sparkspin

A chronicle of the design and construction of a new guitar.

(plus some advice for the novice guitar maker)...

To continue the story of the anti-feedback guitar; the main reason for my building this prototype is that I am going to need some means of recording a good, crisp acoustic rhythm guitar sound, now that I have my dedicated music room set up. This is a first for me - a music room; before, I've always had to live amongst the kit or "get the equipment out", lash it all together and put it away again when finished - and "finished" is obviously yet another whole story in itself!

This year it is my intention to record, as a project, quite a few old standards (mainly '30s and '40s era) as electric blues versions. Acoustic rhythm guitar will form the backbone of the rhythm tracks and, in my experience, acoustic guitar is notoriously very difficult to record. Our ears decode sounds in all sorts of mysterious ways en route to our brains; what sounds cool in an ordinary room, especially when one is the actual performer, comes off tape or disk as a very different animal unless one resorts to expensive anechoic studio rooms and expensive microphones, neither of which beasts do I have to hand.

The main problems are (1)- what to do with the bass boom in the guitar body* (see note below) & (2)- the room reverb. Neither of these things are desirable in a controlled, clean acoustic rhythm track even where the other instruments are up front in the mix; you can kill bass with equalisation but that is not satisfactory when the track needs to be heard solo. It should be obvious that if one strums acoustic first in the bathroom and then snuggled in a deep sofa on a carpeted floor, the perceived sound is very different in each case.

*Conventional wooden-constructed acoustics depend for their volume output upon the individual natural resonances of small sections of the soundboard (table), the whole being divided into these smaller areas by the bracing spars. This is fine until one fits a piezo transducer for amplification purposes and attempts to play at high volume. These individual naturally resonant panels each have a tendency to produce distinct peaks in the frequency response of the guitar and "honking" feedback quickly occurs in the body, usually at the strongest peak first. This is nearly always completely unmusical, being unrelated to any note in the conventional Western musical scale.

One trick I've found for killing the spurious bass peaks is to tape a piece of cardboard over the soundhole but this kills treble, too. What is required is a well-balanced spectrum of frequencies 'though not necessarily acoustically loud. Just sufficient of everything, in equal measure - and this is the key to the anti-feedback format:- there are no peaks in the body response for howl-around to latch on to.

Strumming a conventional semi-acoustic won't do either. Imagine trying to record a Gibson 335 acoustically, for example; it simply wouldn't happen (usefully). Some of those old pre-electric design 'cello f-hole style guitars record the acoustic guitar sound (nearly) that I'm after (viz:- ''Summertime Blues'' or "C'mon Everybody") but again, expensive microphones are called for and with the "trapeze" tailpiece method of string anchorage, there is no proper sustain as there is in the solid-bodied guitar league. You wouldn't use a Rickenbacker to solo on "Crossroads"! This new guitar, as well as sounding full unplugged ('though not acoustically loud), also needs to sustain naturally and be capable of loud electric playing, with both piezo transducer and magnetic pick-up systems (and combined) without erupting into body-honking howl-around.

Many, many experiments towards making a plastic acoustic, starting with complete failure in 1972, resulted in the formula (using Formica® as the table) I arrived at by 1994; inherent in the design was the anti-feedback property of tuned bracing, discovered by accident. The pattern below was used for the bracing which was cut from 1.5mm birch model making plywood. The turned-up end radius (A) was an attempt to get as much sound radiation from the very last bit of (noise active) bracing right next to the ribs (sides) of the sound box whilst retaining a useful degree of flexibility at (B). A latticework of edge-on interlocking braces is obviously extremely strong. Slots like the one at (C) facilitate this:-

It reminded me at the time of those weird shapes on Chinese and Japanese pagoda roof corners; their purpose is to ward off devils and if you consider feedback as a devil, it certainly works well in this context! During experiment, I have found that it is perfectly possible to tune each of the braces to a different specific concert pitch note, thus further disallowing any potential response peaks not relevant to the music; in practice it was found that random tuning was fine anyway and was equally effective at feedback suppression. The proportion was decided by what could be easily cut with the Black and Decker Powerfile!

