Gluing bridge

I am constantly trying to improve my methods. The bridge on an GammelGura made in 2019 had suddenly come loose almost completely, and I had reason to think a bit about how to make the gluing of the bridge better. That the bridge came loose from the string pull was due to two things. The bridge was original and not a new replica, it was narrow and rather weak with airy carvings at both ends. The bridge had also come loose and re-glued earlier, which had torn the top under the bridge. To repair it, I milled a shim under the bridge, but it had finally come loose in the glue joint against the bridgeplate in spruce when the small surface of the bridge that was glued to the top had come loose at the back.

To begin with, I reviewed my way of using hot hide glue. After a chat with Per Marklund, I was verified that he used a thinner mixture of the hot hide glue than the one I usually use, I had started to think in those paths myself. The advantage of this is that the glue becomes thinner and can penetrate deeper into the wood. Now I make the standard glue a little thinner and have also filled a small plastic bottle next to the glass jar in the heater where I dilute that mixture with half water to a very thin glue. When I glue, I wet the wood first with the thin mixture in a few rounds, then on with the standard hot hide glue. When I tested this on a fretboard, you could see that the thin glue was pulled into the wood, it quickly became dry spots here and there. The same thing happened with the underside of the bridge. I think the joint will be stronger that way.

After some thought, I realized that the best solution to repair the top under the bridge is to make a hole in the top where the top damaged. The hole is then filled with a thick piece of spruce that both fills the hole and acts as a bridgeplate on the inside. The spruce that replaces the hole in the damaged top cannot come loose, as the same piece of spruce is glued on the inside as a bridgeplate. In my case, it is very simple, as the middle of my cross-shaped bridgeplate is as made for the purpose. I had a few pieces of extra thick 8 mm spruce top blank that I could test with. I'm lucky to have friends with spruce logs and band saws!

I used one of my ongoing GammelGura tests of the new methods. Also on this one, the bridge was ornate and the bridge had come loose and been re-glued. Unlike the previous GammelGuran, I made a replica bridge in fresh rosewood, on the one that the bridge came off, the bridge was made of softer fruit-wood. There was a lot of damage in the top under the bridge, in addition to two rows of stringpin holes and screw holes.

The bad area was marked out and a hole was cut. The hole should be made as small as possible and always smaller than the bridge on top, what was left of the top under the bridge was not completely flawless, but completely OK.

The middle part of my cross-shaped bridgeplate was formed in 8 mm thick spruce. When placed in place on the inside of the top, it was easy to mark the hole with a sharp pencil.

The excess spruce around the marking was milled down about 3,5 mm (a little more than the thickness of the top) in my simple little milling jig. The spruce is attached with the masking tape & superglue method to the workbench, which works as a double-sided adhesive tape.

This is how the combined bridgeplate and puzzle piece that will fill the hole in the top became. After a bit of polishing, I was able to get a good fit in the hole. This construction makes it impossible for the "shim" under the bridge to come loose!

The spruce slab was leveled against the top with a small, handy aluminum profile with self-adhesive sandpaper.

Inside, the cross-shaped bridgeplate was formed with its two wings, all three pieces were then glued with my new method with hot hide glue.

When I glue, I make sure to tighten properly with strong clamps! I have also started using greaseproof paper as often as plastic wrap as a release agent. The advantage of the greaseproof paper is that it can be reused and that it is more stable than the thin plastic that can sometimes be squeezed into a joint by mistake. Greaseproof paper also lets moisture through, and the warm hot hide glue dries faster.

After gluing, a final leveling was done by scraping with a backed razor blade.

Here we see the bridgeplate at a later stage, where plugs and rounds in bubinga have been glued in.

In the future, I will spend more time getting loose bridges without damaging the top, it is one of the most important details when disassembling the guitar, I have realized. My heat lamp should be allowed to heat the bridge and the glue properly before I start to loosen, it takes a little longer, but it's worth it. On the guitar where the top is already damaged under the bridge, I will use the method I describe here for the best possible durability.

