| Pages:
1
2
3
4 |
jdowning - 6-11-2013 at 11:50 AM
Oddly enough, the on-line chemists generally state that the blue precipitate of copper hydroxide is not only insoluble in water but also will not
dissolve in excess sodium hydroxide. One source, however, adds an exception - "unless the NaOH solution is 'very' concentrated".
I have now prepared a batch of the 'silk reagent' by adding a 40% solution of sodium hydroxide (i.e. 40 grams dissolved in 100 cc water) little by
little to the copper sulphate/glycerol solution until the dense copper hydroxide precipitate first formed is dissolved to a clear bluish/purple
coloured liquid - a rather attractive colour. So presumably a 40% sodium hydroxide aqueous solution (or greater percentage) is the very concentrated
level required to dissolve the copper hydroxide precipitate. Testing the solution with wide range litmus papers indicates the prepared 'silk solvent'
solution to be about pH 13 - so quite caustic and about equivalent to the pH value of the 40% sodium hydroxide solution (for comparison pure water is
about pH 7)
Note that on test, the silk waste did not dissolve in 40% sodium hydroxide solution at room temperature.
The current test is to see how much of the silk waste fabric will dissolve in the solvent. The first batch of 0.5 gram silk waste dissolved in 10 cc
of the solvent in about 30 minutes at room temperature. A further 0.5 grams of silk waste has now been added to the mix. This may be about the maximum
silk to solvent ratio (1:10) for complete dissolution of the silk (as reported by some researchers) - but we will see how it goes.
jdowning - 6-12-2013 at 11:52 AM
Checking through my notes on Japanese silk fishing leader gut mentioned earlier in this thread I find the following comment -
" the Japanese 'gut' is made from spun silk covered with a coating of dissolved liquid silk mixed with several chemicals - or bonded with gelatin".
Unfortunately I did not note the source so will have to keep searching to try to find it again but it likely gives no further information about the
dissolved silk. The gelatin or glue binder is presumably similar to the 'conventional' agar agar (seaweed gelatin) binder previously noted in this
thread?
jdowning - 6-13-2013 at 12:21 PM
While the waste silk is being dissolved in the 'Copper Hydroxide' solvent - thinking ahead a bit.
The dissolved silk solution will be strongly alkaline so cannot be just applied to a silk string without risk of the string being also dissolved or at
least weakened over time (?).
To neutralise the alkaline solution to a point where it will not attack the silk fibroin of the string itself, acid must be added to the solution. As
I am only experimenting with chemicals that anyone can purchase 'off the shelf' from local stores the chosen acid is household strength vinegar (5%
strength Acetic Acid).
The acid was added - little by little - to a small sample of the solvent used for preliminary tests to dissolve the waste silk (so contained dissolved
silk). As the acid was added, the solvent changed colour from deep purple to light blue and then began to 'fizz' or effervesce as bubbles of a gas
were given off. Further addition of acid resulted in eventual ceasing of the gas emmission and sudden clarification of the solution. At this point the
solution tested around pH 7 or slightly less (acidic) so is considered to be neutral. The safe range for fibroin is around pH 4 to pH 7.5 (see page 4
of this thread)
Acid + Alkali = Salt + Water (as I remember it). In this case presumably the salt (in solution) is a complex of copper acetate (from the copper
hydroxide portion) and Sodium Acetate (or Ethanoate) from the excess sodium hyroxide portion? Presumably the gas is Carbon Dioxide?
For information, the attached images show the dissolved silk solution before and after adding the acid. Image A is the neutralised dissolved silk
solution and image B - for comparison - is a neutralised sample of the Copper Hydroxide solvent, pale blue and perfectly clear.
Note that the colour of the neutralised dissolved silk sample is close to that of the Japanese silk fishing 'gut' samples previously posted.
[file]27035[/file] [file]27037[/file]
jdowning - 6-18-2013 at 12:10 PM
Another solvent for Bombyx Mori cultivated silk fibroin is that described by researcher Akiyoshi Ajisawa
https://www.jstage.jst.go.jp/article/kontyushigen1930/67/2/67_2_91/_...
The solvent is made from a solution of Calcium Chloride to which Ethanol (Ethyl Alcohol - the stuff in alcoholic beverages) has been added - the
alcohol reported to greatly increasing the solubility of the silk fibroin. Calcium Chloride is readily available as a chemical used to adjust calcium
levels in swimming pools. Pure Ethyl alcohol is less readily available and costly but is otherwise readily available as a solution in water in
alcoholic drinks such as Vodka and Gin.
Chemists give their chemical mixtures in molecular proportions or moles. In this case the optimum proportion of the silk solvent is given as a molar
ratio of 1:2:8 (i.e. calcium chloride: ethanol: water). Expressed as weight (or rather mass) proportions, the ratio becomes 111: 92: 114 grams
(calcium chloride: ethanol: water).
Vodka as a drink is sold locally as a 40% ethanol/water by volume solution. If I have done my sums correctly (as a non chemist - so great margin for
error!) - a mixture of 46 grams of Calcium Chloride dissolved in 100 cc Vodka (don't want to waste too much of the stuff on experiments!) should give
a solution close to the optimum proportions required for the Ajisawa silk solvent (a bit less ethanol at 31.6 grams compared to the ideal 38.3
grams).
At room temperature the dissolution of degummed silk is reported to be very slow in this solvent. However at 60°C the dissolution is reported as
being complete within an hour.
Unfortunately, I had no success in trying to dissolve degummed Bombyx Mori silk using this solvent. Perhaps my calculations were in error or perhaps
the proportion of ethanol in the mixture is more critical than first thought. I did also try using methylated ethanol (ie 'methylated spirit' or
ethanol poisoned with wood alcohol to discourage consumption of the stuff) in the required proportions without success.
Note that this solvent is reported to dissolve only degummed Bombyx Mori cultivated silk - not species of wild silk fibroin - on which it is reported
to have little effect.
(see research report "Dissolution of 'Philosamia ricini' Silk Film: Properties and Functions in Different Solutions" by Y. Srisuwan and P.
Srihanam)
http://scialert.net/qredirect.php?doi=jas.2009.978.982&linkid=p...
jdowning - 6-22-2013 at 10:49 AM
I was able to dissolve 0.6 grams of silk fabric waste in 10 cc of the prepared 'copper hydroxide' solvent previously reported. However, to neutralise
the solution required the addition of about 50 cc of 5% Acetic Acid - making a dilute solution, probably too 'weak' to be useful for string making as
it stands.
Perhaps the solution might be concentrated by heating it to drive off the water content - but this approach has yet to be tested.
To short circuit the path in making silk solution from scratch, I have purchased some ready made cosmetic grade silk fibroin powder for further
experiments. This is the finest grade of powder (Silk Amino Acid) sold by the New Directions Aromatics Inc. company and is soluble in water - although
the degree of solubility has yet to be established.
The powder costs just under $15 for 100 grams - more than enough for initial experimentation. The powder is triple wrapped in sealed plastic bags and
securely packaged in a padded cardboard box for delivery by courier delivered within 7 days of placing the order. Included was even a 'thank you' card
from the company - now that is good customer service!
Some initial testing with small quantites of powder confirmed that the powder could be made into a creamy paste or a solution in water. The powder
dissolved more effectively in Vodka (40% Ethyl Alcohol and water by volume).
The powder has a 'peppery' aromatic smell and in solution is a clear pale yellow in colour.
So - at this point I am not sure if the use of silk fibroin powder will successfully lead anywhere positive in this silk string making investigation -
but no harm in trying.
jdowning - 7-4-2013 at 12:05 PM
I have just come across another snippet of information that may (or may not) be of some significance in the quest for an historical binder for
workable silk strings.
Earlier in this thread forum member danyel suggested that isinglass (a pure gelatin from the sturgeon (and other species of fish) swim bladders rather
than gum arabic) was the binder component given in the Kanz at-Tuhaf commentary on silk string making.
More recently (page 8 of this thread) the method used by the Japanese for making 'monofilament' silk fishing leaders was discussed as a possible
application for silk instrument strings. The binder for these strings - as mentioned by Humphries - was likely a mixture of animal glue and a seaweed
derived gelatin known as 'agar' (or 'agar-agar').
Interestingly my copy of 'Thorpe's Dictionary of Applied Chemistry' - a wonderful but now relatively out dated (4th edition, 1937) encyclopaedia -
under the entry 'Agar- Agar' (Vol 1, page 162) - describes the material as "Bengal or Japanese Isinglass" and " It is employed as a 'size' (diluted
animal glue) substitute - its gelatinising power claimed to be greater than that of gelatin".
It is also used as a size for silk and other fabrics to add gloss as well as a glue (or gum) - so sounds promising as a poterntial binder for silk
strings.
Agar is readily available as a health food additive so a sample quantity must now be obtained for testing! Costs around $7 for 50 grams in powder
form.
jdowning - 7-27-2013 at 05:44 AM
The sample of Agar powder has finally arrived - so investigation of a Japanese style silk string binder can now proceed.
Agar is an interesting material with a number of applications - including a food thickening agent (a vegetable substitute for animal gelatin) and glue
additive. Agar is considered to be a vegetable gum (like, for example, Gum Arabic). It is compatible with proteins so presumably will combine with the
sericin in raw silk filament.
Agar is made from a variety of red seaweed that occurs worldwide - although the the bulk of Agar on the market is of Japanese origin. The properties
of Agar vary somewhat according to the strain of sea weed from which it is extracted. However its unique property is that it forms a firm gel at low
concentrations (0.5% to 8% by weight) in water dissolving in the water at boiling temperature (100°C). On cooling the mixture forms a gel at around
30°C to 40°C. However, unlike animal gelatins, Agar must be re heated to about 100°C before melting to a liquid. This melting/gelling cycle can be
repeated.
The sample available for testing is food grade sold as a 100% pure "Vegetarian Substitute for Gelatin" - a fine cream coloured, odourless powder.
As a preliminary test 1 gram of the powder was dissolved by heating to boiling point in 100 cc of water (i.e. 100 grams). The sample was heated on a
water bath so its maximum temperature in practice was around 90°C. On cooling the fluid started to gel at around 35 °C and became a soft jelly at
room temperature (21.5°C). On reheating on the water bath, the gel became liquid again at about 80°C.
It is assumed that increasing the Agar concentration will increase the gel strength and firmness - but this has yet to be confirmed. Note that an 8%
concentration is used for making dental impressions for casting - so the gel must be quite hard yet still flexible at this concentration.
The concentration of Agar added to a glue or gum binder (to make it more flexible) has yet to be determined. Note that Agar is apparently used as a
component of glue used for manufacturing plywood. Also it has more recently been mixed with UV hardening epoxy glue - as an adhesive used for
laminating flexible LCD touch screens for computers.
[file]27261[/file]
jdowning - 7-27-2013 at 11:06 AM
As the initial trial using a water bath failed to heat the Agar to boiling point, the gel was remelted and poured into a muffin pan placed directly
on a stove hot plate. This brought the Agar to a simmering boil (measured at 100°C) where it was left boiling for 5 minutes before allowing to
cool.
At around 40°C the surface of the Agar fluid started to gel and by about 35°C the gel had become firm enough so that the pan could be turned on its
side.
At room temperature the gel felt like a firm soft rubber on the surface but it was possible with increased pressure to penetrate the gel with a
fingertip (see the hole in the centre in the attached image). The interior of the gel at this concentration was found to be moist with water present.
This would likely not present much of a problem when the Agar gel is mixed with animal glue to modify its elasticity but that will have to be tested.
A second test will be undertaken at an Agar concentration of 8% (by adding another 7 grams of Agar powder to the 1% stuff just tested and bringing
everything to a boil). By all acounts this should produce a very firm gel - but we will see.
It will be interesting to see how well the Agar might consolidate with silk sericin. It is difficult to degum silk completely as there is always some
residual sericin left combined at a molecular level with the silk filament - so the hope is that with even degummed silk the Agar will 'stick' well as
a binder component.
jdowning - 7-28-2013 at 11:40 AM
Two more concentrations of Agar gel have been prepared - one at 4% by weight and the other 8% by weight - just for experience. Both samples were again
heated on a stove hot plate stirring as the temperature increased.
The 8% solution became a rather stiff glutinous bubbling mass with temperature not exceeding about 85°C.
The 4% solution was a little more fluid but still quite viscous - maximum temperature on heating did not exceed about 95°C.
On cooling both samples gelled at around 37°C. At room temperature the 8% solution had formed a firm but flexible rubber like gel that could not be
penetrated with finger pressure.
The 4% solution formed a stiff gel that could just be penetrated with considerable finger pressure.
Both gels - although firm - could be readily cut or broken into pieces.
At these concentrations the Agar was not fluid enough at maximum temperature so would be unlikely to penetrate the fibres of a tightly twisted silk
string.
The workable concentrations of Agar would therefore fall in the range of 1% to 4% - say a maximum 2%? As the firm gels are easily cut this may confirm
the need to mix the Agar gels with stronger animal glues to form a strong yet flexible string binder.
Humphries (previously posted) states that the Japanese silk string binders were a mixture of animal glue and seaweed derived gelatin (Agar?) - but
does not give information about the animal glue used or the proportions of glue to gelatin.
Animal glues such as hot hide glue deteriorate in strength at temperatures above 60°C so the Agar gel would first have to be melted (at about 80°C)
and then added in fluid state to the hot glue and stirred in.
Agar gel can be reconstituted from a higher to lower concentration by adding water to the gel and reheating to boiling point so the samples already
prepared can be used to determine the optimum concentration for adding to hide glue to obtain a strong yet flexible adhesive as a silk string
binder.
Presumably the gel concentration can also be increased by boiling a weaker solution to reduce water content?
More experimental work yet to follow!
jdowning - 7-30-2013 at 12:15 PM
For a futher test of the adhesive and penetration properties of the Agar gels, the 1% and 4% gels were reheated until liquid and then brushed hot onto
two layers of fine cotton fabric to glue them together. The fabric was allowed to dry (it took several hours in each case).
The dried samples of fabric were stiff yet perfectly flexible (like 'starched' fabric. Not surprisingly starch paste being a gum equivalent like
Agar). The 1% sample being more fluid worked best at penetrating the fabric- the 4% sample was very viscous when melted and tended to gel on the
surface of the fabric when applied.
Trying to peel the two glued fabric layers apart - the 1 % gel showed stronger adhesion than the 4% gel. Adhesion was not as strong as would be the
case with hot hide glue - but lacked the hard brittleness inherent in dried hide glue.
This being the case it might be worthwhile first testing the Agar gel alone as a silk string binder - combined only with the residual (and brittle)
sericin coating on the silk filament.
narciso - 7-30-2013 at 10:57 PM
Agar seems to be a weaker but elastically more supple binding agent than animal glue?
So is the idea that one can create an optimal string binding agent by mixing the two- perhaps a bit like mixing chalk and soap to optimize peg
lubrication?
jdowning - 7-31-2013 at 04:24 AM
That is the general idea narciso.
As Alexander Rakov found when using the natural sericin gum that coats raw silk filaments as a binder, the resulting string is stiff (sericin is hard
and brittle when dry). His solution was to condition a string to make it flexible by winding the string around a wooden dowel to uniformly crack the
sericin. I had the same problem when trying Gum Arabic and hide glue as binders and used the same 'conditioning' technique (see page 2 of this thread)
- however I have never felt that 'conditioning' was the way to go - hence my preliminary trials with hide glue modified to make it elastic reported
earlier in this thread.
