Tuesday, March 8, 2011

Getting some Action.

I recently had the opportunity to inspect and play the early efforts of an enthusiastic new luthier. Our conversation went something like this:
(him) "I'll never use maple again, it sounds awful! There's hardly any sound."
(me) "Are you sure it's the maple?"
(him) "It has to be. The first one I made I used rosewood. It's twice as loud."
(me -resisting the urge to point out that a jackhammer is loud too...) "Let me have a look at it. "

Upon inspection I could see where things had gone off the rails. The finish work was good, as was the action. The bridge however, was very low - only about 1/4" (6mm) high. This, coupled with the beginners propensity to leave things a bit thick (just in case!) had produced an instrument with the resonant properties of a solid-body electric guitar sans amplifier.

We talked about it for a while. He had the makings of a half-decent flamenco instrument, if he wanted to strip the top and scrape the periphery down to make it more responsive. Or, if he was set on a classical sound, the other option involved loosening the bindings around neck end of the back and trying to pry it free from the heel block. With some clamping pressure and a little luck, he might be able to deform the box enough to tip the heel back a degree or two. He said he'd give it some thought.

I think the most vexing thing about the situation was that he'd done nothing wrong. He'd been very conscious about following all the instructions in the manual. Unfortunately, most of the literature available to beginners just doesn't address this issue. "Standard" dimensions are given, and things just sort of come together. Unless they don't. Most of them expect the maker to go ahead and fashion the standard bridge without measuring the geometry of their instrument as it exists. I wonder how many bridges have been discarded when it was found that they produced unplayable guitars?

What are the two factors that should most interest a guitar player? Sound and playability. (In reality cosmetics play a key role, probably more than any of us might like to imagine).

Think about this: the difference between a "flamenco" instrument and a "classical" one is a matter of two millimeters. That's the difference in height of the strings above the soundboard. Classical guitars usually have a height of between ten and 12 millimeters, flamenco guitars - more like 7-9mm. These measurements include the height of the protruding bridge bone.
Interestingly, 2mm is also the difference between "good" action, and unplayable. These measurements are directly tied to one another. Two little millimeters between success and failure.

I have a large collection of reference books. I think it's safe to say that if there's an English-language guitar construction manual I have owned it or have read it. What's amazing is that most describe the design process for the beginner in terms of the face-view of the guitar. They give average measurements for string spacing, scale length, fingerboard and body lengths. None describe a method for predicting the action of an instrument.

If I were teaching a class in construction, I think I would have students begin by drawing a simple plan view of the geometry of the bridge, action, and string lengths. Will there be a dome or arch in the top? How high? Will the fingerboard remain a constant thickness from nut to sound hole? Will the neck be pitched forward or in plane with the soundboard? What are the requirements in terms of action and string height? And what will happen if I change the thickness of my fretboard from 6mm to 5? This is the dynamic geometry that underlies the instrument. In many ways it's more important than bracing schemes or top material.

Here's a sample drawing detailing the differences between my classical and flamenco set-ups:

The tolerances are pretty close. If things are glued up and something did go wrong, it's important to realize that there's only about a millimeter of adjustment in action available to us in the saddle bone. If we're savvy, we can gain an extra half-millimeter by planing the fretboard in one direction or another. Reducing it in thickness by one millimeter at the nut end, for instance, will raise the convergence point with the saddle by a millimeter - the twelfth fret becomes the theoretical pivot point.

What's nice about the Spanish method of construction, using a workboard and having the guitar neck in place, is that this tricky geometry can be locked in and clamped down. Assuming things are kept square and measurements are accurate, there shouldn't be more than .5mm deviation in any direction, well within the adjustment tolerances of the standard set-up.


  1. Replies
    1. 325mm from nut to the 12th fret. The scale length being 650mm (plus 1.5mm compensation)

      In reality, I often add some compensation at the nut as well by back-filing the takeoff points for the G and low E strings.