Ongoing batch

After many regular but time-consuming repairs, I have only worked with the ongoing batch this week. I had already milled carbon fiber rods in all five instruments and glued all the bottom braces. This week, all tops got their braces, the bodies have had cracks glued and repairs done. The top braces are not completely ready, they must be shaped and polished for the best sound as well. The harp guitar needs a bunch of cleats and the neck heel a round rod in birch.

Now that I have basically come halfway, I will choose two at a time to finish and put the most work on one. During certain steps, such as when the bottom is glued or the guitar is varnished, it is good to have a second instrument to work on.

I never took any pictures of the original guitars this time, but here are pictures of the current situation. Two Europeans, two Levins and a harp guitar. Klangolan got an X-bracing, the harp guitar a special ladder bracing.


An advanced repair that I could not refuse was a very special Guitarra, a Portuguese instrument with 12 strings with tuning screws that resembles a medieval torture instrument. I had seen these in the picture and was curious about how and why they work.

It came in a coffin case that was large enough for the instrument, but which was not in the best condition. Fortunately, the guitarra itself was in better condition than the case. After an inspection, I could see that one of the 12 tuning screws did not work. They had tried to fix it, but put a short tuning screw instead of a long one, which meant that three tuning screws could not be used, the wrong short tuning screw was removed. The bottom had two solid cracks in the fine rosewood wood. The top also had a crack, but it was repaired in a perfectly OK way. The rim and neck were whole. The bar frets were not in the best condition. The nut had been tampered with, there were 20 (!) string notches. The bridge, which is probably original, however, had 12 string notches in four pairs, if I remember correctly. The "saddle", in the form of a long bar fret in brass, may have been replaced. The biggest problem in addition to cracks and loose braces, however, was that the top had sunk in at the sound hole and the thick fingerboard in fine rosewood dipped down on the top.

The guitar had a large label inside with the text:

João Miguel Andrade
Corda Instrument Manufacture and Armazem
Guitars, Brilliant Violas, Cavacos
Rua nova de Trinidade
Sole Agents for Great Britain and Colonies
Alban Voigt & Co.,
14, Edmund Place, LONDON, EC

"Fabrica e Armazem de Instrumentos de Corda" means factory (or manufacturer) of instruments and strings. "Guitarras, Violas braguezas, Cavacos" are Portuguese guitars in the style of Braga and probably cavaquinhos. "Rua da Trinidade" is Trinidad Street.

In addition, there were some notes on the label, probably made by the previous renovator.

Online I found a similar one but more lavish guitar with the text:

“I believe this guitarra is from the 1890's because of the book, A Complete Method for Portuguese Guitarra by Havelock Mason, which was supposedly published in 1892 or 1895. This method book was published by the importer of this guitarra, Alban Voigt & Company, at a time when the Portuguese guitar was popular in Britain. The tuning used in this book (gG, bB, dD, gg, bb, dd) was called the natural tuning in 19th century Portugal – as opposd to the fado tuning (dD, aA, bB, ee, aa, bb). It is a tuning which adapted the 10-string 6-course English Guitar tuning to the 12-string 6-course Portuguese guitarra. This method is a basic instruction book which also contains a few fados as well as a polka, a bolero and a romanza. ”

One can assume that this one is also made in the 1890s, bare frets is a tell that it is old. Virtually all manufacturers, except Martin, who used bar frets into the 1930s, switched to the T frets in the years before 1900. The T frets (T stands for "tang") with "barbs" were patented per se by Clinton F Smith first in 1929 in the US, but I have only seen bar frets on 1800th century guitars.

The tricky tuning screws are actually an invention that was originally used on English citterns. They are called, among other things Preston Tuners. There is evidence for similar tuning screws as early as 1766, they have also been used on the German waldcitter, but are now only used on the Portuguese fado instruments. You can still buy similar instruments and tuning screws, including on FolkReps site. One thing you need is a special tool to make loops at both ends of the strings. Luckily I already had one that I bought many years ago from FolkReps. You can also buy new Preston tuning screws of the same type. I also modified a clock key to save my fingertips and ease the tuning.

Well. This is not a regular GammelGura renovation, it would just be playable again. The customer wants to test a different tuning with five double strings tuned GCGCG, which meant that there were enough tuning screws for the 10 strings.