The Ancient Chinese made their strings (for the qin) using the natural sericin and other additions for their binder. One addition was starch (Rice or
wheat flour) that may have added flexibility just as I am hoping that Agar will. The Japanese appear to have used Agar together with animal glue for a
binder resulting in a smooth flexible 'monofilament' (in appearance) type string. Note that the old Chinese silk strings for the qin were all of roped
construction - three or four strand - so any brittleness of the binder may have been less of a concern due to the high flexibility of the roped
construction. The open strings were also much longer than oud strings.
The primary interest here is to investigate the potential workable tonal range of simply twisted strings (i.e. without roped construction). I would
expect the range to easily cover four courses (as the 10th C ouds were strung - tuned a fourth apart) and might even go down to six courses with high
twist construction.
The trick now will be to determine the optimum proportion of Agar gel to glue (be it sericin, hide glue or isinglas etc.) to provide the required
tenacity/flexibility of binder. The first step will be to try Agar alone with raw silk, the twisted silk being boiled (100°C) in the agar solution to
thoroughly saturate the string with binder - followed by trials using Agar/hide glue and Agar/isinglas mixtures.
The method for adding Agar to the glues will be to first liquify the Agar gel (around 80°C) allow to cool to around 60°C (still fluid) and then add
the prepared glue gel (or liquid glue at 60°C) to the Agar maintaining the temperature at 60°C for use. It will be of interest to note how the
Agar/glue mixture behaves on cooling/reheating cycles.
jdowning - 8-1-2013 at 12:13 PM
For the first trial using an agar binder with raw silk filament (i.e. with sericin gum coating) a short length (37.5 cm untwisted) of 64 silk
filaments was made up. The bundle was soaked in a hot 2% Agar gel and then twisted on the string making rig with 123 twists and 2.4 Kg loading.
In retrospect the number of twists applied was excessive (3.3 twists per cm untwisted string length) resulting in slight buckling of the string along
its length (i.e. the string filaments were starting to 'corkscrew'). The number of twists per cm should have been around 2.0 to 2.5 twists per cm.
Also the hot Agar began to gel on the twisting rig as it cooled prior to applying the twists. Application of heat with a hot air gun helped to remove
some of the twisting irregularities by remelting the gel but a lower gel concentration would likely have worked better.
Nevertheless the twisted test string measured 0.84 mm diameter and was quite flexible - resembling high twist gut even in this preliminary and rough
experimental attempt. Compare with Aquila HT gut here:
http://www.aquilacorde.com/index.php?option=com_content&view=ar...
Presoaking the filaments prior to twisting may not be the way to go so the next test will be first to twist the raw silk filaments and then soak the
twisted string in hot liquid Agar - at 1% concentration to allow (hopefully) full saturation of the string filaments.
[file]27298[/file] [file]27300[/file]
jdowning - 8-2-2013 at 12:07 PM
Test #2
A 64 filament raw silk bundle was first soaked in hot water (to soften the sericin gum), twisted at 2 twists per cm under a loading of 1.4 Kg and
allowed to dry on the twisting rig. The degree of twist was retained on removal of the string from the rig - cemented by the sericin gum.
The string was then immersed in a boiling solution of 1.3% Agar. Things immediately began to unravel - the string twisting and tangling around itself
- so much of the original twist was lost. Nevertheless the string was then replaced on the rig under a 1.4 Kg load, straightened out, wiped free of
excess Agar gel and allowed to dry.
The resulting string was quite uniform and gut like in appearance and flexible yet 'springy' - diameter 0.82 mm. Easily coiled into a small diameter
loop without damage (the coin in the attached image measures 1.8 cm in diameter).
To test penetration of the binder the string was unravelled with some difficulty indicating full penetration and good adhesion of the Agar/sericin gum
- similar to the Japanese silk string test previously posted.
This test must be repeated due to the handling problems that allowed the test string to unwind and lose much of its initial twist when immersed in the
hot Agar solution (and perhaps facilitated penetration of the liquid Agar?)
So far so good.
[file]27308[/file] [file]27310[/file] [file]27312[/file]
jdowning - 8-4-2013 at 11:59 AM
Test string #2 has further air dried and become quite 'springy' so that it is difficult to make a tight knot in the string without causing fibre
separation.
Gut lute strings in the 16/17th C were preserved in oil (such as almond oil) to keep them flexible and moisture free.
Out of curiosity, test string#2 was rubbed with almond oil and became noticeably more flexible so that it was possible to then easily tie a tight knot
without fibre separation.
This opens the possibility for strings made this way to be also used for tied frets on a lute.
jdowning - 8-7-2013 at 02:51 PM
Curious about the apparent reaction of the Agar soaked test string with Almond oil, a small piece of 1% Agar gel was left, a few days ago, immersed in
Almond oil. Checking today, the Agar gel seems to have been dissolved. Surprised at this result, a second sample of 1% gel has been immersed in almond
oil - just to verify this observation.
Test#2 has been repeated with a 64 filament silk bundle being soaked in a more dilute 1% Agar gel solution (at 100°C) and then immediately twisted on
the test rig. While this approach results in complete Agar saturation of the string, the almost instant gelling of the Agar when the string is being
twisted - due to the viscocity of the gel - results in a non uniform string twist and slight longitudinal buckling that cannot be fully corrected with
further application of heat and increased string loading.
The next trial will be to fully twist a string bundle prior to immersion in the hot Agar solution to determine if the Agar will fully penetrate the
tightly twisted silk fibres and bond them into a coherent 'monofilament' string.
NOTE! Further testing to determine the solubility of 1% Agar gel in Almond oil shows that the gel appears to be insoluble - which negates earlier
observations to the contrary.
jdowning - 8-8-2013 at 03:32 PM
I am told by my supplier that all commercial silk filament and spun thread is degummed - or rather partly degummed, as an additional degumming step is
usually required before the silk can be dyed to remove any residual sericin.
The next trial is to first twist a string and then immerse the string in hot Agar to - hopefully - fully penetrate the string filaments.
The ancient Chinese processed their silk strings in this way. They wound the twisted string on a tube that was then immersed in a glue binder
concoction. However, their strings were all three or four strand ropes - a stable construction that would not unwind in the hot binder.
For a simply twisted string a tube holder can still be used but the string must be constrained at each end to prevent untwisting.
First step is to twist the string bundle to maximum twist (about 40° filament angle. The bundle is first soaked in hot water (to soften any residual
sericin) and then immediately twisted and allowed to dry for a day. On release from the rig the test string retained its twist indicating that there
is still some residual sericin present.
The test string was then mounted on a plastic tube - metal hooks on each end being clipped over two screws to prevent the string unwinding. The string
was left quite loose to take care of any shrinkage of the string when being processed in the hot Agar solution.
For the first test a 64 filament string was immersed in a boiling (100°C) 1.3% Agar solution for 5 minutes then mounted on the rig under a load and
left to dry. This string was very flexible - more like a common string used to tie up packages - and was easily unravelled. There may have been some
degree of penetration of the Agar but not enough to bind the filaments into a coherent 'monofilament' instrument string.
The following test with a freshly twisted 64 filament bundle was to boil the string (at 100°C) in a 1% Agar solution for 30 minutes. The string was
mounted very loosely on the support tube but on removal from the boiling liquid had shrunk in length exerting such force as to pull the mounting
screws sideways. Upon drying the diameter of the string had also shrunk from the anticipated 0.83 mm to 0.78 mm.
The completed string was quite stiff and hard and made a distinct 'cracking' sound when wound around a small diameter wooden dowel to improve
flexibility. Penetration of the binder appears to have been complete.
When silk fabric is washed, the recommended maximum temperature of the water is 80° C to prevent damage to the silk. When degumming silk optimum
temperature is about 90°C. Heat damaged silk shrinks, hardens and loses its elasticity. So this appears to be the problem with this test. Possible
solution? - reduce the temperature of the Agar solution to about 80°C maximum for safety.
The next test will be a repeat but with a 1% Agar solution held at 80°C for 30 minutes.
[file]27384[/file] [file]27386[/file] [file]27388[/file] [file]27390[/file] [file]27392[/file]
jdowning - 8-10-2013 at 03:17 PM
Test #6
A 64 filament bundle soaked in hot water (to soften the residual sericin) and pretwisted to maximum twist and allowed to dry.
Then immersed in hot 1% Agar held at around 80° C (between
75°C min and 85°C max) for 30 minutes. On removal from the hot Agar bath the string had shrunk in length and was tight on the mounting tube.
Mounted on twisting rig under a load of 1.9 Kg. Wiped with hot Agar solution to smooth the surface of the string and allowed to dry for 24 hours.
The resulting string was smooth, well twisted and uniform but again stiff and which audibly 'cracked' when wound around a dowel to make it more
pliable. The Agar binder seems to have penetrated all fibres of the string quite well. The string is flexible enough to be tied into a tight knot but
with some separation of the fibres indicating perhaps that a more cohesive yet flexible binder is required?
So if the soaking temperature at 80°C is not sufficient to cause heat deterioration of the silk filament (fibroin) perhaps the Agar itself is causing
changes by being adsorbed into the filament (Agar being compatible with proteins i.e. both sericin and fibroin)?
So the quest continues with more potential alternatives to consider and test!
What seems to be certain, however, is that there is no way that a binder of any kind - hot or cold - can be hand rubbed into a freshly twisted string
to penetrate all of the fibres as stated in Kanz at-Tuhaf - so a fundamental step in the procedure would appear to be missing in that historical
record. Perhaps wiping a twisted string with binder solution after it had been boiled was just a final procedure to ensure a smooth finish to a
string? The addition of a little essence of Saffron (Kanz at-Tuhaf) may have just been to protect the string against rotting due to bacterial
attack?
[file]27475[/file] [file]27479[/file] [file]27481[/file]
[file]27483[/file]
jdowning - 8-16-2013 at 05:18 AM
This series of tests with Agar is to gain a 'hands on' appreciation of the properties of the binder and its workability.
Test #7
A fresh 64 thread bundle was pretwisted - after moistening with hot water - to the maximum number of twists (for this diameter string) of 2.5 twists
per cm. under a load of 1.9 Kg - and allowed to dry.
The Agar gel was again re-heated to boiling point to liquify it and then allowed to cool to 65°C. The pretwisted test string was mounted loosely on
the tube holder and immersed in the liquid Agar for 60 minutes with the temperature maintained at about 65°C ± 2°C. This temperature was chosen in
the event that animal glue in combination with Agar might be tested as a binder - the limit for animal glue (hide, isinglas etc) being 65°C.
On removal from the hot Agar the string was again found to have shrunk in length. The string was then left to dry under a load of 1.9 Kg. The string
was quite uniform in diameter - 0.83 mm - except at one end where there was some evidence of the string starting to 'corkscrew' (ie the twist limit
had been reached).
The resultant string was again found to be a bit stiff but on winding around a dowel to make it flexible no 'cracking' noise was heard. The Agar
appears to have penetrated the string fibres well enough although the fibres could be more easily separated by twisting the string in reverse
direction than for previous strings #5 and #6 soaked at higher temperatures.
Note that the same mix of Agar gel has been re-used for all of these tests - simply reheating the gel to boiling point to liquify it each time. This
has the advantage also of sterilising the gel. Water had to be added for each test to rplace evaporation losses so the gel concentration at about 1%
is approximate.
The specific gravity of each test string was determined by weighing each sample, measuring diameter and length to calculate volume and calculating the
S.G. which ranged between 1.24 and 1.27 - confirming that the silk thread used for these trials was partially de-gummed. The Agar itself - being
mostly water (SG = 1) - adds nothing the the specific gravity of the completed string.
Each test string sample was then loaded to breaking point to determine the Ultimate Tensile Stress. Load was applied using a spring balance.
Unfortunately with each impact of string breakage, the pointer on the balance moved to a different zero point so the indicated breaking load had to be
estimated each time as there is no reset facility on the balance. So - for what it's worth breaking loads ranged from about 10 Kg to 12 Kg.
Calculated UTS for the test strings then ranged between 0.20 GPa (Giga Pascals) to 0.22 GPa. This compares with average commercial values for Bombyx
Mori silk of 0.5 GPa for raw silk and 0.65 GPa for degummed silk - values, however, that apply to untwisted single silk filament under load.
The lower UTS values for the twisted test strings may have been partially due to heat damage to the silk but probably mostly due to the high twist
that was applied to the test strings - a reason why high twist silk strings (as well as gut) are not suitable for use as treble top strings - they
will just break under tension. Top strings must be of low twist construction.
So if the test string calculated UTS is about right (despite the potential inaccuracy of the measurements) at say 0.21 GPa a top string measuring say
0.45 mm diameter would break at a string tension of 3.3 Kg (in practice a factor of safety of at least 2X (ie breaking load of 6,6 Kg) would be
required to prevent breakage).
Note that the Agar binder likely does not add significant strength to a string except perhaps in the case of strings made from spun rather than
unbroken filament silk.
Clearly there are an infinite number of possible combinations of these variables:
- temperature
- soaking time
- Agar concentration.
- degree of twist
- string diameter
- string loading during twisting and drying
The simplest way to soak a string in liquid Agar is at boiling point (100°C) as this does not require constant measuring and adjustment of
temperature throughout the soaking period. Reducing the soaking time to somewhere between the 5 minutes and 30 minutes already tested (say 15 -20
minutes) might be a good compromise to achieve full Agar penetration without any associated significant heat damage (hardening) of the silk
fibroin?
So next to make a full length 64 bundle, Agar binder string for testing on a lute.
[file]27571[/file] [file]27573[/file] [file]27575[/file] [file]27577[/file]
narciso - 8-22-2013 at 03:39 AM
jdowning, You mention above that you consider agar to be compatible with the proteins sericin and fibroin. Do you mean this only in the sense that it
does no dissolve them? Or do you mean that agar's penetration into the protein matrix is particularly strong ?
Perhaps the strength of capillary force governed penetration for different binding agent candidates could be roughly appraised by measring the
spreading circumference of a controlled drop on silk fabric. Your test for adhesive strength of agar on cotton fabric is along similar lines, but
agar-cotton is essentially a starch-starch interaction surely; as opposed to the starch-protein interaction relevant for agar-silk?
Do you anticipate that a significant additional degree of mechanically forced penetration will occur when you combine the drying process with a
sizing die ? (which I assume you will be doing at a later stage as you aim for uniform diameter)
Your mention of a possible antibacterial agent used by the old masters is intriguing. WOuld this have been added to the melt or rubbed on later after
drying do you think?
Thanks again for this fascinating series of posts
jdowning - 8-22-2013 at 05:08 AM
Thanks for your comments and suggestions narcisco.
I am not sure at present how agar might associate with the proteins sericin and fibroin of silk. Checking out the general properties it is said that
agar is compatible with protein whatever that may mean. The sericin coating of raw silk filament is apparently a four layer substance. The outer layer
being soluble in hot water is easily removed - the last layer being difficult to remove being combined with the fibroin at a molecular level. To
de-gum silk to a level required for dyeing or spinning, chemical additives (such as soap) must be used.
My thought here is that the Agar might be adsorbed by the outer sericin layer when heated to form a strong bond between the string silk filaments -
both sericin and agar being in solution at 100°C. As commercial silk is partially degummed I cannot at present fully test this idea. I am able to
purchase raw silk waste (short lengths) that requires degumming before it can be spun into thread but I do not have a source for raw reeled
(continuous filament) silk that I need.
