is lifted off the frets by the fingerboard-end rising slightly. Often it will. In mild cases, tightening the truss rod nut will average out the problem. However, if the end of the fingerboard rises in the extreme, you might as well stop the test right now. This is usually a tell-tale sign of a bad neck angle. But this is the extreme case. We're talking about mild, borderline cases: the kind that require fine judgement calls.

With the straightedge thus placed on the frets, its far end should just "tickle" the top of the bridge on a well settled-in, well-used guitar. On a brand-new guitar that has yet to settle in (or on a 12-string guitar, which is expected to always settle considerably under tension) the straightedge should clear the bridge by not over 1/16-inch (1.5mm). There is an exception: on new 6- and 12-string guitars which are intended to carry medium gauge strings exclusively, the straightedge should clear the bridge by 3/32" (2.5 mm). On a classic guitar, the straightedge should not clear the bridge, but rather should touch the front of the bridge about 1/16-inch BELOW the top edge of the front lip, since lower tension and higher average actions are the norm on classicals.

Usually on a guitar with a workable, rather than ideal neck angle, the straightedge will just bump into the bridge a hair below the top edge. In that case (following the guidelines indicated above) a gentle recountouring of the bridge will make everything shine--on the cheap.

But many guitars fail this neck-angle test. Unfortunately, a good neck angle is a rare thing, indeed -- even, alas, on many brand-new guitars hanging in the average music store.

I recently visited a major guitar factory to assist a good friend in selecting a guitar (he was an ace dealer and his reward was to pick one out). Only one guitar in a dozen had anything even approaching an ideal neck angle. He couldn't believe it. The guys in the factory got most upset at me. I felt like the messenger with bad news that was about to be killed. Most of the guitars we examined (which, indeed, were ready for shipping), had string actions higher than would be comfortable for the ordinary player, and saddles which invariably were too low.

When a new guitar like that reaches its destination and its already-low saddle is adjusted (by the music store techie) down even further in order to remedy the stiff action, there will be nothing left to adjust over the guitar's long future of progressive settling-in. Virtually all the guitars we examined in the shipping room of that guitar factory were candidates for an expensive neck reset barely five or six months down the road -- if not right then and there. After examining about fifty guitars, we found one that happened to have nice low action, and a full-height saddle. I don't want to go through that ordeal again.

Ideally, after a guitar receives its first readjustment following an initial settling-in period, there ought to be sufficient saddle height left for a long lifetime of downwards adjustments. On all these guitars, a better neck angle would have assured this "quality." Sadly, the production engineers at that factory missed a golden opportunity to impart this additional measure of value into their instruments.


And more...