The bottom came loose without problems. All braces at the bottom were more or less loose, the hot hide glue does not adhere as well to rosewood as to spruce. In the top, the bracing were fixed as they should be. When the bottom braces were completely loosened, the bottom was in two parts, one of which had an almost continuous crack. The cracks were cleaned and glued together with a long spruce cleat on top. For once, I antiquated the new wood with potassium permanganate, which quickly oxidizes wood and makes the spruce darker and orange-yellow, which matched the center stick. All bottom braces were re-glued with hort hide glue.

All the braces in the top were firmly glued. But the top brace was loosened so that I could wet the top and put the top plus fretboard in pressure for XNUMX hours. It worked well, after pressing the top, it was not as sunken and the fretboard straight. To keep the top in trim, made a new and stronger brace with a curved underside, the tips of that brace were secured with two abutments that were glued to the inside of the rim. In addition, I made an A frame to reinforce the area around the sound hole.

The bar fret was not in good condition, but luckily they were thin. My old Dunlop brass frets, with a little thicker tang, fitted perfectly in the fret grooves. I also checked the location of the frets, and they were in just the right place. The very thick fretboard had an impressive radius. My tool for bending frets to a radius maxed out at about 6″ radius, but the fretboard had about 3″ radius! With a few feeler gauges, I was able to adjust the fretbender so that I got a circular fret string with almost two full circles at 15 cm in diameter, which corresponded to the radius of the board.

The tuning screws were oiled and the bottom was glued back without any problems. With a straight fretboard, the geometry towards the bridge was perfectly OK. I had to shim up the feets of the bridge a couple of mm to adjust the string height correctly. The top also had a radius, it took a few hours to shape the bridge feet against the top. A new bone nut was made, a brass fret notched for five double strings replaced the original saddle fret.

A charming detail was the 11 mini-string pins in bone that sat on the rim at the back as a bracket for the strings' loop. All of them were more or less loose, and I glued them with hot hide glue. The missing "string pin" was made of a piece of bone with the help of my little cordless drill that had to act as a lathe. It took at least an hour to do it. At one point, before I glued them down, I discovered that another one was missing. Unlikely enough, I could find it on the floor, even though I had used with the vacuum cleaner before…

As usual, small pieces of frets were temporarily mounted along the middle of the fretboard and the guitarra was strung up. It was a bit new that each string had to be made before you could string it on. The tool for making loops was not directly of the highest quality, but it was possible to use it. The problem was the thickest spun strings, if I made the loop too tight the string broke. I also managed to cut off the wrong side of the string at the last loop, the string itself and not the abundance! I found another string with the same thickness and was able to string up all 10 strings. The neck was allowed to settle in a couple of days during the vibration session.

With my new Stewmac jig, the guitar was fixed in the position it has with tensioned strings, the temporary band pieces along the middle were loosened and the fretboard was leveled and given a little relief. The board was then banded with Dunlop brass tape. I could not use my usual belt press as I did not have a suitable 3 ″ bracket, I had to work with hammers, super glue and clamps. It went perfectly OK even some of the bands had to be redone as they did not sit properly as they should. The guitar got another pass in the Stewmac jig for crowning and polishing the bands.

The “interesting tuning” made it difficult for me to play it, but my violinist Björn Sohlin was fascinated by it and improvised medieval melodies that matched the appearance of the tuning screws!

The tuning screws worked well, of course it takes a while before the two loops stretch out and the tuning becomes stable. The clock key made the tuning easier, but it was also possible to tune with the fingertips, even if it was harder to do.

The whole instrument, as big as a larger mandolin, was given a couple of rounds of spirit varnish which was then matted down with steel wool. The small format puts a limit to the bass tone, but with this tuning it had a good timbre and long sustain.

Pythagoras' theorem

When I measure the intonation point on the nut from the back of the first fret, you get a right-angled triangle with the caliper with a longer hypotenuse. By using Pythagoras theorem, I realize that you should add 0.05-0.1 mm to the measurement I got at the measurement, as the nut to be manufactured is always higher than the one mm that is the height of the fret piece.

Now I will measure as in the picture, or possibly cut down one of the tips of the caliper (have to think about that). The question is whether the small difference is audible. Not wrong to do it more right, anyway!

