Back to the index page
|
While the double strung Irish Bouzouki has become very popular in the last twenty five years, especially in the Celic music community, little has been written in the instrument, its origins and the way in which they are built. This is really two articles in one, the first part describibg the variety of instruments which get called bouzoukis (or citterns or octave mandolins) and giving some idea of how they get put together, and the second part on the method I have developed for making molded and carved arched soundboards which I use on bouzoukis, mandolas and mandolins.
The Evolution
of the Irish bouzouki While the Greek bouzouki is tuned CFAD (two semitones below the top four strings of a guitar, with octave strings on the third and fourth courses) the Irish players evolved a GDAD tuning (an octave lower than the mandolin with the top strings tuned down to D and unison strings for the bottom two courses). Irish music is mostly in the keys of G and D, this tuning meant that both the first and fourth strings could be played open as a drone, and allowed chords and melody playing around them. As anyone who has ever played one knows, the bowl back of the Greek instrument makes it quite hard to play standing up, and someone came up with the idea of building them with a flat back. So the Irish bouzouki was born. The Greek instrument commonly has a 660mm(26") or so scale and many of the builders of the new flat-backed variety kept this scale length along with a flat soundboard with a low, floating bridge, and a mandolin style tailpiece. The 660mm scale is the usual maximum, but builders usually offer a variety of string lengths, down to around 510mm (21"), which is about the shortest you can go with a GDAD tuning without really heavy strings. Longer string lengths seem to give more richness and texture to the tone, to my ears at least. The bottom of the fingerboard is usually in the same plane as the top between the neck/body join and the bridge (or kicked back only slightly), making the bridge quite low (usually 10-15mm) and the break angle over the bridge rather shallow. The top bracing is variable, from a couple of simple transverse braces to a guitar style X. Bodies are usually a teardrop shape of one sort or another, often scaled up mandolin shapes. Mine are 400mm long and 350 wide, and 90-100mm deep, which for a 660mm scale means a body join at the 17th fret. I use a 75mm diameter soundhole for no real reason other than it looks right with that size body. There is a lot of variety in body shape and size, and mine have certainly grown in size over the years. The teardrop shape is the most common, but there is no reason why any shape cannot be used. I have built quite a few with a small guitar type body which, having a waist, has the advantage of being much easier to rest on the leg. Having a long neck, and eight ot ten machine heads, makes these instruments neck heavy, so a longer body with a waist is of some advantage. Neck profiles are similar to a steel string guitar, though usually narrower, and an adjustable rod is usual. For a four course instrument my neck width at the nut is 34mm and 44mm at the 12th fret, which seems to suit most players. This is about the same flare as a standard acoustic guitar. For reasons I have never understood, several of the Irish builders keep the strings almost parallel, but I think it looks strange. Alongside the bouzouki with its four courses of strings, has evolved a five course instrument which luthier Stephan Sobell dubbed a cittern, after a 16th century wire strung instrument. The original idea for these instruments owes a lot to the Portuguese guittara, a short scale 6 course instrument with a floating bridge. The modern citterns are usually a shorter scale - 510-610mm (21-24") - than bouzoukis and a range of tunings are commonly used although commonly utilising 5ths between courses. There is ongoing discussion about what to call this family of instruments, but there now seems to be a slowly forming concensus that long scale four course instruments are 'bouzoukis' and shorter scale five course instruments are 'citterns' and anything else gets called whatever the maker wants to call it. I'm the first to admit that my understanding of all the various resonance modes of soundboards is imperfect, and while there has been lots of work done with guitars and violins, I don't know of any experimental work done with bouzoukis and citterns. These instruments are somewhere between guitars and violins in size and tuning, so I figured that if at least the basic air and top resonaces were somewhere between fiddle and guitar resonances I might be on the right track. The teardrop bodied instruments seem to work best with a resonance between C# and E (somewhere around the open D string) again with the air resonance a little lower. What I don't know is what is happening with the higher vibrational modes of the soundboard (have a look at Graham Caldersmith's articles in back copies of AL for the most understandable explanation of this), although the noted Irish musician Donal Lunny mentioned to me that just about all the teardrop shaped instruments he has played over the years have a resonance peak around 220hz. This would be the second top resonance mode, I suppose, with undoubtably lots of other at higher frequencies.