The overall effect of the assembled bracing latticework is to produce something very like "white noise" (the sum of all frequencies). Hence:- feedback? No chance...

How will the neck be constructed?

For those uninitiated in the methods of guitar neck construction, there may be a few surprises here; luthiers - please forgive the egg-sucking tutorial...

It was a pleasant surprise to find that my local builders' merchant (Jewson) holds a considerable stock of hardwoods, Sapele or Mahogany being the favoured stock of choice for this project. Browsing the racked timber threw up a couple of promisingly straight-grained pieces. One has to bear in mind the sad fact that there is still so much rainforest timber being consumed that there is absolutely no chance of buying anything properly seasoned; the stuff stays on the shelf for only brief periods. The only option was to buy something way oversize and force-dry it in a warm environment for some weeks, then to allow it to rest for a few weeks further, prior to shaping (gone are the days when I held stocks of controlled stability woods; we are talking strictly retirement hobby and garden shed now!)

The other main neck component has to be the fretboard and without access to my old fretting jigs, the sensible option was to order a pre-slotted fretboard blank from a luthiers' suppliers. One such supplier that I have used frequently before was Touchstone Tonewoods of Reigate, Surrey - (sales@touchstonetonewoods.co.uk) and from them I ordered a Rosewood pre-slotted blank, together with some big-section fretwire. Much as I love working with Ebony (contrasting inlay work really sets off a treat), I've found it in the past to be dreadfully unreliable because of its hygroscopic tendencies; it sucks up moisture like a sponge from the atmosphere, making for dimensional instability in great measure. So it had to be Rosewood - still a beautiful timber with which to work.

Truss rods are necessarily curved (or they don't do anything towards counteracting string tension). 3/16" mild steel is adequate for this and a 2B.A. thread serves for the adjustment nut. The slot is cut in the neck block blank and a shaped fillet piece of hardwood is glued in on top of the rod. Then the fretboard blank is glued over the (perfectly flat) top surface of the neck block blank. The plane for the front of the headstock is cut prior to fretboard glueing; this can be very awkward if attempted afterwards.

For many years I used multiple G-cramps for this job - until I learned a trick form an archery bow maker:- strips of car tyre inner-tube spiral wrapped bandage-like work much better as the whole "Mummy" is self-centering; the problem of parts slipping around on the lubrication of the wet glue is solved at a stroke. Each wrapping around of the rubber strips stacks up the hand pressure and there is nowhere for the fretboard to go but dead center and with a dead flat joint.

Recently published experiments have disproved the old myth that wood should be "roughed up" or "keyed" prior to glueing; as smooth as possible is now the norm.

The next job is to establish the center line and bandsaw the taper of the glued-together neck assembly. There is no doubt in my mind, after building several hundred guitars, that the best proportions for a fretboard are:- 42mm at the nut and 55mm at the 21st fret

- so that is the way to go for me on this instument. Several times, I've experimented with wider fretboards because I like the classical guitar spacing format but it's re-inventing the wheel, really; Leo Fender and Orville Gibson did their homework and you just can't argue with that (but the 25½" scale just feels right to me (and countless thousands of others).

Inlays and fretting:-

Jim Dunlop makes great fretwire and I happen to like it big - as big as possible. The effect of this type of fret is that there is minimal contact between the finger tip and the fretboard (rosewood in this case) - big frets don't seem to slow Ravi Shankar down much... but before we get to the fretting job, there's the position dot inlays to sort.

Old fashioned plastic knitting needles are getting harder to find but they are ideal for dot inlays. Cutting the camber (shaping the fretboard into a curved surface) can be done without the aid of a large flatbed linishing machine merely by use of a long sanding block and the position dots will go in and be left proud to be sanded off flush to the fretboard surface...

making a guitar tips continued...

Keep coming back for progress reports. Last updated 3rd Feb. 2004.