Repair of a Söderman lute dated 1920

Lately I have been doing a lot of repairs. Not all are worth mentioning, but some are interesting, such as this Söderman lute. It was to be repaired to a playable condition, but when it turned out to be an unusual and early Söderman, I put my soul into it. It was a half GammelGura conversion / renovation where none of my inventions or improved bracing patterns were used. On the other hand, it got a carbon fiber rod in the neck, a new fingerboard with EVO straps and a new bridge with intoned saddle. The tuning screws were replaced with better, time-typical originals that have been left over after previous renovations.

When I first got it in, I thought someone had been inventive and fitted three extra tuning screws by sawing off the head. Looked amateurish. It came with a Levin tailpiece in sheet metal from about the 1940s and a clumsy wooden saddle on top of the pinbridge. Someone had been on it.

With the bottom off, I got a nice surprise in the form of a dating to 1920 (the oldest Söderman I have seen so far) and also numbered to the 198th instrument. When he, according to an article, started building lutes and guitars In 1917 he kept a high tempo, one can perhaps count off the 50 violins he started building when he was 12 years old from the number 198. He have manufactured at least 2500 guitars and lutes until 1971, violins and mandolins not included.

After inspecting the details, I realized that the extra tuning screws were original and that Söderman himself mounted them. The wooden nut was clearly original for nine strings, the bad and ill-fitting tuning screws that were on were replaced. No wonder, really, he was both inventive and did not follow ordinary conventions. I think this was one of the first attempts to build a lute with more than six strings. Later similar lutes from the 1930s have the same short string length, 59,5 cm, but only eight strings with two tuning screws intended for a mandolin. It undeniably looks more professional, and he may have come to the conclusion that a doubled G and e string is enough.

Unlike later lutes, this one had solid birch at the bottom and sides, later lutes have a three-layer laminate.

The fretboard was made of bad walnut with a 45 degree run-out and could not be loosened in one piece. An odd detail was that the end of the fretboard was floating on the top with an approximately 3 mm thick fitting piece between the top and the underside of the fretboard. The shim was carelessly made and only covered the middle of the fretboard, leaving a large air gap around it. The bridge, also in walnut or something similar, had cracked along the length and been glued together. I chose rosewood blanks for a fretboard and a bridge with a similar color as the originals.

The milling for the gluing of the carbon fiber rod in the neck was done with my usual milling jig which works well even with the neck glued to the body.

The fretboard was given a 16″ radius and the decorations in celluloid were transferred from the original board. Gottfrid had fitted the decorations nicely, a little better than what I managed in the new board actually! The beautiful shape of the end of the fretboard was copied, but I did a better job with the fitting piece between the top and the underside of the fretboard. The bridge got a similar shape as the original, but also an intoned saddle.

The top had sunk in around the sound hole and the neck was bent. With the fingerboard off, I could bend the neck with heat straight before the carbon fiber rod and the fingerboard were mounted.

Under the bridge, the top was broken, and a shim was glued in.

In order to be able to reshape the top and repair the long crack on the bottom in the best way, the bottom was loosened. In addition to the dating and a short pencil text at the back, it was also revealed that the good Gottfrid had miscalculated the placement of one of the braces, the holes for the stringpins rubbed along the edge of one of the braces. One brace was replaced and a bridgeplate in spruce with hard buttons around the stringpin holes was mounted.

All bottom braces were more or less loose, however, the braces in the top sat well. To be able to straighten the sunken top, all braces were loosened. The braces in the top had the wrong bend in the contact surface towards the top and had to be reshaped in the other direction to push up the top, they at the bottom had retained their shape and only needed to be re-glued.

The crack in the bottom was glued and secured with a long cleat. A very practical tool for cracks in a bottom is a special clamp used by violin builders, an "edge clamp", with which the edge can easily be forced together during gluing.