Not sure if Agar is classified as a starch as you suggest?
As I have some old silk fabric I shall test some samples with Agar to see if there is better adhesion to the silk fibres than with cotton fabric. Silk
fabric is of course de-gummed so this will be an interesting trial.
Note that degummed fibroin will readily absorb not only dye but other chemicals. One that particularly interests me (reported earlier in this thread)
is Tannin (used to cure leather among other things). Tannin can be absorbed to a high degree and adds mass to the fibroin - so is a weighting agent.
Perhaps it will also combine with Agar?
I am not sure if sizing dies will be necessary to create greater uniformity or smoothness than can be obtained by twisting the filaments but I doubt
if use of a sizing die would increase further penetration of a binder other than just being a surface finish procedure. The pratical test for adequate
uniformity will be to try the strings on an instrument.
Agar held at 100°C for a period of time would be sterilised by the heat but for good measure an antibacterial coating rubbed on the surface of a
completed string might be of benefit. Once the Agar has fully dried there should be minimal danger of destructive bacterial attack?
The directive in the 14th C Persian work Kanz at-Tuhaf is that "a paste of moderate consistency of gum and a little essence of Saffron is rubbed on
the strings with a piece of linen until it has penetrated into all the parts ". I know that simply rubbing a binder solution onto a fully twisted
string will result in no more that a superficial surface penetration. As it would be impossible to know if this mixture had fully penetrated into all
the parts of a string (without pulling the completed string apart!) then I take this passage to mean that the mixture is rubbed onto a string so that
the coating is uniform, covering the entire outer surface - i.e. it is the equivalent of a varnish coating that string makers today often apply to
their gut strings for additional protection and durability. I also reckon that the Saffron may impart some antibacterial protection.
Bear in mind though that gut and silk strings have a finite useful shelf life - so will eventually deteriorate on exposure to the atmosphere even if
left unused.
jdowning - 8-22-2013 at 11:55 AM
Following up on the question of whether or not there is some interaction between Agar and Bombyx Mori silk proteins, I ran some quick tests this
afternoon with a silk cocoon - that should be sericin coated raw silk (but may not be as the cocoons are generally purchased by those wishing to dye
them various colours for decorative purposes and so may already be degummed. This will have to be checked out with my supplier) - and with pure
fibroin silk powder.
The cocoon was cut in half to remove the sad remains of the pupa. Two glass containers were heated in a bath of boiling water - one containing water,
the other liquid Agar - with a cocoon half floated in each like little boats. The cocoon halves remained in the fluids for 15 minutes the temperature
maintained between 85°C and 90°C. The cocoon halves were then removed and allowed to dry and will be examined for penetration of the Agar
tomorrow.
The hot Agar solution was then spread on two pieces of silk fabric to test strength of adhesion later once the Agar has fully dried.
A level teaspoonful of pure silk fibroin powder was then added to about 15 cc of the hot Agar solution (still heated on the water bath) and stirred
into the fluid to see if it would dissolve. After a minute or so the mixture suddenly began to bubble, as if boiling, and the powder appeared to
dissolve completely into solution creating an almost transparent golden orange coloured soft gel (gel point temperature about 50°C). The resulting
gel is clearer than 1% Agar gel that is a milky coloured and translucent (or semi-transparent - see attached images).
This is a very interesting result and suggests that hot Agar solution does combine with fibroin.
I shall repeat this test again (reconstituted fibroin powder) using a new, clean glass container (just in case there has been some cross contamination
with chemicals from previous trials). I will also try to dissolve some degummed spun silk thread in the same manner.
To date I have only been able to successfully dissolve silk fibroin in the standard copper hydroxide solution used as a test for Bombyx Mori silk - as
previously reported in this thread.
[file]27621[/file] [file]27623[/file] [file]27625[/file]
jdowning - 8-24-2013 at 11:52 AM
Although the reconstituted silk fibroin powder does not dissolve in cold water further tests show that it readily dissolves in water at boiling point
to produce a clear golden coloured liquid. So the presence of Agar in the solution is not a factor in the dissolution process as previously thought -
although it would seem that the Agar is compatible.
A smear of water/silk fibroin powder solution on a plastic plate was left for the water to evaporate leaving a dry powdery film - presumably of the
reconstituted silk powder.
Further attempts to dissolve degummed spun silk thread in hot 1% Agar solution over time failed. So (perhaps not surprisingly) it can be concluded
that silk fibroin in filament form has different properties from silk powder reconstituted from silk fibroin filament. Still not sure at this point if
silk powder might have a useful application in the string making process. In solution though it has a nice 'saffron gold' colour.
The cocoon halves soaked in hot water and hot 1% Agar upon drying were both so tough that it has been impossible the tear them apart with the fingers.
As far as I can tell the Agar solution has fully penetrated the 'walls' of the cocoon that are about 0.5 mm thick.
Silk fabric pieces saturated with the hot Agar solution - upon drying - showed very little tenacity and were easily peeled apart although the silk
pieces were stiffer than before.
The bundle of degummed spun silk threads - saturated in the hot Agar solution in a further unsuccessful test to dissolve the fibroin - was placed upon
the twisting rig, twisted to a low twist level and allowed to dry under a 1.9Kg load. The string (0.45 mm dia.) failed overnight while drying
suggesting that for strings made from spun silk a stronger binder than Agar is required - at least for the smaller diameter low twist strings. Or,
perhaps only reeled (i.e. continuous rather spun from short lengths) silk filament should be used for the smaller diameter strings where maximum
breaking strength is required.
jdowning - 9-8-2013 at 02:43 PM
Now that it has been found that silk fibroin powder readily dissolves to solution in hot water, a 64 thread silk string test sample has been soaked in
a hot 8% solution of silk fibroin solution prior to being twisted. The immersion was for 30 minutes at 90°C then twisted to maximum twist under a
load of 2.4 Kg. The dried twisted string sample was then immersed in the boiling (100°C) fibroin solution for 15 minutes and then allowed to dry
under a loading of 2.4 Kg.
The resulting dried string sample, measuring 0.81 mm diameter, was quite stiff so was 'relaxed' by winding around a wooden dowel of 1/2 " diameter.
There was no distinct 'cracking' sound as this was done (unlike the hot agar soaked sample previously reported). The relaxed string sample was 'gut
like' flexible yet still springy and could be tied into a tight knot. Compare this to the broken lute string depicted in Holbein's painting the
'Ambassadors', 1533 - National Gallery, London.
So, silk fibroin solution may potentially be another binder for making silk instrument strings.
jdowning - 9-9-2013 at 12:14 PM
The other potential binder for silk strings is isinglas (glue made from fish swim bladders).
Isinglas - like hot hide glue - may be prepared by soaking in water and then heated until liquid before being applied hot.
Like hot hide glue, once prepared this way, it has a limited 'shelf life' and soon subject to decay.
Charles Holtzapffel in book 1 of his five books on 'Turning and Mechanical Manipulation', London, England 1843 describes a very strong glue (or
cement) for ivory work - gluing ivory to ivory - otherwise known as 'Diamond Cement'). This glue was made from isinglas dissolved, not in water, but
diluted spirits of wine (ethyl alcohol) or more usually common gin. The two are mixed together and gently simmered on a boiling water bath for about
an hour until the isinglas has fully dissolved and is ready for use. When cold, the glue is an opaque, milky, hard jelly that does not decompose over
time due to bacterial attack. The glue must be liquified by reheating on a water bath before use.
In the 19th and early 20th C when the general population of Britain and North America/Canada were more self sufficient (out of economic necessity)
than today - there were a number of publications - recipe books containing formulae for making everything of use in the home and industry - including
adhesives for a variety of applications.
For example the 'United States Practical Receipt (sic.) Book', Philadelphia, 1844 gives a formula for making 'Armenian Cement' (aka 'Diamond Cement').
This cement is made from isinglas dissolved in alcohol with the addition of gums - presumably to add more 'body' to the glue and perhaps an anti
bacterial element (gum ammoniac or galbanum?).
The point here is that isinglas is compatible with vegetable gums dissolved in alcohol. Agar is also classified as a vegetable gum and can be
dissolved in alcohol.
So more potential combinations of glue/gum binders (in alcohol) may be possible!
[file]27833[/file]
jdowning - 9-10-2013 at 12:17 PM
Another 'twist' - Spider Silk!
The ancient Chinese used silkworm moth filament for their instrument strings - we know because they say so in their historical record. We do not know
what kind of silk was used for the early oud strings - although, most probably, it was from the silkworm moth.
The alternative - spider silk filament has been used for making thread and textiles for centuries among more 'primitive' civilisations. Europeans
since the early 18th C have investigated the commercial possibilities of spider silk as opposed to silk moth silk - attempts that have generally
failed due to the difficulty of domesticating and controlling spiders in order to extract their silk.
In more recent times there has been renewed interest in spider's silk as a bio-engineering material and research into making synthetic spider's silk -
so avoiding the difficulties of using spiders - is showing some promise.
Spider's (non sticky 'dragline') silk is very strong (stronger than some steels by weight and much more elastic). A spider is able to adjust the
composition of its silk to suit the purpose of application - so, for example, dragline silk used as the foundation for web building is physically
different from the silk used to build their nests, or sticky entrapment threads, or for wrapping insects caught in a web.
Early attempts (18th C) by Europeans to harvest spider webs for spinning and weaving into textiles involved boiling spider's nests in Gum Arabic as a
binder.
Can spider's silk be used to make instrument strings? The answer is 'yes' as recently demonstrated by Japanese researcher Shigeyoshi Osaki from the
Nara Medical University who has made a set of violin strings from the silk of a large (as big as your hand) golden orb weaving spider Nephia maculata
- a species found across the world in warmer climates from the Southern States of America to Australia. The largest diameter G string is made from
15,000 silk filaments twisted in a three strand roped configuration.
Being sensitive to insect bites, I do not plan to spend time attempting to make instrument strings from spider silk - the largest spiders locally
being the relatively small but colourful Golden Orb garden spider (Argiope aurantia - 'guilded silver face') - see attached image - overall length,
legs included about 4.5 cm. However, for those brave souls who might want to try, attached is some more information about the history of spider silk
and how to process and obtain the silk.
Also, for information, a recent textile made entirely from 'golden' spider silk - just to demonstrate the potential of this remarkable material.
http://www.vam.ac.uk/content/articles/g/golden-spider-silk/
[file]27859[/file]
[file]27860[/file]
jdowning - 9-15-2013 at 12:10 PM
As previously reported, Franz Jahnel in his book 'Die Guitar und Ihr Bau', noted that some experiments to make viable instrument strings (prior to the
commercial availability of nylon strings in the 1950's) involved dissolving the outer layers of a twisted string made from silk fibroin filament with
a solvent such as caustic soda (sodium hydroxide). The idea was to consolidate the outer layers into a smooth, uniform surface.
Curious about this possibility, the test string - previously reported - made by boiling the string sample in silk fibroin solution, was remounted on
the test rig under load and then wiped with the silk copper hydroxide solvent (see previous posting) and allowed to dry.
The coated string sample was stiffer and the filaments more tightly cemented together than before but still flexible. A knot tied in the sample string
was bound much tighter than in previous trials - so promising for making frets. String diameter 0.81 mm
Copper hydroxide solvent will completely dissolve a silk string of this diameter in about 1.5 hours at room temperature.
The next trial will be to dissolve some scrap silk filament in the solvent and then to test this silk saturated solution as a string coating for
possibilities as a binder.
[file]28953[/file]
jdowning - 9-18-2013 at 12:03 PM
The test string has been given two coats (wiped on) of a saturated solution of silk scraps dissolved in copper hydroxide solvent and allowed to
dry.
In the resulting string, the filaments are more tightly bound together (but not fused together) with diameter reduced to a uniform 0.77 mm dameter.
The string was flexible but with a lot of surface friction allowing a very tight simple knot to be tied. Examination of the string under magnification
revealed surface 'hairiness' - which would account for the increased surface friction. Rubbing the string with a piece of soft chamois leather - in an
attempt to polish out the hairs - only made matters worse (see attached image - scary huh!). Note, however, that the 'hairs' are relatively
microscopic in diameter and may be removed by (quickly!) passing the string through a clean alcohol lamp flame to burn off the surface hairs. This is
the solution that was resorted to by 19th C violinists using silk strings but is not really satisfactory.
The hairiness is caused by the solvent partially dissolving the silk filaments at the string surface - but clearly has not resulted in a smooth finish
to the string as hoped for.
The next test will be to immerse a string test bundle in the silk saturated solvent for, say, 5 minutes prior to twisting. After drying the twisted
test string will then be boiled in agar solution - just to determine if this will result in tightly consolidated silk filaments (due to the brief
dissolving action of the silk solvent) but modified to a smooth surfaced uniform string by the influence of the Agar gel.
rootsguitar - 12-9-2013 at 11:26 AM
Smooth Thread -- and a lot to take in at once, will read more soon
I'm curious if the diameter of the silk strings suggest that they may have been originally manufactured for another use and then adapted for stringing
an oud.
At any rate, the diameter of the bridge holes and peg style of ouds/lutes allow for many kinds of test stringings using a variety of materials.
Although some of these experiments don't pan out it is plausible that oudists of distant times may have been forced to make due with what they could
find in order to keep their instrument strung...
especially when on a far ranging travel.
Thought provoking work,
best------------T
jdowning - 12-9-2013 at 03:51 PM
Musical instrument strings historically were made from any fibrous material - animal (e.g.silk filament or intestines) or vegetable (e.g flax or hemp)
- strings that had a number of commercial general applications apart from instrument strings.
Some of the earliest recorded accounts of instrument string materials date to the 9th C (Ziryab) and relate to the oud. They specify only silk or gut
strings - the preferred animal gut then being that of young lions (the origin of the term 'cat gut' as well as sheep's gut - the latter domesticated
animal source being more readily available (and hence lower in cost) than the wild feline variety.
As for the European lute, there are very few surviving references to animal gut being specifically mentioned as lute string material (and none for
silk) although - if the lute is supposed to be directly descended from the oud - one might expect that at some point in the history of the lute both
silk and gut strings were used.
The key factors in string material selection are physical characteristics - strength, elasticity, durability etc. and raw material availability.
In 16th/17th C Italy the animal intestine of choice for instrument strings was that of the sheep (or rather lamb) primarily because meat of these
animals was in popular demand - particularly during religious festivals - and so there was a ready supply of intestines for the string makers.
Furthermore, baby lambs, being favoured as a culinary delicacy, their intestines - being small in diameter - went whole into making the very best thin
gut treble strings.
Nowadays, historical string makers must of necessity use larger cattle intestines that must be split in order to be made into the thinner gut strings.
Both cattle gut and splitting of intestines for making instrument gut strings was considered bad practice (and so banned by the trade organisations)
at one time.
Another problem for the modern 'historical' string maker is that the domesticated breeds of sheep and silkworm once available in the 16th/17th C and
earlier are now extinct so it is not possible to exactly replicate the physical properties of early strings.
Here, for information, is an article that I wrote for FoMRHI about 2 1/2 years ago that may be of some interest related to this topic.
rootsguitar - 12-15-2013 at 11:05 AM
The paper is excellent and introduced some new ideas to me, reinforced a few others and also led me to speculation about how the pairing of gut
strings with silk strings may have influenced a tuning regime.
I share these ideas to pass the time on this winter night and hope that if nothing else it will keep the motion, the fingers, and the instrument a
part of this discourse on strings.
