Design and Construction of Acoustic Steel Stringed Instruments

In this series of articles I want share with you my views and philosophies that guide and form the way I make acoustic steel string instruments. I have been making them for a little over ten years now and the main thing that I have learned is that there is no absolute "right" way of doing things - there a lots of ways that work and many of them are completely contradictory to other methods that also work. I've also learned to keep an open and inquisitive mind - you can learn new things from "old lags" that have been building guitars for decades but also from brand new "kids on the block" who approach things sometimes without the baggage that that brings. By and large the steel stringed instrument building community is a very open and sharing one taking the view that by pooling knowledge everybody gains and you create an innovative and evolutionary environment - this is very liberating indeed. I have benefited from this and in these articles hope to add and put back something into this knowledge pool but also want to stress that these are just my views and beliefs and not an attempt to set them up as the "one and true way". I'll be aiming to add a new article each month so if you are interested keep checking back


1. Cunning Plans.


I'm going to start with "Cunning Plans" and the importance of the understanding of first principles, but first a little about me and how I came to make steel stringed instruments. Since the age of 12, when I mimed to Beatles records using a cricket bat, I knew I wanted to play guitar. When I got my first one that year, a Musima nylon string that I still have, I was in heaven. Over the years I progressed onto steel string acoustic instruments, and guitars and music have been my lifesavers and stress relief - along with Irish flutes/whistles where I "dabble" from time to time. As my guitar collection grew I became increasingly more fascinated with the beautiful woods they were made from and what made them sound and play the way they did. I wanted to see if I could create instruments of my own and take and shape a sound of my own. I then spent a good few years researching all I could find on the internet about making guitars and reading all the guitar making books I could get my hands on. I bought an Appalacian Dulcimer kit and that had me hooked. I always do things my way and learn by understanding the principles so I did not want to buy a kit or plans and wanted to design my own instruments. My first two creations in 2002, "Nancy" and "William" were conceived as a matching pair of small bodied guitar and guitar-bouzouki and their naming comes from the frequency of appearances in traditional English songs and their design was dictated by the size of some wonderful pieces of Old Rio Rosewood I found in Isaac Lord in High Wycombe plus two spruce guitar top sets. The backs were six joined pieces and the necks had 16 frets clear of the body so you could say they were unique but there were influences there from the great maker Stefan Sobell who I very much admire.


Since then I have only ever made one instrument from a set of plans that I have not drawn up myself and that was a copy of a 8 course Heiber lute that I made under the guidance of my good friend Colin Symonds so that I could properly call myself a "luthier". A lot of would be guitar makers when first starting out look for sets of instrument plans as do established makers when they venture into new territory such as Baritones, Acoustic lap slide guitars, Harp guitars or multi-scale instruments. My question has not been "where do I find plans for . . ." but "what do I need to understand to make plans for . . ." and by taking this route you learn the first principles that dictate the instruments design rather than follow set plans. Why is this so important - well basically you are free to design and build any instrument that you want having decided what it is you want the instrument to do using these principles. I can understand that when starting out it's a scary enough prospect having to make your first instrument and a set of plans is one less thing to be afraid of but if you don't understand why the plans are as they are and develop a mindset that thinks the instrument will no longer work if you stray a micro-millimetre away from the plan I think it is an opportunity missed. You don't need detailed draftsman's plans either to make an instrument - mine are basic "skeletons" sketched out on lining wallpaper and they will sometimes get changed in the process of making the instrument.


So with a pencil and blank piece of paper in front of you where do you begin and what is the first of the "first principles" that you need to understand? In the next article I'll talk about scale length which is where for me it all begins.


2. Scale Length.


At its simplest scale length is the maximum length of the vibrating string before any intonation compensation adjustments are made at the saddle or nut. I'm not going to talk about intonation in this article but for an in depth explanation of intonation and the theory of different musical scales Mike Doolin's articles are well worth reading at your leisure. Musical octaves generally divide into twelve intervals or notes and come in a number of different "temperaments". The Twelve Tone Equal Temperament is the one used on the vast majority of guitars and stringed instruments. In an equal temperament, the distance between each step of the scale is the same interval which means that any key, modulation or tonality works as well as any other. If you delve into Mike's articles you will see that this does not exactly match the "harmonic series" derived from the harmonics naturally produced by vibrating strings and is a compromise as in all keys the guitar will always be out of tune with itself somewhere. This is an important principle to understand.