Neck and neck blocks are in one piece on a lute like this. It is not possible to change the neck angle in the usual way. But it turned out that the angle with the top in the right shape, a new lower bridge and a straight neck was perfectly OK if the rim was squeezed together a couple of mm lengthwise. A special jig in the form of a threaded rod through the endpin hole was used to pull the rim together to the correct shape for the neck angle before the bottom gluing. Normally I fasten it with a screw on top of the neck block, but here a small wooden block was glued on for better alignment of the bottom with the rim. With the help of a metal beam, which was attached on top of the fretboard with clamps at the neck and a metal screw through one of the stringpin holes, I was able to fix the angle between the fretboard and the bridge top at the bottom gluing. When done, the bottom had to be cut to the rim a few mm in length.

As the bottom shrank in width, it was impossible to get an exact fit to the side, especially when the side was squeezed lengthwise. In my jig, I could have forced everything in place, but then you build in tensions that are guaranteed to result in new cracks over time. I hesitated if I would mill a binding around the bottom, but I let there be a small gap between the edge of the bottom and the side.

The kerfing on both top and bottom was very thin, to get more glue surface for the bottom I glued on an approximately 2 mm thick kerfing in linden from a wooden shutter. The bottom was glued with hot hide glue with urea for longer opening time. This time I applied the glue along the entire edge and heated the glue as I forced the bottom with a heat gun. On the neck and bottom block, I use as before normal hot hide glue, you have time to clamp the two ends of the bottom before the glue coagulates if you heat the wood properly just before gluing.

The most special thing about the instrument is otherwise the three extra strings. I added Newtone Heritage 0.11 strings and doubled the unspun e and b strings with loose matching NH strings. The octave string on G had to be a 0.09 string that I happened to have at home. Two grooves in the string pin holes for e, b and G strings with solid stringpins were filed to make room for two strings in one and the same hole. It's a little trickier to re-string with two strings in one hole, but it's not a big problem. As the strings end up in each edge of the stringpin hole, you get the approx. 3 mm spacing between the strings you need.

The wooden nut original was copied, but made of bone and not wood. The intonation for all strings was done only on the saddle according to the usual 12th fret method.

The lute was given a coat of spirit based varnish on the entire instrument, fresh wounds and dings were dampened with potassium permanganate and Herdin's "Carl Johan" water based stain before varnishing. The super-shiny new varnish was then dampened down with 000 steel wool.

The fretboard was fretted with EVO tape after a first sanding of relief of the fretboard. The crowning of the frets was also made to a relief of 0.15 mm. When I use my new jig from Stewmac, not much metal has to be ground down on the frets.

It turned out that the lute worked very well with the extra tuning screws. An extra bonus is the three skewed treble tones you can pick up on the three short extra strings above the upper saddle! I quickly learned in what order I had to to tune. It was a little more difficult to tune the extra tuning screws, but even there I quickly learned a good method. The sound is like a twelve-string guitar, but with a firmer base and less jingle & trebles. Since I'm not fond of all the rattles from a 12 string guitar, I think it sounds better with 9 than 12 strings. It sounds and plays really good!


GG191, Levin 1926

The week before the fair in Umeå, I realized that I needed an GammelGura to sell. I chose one of all the parlor guitars I have hanging on the wall in the apartment, a Levin 1926 in good condition so as not to have to spend a lot of time on repairs.

The old parlor model with a slotted head was on its way out in 1926, the youngest I have seen I think was from 1929. Around 1930 the old models were phased out, and a flat head replaced the slotted one. I think that during the transition, they made sure to use up all the old material that was in stock, including the rosette with mother-of-pearl as on this one. In the early 1930s, a layer of extra fine flamed birch was also used on relatively simple parlor guitars with flat heads and floating bridge. They are very beautiful, but basically cheap guitars.

This one also had a celluloid binding around the bottom which became common on the newer models.

Well. I worked hard with it to get it ready on time, there were a few extra evenings in the shop and smart planning so that glue and varnish would have time to dry and the neck settle. Now, of course, the neck was in hard birch and did not give in after more than a couple of tenths at the 12th band with strings at tension a couple of days.

I took some pictures of the gluing of the bridge. The bridge was glued, and the top was broken under the bridge, so I glued a shim in spruce over almost the entire surface so that the stable would sit properly. Nowadays, I make four plugs, one hard in birch on thick E and end wood in spruce for the other spun strings. The treble strings b and e are dampened with the soft spruce in the bridgeplate and top. My new favorite tool, a toothbrush on an 8 mm birch rod, was used to clean off excess hot hide glue.