“…so the sheep intestines of Al-Kindi’s time were not strong enough to make the small diameter, bright sounding top strings.” (FoMRHI Comm.
1937 (April 2011, John Downing)
…which were of silk.
I wish to share how one string capable of higher tension but not tuned higher, might influence a tuning and subsequently the hand motions and sound
produced by a solo lute.
In my experiments with an 8c lute I have chosen to split- tune the regime, by which I mean that I treat the stringing as two separate groups on a
single neck, separated by a blank course.
In the attached link, the close up videos of the right hand show this clearly.
The relationship of the strings to each other is of interest to a player. Using only strings of the same thickness, higher tensions are not available.
Courses seem to favor thirds, seconds and sympathetic strings.
After reading ‘CATGUT’ Instrument Strings Revisited,
I decided to add a wound string…very unlike the timbre of the fishing line strings I have been using.
This single, more resonant string adds a unique color to the music and is not tuned higher.
This thicker string was made for classical guitar and was tuned to the highest pitch of the weaker strings( F#) and placed on the top course.
I hope the videos are interesting to this discussion.
Thanks and here’s a link for the lute minded:
http://youtu.be/fW5Po7Htg5o
rootsguitar - 12-17-2013 at 10:03 AM
The link between stringed instruments and bows is also worth a look.
Though bows are uncommon to most cultures in modern times, it is plausible that as the bow evolved from a fire starting and hunting tool to a weapon
produced in large numbers, string making techniques and the materials used to achieve extreme tensions were cross pollinating musical advances all the
while.
Especially within a Royal Court where an awareness of expansion/defense played a key role.
This is a random quote from a primitive archery forum:
" The Traditional Bowyer's Bible, Volume 2 has a long detailed chapter devoted to making strings, and would be a great reference point. I have been
using B-50 string material that I ordered from 3-Rivers Archery. For the price of a store bought bowstring, you can buy a spool of string material
and make a dozen strings. I believe that linen strings, twisted up from flax fibers was the European traditional string. Native Americans favored
sinew and rawhide, as well as types of plant fiber. "
The alatal/bow timeline may be an interesting comparison with that of harp/lute like instruments.
Just a thought....------T
jdowning - 12-17-2013 at 01:16 PM
I started my investigation into historical instrument strings nearly 20 years ago and published a number of articles on the subject in FoMRHI from
1995 to the present day. Early studies included research into 'alternative to gut' string materials such as silk, sinew and other fibres that led me
to look at other early technologies such as archery bow strings, siege catapult cordage etc. These numerous articles are now available for free
downloading at fomrhi.org.
The published subject matter is too extensive to post here but there is now a PDF search facility on the FoMRHI home page. So to locate my work on
strings (and other topics) over the years search 'Downing' and just plough your way through the entries!
I also have an interest in archery bows - particularly the traditional powerful Turkish reflex composite bows. Strings for these bows must be made to
with stand extreme tensile forces whereas instrument strings must be flexible enough to work well at much lower tension - requiring a twisted
construction that also lowers the tensile strength - the greater the amount of twist the lower the tensile strength. Turkish bow strings made from
silk filament were laid with the fibres as straight as possible (minimal twist) for maximum strength, the string bundle being held together being
tied at several points along the string (such as the arrow nocking position at the string 'centre').
rootsguitar - 12-24-2013 at 05:58 PM
Catapult cordage! What a great lens to look back through.
I look fwd to chkn more of your work @ FoMRHI.
Thanks for keeping the ideas moving
--Best solstice wishes
jdowning - 3-11-2014 at 12:14 PM
Time to investigate further the glue and glue/agar binder possibilities so the first step is to make (or try to make) some hide glue from rawhide. I
will be using this glue, if successful, for another non related project (rawhide backed reflex archery bow) later this year.
Much of the readily available so called hot 'hide' carpentry glue is made from bones not animal skins so - although fine for woodworking - is not what
I want for a silk string binder (or bow making)
Rawhide may be prepared directly from the skins of freshly butchered animals but there is quite a lot of preparatory work involved in preparing and
de-hairing the skin.
A more convenient source can be found in pet food stores sold as snacks for dogs - made up to look like bones. The hide usually comes from cattle and
as purchased is hard and dry so the 'bones' must be soaked in water for a day so that the hide becomes soft and can be flattened out ready to be cut
into small pieces about 1 cm square. I add sodium carbonate to the water (a mild alkali sold in hardware stores as a swimming pool water acidity
adjuster) to further degrease and sterilise the hide during soaking
The small 'bone' selected for this trial has yielded about half a cup (125 cc) in volume of rawhide pieces when dry.
Essentially to make glue, the rawhide is just heated at about 60°C in water for some time (hours) in order to extract the collagen and other
components of the rawhide. The resulting fluid in then reduced in volume by heating until the required strength of glue is obtained. Sounds pretty
straightforward!
Next - to the kitchen for some cooking.
[file]30744[/file] [file]30742[/file] [file]30738[/file] [file]30740[/file]
jdowning - 3-12-2014 at 12:09 PM
The dried raw hide pieces were placed in a pan of water - water volume 375 cc (i.e. about 3X the volume of rawhide) and brought up to a steady
temperature within the range 60 ° C to 65° C. At this temperature the liquid is just simmering - the surface of the liquid moving slightly with an
occasional small bubble rising to the surface. After 1.5 hours at this temperature the volume of liquid in the pan had reduced so another 150 cc of
water were added to make up the loss. The liquid in the pan was stirred every 10 minutes or so to avoid a skin forming on the surface.
After another hour of 'cooking' a further 150 cc of water was added and the liquid again brought up to simmering temperature before being poured
through a fine kitchen strainer into a clean glass jar. The liquid (now gelatin glue) was allowed to stand so that any fine particles would settle to
the bottom of the jar and then decanted into another clean jar. The glue was further reduced in volume by heating on a water bath for another 1.5
hours and then set aside to cool.
A second batch of glue was prepared from the material still remaining in the pan. Again 375 cc of water was added and maintained at a simmering
temperature of about 65° for 1 3/4 hours until the volume of liquid in the pan had reduced to less than 50%. This time around no additional water was
added prior to straining the glue into a clean jar to cool.
After this batch little rawhide material remained for further processing so was discarded.
The relative strength/tackiness of the glue was assessed by rubbing a few drops between thumb and forefinger.
The glue will be further evaluated tomorrow.
[file]30746[/file] [file]30748[/file] [file]30750[/file]
jdowning - 3-13-2014 at 12:11 PM
After standing for several hours in a warm kitchen both batch #1 and #2 remained fluid but gelled when placed in a refridgerator. The gel point of the
more dilute glue (batch #1) is about 20° C and about 25° C for batch #2.
To prevent spoiling of the glue in the liquid state, powdered Borax (sodium borate) has been added at a rate of 5 grams/100 cc (or 3cc borax
powder/100 cc) and stirred in until dissolved. Borax has been known since at least the 8th C in Persia the name, derived from Persian and Arabic,
meaning 'white' - so is an 'authentic' historical chemical additive.
The advantage of the current fluid state of the glue is that some further clarification of the glue can be realised by allowing the batches to stand
to let any fine suspended particles settle to the bottom of the storage jars so that the cleared liquid may then be decanted off.
After standing for a day or two in a room temperature of about 25 °C the decanted glue solutions will be concentrated by allowing the water content
to evaporate to form (hopefully) dry glue flakes that should keep well and will be more convenient to use.
jdowning - 3-14-2014 at 05:40 AM
The clarified glue batches have been decanted onto flexible plastic food container lids to dry. There was some particulate residue in batch #1 but
none visible in batch #2.
At this time of year, with particularly cold wintery conditions outside, the woodstove heated kitchen is at a temperature of about 25° C+ with
relative humidity of around 45% so natural evaporation of water from the glue should proceed fairly quickly. It is best to avoid using artificial heat
that causes the glue surface to skin over preventing uniform drying of the glue.
The glue is quite clear but not perfectly transparent. Commercial glue makers go to some lengths to produce perfectly clear glue but that is just a
cosmetic appearance matter. There was no sign of any fats on the glue surfaces so the rawhide must have been well treated before being put on sale.
In the meantime I have recovered some dried glue flakes from the side of the pan used to heat the rawhide pieces. The glue is quite brittle as
expected. This will be designated batch #3 for the purpose of these trials.
jdowning - 3-18-2014 at 11:40 AM
In order to allow the glue to dry more efficiently the still liquid glue (at warm room temperature) has been transferred to low cost 'throw away'
aluminium pie dishes - increasing the surface area, reducing the depth of the glue and transferring heat through the heat conducting metal. This seems
to have worked quite well - the glue layer, shrinking as it dries, gradually and conveniently detaches itself from the dish. From liquid to dried glue
has taken three days.
The resultant glue is very brittle and hard - glass like - as it should be. Broken into small pieces and kept in a dry place it should last for a long
time without deteriorating - particularly as Borax has been added.
Next to make up a sample and first assess the glue for strength and quality (as a hot hide wood glue).
The dry glue yield from the rawhide (using this method) is about 50% by weight.
[file]30894[/file] [file]30896[/file]
jdowning - 3-24-2014 at 04:19 PM
To test the hide glue strength a sample was made of 10 grams of the dried glue flakes dissolved in 15 cc of water at 60° C heated on a water bath.
This glue concentration remained fluid at around 25°C.
Some dry cedar strips were glued together and clamped for 24 hours before testing.
The test strips were held in a vice and loaded as shown in the attached images - the glued area A being subject to torsional loading (load X moment
arm). Both test joints failed at a calculated ultimate torsional shear stress of about 7,500 p.s.i. - the joint failure being in the wood rather than
the glue. So a promising glue for instrument making at least - if not for a string binder.
The glue being liquid at 25° C would be convenient as a binder for string making but its hardness on setting will likely be a problem. The objective
now is to determine if a glue/agar gel mix binder might be a more flexible yet strong alternative.
A hot 2% agar liquid gel was added to the liquid glue at 60° C but an insoluble mass of gel immediately formed. The gel appeared to be agar coming
out of solution as, when removed. the remaining fluid appeared to be only glue (tested to be of the same strength as in the above untreated
sample).
As gelatin glue and agar gel are supposed to be compatible a further test will be undertaken - this time to try to dissolve the glue flakes directly
into the hot liquid agar solution.
[file]30972[/file] [file]30974[/file] [file]30976[/file]
jdowning - 4-3-2014 at 04:13 AM
The second attempt to make a mixture of hide glue and agar also failed.
The 2% agar was first transformed from a gel to liquid by heating on a water bath to 90° C and was then allowed to cool to 65° C at which point the
dried flakes of the prepared hide glue were stirred in until dissolved - the liquid temperature being maintained between 60 and 70° C throughout. As
glue flakes were added the mixture became stiffer so water was added to maintain a hot glue like viscocity. However, on cooling to 25° C a lump of
gel had formed within the liquid - the gel most likely being the agar separating from the liquid glue.
So, prepared in this manner, the glue and agar do not seem to be fully compatible and will not form a homogeneous solution.
The remaining solution tested as quite a strong glue on wood samples as well as on cotton fabric.
It will next be tested as a binder on a small silk string sample to determine if the end result is sufficiently flexible. Perhaps a portion of the
agar does combine with the glue to provide some degree of flexibility to the glue?
jdowning - 5-1-2014 at 06:09 AM
Again a test with the glue/agar binder failed when tested on a silk string sample. The sample was soaked in the hot glue/agar binder solution at 60°C
and then twisted under load but the binder immediately gelled and made uniform twisting of the string impossible. Heating of the string with a hot air
gun failed to re-melt the binder. This suggests that at 60° C some agar does go into solution with the glue only to form an intractable gel on
further cooling. This test also reconfirms that a binder must be applied to a string after it has been fully twisted - as shown earlier in this
thread.
Hide glue is hard and brittle when dry. A method used to make a flexible glue involves adding glycerol to the glue. However, as glycerol was
discovered in the 18th C it could not have been used to make a flexible binder in earlier times (if indeed a flexible binder was ever used).
An alternative used by art conservators to make isinglas glue flexible is to add honey - apparently a very ancient method. Having a few grams of the
hide glue to hand, a brief test was made to confirm if adding honey might produce a flexible binder. It does!
The proportions used were dry hide glue 12.5 grams, water 25 grams (25 cc) and honey 12.5 grams (10cc) all heated to 60°C. Not sure if this is the
optimum proportion of ingredient. The glue used was not treated with borax as a preservative.
antekboodzik - 5-4-2014 at 04:41 AM
Sheep gut for sausage casings is cheap and readily available... Many times I was thinking of try to making some strings of it 
jdowning - 5-4-2014 at 07:36 AM
Good luck! - making historical instrument strings from any animal intestine is whole subject in itself quite apart from silk string making.
rootsguitar - 8-25-2014 at 05:42 PM
• In Zhu Fan Zhi (Description of various barbarians) by Zhao Ru Kua (1170-1228 AD) in 1225 AD, there is a chapter on “Coral Tree” which says
that this is the product of Venice. He describes in detail how Venetians would drop five claw iron anchors tied to long silk rope and lead weights
into the sea to root up coral. The rope was tied to the side of the boat, pulleys were used to raise up the coral. This is a clear indication of
China trading with Venice in the Song Dynasty
great project you have going! keep thinking of the mariners that relied on cordage of all kinds...
The silk leaders you mentioned on Japanese fishing gear make a lot of sense too.
I would think that the silk-ropes were probably a combination of materials like hemp.
Also interested if you think horsehair may have been twisted into silk strings...
jdowning - 8-26-2014 at 05:49 AM
Curious as to why the Venetians would use silk ropes for this purpose rather than the conventional cable construction hempen rope that has been used
for hauling anchors in the European maritime tradition for centuries. Why would they combine silk with hemp fibre for this purpose?
The maritime rope/cord making traditions and applications have already been explored in the hope of finding some connection to lute string making. It
was once proposed in the 1970's that the large diameter lute bass strings of the 16th/17th C - referred to as 'Catlines' - was a name derived from a
special flexible rope (line) used for 'catting' (catching) a ship's anchor so that it could be securely tied alongside the 'cathead' of a sailing
vessel after being hauled up by the anchor cable. So it was concluded that 'catline' lute strings were of this special roped construction to give the
strings the required flexibility and elasticity essential for adequate acoustic performance. Unfortunately there is no historical record of such a
rope called a 'cat line' so that derivation theory has been proven false. Unfortunately, the idea still persists and it is sometimes quoted on the
Internet that lute catlines were made like some kind of a nautical rope! The lute catlines may well have been of some kind of roped construction - we
don't know for sure - but not derived from any nautical tradition.
Horse hair is recorded as once being used as a fibre for instrument strings as well as for fishing lines (at least for the leaders). Horsehair is a
relatively short fibre so I am not sure how the strings would have been made for practical use on instruments - presumably by spinning the fibres or
by braiding. I have not come across any historical record that states or even suggests that horsehair fibres were ever combined with silk filament for
string making. Why would (instrument) strings be made with that combination - horsehair being much coarser than silk filament?
rootsguitar - 8-26-2014 at 10:39 PM
good questions.
I'll first give a reason that a person may have cut the long tail hairs from a horse and wove them together with strands of silk to create strings
that would fit through a bridge hole and then fit the tuning peg hole of the wooden lute.