So how do you calculate the fret positions using the Twelve Tone Equal Temperament system? Well there are twelve semitones that take you from open string to octave and at the octave the frequency is doubled, so the strings pitch will be increased by one semitone when the strings length is changed to the scale length divided by the twelfth root of two. From this principle the distance of the nth fret from the nut can be calculated by multiplying the scale length by the following factor Fd(n):


Fd(n) = (1-1/(2^(n/12))


Simples !! You will notice (if you are still following this) that at the twelfth fret the factor has the value 0.5 - half the scale length and that the factors are independent of the scale length chosen. So using these factors you can draw out and cut an Equal Temperament fretboard for any scale length that you want. Also if you take an Equal Temperament fretboard and cut it off at the first fret you end up with an Equal Temperament fretboard with scale length equal to the original minus the first fret segment you have cut off - essentially as you fret a note you are creating smaller scale length Equal Temperament systems.


It's important to know these principles but where does scale length fit in to steel stringed instrument design? Basically when matched with appropriate gauge strings and tensions, the musical range - note frequencies - that the instrument will play in determines the scale length. If you think of stringed instruments in an orchestra the highest range is covered by violins, then come violas, cellos and finally double bass with scale lengths and string gauges both increasing as you progress through the range. For steel stringed instruments a similar analogy might be mandolins, octave mandolin/mandola/bouzoukis, tenor guitars, terz guitars, concert guitars, baritone guitars


Other considerations in choosing scale length are the desired timbre and tone and playing ergonomics. With modern strings you could make guitars of similar sized bodies that are tuned EADGBE with similar string tension with scale lengths ranging from 610-660mm. These will all have a different feel and sound to the player partly because the differing string gauges will produce subtly different series of overtones but also due to the different physical stretches of the fretting hand - especially important if the player has problems with arthritis, tendonitis and other similar hand/arm issues. Alternative tunings add another element into the mix as they generally involve tuning strings down and longer scale lengths particularly on the lower strings can help here.


Then we come to the interesting area of multi-scale instruments. Technically "normal" guitars are multi-scales as the strings don't run perpendicular to the frets (string separation is wider at the saddle than the nut) and so each string has a minutely different scale length but that's nit-picking. Multi-scale instruments have a very long history - the Orpharion was around in the 1500's - but are becoming more common in acoustic guitars today. Interestingly some people call them fan-frets implying that if you draw the lines of the frets out long enough they converge into a point like a fan, but this only happens if the scales are drawn out on a parallel fretboard - with a tapered fretboard there is no single convergence point and you get a pretty spirograph type pattern. As "Fan fret" is a trade mark, most makers call them multi-scale, which is what they are.


The basic design parameters of a multi-scale instrument are the differences between the shortest and longest scale lengths and which fret is the "orthogonal" one (perpendicular to the centre line of the fretboard) - this need not be an actual fret but can be a "virtual fret" in between two real ones. I personally think 30mm is the maximum in terms of the scale length differences and beyond that normal playability becomes an issue as well as structural and aesthetic positioning of the bridge. Most of my multi-scale instruments have been 30mm scale difference with the 9th fret orthogonal. Recently I've been making more gentle scale differences of 12mm with the 14th fret orthogonal which puts most of the slant at the nut end where and allows for a more normal bridge shape and angle. Here are two guitars that show the differences:



So how does multi-scale help? Firstly for hybrid instruments - for example a mandolin is tuned GDAE and a mandola has a longer scale length and is tuned lower in CGDA. So if you make a multi-scale instrument with five courses of paired strings tuned CGDAE and a body size somewhere between a mandolin and mandola you get a hybrid. Secondly for instruments such as 7 or 8 string guitars, Harp guitars and Baritone guitars where you are looking for extra string length for the lower note ranges. With a Baritone guitar the top five strings are in the same frequency range as the bottom five notes of a Concert guitar but to go down to low A or G acoustically you need to increase the scale length quite a lot. If all six strings have to have the longer scale length then the timbre of the other notes will be different - multi-scale makes a lot of sense here. This logic also works for guitars that are used a lot for modal or altered tuning where you can combine the slightly "sweeter" harmonics of shorter scale lengths on the treble notes with longer scale length string tension for the tuned down lower notes. This gives a more balanced sound and string tension feel to the player across the guitar strings - it doesn't make for better intonation though as is sometimes argued for multi-scale instruments, each string is treated based on it's own scale length for intonation compensation for both fixed and multi-scale instruments and with the appropriate string gauges they should both intonate as well. Even if you play in EADGBE tuning, a gentle multi-scale can bring similar string feel and sound benefits between treble and bass strings.


So having settled on the instrument's scale length(s) now gives us two very important points to draw in on our blank piece of paper after allowing for intonation compensation - the nut position and the saddle position. So what is the next design consideration? In the next article I will talk about body shape and size.