Bottom and top, before bottom gluing the bottom, with mounted K&K mic.


Instead of fish glue to glue the bottom, I tried for the first time to mix the usual hot hide glue with 10% (of the volume of the hide glue granulate) with urea to extend the gluing time. Fish glue is good, but genuine hot hide glue with urea is better. By spending some time heating the joint and the glue at the bottom when gluing, it worked beyond expectations. It got a little sticky the first time when I did not apply the glue first but applied with a spatula as it goes. It becomes less smudged by first applying glue along the entire edge and then heating the glue liquid again as I did on later bottom glues. Very good!

GammelGuran was sold to a very satisfied customer at the fair in Umeå 🙂

Apprentice with Levin and tenor

I recently had an apprentice in the shop. It was a nice week as usual, this time we worked with a Levin parlor from 1912 and next to a newer Orpheum tenor guitar from around the 1950s. I did not take many pictures while we worked (the apprentice took all the more!), But I can report the end result. Levinen, GammelGura 190, was almost finished and needed final adjustment when he left, but I had to add a few more days to the tenor guitar.

The Levin was in perfectly OK condition, just a little more cracks in the top than usual. The job went as it should, the only problem was that the neck in soft poplar did not get the chance to shape itself with tightened strings before we fretted. The result was that the neck bent up close to the nut when everything was ready, and I had to redo the grinding of the fretboard and fretting.

The original tuners were in good condition and were allowed to remain, the plaque on the back of the head had a hint left of the printed text. A new bridge and a new fretboard with EVO frets were manufactured. The nut and the segmented saddle were intoned. A K&K mic mounted. The bottom was really shrunk and a center stick in rosewood was glued in. Was then sent in a new Gator 3/4 case.

The tenor guitar was a bit unusual, in mahogany and basically a dreadnought body with a narrow tenor neck and four tuning screws. Fine rosewood in both fretboard and saddle. Maybe a little lame in the design language.

The clumsy and screwed-on bridge was reused after a bit of planing, the screws were replaced with pearl dots (the bridge was actually a regular guitar bridge, the two outermost string holes had been filled with white celluloid dots!). The neck was reset, it got a carbon fiber rod in the neck, and the flat fretboard was given a small radius and new frets. The ugly pick guard was allowed to remain at the apprentice's request. The top had been deformed, the bottom was removed and the original bracing in the top were reshaped and glued back to push the top up after soaking and clamping to flatten the wood. The bridge plate was changed to spruce with plugs and bubinga rounds, but no major changes inside. A nut inttonation and a segmented saddle were done. The original tuners were lousy and were replaced with new ones.

The bottom and side were a three-layer laminate with thin mahogany on the outsides. At the bottom, it was easy to split off the top layer of mahogany down to the reasonably thick middle part of the maple. Became both lighter and better sound in the bottom piece!

The biggest job was the bottom celluloid binding. It had an unusual and very light tortoise color that is no longer available. Instead, I mounted a wooden binding that was similar to the old celluloid. A K&K mic was mounted. When it was finished and vibrated, it sounded much better than before the renovation. Sent in one Thomann Elite Case Classic 1 case.

The bottom and top were thinned on the Levin in the usual order, both got a couple of rounds of spirit varnish on top of the old varnish.

Some pictures while we worked!

For sale: Old GammelGura

An early GammelGura, “The rose” number 28 from 2014, is for sale. It has one of the first oversaddle intonations I made, but lacks later inventions. A very beautiful guitar, especially the stable! Unique in shape too, all indications are that it was built by a solitary builder and not in a conveyor belt. It comes with a used Gator 3/4 case and mounted K&K mic.

When I got it in, I thought I would have to work with it as it had been in a dry storage room for a year or so, but to my surprise, everything was just as it should be! Correct string height and no cracks. The only thing I did was to file off the tops of the bands that protruded a little bit and put on new strings 🙂

The sale price of it is SEK 8700, the same amount as it was sold once upon a time. Take care!