The first reason may be that silk cocoons were a controlled trade item in parts of china and it may have been difficult for a traveling foreigner
there to access much of the actual thread. The strands may have been in short supply, but still proved interesting to blend with the tail or mane
strands to create a string that could be tightened and not break as easily as all silk. Assuming their lutes were tuned to steps of some fashion where
higher registers were desirable.
Also, in all seriousness, getting a gig with a lute strung with such a symbolically charged material like; young lion gut or strands of silk, was
probably a good way to pitch (advertise) a performance opportunity in a new place. As far as the actual sound they could produce, one would hope such
a player could make an interesting presentation to maintain good will with whoever's audience they sought. Maybe even to begin a relationship for the
purpose of selling silk or other trade items.
A second way of looking at it may be that there were lots of silk cocoons or mass produced silk strands available to a certain kind of lutenist and
the strings were capable of high enough tension to create a vibration that would collectively pass through the all wood instrument in a favorable
way...but they still chose to add a coarser material like horse hair to provide a theoretical opposite of the smooth silk string. Ideologically
such a commitment to yin/yang in process may not be such a stretch in behavior by someone concerned with the making of an art music or, a distinctive
pattern of sounds they relied on to initiate trade through language barriers
The environment of a maritime lutenist may have been much different than one stringing lutes in an overland setting around pack, travel, and luxury
horses.
Also, from a practical point of view, could the coarser hair of the tail act as a binder to the silk strands?
[file]32396[/file] [file]32398[/file] [file]32400[/file] [file]32402[/file]
rootsguitar - 8-26-2014 at 11:38 PM
I found it really useful to review the piece you
posted on 7-31-2013 at 04:24 AM
in this same thread.
also:
Modern borax and chew toys for rawhide is an excellent use of what's readily available.
Good stuff
jdowning - 8-27-2014 at 06:33 AM
The history of the silk industry is well documented. For the historical time period that concerns us here about stringing of the oud and lute (records
dating from about 800 AD onwards - the time of Ziryab) travellers did not have to go to China to obtain silk filament for string making. Since
Byzantine times silk - in whatever form - came to the Romans via land and sea trade routes starting around 300 BC. The Romans were able to start their
own silk production around 522 AD as did the Arabs at about the same time. So silk filament or cocoons would have been readily available as a basic
raw material and likely were not that costly. The rarity and high cost of silk products (like fabric or clothing) is in the skilful labour intensive
work required to make those products - the 'value added' element - not in the raw silk itself.
Zyriab used both 'cat gut' strings (from the intestines of baby lions - felines) as well as from silk filament as did those who followed after. His
silk filament was likely produced at home in Muslim controlled Iberian Peninsula (later Spain) where sericulture was to become a significant part of
the agricultural industry of that region.
I know of no account indicating that horsehair was blended with silk filament to make strings for an oud or lute - or indeed that horsehair was ever
used to make oud or lute strings.
Athough spun silk can be used to make strings, the use of continuous reeled filament is most likely how the strings were made either as a twisted rope
like construction or (possibly) by braided construction (more about braided construction to follow). The use of continuous filament makes for a much
stronger, smoother string.
The silk filament of an instrument string must be tightly bound together to form a durable cylindrical homogeneous whole and minimise any internal
friction losses that might affect acoustic efficiency. A penetrating glue concoction is the binder - or alternatively the natural glue (sericin) that
coats the raw silk filaments. Horsehair would not serve in this function.
Reeled silk filament is relatively strong compared to other materials - animal, vegetable or mineral (tensile strength depending upon the strain of
silkworm and how they are fed and raised). The filament does not break easily. I don't know how horsehair compares in tensile strength but again that
would no doubt depend upon the breed of horse, how it was fed and raised.
jdowning - 8-30-2014 at 12:08 PM
The larger diameter lute (and oud?) strings (7th course and lower) of the late 16th and 17th C were referred to as 'Catlines', 'Catlins' or 'Catlings'
in some of the 17th C English publications about the lute. The name given to a lute string might refer to where it was made, or where it was obtained
(market source) or from its appearance.
So the fictitious special nautical rope (cat line) for 'catting' and securing a ship's anchor - proposed for a while to be the origin of the catline
lute string in both name and construction (rope) - is not valid. Another theory (and there have been a number of others based upon, as yet, unproven
etymology) is that the Catline lute string was originally made in the 'Spanish' Catalonia region ('Catline' suggested as being a corruption of
'Catalin') and exported to Europe during the early 16th C as costly gut strings of roped construction.
The main agriculture of 'Spain' by the 16th C. - established during the long standing Muslim occupation of the Iberian Peninsula - was sheep farming
and sericulture (silk farming) with the associated wool and silk fabric industries as well as specialised trades (no doubt?) such as gut and silk
instrument string making. Catalonia was a major centre of the silk industry (not sure about the woollen industry, however). So it would seem possible
that if is true that superior lute (and oud) bass strings were being made in that region then the material of construction could easily have been silk
filament rather than gut. Furthermore, if this was the case, then the construction of these strings may have been braided rather than twisted like a
rope - there being a long tradition of making braided cords from silk filament. Braided strings being labour intensive to make, would have been
relatively costly.
More to follow.
jdowning - 8-31-2014 at 05:32 AM
If Spain was once the source of high quality lute bass strings this situation did not prevail far into the 16th C as the ruling government embarked
upon a campaign of expelling Muslims and Jews - groups with expertise in the main agricultural activities of the country and their associated trades
(wool and silk). The silk industry in Catalonia was primarily in the hands of Jewish artisans. The eventual decline of these industries in 16th C
Spain was further reinforced by application of punitive taxes.
Barcelona in Catalonia was by then producing a reduced range of products including silk trimmings, and other clothing accessories such as braided
cord.
We know that many of those escaping the Inquisition were evacuated in 1492 by the Ottoman naval fleet operating in the Mediterranean to be resettled
in areas of the Ottoman Empire under the Sultan Bayezid II where their expertise in many fields (including sericulture and silk fabrics) contributed
to the economic development of the Empire.
Italy and France had also established their own silk industries - the main centre in France being Lyon. By the late 17th C some of the best bass
strings (for 11 course lutes) were made in Lyon (called 'Lyons') as well as those of Italian manufacture. There were also mid range strings known as
Venice Catlines.
Nothing is known about Catline strings except that they were 'smooth and well twisted' and elastic enough to work - so were presumably of some kind of
roped construction as well as being loaded with metal or metal salts to increase their specific weight. It is assumed that these strings were made
from gut but they could just as well have been made from silk - of braided construction for maximum elasticity and a smoothness not to be found in a
lumpy roped construction.
An earlier version of bass string is the 'Gansar' mentioned in the Italian 'Capirola' manuscript (c. 1520) as well as later by Dowland (1610) who
describes them as 'a kind of strings of of a more fuller and larger sort'. Nothing more is known about this kind of string but Mimmo Perrufo, some
years ago, suggested that this string may have been so named because it was like a flexible French decorative cord called a 'Ganse' used in the
clothing trade. The Ganse was a small diameter cord of braided gold, silver or silk thread - not a twisted rope construction. There is no mention of
gut strands being used in the manufacture of Ganse cords.
Assuming the Gansar string used by Capirola was for his lowest sixth course (he does not say that it was) then - with octave course pairing - it was
still good enough acoustically to be tuned down a further two semitones (as it was for some of the pieces in his book of tablature). Re-tuning the 6th
course is something of an inconvenience for the player and perhaps led to the adoption of a seventh course tuned two semitones below the 6th course?
However it does indicate that string technology had by then advanced sufficiently for the basses to be tuned lower than the 6th course limit for
plain, unloaded gut or silk strings with no significant acoustic penalty.
As mentioned earlier in this thread(!), making braided cord of silk is an ancient craft practised by the Chinese and Japanese - crafts that are still
popular today - and is associated with clothing and other decorative applications. This popularity together with the invention of modern high speed
braiding machinery means that small diameter braided cord (Chinese Knotting cord) is readily available at relatively low cost in diameters ranging
from 0.4 mm to 2 mm or more - a range of diameters making the cord suitable for those wishing to experiment with braided instrument strings. Another
source of low cost small diameter braided cord is fishing line available as plain braid (with or without a filament core) and also loaded with a heavy
metallic core. The majority of these cords use synthetic filament such as nylon for braiding although braided silk fishing lines are made but are very
costly.
jdowning - 9-3-2014 at 04:53 AM
Author Luca Molà in his book 'The Silk Industry of Renaissance Venice' provides a well researched, detailed account of the development of the silk
industry in Italy and the rest of Europe.
The Italian silk industry - for centuries a major industry in the peninsula - was established like that of Moorish 'Spain' by the Arabs and Jews
around the 9th C. By the late 14th C the weaving of silk fabric was confined to the cities of Genoa, Venice, Bologna, Lucca and Florence in the North.
Other towns such as Pistoia, and Livorno were later to become silk production centres. The attached map of Northern Italy shows the main silk
producing area
The main product of the silk industry was silk fabric the production of which was divided into specialist activities - silk farming, silk reeling and
spinning, dyeing and weaving. The best quality silk fabric was often interwoven with gold thread. A secondary but significant part of the industry
often carried out by family concerns involving women in particular was the making of silk accessories for clothing - French 'passementerie' - which
including braided cords or 'ganse'. This was the trade not only of silk spinners but also the associated trade of gold and silver thread spinners -
all part and parcel of making decorative braid.
A significant part of the Venetian silk industry - apart from the manufacture of costly silk fabric woven with gold thread - was in the production of
silk trimmings and 'passementerie' for clothing.
So what might be the connection with lute bass strings of the 16th and 17th C?
[file]32471[/file]
jdowning - 9-3-2014 at 05:19 AM
Italian artisans of the silk industry were employed in setting up operations in Lyon, France in 1536 then to become a major centre for the
manufacture of silk products.
Germany was a latecomer to the silk industry - established during the second half of the 16th C. in cities such as Nuremburg and Frankfurt.
jdowning - 9-3-2014 at 10:59 AM
So what did the writers about the lute in the 16th C and 17th C have to say about their lute bass strings. Not a lot.
Italian lutenist Vincenzo Capirola (c 1520) used gut trebles that suffered intonation problems due to their natural taper but less so the strings made
in Munich, Germany. He mentions that the intonation problem was less with thinner strings than with thicker strings and then mentions a string called
'da ganzer' where the problem was even more acute. As mentioned earlier in this thread a 'ganse' was a braided silk cord used to decorate clothing.
Not only does this suggest that Capirola's 'ganzer' string was large in diameter but that it was of roped construction for the necessary elasticity
for use as a bass string. He does not say where the strings were used but Dowland at the beginning of the 17th C does - 'strings of a more fuller and
larger sort than ordinary which we call Gansars. These strings for the sizes of the great and small means (4th, 5th and 6th courses) are very good'.
Neither Dowland nor Capirola say where the 'Gansar' strings were made or from what material.
Let's speculate and assume that these strings were made not from gut (as invariably assumed by present day string makers) but from silk braid with
some kind of binder surface coating (glue?) rubbed in as protection against wear and to provide a consolidated smooth surface. Such strings would most
likely have been made by artisan silk spinners.
A similar larger diameter braided silk cord but with a spun central core of silver or gold - with its substantially increased specific weight and
elasticity would have made good bass strings for lower pitched courses (7th to 11th) on lutes. These basses were called Catlins and - if constructed
in this manner -would again have been made in and exported from the silk producing regions of the world.
Robert Dowland (1610):
Gansars has already been mentioned above. Dowland goes on to say 'there is another sort of the smaller strings (i.e.gansars) made in Livorno,
Tuscany.
For bass strings, some are made at Nuremberg. The best strings of this kind are made at Bologna in Lombardy and from there sent to Venice ... commonly
called Venice Catlines.' He then mentions that the best quality strings made in the Summer are sold at the Frankfurt and Leipzig markets.
Note that Bologna, Venice, Nuremberg, Frankfurt and Livorno were all centres of silk production at that time.
jdowning - 9-3-2014 at 12:04 PM
Dowland also mentions that some strings were coloured (green, red and blue). The dyeing of silk filament was another trade within the silk industry as
mentioned previously. He also cautions about buying strings that show signs of 'hairiness' ('faseling with little hayres'). Hairiness is a
characteristic of worn strings made with silk filament.
Lutenist Michelangelo Galilei wrote in 1617 from Munich, Germany to his brother (in Italy?) asking him to obtain 'four thick strings from Florence for
his own and his pupil's needs'.
With Florence being a main centre of the silk industry in Italy those 'thick strings' may have been braided silk catlines?
If not, why would Galilei request these strings when he could have obtained the best gut basses that, in the opinion of Adrien LeRoy, London (1574)
came from Munich?
Mary Burwell in her lute book (1676) writes that good strings (i.e. treble strings - of sheep's gut or cat's gut) were made at Rome but that the great
(bass) strings and their octaves were made in 'Lyons at France'. Yet again a connection with a major centre of the silk industry.
Thomas Mace (1676)
Mace does not mention from what material lute strings are made.
He says there are 3 sorts Minikins (trebles) Venice Catlines for the 4th and 5th courses and most of the other octave strings and Lyons for the
basses. He also mentions that 'there is another sort of string(s) called Pistoy Basses that he considered were none other than thick Venice Catlines
which are commonly dyed a deep dark red colour. These (Pistoys) are the very best being smooth and well twisted but hard to come by'.
He also mentions several sorts of coloured strings - green, red, blue and yellow.
Mace also tells us that the Venice Catlines were very strong and might be broken only with difficulty by the strength of a man.
He, like Dowland, mentions use of graduated fret diameters (reducing in diameter towards the bridge) to minimise intonation problems with the thicker
strings.
So once again there is a connection by association with the silk industries of Venice and Pistoia and the silk dyeing trade.
The final historical step involving silk filament in instrument strings is, of course, the close wound string - copper or silver wire wound around a
core of silk filament - an innovation probably too late for the lute but finding general application for the oud, guitar and other instruments up
until about 1960 when synthetic nylon filament replaced silk as a core material. Wound strings would have been much cheaper to make than the earlier
silk braided strings - with or without spun metal cores - but sounded much different. Although the first evidence of wound strings (open windings?)
dates to the late 17th C there is no evidence to suggest that these were generally adopted by the lute community. Perhaps they had just too much
sustain and metallic sound for the late 17th/ early 18th C ear to accommodate?
jdowning - 9-4-2014 at 06:25 AM
Turning to the early oud strings made from either silk or gut, their physical limitations would have been the same as the strings for lutes.
Apparently nothing is yet known about how the oud sixth and seventh courses were made or from what material - or even how they were tuned (a fourth
apart perhaps?)
What we do know from the earliest sources dating from the 10th C (the Ikhwan al-Safa) is contradictory - as discussed on page 8 of this thread. The
Bretheren give the number of silk threads required to make each string of a 4 course oud that provide the desired 4/3 diameter increase between each
course tuned a fourth apart - Zir 27 threads, Mathna 36, Mathlath 48 and Bamm 64 threads. However, assuming the silk threads are all of equal diameter
and the zir string is 0.45 mm in diameter, simply twisting the silk threads into a string does not give the required diameters for the remainder of
the strings. The solution previously suggested was:
Zir string - simply twisted, minimal twist - 0.45 mm diameter, tension 35 Newtons.
Mathna string - simply twisted, maximum twist - 0.56mm diameter, tension 31 Newtons.
Mathlath string - simply twisted, maximum twist - 0.69 mm diameter, tension 27 Newtons
Bamm string - four strand roped construction - 0.9mm diameter, tension 25 Newtons.