3.Body Shape and Size.


Having chosen a scale length the next decision is how many frets on the neck will be clear of the body, the shape and size of the body and where the bridge will be on the lower bout. In the development of the steel string guitar a number of distinct body sizes and shapes have emerged as "standard". One of the more well known are the Martin "nomenclature": Size 5, O, OO, OOO/OM, and Dreadnaught but I prefer the more descriptive Parlour, Concert, Grand Concert, Auditorium, Grand Auditorium which gives a better indication of the size and where they were designed to be played. There is also the Jumbo and Small Jumbo (oxymoron?) from Gibson and specialist instruments such as Piccolo, Terz and Baritone guitars.


As you go up in size you generally get a larger top surface area, a bigger volume of air inside the box and a deeper guitar. The scale length tends to increase with body size but not always - this also depends on how many frets clear of the body are chosen. The larger top surface area means that the top has a lower main resonance as does the box with the larger air volume and this will tend to support bass notes better. The fullness and complexity of the sound also increases with body size with more natural reverberation present in the sound - you can hear this with hollow arm harp guitars even without their sub-bass strings, the main strings sound as though you are playing inside a church or cathedral. This does come at a cost though and you need to work at maintaining string clarity across the strings particularly for strumming - take the example of the large O series Lowdens strummed compared with the smaller F and S series. Conversely with smaller guitars you have to work at bringing out the bass and making the sound as full and complex as possible. One of the signs of good makers in my view is the ability to make small full sounding guitars that surprise and change the perception many players have that these instruments are "boxy".


You can also shape the sound by altering the depth and taper of the box - a deeper box can help with bass and complexity and a shallower box can give an instrument more punch and mid-range focus and less susceptible to feedback when plugged in - useful in some live performance situations. Again there is a trade off though as the air movement inside the box is governed in part by the interplay of the back and top movement and when you go beyond a certain depth you loose a degree of clarity and focus across the strings. It's also worth mentioning here that bigger guitars are not always going to be louder guitars. Loudness or perception of loudness is one of those interesting phenomena that occur in our brains. These have evolved to hear better in certain frequency ranges (usually related to hearing other human voices) and like vision, the human brain hears certain things and then "fills in" the rest. It is possible to make it hear a bass fundamental that the instrument doesn't generate via a series of harmonics that the instrument does generate that are associated with the fundamental, or by accentuating the power of the mid-range frequencies - I'll explain in more detail in a future article about Baritone guitars. It also explains why higher pitched instruments (like small yappy dogs) will cut through a mix of instruments and be perceived as loud. The way our brains "hear" also explain why it's best to use you ears to tune a guitar rather than a precision tuner unless you are playing for an audience of Peterson strobe tuners.


Where the waist is positioned on the guitar and how "pinched" it is also shapes the tone. Guitars with curvy waists tend to have more focused mids and trebles whereas guitars with very little waist - like dreadnoughts - have a more bass focused sound. This is partly due to there being more air volume but it is also due to how the waist changes the motion of air inside the box. Where the soundhole is placed on the soundboard will also impact as the further towards the fretboard it is (which usually also goes with where the waist is positioned) the larger the lower bout active area of the top will be and the lower the Heimholz resonant frequency of the box. There is a trade off here as the total number of frets that can be on the neck is less. Finally the relative size of the upper bout to the lower bout will also shape the sound both in terms of box volume and air movement patterns - Dreadnaughts tend to have wider upper bouts in relative to the lower bout. The Modified Dreadnaught developed in the 1970's by Ervin Somogyi took the dreadnaught shape, made the waist more pinched and moved the waist and soundhole towards the fretboard to make the instrument more suited to the Wyndham Hill fingerstyle guitarists and there type of sound.


This brings me neatly onto the positioning of the bridge in the lower bout area and the discussion about "sweet spots". There tends to be a view that positioning the bridge close to the middle of the lower bout at its widest point is a "sweet spot" to drive the top. This is not necessarily the case - it depends on the top bracing. The best spot is likely to be the centre of the "active" part of the top and this depends on where the bracing is put. On classical guitars where the energy available from the strings is much less than on a steel string, the active top area is generally reduced to the soundhole downwards by putting a large transverse brace there as well as the one on the upper bout, and so classical guitar bridges are further down in the lower bout than on steel strings. With steel string guitars the large upper transverse brace in the upper bout will generally define the active top area and with makers such as myself that believe in the importance of the upper bout to tone this is a tonal rather than structural brace.