Some developmental ideas One of my thoughts in using a guitar shaped body for a bouzouki was the idea that I was wanting an instrument that was going to have the same sort of resonance and string-to-string balance as a guitar and the bottom string on a bouzouki is usually tuned to the G (which is the third fret of the bottom string of a guitar) so it seemed reasonable to be using a body much the same size and air volume as a guitar. Good guitars usually have their basic soundboard resonance around G# on the bottom string, and so it seemed reasonable that a guitar bodied bouzouki should work with its resonance in the same relative position, four frets up on the bottom string. So I aim for a resonance around B-C on the bottom G string on a guitar bodied bouzouki with the air resonance a little below that. The smaller boded teardrop shaped instruments I look at differently. They are going to have a thinner, brighter sound, without as much richness in the bottom string. This often the sound people want, because it allows the bouzouki to be played with a guitar, with the two instruments having quite distinct timbres (listen to the Bothy Band recordings from the late 70s), but there are others who want more of a well rounded solo instrument. Necks are attached in the same multitude of different ways that people put necks on guitars. I like the idea of the neck/fretboard being built as a separate unit to the body, preferably with the neck extending the full length of the board. My current way of attaching necks is using a mortice and tenon, with barrel nuts inserted into horizontally drilled holes in the tenon, and mating up with machine bolts that are inserted through matching holes in the headblock. Bridges for archtop fretted instruments are something for which no-one has a definitive answer. There is the standard two footed Gibson style bridge, which I always felt is not very acousticly efficient, and the top section always seems to eventually sag in the middle. Bob Benadetto in Making an Archtop Guitar suggests that a light violin type bridge might be the answer, but the article a year or so back in AL on James D'Aquisto's instruments showed him using a much more massive bridge with a footprint much the same as a flattop bridge. It is one of those areas that needs some scientific analysis of what bridges of various size and materials do to the sound. One day..... I use a two footed bridge, and make it in top and bottom sections, with the top part located by two brass pins, but otherwise fully in contact with the bottom part. If it does need it's height adjusted it can be shimmed or planed down, and it is also then practical to use an undersaddle pickup (one of the full width sensing ones) by routing a slot in the upper section of the bridge and fitting a bone saddle. The output lead of the pickup can then run out of a curved slot in the bottom part. Tailpieces should be massive enough so as not to absorb too much of the strings' vibration. Most makers use either a a metal mandolin or wooden archtop guitar style tailpiece. The main problem I see with the mandolin style tailpieces is that they tend to take only loop-end strings, and it is much easier to find ball-end single strings than loop-ends, bouzoukis and citterns needing sets of strings that have to be made out out of the singles box in the music shop. I make my tailpieces from rosewood or ebony, and attach them to the instrument with stainless-steel hinges obtained from ships' chandlers. 2mm thick and as solid as you want.
Bouzouki soundboard construction Over the years there have developed two main approaches to the way in which Irish bouzoukis (and citterns) are built. The first is with the flat soundboard (usually somewhat curved laterally though bracing) and a fairly low floating bridge mentioned above. The second method uses a carved, arched soundboard, again with a floating bridge, but around Gibson mandolin height (25mm). There are also a couple of other ways of going about it. One approach is to build the instrument with a flat top with a pin bridge (like 2/3 of a 12-string). The English company Fylde has built bouzoukis and citterns like this for many years, and it does make putting them together a lot easier (especially if you have built guitars), and the sound is not very different from a floating bridge. There is also the bending, or cranked top, method where a transverse bend is put in the soundboard just behind the bridge as is usual with Neapolitan mandolins. I built a few of these some years back, routing a groove on the inside of a glued and thicknessed soundboard, and then bending it over a hot pipe. Transverse braces with a reasonable arch were then glued on, and the top attached to the sides with tentalones. Getting the sides and top lined up was a bit tricky, but that's why they invented bindings. (As an aside, I was interested to note that in Robert Lundberg's recent article on mandolins, he says that the old makers would bend each half of the soundboard separately and then glue them together. The only excuse I can offer for bending the whole soundboard at one time is that I didn't know any better.) The main reason I experimented with the cranked top was for strength. Putting that bend in the soundboard adds a great deal of strength to offset both the downwards and longitudinal compressional forces at work.