This being the case the Bretheren were telling us not only how the strings were constructed but the need for diminishing string tension from treble to
bass.
The other solution for construction of the Bamm string might have been braided silk. From some preliminary trials in hand braiding thin cord (Japanese
Kumihimo) I have found that the diameter increase of the braid compared to the same number of cords simply twisted to maximum twist is about +30% for
a fairly loose braid. The percentage increase would be less for a tighter braid. So braiding would give the required diameter increase even without
string tension reduction.
Note that braiding must be done in multiples of 4 strands so could also apply to the Mathna and Mathlath strings but not the Zir string that could
only be simply twisted (required for maximum strength anyway).
Later the 14th C Kanz al-Tuhaf also gives the number of silk threads for a 5 course oud. All of the thread numbers are divisible by 4 so the strings
could all be of braided construction (unlikely though?). The finished strings were rubbed with a glue. Tests have confirmed (reported earlier in this
thread) that the glue by this procedure will only consolidate the surface fibres and not completely penetrate a string. This would work as described
only for a silk string braided around a core - the glue consolidating just the surface braiding.
jdowning - 9-6-2014 at 04:13 AM
The early thick lute bass string(s) necessitated being paired with a thin string tuned an octave higher (octave tuned courses). The thin string
provided the upper partial tones missing from the thick string and so brightened the overall sound of the course.
Robert Dowland (1610) mentions that in his time the basses were unison tuned ('the bases must be both of one bignes'). He says that the general custom
has been to use octave tuned basses ('to set a small and great string together') but that this practice has been abandoned by musicians as being
contrary to the rules of music. So bass string technology, by the beginning of the 17th C, had advanced to the point where an octave tuned bass course
was unnecessary. However, the use of octave tuned basses continued to prevail and are mentioned by Thomas Mace in 1676.
Why did the use of octave tuned bass courses continue (possibly until the demise of the 13 course lute in the German Baroque period by the mid 18th
C)? Perhaps the sound of the octave tuned basses was preferred to that of unison tuned basses because they (the improved basses) were too loud
compared to the other strings? Perhaps those improved strings (braided silk over silver or gold core?) were very expensive so octave tuning continued
to be popular as a way to economise on string cost?
We know nothing at present about the construction of the early oud bass courses (6th and 7th) except that there seems to be no tradition of using oud
octave tuned basses. So does this imply that the oud basses by the 15th C were already acoustically superior to those made by the European string
makers at the time?
The Spanish vihuela - popular in Spain during the 16th C - appears to have been unison strung throughout (up to 7 courses) before falling into disuse
by the third quarter. Perhaps those strings - a legacy of the Muslim string making tradition - were no longer being made in Spain following expulsion
of the Muslim/Jewish string makers and their expertise to the Ottoman Empire and had to then be imported at great cost? Indeed the vihuela gave way in
popularity to the little 4 course guitar (said to be nothing more than a six course vihuela with the 1st and 6th courses removed) that did not require
either costly bass strings or costly lute first courses (but did use an octave tuned 4th course).
Were these superior bass strings constructed of braided silk (with or without heavy metal cores) first exported to Europe from the Ottoman Empire
(through Venice for distribution to the rest of Europe). Were these superior strings examined by the Italian silk industry who then figured out how to
make them? Or did someone, curious to find out if braided silk ganse cords would make good lute basses, carried out a few tests (just as I am now
doing) - and so the lute Catline string was born?
Note that the traditional silk braided cords used by the Chinese for making decorative knots were about 0.8 mm and 1.5 mm in diameter braided around a
silk core - a good starting point for experimentation without having to go through the trouble of learning silk braided cord making skills.
rootsguitar - 9-9-2014 at 12:03 PM
excellent thread!!
just copied & look fwd to reading later...( rural location again, wifi only on supply run)
thanks again for sharing your work
jdowning - 9-10-2014 at 11:42 AM
Glad that the topic is of interest rootsguitar.
I hope to spend some 'hands on' time in exploring the practical possibilities of using silk braided construction for oud and lute middle and bass
strings (5th course downwards) - 'Catlines' or 'Pistoys' as the English called them.
As braided cords - made from synthetic filaments such as Nylon on high speed braiding machines - are readily available and inexpensive in the
diameters of interest for instrument strings, these will be used to initially explore the possibilities before moving on eventually to hand braided
strings of silk filament if the preliminary experiments are reasonably promising.
Braided cord (machine or hand made) may be constructed with:
1) No core (solid).
2) Hollow no core (more elastic).
3) Cored - synthetic or silk filament or metal filament core (for greater specific weight).
Synthetic braided cored strings are available on the market as 'Chinese Knotting Cord' in diameters from 0.4 mm to 3.0 mm - dependant upon quantity
purchased (10 yard minimum) - from about 15 cents to 30 cents per yard or metre.
Metal cored braided strings - lead core/nylon or Dacron braid - are available as trolling fishing lines in various diameters (specified also as
breaking loads) at reasonable cost (say $10 to $20 per 100 yards).
Initial trials with 'Dacron' braided/lead 7th course string is currently being reported here:
http://www.mikeouds.com/messageboard/viewthread.php?tid=11998
Note that hand made 'historical' gut 'Catlines' of roped construction can each cost up to $50 or so, for the history buffs, an alternative synthetic
string with comparable performance - costing only a few cents - might be of interest!
jdowning - 9-10-2014 at 11:57 AM
As silk braided cord is not readily available, the ultimate objective here is to hand braid silk filament strings - Japanese style. The apparatus
required is simple and easily made - a square or round wooden table - on legs like a stool - with a central hole and the silk strands on weighted
spools - the weight of each spool providing the required tension for compactness of the braid.
The braids may range from fairly straightforward to very complex. However for a round braid of small diameter an 8 strand braid might be the
economical way to go?
Here is a video of a skilled braider at work making an 8 strand round cord. Not sure how long it would take to make a braided instrument string a
metre in length but likely not too long at this speed:
http://www.youtube.com/watch?v=pnCggGQ-z3A
..... and for an unskilled amateur string maker the braided string making process does not have to be completed in one session!
jdowning - 9-17-2014 at 11:27 AM
"The Silk Industry of Renaissance Venice" by Luca Molà - mentioned earlier in this thread as relevant to this topic - is now available from John
Hopkins University Press, a $64 book on sale for $18 plus shipping. I have just received my copy - cloth bound, 480 pages - a very nice edition to
read in armchair comfort.
Very good reviews so essential reading for those interested in the early development of the silk industry.
https://jhupbooks.press.jhu.edu/content/silk-industry-renaissance-ve...
jdowning - 9-20-2014 at 04:43 AM
Preliminary trials with varnished lead cored/Dacron braided fishing line have been promising when tested on a 7 course lute - so may indicate the
viability of this form of construction for lute bass courses of the late 16th C - Catlines and Pistoys.
This type of 'ready made' low cost trolling line of braided construction is made by a number of companies in various diameters and sized according to
breaking load. Standard breaking load range is 12 pound (5.5Kg), 15 pound, 18 pound, 27 pound, 36 pound, 45 pound, and 63 pound with approximate
outside diameters of 0.66 mm, 0.69 mm, 0.74 mm, 0.86 mm, 0.94 mm, 0.99 mm and 1.07 mm respectively. So there is ample scope for using these modern
cords for testing as instrument strings at low cost.
If solid metal cored strings of braided silk construction were used historically it is unlikely that pure lead would have been used (it is not strong
enough to be drawn into fine wire) but pure gold, pure silver and perhaps copper may have been the alternatives as the drawing of these metals through
dies into fine wires is ancient technology dating back thousands of years.
With the exception of copper, pure gold, silver and lead remain in a soft malleable condition when worked (by hammering or drawing into wires) - i.e.
they do not work or strain harden like many other metal alloys. Work hardened copper, however, may be easily returned to a softened state by annealing
- by heating to red hot and then quenching in water.
Lead is particularly interesting as a potential core material as it is low cost, denser than silver or copper but also will slowly stretch (creep)
under low loads at room temperature. So as a core in a braided line under load it will stretch until only the braid is carrying most of the tension.
Lead being so soft has a low tensile strength so could not be used alone as an instrument string. The same applies to pure gold and silver both with
low tensile strength.
A downside of lead as a metal is that it is toxic if ingested as dust or particularly as soluble salts of lead (although endemic in the environment of
modern society) - so should be handled with respect (wear plastic gloves when handling and dispose of properly as a hazardous material).
As the test strings in these trials are varnished and string use as a 7th course infrequent, abrasion wear of the braided jacket is low so that the
lead core is essentially safely sealed inside the string. Also the lute 7th course is usually played open and rarely stopped on the fingerboard.
For the experimenter, braided lines have a further practical use as it is a simple matter to remove the lead core to leave the hollow braided sleeve.
The lead core may then be replaced with alternative cores of solid wire gold, silver or copper. With 22 gauge gold wire (0.64 mm diameter) costing
today from about $100 to $200 a foot length (30 cm) pure gold wire as a core material will not be used in these trials. Pure silver (0.999 purity)
costing about $3/ft for 22 gauge wire is a possibility but can wait until trials with lead and copper cores yield positive results.
jdowning - 9-21-2014 at 04:23 PM
It is a simple process to create modern 'Catline/Pistoy' bass strings from braided Dacron/lead cored fishing lines. For this trial a 27 pound breaking
load line - costing less than 10 cents a metre - is being used.
A length of line was first tensioned with a 3 Kg load and then wiped with two coats of a thin varnish - in this case TruOil. The varnish penetrates
the braided sleeve and consolidates it with the core - as well as providing additional surface abrasion resistance and sealing of the lead core.
In retrospect the varnish should be thinned further (50/50?) to ensure a uniform coating - perhaps with more coats.
On the string test rig , for a string length of 60 cm and diameter of 0.77 mm the following results were obtained :
Pitch standard A440
F (87 Hz) - tension 2.94 Kg.
E (82 Hz) - tension 2.59 Kg
D# (78 Hz) - tension 2.33 Kg
D (73 Hz) - tension 2.08 Kg
C# (69 Hz) - tension 1.85 Kg
C (65 Hz) - tension 1.65 Hz
String sustain (judged by my aged ear) ranged from about 12 seconds+ at the highest tension to about 8 seconds at the lowest tension.
Mounted as a unison pair 7th course on an F tuned lute - tuned two semitones (78 Hz) below the sixth course - the strings took a while to stabilise in
pitch but are now performing quite well in my opinion. So far no wear of the braided Dacron sleeve is in evidence.
Attached, for information, are a couple of audio clips of the unison pair sounded alone and as part of a 16th C dance composition for lute - played
with soft fingertips.
Next to run some trials using copper as an alternative to the lead core.
[file]32613[/file]
[file]32614[/file] [file]32615[/file] [file]32617[/file] [file]32621[/file]
[file]32623[/file]
rootsguitar - 9-22-2014 at 01:27 PM
Interested in the varnishing process you've been using, do you coat a large length of material then let it dry and cut it into string lengths?
Just got a hold of some waxed linen kite twine ( also some of dacron ) and thought to try your binding idea to make strings for my 8c.
Also someone gave me some nylon harp strings to try as well.
Curious if you think established harp string construction informed lute/oud set ups in the far past.
---Thanks
jdowning - 9-22-2014 at 04:07 PM
I am just varnishing each string length (say about 100 cm). The varnish is thin enough to soak through the braided sleeve to bind it with the core.
The second varnish coat then fills in any surface hollows to produce a smooth, uniform string. During this process, the string length is placed under
tension so that some equilibrium between braid and core is achieved prior to varnish application.
Note that twine (twisted fibre construction) is not the same as braid so a varnish may not fully penetrate as a binder (if that is the intention).
Also varnish will not stick to wax.
Modern monofilament nylon harp strings are no different to modern monofilament nylon lute strings as far as I know.
Harp strings are somewhat less critical than lute strings in that all harp strings are played 'open' whereas the first six courses of a lute must be
uniform enough to remain 'in tune' with each other when stopped on the fingerboard.
I know nothing about harp gut string construction in the dim and distant past but imagine that the gut string makers supplied strings to both lute and
harp players on demand (as they did for many other, non musical, applications of gut strings).
Some instruments like the old Irish harps were strung with metal (brass or silver) strings and played with long fingernails as plectra - the precise
drawing of small diameter metal wires being ancient technology. The bass strings of those instruments were made by twisting two wires together to
provide the required flexibility necessary for best acoustic performance - equivalent to bass strings of twisted filament construction.
jdowning - 9-25-2014 at 03:35 PM
As the lead core is not tightly bound to the Dacron braided sleeve it is a simple matter to remove the lead core and substitute a core of another
material for testing - be it copper, silver or gold.
To remove the lead core just rub the last couple of centimetres of end of the line with a smooth hard rod (e.g. a metal rod like a screwdriver) to
compress and squeeze out a bit of the lead core. Grip the end of the core with pliers and hold the end of the sleeve (wearing plastic gloves) and just
pull to remove the core. Place the lead core in a sealed plastic bag for safe disposal as toxic waste.
For this next trial a pure copper core will replace the lead core. The most convenient and economical way to obtain small quantities of small diameter
copper wire is from domestic appliance flexible electrical cords - just strip the plastic insulation covering from the copper conductors with wire
strippers. In this example I am using wire stripped from some scrap telephone wire - the conductors measuring 0.64 mm diameter (22 AWG).
As the copper wire is in a semi hardened 'springy' state it must first be annealed to make it soft. To do this the wire is made into a small diameter
coil and heated with a small propane torch until red hot and then dropped into a pan of cold water to quench it.
Working with a metre length for convenience, the end of the soft, stripped wire is first made smooth with fine sand paper and then pushed into the
braided sleeve. The sleeve is then pushed little by little on to the core - the sleeve opening up in diameter as it is pushed over the wire - until
the core is completely covered by the sleeve (it takes a bit of patience to complete!).
The covered string is then suspended under a load of about 3 Kg and wiped with two coats of TruOil varnish to bind and seal the Dacron sleeve to the
copper core.
Next to test the copper cored string on my test rig.
jdowning - 9-26-2014 at 11:41 AM
The varnished copper cored string measures 0.89 mm in diameter with the solid core measuring 0.64 mm. Even after annealing the copper core the
finished string feels too stiff and lacking in flexibility.
These are the results of the string test for information:
F (87 Hz) - tension 3.5 Kg
E (82 Hz) - tension 3.1 Kg
D# (78 Hz) - tension 2.8 Kg
D (73 Hz) - tension 2.4 Kg
C# (69 Hz) - tension 2.2 Kg
C (65 Hz) - tension 1.9 Kg
The stiffness of the string made it very sensitive to tune on the lute (D#) so the trial was terminated. It is possible that a multi strand copper
core (salvaged from electrical appliance flexible power cords) might give better results so this alternative will be tested later.
Note that although the density of copper is 8.96 gm/cc the equivalent density (according to the Mersenne-Taylor law) of the braided composite string
is only 5.0 gm/cc the Dacron sleeve being about 1.4 gm/cc. Likewise for the lead cored string the equivalent density is only about 5.7 gm/cc compared
to the density of lead at 11.4 gm/cc.