So once the maker has the scale length and view of required body shape/size/volume and has determined the best tonal position for the bridge based on his building style the number of frets clear on the neck usually comes out in the wash as it were.


With both small and big bodies there can be issues of ergonomics and player comfort. Generally fewer frets clear of the body tends to feel easier for people to fret but there are issues of upper fret access that cutaways can help to address. Small people can find bigger bodies more awkward to play and so features like the Manzer style wedge (where the treble side is wider than the bass side of the guitar for the same body volume), armrests and bevels can make them more manageable. Some bigger players can envelop smaller guitars and straps that attach to the guitar and rest on the players knee holding the guitar in the "classical" position can also help here.


One thing to bear in mind when designing different body styles - as well as aesthetics - is the availability of hard cases that will fit the instrument. It helps to design around commercially available case sizes unless you want to go down the custom case route. Also for a maker that uses heating blankets and bending forms together with external moulds it helps to have a manageable number of body shapes and sizes that can be used with different scale lengths and frets clear of the body to produce a wide range of models. This is what I now do and have five basic body sizes for my guitars. You can read more about this here on my website.


Finally here's a photo illustrating my five main guitar body sizes - left to right "Treebeard" (baritone), "Buachaill Mór" (Modified Deadnaught), "Lughnasa" (Small Jumbo), "Samhain" (Grand Concert) and "Féileacán" (Concert):



4. Defining Moments.


The next logical step after body size and shape is to talk about bracing designs and principles but first a little digression is in order and I'll talk about some "defining moments" that have influenced my building techniques. Changes in my designs and techniques have mostly been an evolution as I learned more but there have been five "defining moments".


I've never glued the fingerboard extension to the guitar top as it never made any sense to me. Then I learned about how American builders Mike Doolin, Rick Turner and Howard Klepper were using "flying-buttress" braces inside the guitar body to deal with the string tension forces on the neck-block enabling a free floating fingerboard and freeing the upper bout area of the top to become a tonal rather than a structural area. Many builders dismiss the top's upper bout area as having no role to play tonally - they couldn't be more wrong. Using carbon-fibre flying buttresses, the rim-set becomes incredibly stiff and can be used as an efficient way to help support the string forces of the neckrather than using a heavily braced area of the top's upper bout.



The next "defining moment" was when I changed my neck design so that the neck shaft runs all of the way up to the upper transverse brace on the top and then added a final "tweak" on my neck/neckblock design based on the system developed by my friend and fellow UK builder Colin Symonds. This combined with using two carbon fibre bars in the neck either side of the truss road gives a similarity of tone as you play notes up the neck and across strings and capo up the neck and also eliminates the neck/body join "hump" problem.


The third "defining moment" came when I had the good fortune to read about the American Scott van Linge's work and theories about top bracing and had a long exchange of e-mails with him. Scott "annoys" many builders due to the "scientific theory" he has put around his methods. I don't agree with his theory but this doesn't matter as his methods work for me and turned on the light for me on to how to "voice" my tops and backs to give the sound I want and changing my bracing profile to "tapered". For this I am eternally grateful to Scott.


The fourth "defining moment" was using side sound ports. My instruments have always been loud and project well but this is not always good news for the player who would like to hear what the audience hears. Side sound ports let this happen without compromising the forward projection and the results for drilling a small hole in the guitars side are fabulous. They act as a personal "monitor" to the player giving more feedback and the impact on the forwards projection and sound is minimal - in fact I believe that the extra port helps the top move in a more interesting and free manner.


The fifth "moment" was discovering multi-scale instruments. I came to multi-scale by a strange route. I play a lot in dropped/modal tunings and this is where I want my instruments to thrive. I had in my head the mantra that you needed longer scale lengths to do this to give better string tension as you tune down. Then inspired by American builder John How I made a ladder braced guitar with a short scale length - 630mm. I was surprised firstly by how sweet the trebles sounded and secondly by how well the instrument played in DADGAD with light strings. So in the quest for the ideal DADGAD guitar I decided on a 630-660 multi-scale length giving sweet trebles and more tension on the bass strings. This string tension balance plus the fact that the slanted frets suit the natural angle of the fretting hand are the main contributions to my instruments, even if you play in standard tuning. The other area where multi-scale make great sense is Harp guitars where the transition to lower sub-bass strings calls for longer scale length string by string. With a multi-scale this means that the bridge design and the way it drives the top is much more efficient. I've done this now on one harp guitar and a Tarropatch Harp Ukulele. The same principles would apply to Baritones. I'm a believer.


Actually my friend Leo Roberts has discovered my main little secret as this song of his explains.