The problems with building bouzoukis There are a couple of inherent structural problems with the common flat-top bouzouki. All the stresses on the body are quite different from a guitar. There are two main stresses at work, the pressure from the tailpiece trying to collapse the body longitudinally, or at least pull the tailpiece up and over towards the machine heads, and the downbearing from the bridge trying to cave-in the soundboard. What tends to happen often is that the soundboard distorts into an 's', with a hump just in front of the tailpiece, a dip underneath the bridge and the headblock rotating bringing the neck up. Building them with the neck in the same plane as the top (or kicked back only a degree or two) and useing a tailpiece, the bridge must necessarily be quite low, and the break angle over the bridge quite shallow. This will mean that the down-pressure on the soundboard is minimised, with the downside of this being that soundboard isn't being driven as hard as it might be, and so volume often suffers. (The loudness of the instrument is often an important consideration for the players, especially of Irish music, because they often spend a good amount of their playing time in sessions trying to hear themselves over the button accordion and bodhran player on either side of them.) Both archtop guitar and violin builders work on a 15° break angle over the bridge. This just puts too much down pressure on a flattop instrument, and some degree of structural deformation seems inescapable. Over several years of building flat-top bouzoukis I tried all manner of approaches to this problem of compressional forces. I played around with the neck angle and bridge height to see what would happen. I took up Greg Smallman's idea of a plywood rim replacing the usual linings and various sorts of flying braces from the tailpiece forward to stiffen the instrument. I think I finally decided that flat tops, at least with a floating bridge and tailpiece were not the way to go.
The idea of molding soundboards The answer to all these structural problems seemed to be carving the tops, as archtop guitars. mandolins and violins don't have anything like these structural weakness, but the cost of blocks of spruce big enough for bouzouki tops was pretty scary, and I didn't have anything in the way of the necessary gouges and carving chisels, let alone those nifty little violin maker's planes (which I've since bought and consider the most wonderful tools I have ever spent money on). Then I came across an article in the English magazine, Early Music, (I think from about 1976) describing the restoration of a 16th century viol, and the repairer's discovery that the soundboard appeared to have been bent and then coopered together in four sections, rather like staves of a barrel. He found what looked like scorch marks on the inside where the most radical bends had been made. The idea of bending the soundboards was appealing for a couple of reasons. Using timber 6mm thick rather than 18-20mm was an obvious cost saving, to say nothing of a more efficient use of resources, and the fact that the grain would be running along length of the soundboard's arch would seem structurally stronger. It might then be possible to achieve the same structural strength and stiffness with less weight, and so more efficiency in the way the instrument works. (If a string has to move less mass in converting the mechanical energy of the string into acoustic energy, then more of that energy can be turned into sound. This is one of the ideas behind Greg Smallman's very light, but stiff classical guitar soundboards.) I had a couple of attempts at building these coopered soundboards, but it was quite tricky getting the wood that was around 6mm thick to bend in the quite complicated three dimensional way that was required over a hot pipe, and the shapes were never quite right. This technique was put into the 'interesting, but too hard' basket, and I continued to ponder on ways to form 6mm thick pieces of spruce into arched soundboards. The two things used to bend timber are heat and/or moisture. I couldn't think of an easy way to heat up pieces of spruce big enough for bouzouki or, for that matter, guitar soundboards while keeping them wet, but I could certainly keep them wet in a large fibreglass tray that I had in the workshop. So I soaked two soundboard halves for a couple of days and clamped them down on a piece of plywood with a 18mm ply hump in the middle, using a ply caul in the shape of the body outline and extending in about 20mm inside the body shape almost to the edge of the hump. After letting it dry for a few days I took the clamps off, and the arched shape remained, although the spruce sprung back so there was only a 8-10mm internal arch instead of 18mm. There were still lots of problems to solve, like joining the two halves, and discovering that it was a good idea to plane what would be the inside surfaces before bending. I still wanted to get more arch in the tops, somewhere between a guitar and a mandolin I figured, so longer soaking periods were tried, a double soaking and clamping and a higher hump to bend the wood over. The springback was still there, and trying to bend over the higher hump started to cause some noticeable compression marks in the spruce at the inside edge of the clamping caul where the upward curve started. Heat was what we wanted as well as the moisture, and I eventually thought of oven bags. They are made by a company called Glad out here in Australia, and are made of a stiff clear plastic and are used to cook things like chickens and turkeys in the oven to keep them from drying out as they cooked. The bags are good to 200°C which seemed fine for steaming wood. The small problem was that even the biggest ones were not really big enough for a bouzouki soundboard, but they could be cut and opened up into sheets and masking tape held them together and sealed them quite nicely. The size problem was sorted out when I rang the manufacturer, and their PR man said I could have the end of a roll of the material they make the bags out of. The roll was about 400mm wide and doubled over, so a soundboard half fitted in just right, and when the plastic was cut of to a length 100mm or so longer than the soundboard half there was plenty to fold over and seal with the masking tape. The oven in the kitchen was just big enough to hold the wet bits of spruce in the their bags, so I fired it up to around 150°C and left them for 20-30 minutes, until I could feel them getting flexible and then onto the mould and clamped up. And it worked even better. There was still some springback, but it was considerably less, and this is the way that this technique has been evolving over the past couple of years.