I have just received some samples of braided Chinese knotting cord so will report on these next.
jdowning - 9-30-2014 at 11:54 AM
The company 'Tangles n Knots' stock a range of Chinese braided knotting cords and kindly sent me a free sample card of their products in the following
(approximate) diameters - 0.4 mm, 0.5 mm, 0.8 mm, 1.0 mm, 1.4 mm, 1.5 mm, 2 mm, 2.5 mm 3.0 mm. The cords are braided in nylon filament around a simply
twisted nylon core and come in a wide range of colours. The cords are sold in 10 yard lengths for just over 30 cents a yard or in longer lengths for
lower cost per yard.
The cords are very flexible and the braided sleeve appears to be of relatively loose weave (?) - the attached image is the 2 mm diameter braided cord.
The cores are easily removed so could be replaced with cores of other materials. Some kind of core would seem to be necessary to ensure uniform
roundness of the string under tension. Twisted silk filament might be a better core material than nylon and would ensure binding of the nylon sleeve
to the core when the string is varnished
One interesting alternative core possibility might be worn wound strings (smaller gauge Pyramid lute strings perhaps?) - to give them a new lease of
life as a heavier gauge bass string.
So, will order some 10 yard lengths for experimentation and see how it goes.
jdowning - 10-12-2014 at 12:01 PM
The tests with braided Dacron/lead core trolling line previously reported were for Sunset brand Tel-a-Depth line of 27 pound breaking strength (12 Kg)
- lead core diameter (unloaded) is 0.53 mm. Line diameter loaded 0.75 mm.
Continuing to investigate the potential of braided lead core fishing line a 100 yard spool of 63 pound (29 Kg breaking strength) Sunset brand
Tel-a-Depth nylon braided lead cored trolling line was purchased on EBay for $18 (including shipping). This is half price because the line colours are
not perfect - a cosmetic fault that does not affect the physical properties.
To prepare the line a metre length was loaded with 4.4 Kg and then wiped with two coats of TruOil varnish.
Diameter under load is 0.94 mm and core diameter unloaded is 0.45 mm.
Surprisingly the lead core of this larger diameter 63# line is less than that of the 27# line - I was expecting the core diameter to increase with
line diameter but perhaps it is a case that more 'weight' is required for the smaller diameter fishing lines to sink through the water when being
trolled?. The braid is somewhat coarser because it is the braided sleeve that carries the load not the soft lead core that will stretch under load so
that it carries little if any of the string loading. This means that the larger diameter 63# line will have a relatively lower linear density than
that of the 27# line so will not perform as well at the lower pitch bass frequencies as the 27# line as we may see on the string test rig trials. This
may be a good thing for bass string performance - the smaller diameter - all else being equal - the better.
The attached image shows the varnished (2 coats) 27# test line (red colour) alongside the varnished (2 coats) 63# line (blue colour) for comparison.
It can be seen that 2 coats of varnish is insufficient to completely seal the braided sleeve so more coats will be applied until a smooth surface is
achieved before testing.
A 100 yard spool of 12 pound (5 Kg breaking load) of another brand of braided/lead core line by Sufix is on order from EBay suppliers for future
testing. No doubt different manufactures will have differing sleeve breaking strength to core diameters in the design of their lines. The range 12
pound to 63 pound lines seems to be the manufactured 'standard' range commercially available.
[file]32790[/file]
jdowning - 10-15-2014 at 11:45 AM
The preliminary testing of the larger diameter (blue) lead cored line has been completed with consistent results.
The string was given another coat of TruOil varnish (total 3 coats) that has produced the required smooth surface finish. String diameter after
varnishing and under load is 0.94 mm. Test string length is 600 mm and string tension range limited to between about 2 Kg to 3 Kg - the approximate
range for a workable bass string.
Here are the results obtained on the string test rig - pitches at A440 standard, string tension in Kilograms force.
G 98 Hz - 3.2 kg - Sustain about 8 seconds
F# 93 Hz - 2.9 Kg
F 87 Hz - 2.6 kg
E 82 Hz - 2.3 Kg
D# 78 Hz - 2.0 Kg - sustain about 7 seconds, temporary pitch increase when plucked
D 73 Hz - 1.8 Kg - ditto
The approximate density of the compound string is 3295 Kg/m³ compared to the previous (red) string density of about 5650 Kg/m³. So - not
surprisingly - this larger diameter string being of lighter composite density (more nylon filament to lead core ratio) cannot perform as well as the
red string at lower bass frequencies.
The 63# test string is a manufacturer's second quality product said to be due to cosmetic colour deviations but I have to wonder if a wrong lead core
diameter was also a reason for the downgrade to second quality?
Awaiting delivery of the 12# line for testing.
rootsguitar - 10-18-2014 at 11:50 AM
This link may be of interest mainly because of the string coating process( The specifics of powdered glass added to string for kite fighting is of
course irrelevant).
It includes:
Charkho = Using a spinning wheel to coat kite string.
Entangled: A Documentary Film by Aditi Desai & Kai Fang
http://entangledmovie.wordpress.com/author/aditivdesai/
The lower photo is titled: coating the string (manja) by hand
[file]32950[/file]
[file]32952[/file]
rootsguitar - 10-18-2014 at 12:18 PM
Kite string coating in India:
jlmsika - 10-18-2014 at 12:47 PM
Twenty years ago, they sold in Paris oud string sets with the trebles in natural gut (the highest three strings, i.e. 0.61, 0,79 and 1,04 mm) combined
with the three lowest strings wound around a red natural silk core. The brand was "Triomphe, Cordes pour Aoud". These sets were balanced and delicious
sounding on a good oud. They were the ORIGINAL string set for ouds, for centuries...
I wander if we couldn't get a string maker somewhere to make them again, since oud playing is now very popular all over the world, with a great
potential market for these sets which would be unique in the business?
jdowning - 10-19-2014 at 04:08 AM
This recent topic by Yaron Naor about the restoration of an old oud owned by Yair Dalal includes information on the inspection of string fragments
found on the instrument.
http://www.mikeouds.com/messageboard/viewthread.php?tid=13030
The strings were gut trebles and wound/silk core basses - the silk filaments being dyed a red colour. Perhaps from the same French manufacturer?
Interesting that string sets of this kind were still available 20 years ago (1994) as I imagine that they quickly went out of fashion when modern
nylon monofilament and wound nylon filament core strings became generally available at much lower cost after the 1960's.
I purchased gut and wound silk filament oud strings in Cairo in the early 1960's and they too were of French manufacture as I recall.
Is there any more detailed information about the French manufacturers of this type of string? That would be useful historical data.
No doubt there are string makers who might be persuaded make custom sets of this kind of string - for a price - but I doubt if there would be any
great demand among modern oud players.
[file]32962[/file]
jdowning - 10-24-2014 at 11:12 AM
The 12 lb (5.4 Kg) breaking strength braided lead core fishing line has arrived for testing. Cost including shipping - $20 Can. for 100 yards (91
metres). Made in Taiwan - China and Taiwan appear to have the expertise and high speed braiding machines to make these small diameter braided cords at
low cost.
Line outside diameter unloaded is 0.55 mm and lead core diameter 0.40 mm. Brand is 'Sufix'.
A length of line has been prepared for testing with 2 coats of TruOil varnish with the string loaded with 2.4 Kg. Diameter under tension is 0.52
mm.
Test results next.
(After the attached image was taken, my Canon PowerShot A470 camera partially failed so it has not been possible to provide a macro image of the line.
The problem may be due to a failed ribbon cable that controls the lens shutter so may be too costly for a Canon repair. Ribbon cables are available on
E Bay from China for about $4 so I may try to repair the camera myself. Pity as this has been a useful camera for macro photography over the past few
years).
[file]33007[/file]
jdowning - 10-27-2014 at 11:54 AM
Here are the results of the tests of the varnished 'Sufix' 12# braided nylon/lead core line.
Vibrating string length 60 cm, pitch standard A440, string outside diameter 0.52 mm.
Approximate tension range measured for bass string is 3 Kg to 2 Kg.
B 124Hz - tension 3.1 Kg
A# 117Hz - tension 2.8 Kg
A 110Hz - tension 2.4 Kg
G# 104Hz - tension 2.2 Kg
G 98Hz - tension 1.9 Kg
Sustain ranged from about 15 seconds at the higher tension to about 12 seconds. Bright sound.
This is a compound string so the linear density works out to about 6450 Kg/m³. (This compares with 5650 Kg/m³ for the 27# braided/lead core and 3295
Kg/m³ for the 63# braided/lead core strings previously tested)
As repair to my A470 camera is not now an economic proposition it has been replaced with a Canon PowerShot Elph 135 'point and shoot' (the 'older'
cameras are now coming on the market at reduced cost). This camera focusses down to 1 cm but does not seem to be as good for macro work as the older
A470 even with its higher resolution sensor but I still need to mess around with lighting etc. to obtain best results
Here, for what it is worth, is a hand held macro shot of the varnished 12# varnished string (0.52 diameter with some white dust!). Not as uniform as
the previous test strings so perhaps only a single coat of varnish would have been sufficient due to the relatively thinner braided sleeve?
[file]33018[/file]
jdowning - 10-31-2014 at 11:53 AM
With the 12# varnished line removed from the test rig, a few more attempts were made to produce a reasonably clear macro image of the string using my
relatively low cost equipment. The Canon Elph 135 although capable of focussing down to 1 cm (lens to object) only a relatively small central section
of the image (about 1/3 of the full image width) is in sufficient focus to be used by cropping the full size image. Rather disappointing.
Note that the string outside diameter in this image is only 0.52 mm.
I had noticed some 'falseness' of the string on the test rig (i.e. with the string sounded open or unstopped) judging from the slightly uneven string
vibration along its length.
The attached macro image of the coiled string may show the reason why - although it looks a lot worse than would otherwise appear to the naked eye.
The bottom string section exhibits 'hairiness' due to surface fracture of the filaments in the braid. This may have occurred when the string was wiped
prior to removal from the rig - or the filament breakage occurred when the string was loaded and varnished - not sure. The middle string section in
the image seems to be flattened and not perfectly round. This may be an optical illusion but I suspect is an actual deformation perhaps occurring
during manufacture or when wiping varnish on the string?
Or it may be that the oil varnish coating was not sufficiently dried before handling the string? All factors to be examined when making the next -
hopefully more successful - test string of this size!
[file]33066[/file]
jdowning - 11-13-2014 at 01:03 PM
There are two early sources previously mentioned in this thread that provide some information about the construction of silk oud strings - 10th C
'Ikhwan al-Safa' and 14th C Persian 'Kanz al-Tuhaf'.
The implication of the silk thread count per string given by the Ikhwan al-Safa for their 4 course oud (27, 36, 48 and 64 threads) has already been
discussed. The silk thread count increases by the sacred harmonic ratio of 4:3 as does the pitch increase between courses (a perfect fourth apart) for
plain silk strings at equal tension. However, simply twisting bundles of silk threads of equal diameter with those thread counts does not give the
required string diameters except for the first two courses. As a consequence, it is proposed here that the third and fourth courses might have been of
braided construction (without a metal core) in order to achieve the required string diameters according to the Mersenne-Taylor Law (a mathematical
relationship understood by the IKwan al-Safa).
Looking again at the 'Kanz al-Tuhaf' translation provided by Dr G.H Farmer the string thread count given for a 14th C five course oud is: 16, 24, 32,
48, 64. In this case the thread count increase from treble to bass string is 3:2, 4:3, 3:2, 4:3 - so introduces that other sacred harmonic ratio of
3:2. From a numerical perspective there is yet another sacred harmonic ratio of 2:1 (i.e. 16:32, 32:64, 24:48) as well as 3:1 (48:16), 4:1 (64:16),
8:3 (64:24). The Persian oud theorists of that time were numerologist in their beliefs!
There is no mention of string diameter in the KAT but the 5 courses of the KAT oud - like the 4 course oud of the Brethren - were tuned a fourth apart
so, consequentially, in order to achieve the required string diameters with the silk thread count per string given and assuming equal thread diameter
and equal string tension, the fourth and fifth courses must have been of a construction other than plain twisted. A plain braided string construction
(i.e. without a metal core) on the other hand would meet this requirement.
jdowning - 11-15-2014 at 03:28 PM
Good images are an important part of forum communication where a picture may go a long way towards clarifying a particular subject matter where many
words might fail - especially in a forum where English may not be the first language of many members.
So - as a small diversion and for general information - here is a latest attempt to improve on the previously posted close up images of the 12#
braided string test sample using my new relatively low cost 'point and shoot' Canon PowerShot Elf 135 digital camera. Not as easy to use to obtain
reasonable 'super macro' (extreme close-up) images as my now broken PowerShot A470 but perhaps good enough in detail for posting on the forum.
The attached two images show first the 100%, hand held, (compressed to 480x680 format) macro image of the 12# braided string simply held in my fingers
and then the cropped image to show greater 'super macro' detail. The images - taken under incandescent lighting - have not otherwise been 'improved'
by photo editing. So not bad at all for a low end point and shoot camera and hand held macro image.
The camera also performs well enough in copying 35 mm microfilm images of text (when set to grey scale recording) to produce clear full page 8.5 X11
prints.
[file]33341[/file]
[file]33343[/file]
jdowning - 11-16-2014 at 11:02 AM
An advantage of using the ready made braided lead cored fishing line for these trials is that the lead core may be easily removed from the braided
sleeve and replaced with an alternative core material to modify the vibrating string characteristics.
Alternative core materials to be experimented with are flexible stranded copper wire, and worn wound strings. Pure silver is a possibility but has no
advantage over lead as the metal densities are nearly the same. Gold is now out of the question as since 1973 - once on a gold standard fixed price of
$35 an ounce since 1945 - its crazy price (currently $1,200 or more an ounce) is now determined by free market speculation.
The 63# line as previously reported only has a lead core of 0.45 mm diameter and outside diameter 0f 0.94 mm. However, the lead core diameter may be
increased by removing a double length of core from the line and twisting it like a rope to make a flexible two strand core measuring 0.78 mm diameter.
This twisted core may then be re inserted into the braided sleeve to produce a string of about 1.2 mm diameter that should have a good bass string
performance.
A string of this construction will be made next for testing.
The attached image shows the two strand twisted lead core inserted into the braided sleeve - with room to spare.
jdowning - 11-17-2014 at 12:59 PM
For information - small diameter pure lead wire - suitable for the cores of braided string - is readily available from stores catering to fly
fishermen. The wire is used to weight fly fishing hooks and is sold in small spools of 13 ft length (4 metres) or 1 pound in weight (0.45 Kg). The
length of wire per 1 pound spool will, of course , depend upon wire diameter - cost ranging from about $18 to $35. Cost per small spool is about $3
regardless of wire diameter.
Wire diameters available are 0.25 mm, 0.38 mm, 0.50 mm, 0.64 mm, 0.76 mm and 0.89 mm.
These are the same wire diameters used to manufacture braided lead core fishing lines that range in cost from about $10 to $30 for 100 yards (91
metres) so is likely the cheapest source of lead wire (and comes with ready made braided sleeve).
jdowning - 11-20-2014 at 12:50 PM
To make a two strand twisted lead core line for testing, two 1 metre lengths of the 63# braided line were prepared by first removing the cores as
previously described. The lead cores measure 0.45 mm in diameter. It can be seen from the attached image that the core is not smooth but has been
indented by the braided sleeve during manufacture.
The two cores were then twisted together on my string making rig applying 403 turns until the twist became uneven (more compact) at the driven end of
the rig indicating the the twisted core was about to break. At this point the length of the twisted core had reduced in length by 3 cm and increased
in outside diameter to about 0.8 cm. The attached image shows the compact twist for a short distance at the driven end compared to the more open,
uniform twist along the rest of the core. The roughness of the lead strands no doubt prevented a more desirable compact twist forming along the
twisted core.