The bending of soundboards - step by step 1. The
Mould The clamping caul is also made from 18mm ply, and faced with cork. It is 40mm wide, half outside and half inside the body outline and this gives a little room on the inside before the spruce is forced into the curve. I used offcuts from cork floor tiles about 6mm thick for the facing. This lessens the chance of crushing the spruce, especially at the initial bend point into the arch. The caul should also have a centre line marked on it. Another curved caul is useful to clamp over the length of the centre joint. I made this from a piece of 50mm x 50mm wood, a little shorter than the length of the hump, cut to the longitudinal curve of the hump and again faced with cork. This keeps the timber from 'lipping up' as it dries. 2. Soundboard
preparation Soak both halves of the wood in water for at least 24 hours. If you have some way of getting them into continuously hot or boiling water this time would probably be reduced, but the cold water seems to work fine. After soaking, seal each piece into an oven bag with a little extra water. Ordinary masking tape will seal the bags effectively enough, and you want the bags fairly tightly around the wood. You want the bags fairly well sealed so that the oven doesn't boil the moisture out of the wood, because it isn't nearly as flexible when that happens. Place the two bags in the oven on its lowest setting, usually about 150°C, and leave for about 20 minutes or until the bags have swollen up from the steam, and the wood feels flexible. Gloves are useful at this stage. 3. Forming
the arch At this point you can cut as much of the bags away as is feasible without disturbing the caul and clamp on the secondary longitudinal caul. Leave the whole thing to dry for at least four or five days in a warm spot. After three days or so it can be disassembled, the remnants of the bags removed and then reclamped until fully dry. The spruce will shrink appreciably across its width, with a gap of 2-3mm at each end and more in the centre. When finally disassembled there always is some springback, but you should hope for close to 15mm of inside arch. 4. Jointing
and gluing the soundboard To glue the two halves together I use the traditional Spanish method of 50x25mm timber bars, rope and wedges (again as per Jim Williams), but leaving the top bar off in the centre and using the centre wedge to just push the centre of the joint together if required. After gluing, roping and wedging make sure that the inside edges of the soundboard are as aligned as possible, pushing them into place as necessary. 5. Graduating
the soundboard 8mm wide X-braces are then glued on, cut to the inside curve of the arch with a height of 15mm at the crossing point, curving down to 3mm at the ends. Start with brace stock 20mm x 8mm to allow for trimming and fitting. I end them 40mm from the soundboard edges which is in proportion an archtop guitar X-brace as described in Bob Benadetto's book) . I use a round soundhole 75mm in diameter, but there should be no reason why F-holes shouldn't work. I just haven't got round to trying them yet, and I think bouzoukis look right with a round soundhole. 6. Resonances
and tuning
Possible further development This method of bending an arch into a soundboard is decidedly low-tech and cheap, but it works. I'm waiting for someone to tell me that it doesn't and why, but in the meantime I'll keep doing it. I am sure however that it is a method which can be improved, refined and made more efficient. I have thought that a cast aluminium mould, which could be heated from below would get rid of the oven bag idea, and vacuum forming would give a more even pressure to the process, but these are ideas that I hope someone else might have a go at. I've never tried this for an F-hole guitar (mainly because no one has ever asked me to), but I would be delighted to hear from anyone who tries it, or for that matter anyone who has some thoughts on the ideas in this article. |