The two strand twisted core was then fed into the braided sleeve little by little as previously described - easily done but rather tedious taking
about 20 minutes or so to complete. A perfect task for members of the Dull Men's Club!
http://www.mikeouds.com/messageboard/viewthread.php?tid=15251
The braided sleeve expanded to accommodate the larger diameter twisted core so reduced in length to 89 cm. So starting with the two unmodified 1 metre
lengths of 63# braided string the final working length of the new completed twisted core string is 89 cm.
The attached image shows the unvarnished braided string with two strand core inserted and compared with the original single strand core 63# braided
line. Note the minute surface hairs of filaments broken during the abrasive process of inserting the twisted core into the braided sleeve.
A first coat of 'TruOil' varnish has been applied. As might be expected the string absorbed a fair amount of varnish as it soaked through the braided
sleeve to fill the gaps in the twisted core.
Next to test the string on the test rig once application of the varnish coats is complete and dried. The string outside diameter will be measured at
that stage.
[file]33399[/file] [file]33395[/file] [file]33397[/file]
[file]33401[/file]
jdowning - 11-30-2014 at 12:08 PM
The diameter of the varnished 63# braided string with the 2 strand twisted lead core is 1.17 mm under tension.
The test results are as follows for the approximate 2 Kg to 3 Kg working range, A440 standard pitch:
D 73Hz - 3.2 Kg sustain 10 seconds.
C# 69Hz - 2.9 Kg.
C 65Hz - 2.5 Kg.
B' 62Hz - 2.2 Kg.
A'# 58Hz - 2.0 Kg sustain 6 seconds.
The equivalent linear density for this composite string is about 3750 Kg/m³ - not a great increase compared to the 63# string with only a single
strand core of 0.45 mm diameter (3295 Kg/m³). This is because of the open winding of the 2 strand core - the gaps in the core being filled with
varnish. A 3 or 4 strand twisted core would be more compact and have a lower frequency range.
The length of empty braided 63# sleeve left over from this trial is to be varnished and tested next - to gain some appreciation of the performance of
a plain braided string (without metal core).
jdowning - 12-5-2014 at 12:44 PM
The 63# braided sleeve without core has been prepared for testing by tensioning with a 1.4 Kg load and then saturated with two coats of TruOil
varnish.
The diameter of the completed string is 0.88 mm under tension.
Testing in the approximate 2.0 to 3.0 Kg tension range gave the following results.
Standard pitch A440.
g 196 Hz - 3.5 Kg tension - sustain about 6 seconds
f# 185 Hz - 3.1 Kg
f 175 Hz - 2.8 Kg
e 165 Hz - 2.5 Kg
d# 156 Hz - 2.2 Kg
d 147 Hz - 2.0 Kg tension - sustain about 5 seconds
The string density works out to about 1010 Kg/m³ so is a bit lower in density than mono-filament nylon but the string is less stiff and more elastic
so should perform better than plain nylon.
This string will next be tested mounted on my 60 cm string length lute tuned in F (A440 standard) on the 4th Course (d#) so will be tuned an octave
above the 27# braided lead core 7th course that is still performing well under test - so should be an interesting comparison.
The attached 'macro' image of the string shows it to be nicely 'smooth and well twisted' - albeit with some interesting multi-colour dye particle
inclusions.
[file]33551[/file]
rootsguitar - 12-5-2014 at 03:09 PM
Interesting as always...to clarify, the tension described is from the pegbox to the bridge?
Scordatura driven by circumstance is always in the back of my mind...curious if the overall instrument bracing for a lute/oud seems suited to the last
few stringings you have documented. Also does the grouping have a collective sum?
--regards
T.Robb
jdowning - 12-5-2014 at 05:05 PM
More precisely the tension described in the tests is from the front edge of the nut to front edge of the bridge i.e. the vibrating string length
I do not understand what you mean when you wonder if the last few stringings I have documented are suited to the overall bracing for a lute/oud or if
the groupings have a collective sum?
The string test results simply provide pitch/tension data for each string of a specific construction, diameter and vibrating string length. For a lute
medium to bass string (the kind of string currently being investigated) - the working tension historically would be in the approximate range of 2 to 3
Kg dictated by the breaking tension of the treble gut or silk strings and the perception of the feel under the fingers of 'equal' string tension
(tension usually reducing from treble to bass rather than strictly equal throughout).
rootsguitar - 12-7-2014 at 06:14 AM
When I adapted found materials for instrument strings they usually could not be broken by turning the tuning pegs ( copolymer mostly).
The limiting factor then becomes the ability for the pegbox/tuner to hold the tension…the peg might hold momentarily, only to slip & spool off
rapidly.
Of course if the string material would break under tension this would not happen.
I wonder if there is a way to quantify what tension a wood on wood peg can hold?
I had been warned in the past that even if a grouped stringing could be successfully set up, over time the strain may deform the instrument if the
collective tension was outside of what the luthier intended.
I began to rationalize that if, for example, a 15-string lute was strung with only 6 or 7 strings could it handle a higher tension since the sum of
the missing strings would not be applied against the bracing?
Also The kite strings I tried broke easily as I tested them. They would be useful for lute strings only to someone far from supplies & with a
strong desire to explore their instrument.
It was a worthy project to varnish them though, a direct influence of this thread. I coated them & hung them from rafters with rebar tied on the
ends.
The rebar occasionally knocked together as I worked, making a chime-like sound.
At first glance I thought your figures were listed as a full stringing set up, that might explain my post as well…
g
f#
f
e
d#
d
________
D
C
C
B
A'#
_________
B
A#
A
G#
G
The material your working with seems to show good potential for instrument strings...
(attached images of pegbox views, fretted & fretless)
[file]33603[/file]
[file]33605[/file]
jdowning - 12-7-2014 at 12:49 PM
A good analysis of the historical aspects of lute strings is 'The Lute in its Historical Reality' by Mimmo Peruffo (Aquila Strings) here:
http://ricerche.aquilacorde.com/wp-content/uploads/liuto-en.pdf
The string technology applies equally well to ouds.
The plain braided fourth course string has been mounted on my 7 course, 60 cm string length lute for testing. It has slightly less sustain and fewer
higher partial tones than the equivalent wound Pyramid string that it has replaced but is otherwise quite acceptable in my opinion. When stopped in
higher fret positions it tends to sound slightly sharper in pitch than the equivalent wound string - not unexpected however - easily corrected by
adjusting fret positions and diameters.
For information the attached audio clip is the sound of the test 4th course string followed by the braided lead core 7th course (unison pair) sounding
an octave lower. The braided 7th course strings have been in operation for several weeks now with no perceptible wear and I rather like the full bass
sound.
Also attached is a short audio clip to demonstrate how the braided test strings blend with the modern Pyramid lute strings - 'Pergamasco' by G.L.
Fuhrmann from his lute book 'Testudo Gallo-Germanica' , Nuremberg, 1615.
(Pergamasco or Bergamasca - a rustic dance from Bergamo in Italy - was a popular setting for 16th/17th C lute composers)
[file]33607[/file]
[file]33608[/file]
rootsguitar - 12-8-2014 at 12:27 PM
Hey I look fwd to reading that soon...the clips sound good too. I get the 2-3kg range now, missed it earlier. Good stuff, thx.
jdowning - 12-10-2014 at 01:02 PM
The 63# braided twisted two strand lead core was previously reported. Two strands gave quite an open core so the viability of a four strand core has
been tried.
Two 63# lengths of line were stripped of their single cores and the cores doubled over to make four strands. The four strand bundle was then twisted
under light tension to form a twisted core measuring 1.10 mm in diameter - the individual strands being 0.45 mm diameter.
The twisted core was then fed into the 63# braided sleeve to form a string measuring 1.4 mm diameter. However, it was not possible to insert the core
into the sleeve for more than a couple of centimetres so the expansion limit of the sleeve has been exceeded at this core diameter. The core diameter
limit for the 63# sleeve may be between 0.9 mm to 1.0 mm. A four strand core of this diameter might be made from 0.4 mm diameter core stripped from
12# line.
The attached images show the twisted four strand core, the 63# sleeve without a core and the 63# filled to the limit with the four strand core - all
unvarnished.
(I now once again have familiar good macro imaging facilities having purchased a broken Canon PowerShot A470 camera on Ebay and used the parts to
restore functionality to my broken A470).
[file]33639[/file] [file]33641[/file] [file]33637[/file]
jdowning - 12-18-2014 at 12:03 PM
For completeness the Sufix 12# lead core line previously reported on 10-27-2014 has been mounted for testing on my lute together with an equivalent
Pyramid wound string at the fifth course. Pitch A# (117 Hz), tension 2.76 Kg, string length 60 cm, string diameter 0.52 mm.
The sound is bright and 'brassy' - similar to a new Pyramid wound string - with good sustain. In tune when fretted up to the 8th fret.
The attached audio clip is the 12# string played open followed by the wound string for comparison.
As the lute now needs to be completely re-strung and re- fretted the 63# braided lead core string previously reported will be the last string to be
tested on the lute at the 6 course position tuned to F 87 Hz - again compared to an equivalent Pyramid wound string.
jdowning - 12-20-2014 at 12:34 PM
For a final test, before restringing and refretting the lute, four of the braided lead core strings (previously described and tested on the string
test rig) have been mounted as follows:
7th course - 63# braid with 2 strand twisted core, diameter 1.17 mm, pitch A'# 58Hz.
7th course -27# braid 0.53 mm lead core, diameter 0.75 mm, pitch D# 78 Hz.
6th Course - 63# braid 0.45 mm lead core, diameter 0.94 mm, pitch F 87 Hz.
5th course - 12# braid 0.40 lead core, diameter 0.55 mm, pitch A# 117 Hz.
String length 60 cm.
For information, the attached audio clip is the sound of the strings played open and in sequence from 7th to 5th course. The brightness of sound for
each string depends upon the ratio of the core linear density to nylon braid linear density (and string tension, diameter and length being constant)
so may be modified by adjusting these two variables as necessary.
Note that the 27# braided strings, as a unison pair for the 7th course, have been in service daily now for three months and show no signs of wear.
Lute strings of the 16th and early 17th C were sold in long lengths bound up in knots as shown in the attached engraving of the period. By way of
comparison and demonstration, a length of the 27# braided lead core string has been made up in the same fashion.
[file]33770[/file]
[file]33772[/file] [file]33773[/file] [file]33775[/file]
jdowning - 12-25-2014 at 09:37 AM
Here is another engraving of a 'knot' of strings taken from Hans Gerle's tutor for lute and viol - 'Musica Teutsch', 1532.
Bound up like a bundle of modern boot or shoe laces - further support for braided construction?
[file]33830[/file]
jdowning - 1-26-2015 at 12:33 PM
I have picked up another 100 yard spool of lead cored braided nylon line - this time a 45# line by 'Mason' - for about 1/4 normal retail price as it
is old stock.
The line measures about 0.88 mm outside diameter with a lead core measuring 0.5 mm so performance will be similar to the Tel-a-Depth 27# tested
earlier.
A metre of line was loaded with 3.5 Kg and then varnished with three coats of TruOil wiped on with a cloth (in retrospect two coats should be
sufficient). After varnishing and under load the line measured 0.78 mm outside diameter.
The test results are as follows for a string length of 60 cm:
F# 83 Hz - tension 3.33 Kg - sustain 12 seconds
F 87 Hz - tension 2.96 Kg
E 82 Hz - tension 2.65 Kg
D# 78 Hz - tension 2.32 Kg
D 73 Hz - tension 2.10 Kg
C# 69 Hz - tension 1.82 Kg
The average linear density for the composite string is 5507 Kg/m³.
This string should work for either the 6th or 7th course of my lute. I will test it to see how it performs (intonation) when stopped up to the 8th
fret. Cost about 20 cents (shipping and handling from the USA in this case cost more than the spool of line otherwise cost would be about 10
cents!)
There are many brands and sizes of lead cored line on the market but it is not know how much variation there might be between the various makes. These
four samples tested suggest that the core diameters used may be fairly limited regardless of breaking load - say 0.4 mm and 0.5 mm? Breaking loads for
this type of line are 12#, 15#, 18#, 27#, 36#, 45# and 63# most far more than the usual breaking load of a bass string in the 2 Kg to 3 Kg range.
However - as has been demonstrated - the cores may be easily stripped from the braided sleeves and replaced with cores of another diameter. The
replacement lead cores may be solid wire or cores made by twisting together 2, 3 or 4 solid lead wires of any chosen diameter.
If required solid lead wire may be purchased in spools in standard nominal diameters of 0.38mm, 0.51mm, 0.64mm, 0.76mm and 0.89 mm.
An alternative core material might be annealed copper wire filament twisted to form a cylindrical core. Fine copper wire is readily available at low
cost stripped from flexible domestic electrical power cords
Pure gold solid wire would be the best core material but the cost these days is prohibitive.
[file]34199[/file] [file]34201[/file]
jdowning - 3-11-2015 at 11:55 AM
One interesting and common style of braid is the so called 'Maypole' braid because it replicates the movement of one of the maypole dances where two
circles of dancers moving around a maypole each in opposite direction - the dancers holding a ribbon attached to the top of the pole - weave in and
out of each other. This action forms a distinctive braid or plait around the pole as the dance proceeds.
The action is clearly seen here in an antique shoe lace braiding machine (the braid in this case is without a core).
https://www.youtube.com/watch?v=4Yjpu64Y_oc
The maypole tradition in Europe (held either on May 1st or Mid Summer) is of ancient but obscure origin. It is not clear if the popular maypole dance
routines of today also date to ancient times. If they do, however, perhaps this particular version may have been the inspiration for cored braided
strings? A tempting speculation!
Lots of examples of maypole dancing have been posted on YouTube - but mostly pretty chaotic in their execution!
jdowning - 4-3-2015 at 10:16 AM
Having recently purchased a low cost 'off lease' desktop PC loaded with Windows7 Pro 64bit operating system I found that I could not access Arto
Wikla's handy 'on line' string calculator as it was blocked from loading due to the low security level of the Javascript required for its operation. I
have found the software very useful as a tool for interpreting data collected from my string test rig results and although I am still able to access
Wikla's software via a Linux operating system on another computer I looked around for an alternative program that might do the job.
Luthier Oliver Wadsworth has provided just that with his StringCalc32 - a stand alone program offered as a free download here:
http://www.wadsworth-lutes.co.uk/software.htm
In Windows7 Pro 64bit the software installed without need to make special arrangements for operating in 'Windows 98 compatible mode'.
I have tested the software in 'Single String View' and it provides all of the facilities provided by Arto Wikla's program and more. Nice easy to use
interface. Recommended.
Note that the 'Tone' button that is supposed to give the sound of the selected frequency does not work in any of the Windows operating systems.
[file]34920[/file]
Jody Stecher - 4-3-2015 at 10:42 AM
Thanks very much for this! Note that the tone generator will crash a computer using XP home. I have written to ask if the calculator will work on
Mac computers. I will report back when I've received a a reply.
Jody Stecher - 4-3-2015 at 10:44 AM
Update, 20 seconds later. My email to Oliver at the address given on the website bounced back as User Unknown.
jdowning - 4-3-2015 at 12:10 PM
I sent an email to Oliver earlier today to confirm that his software works on Windows 7 Pro 64 bit but, so far, have not received a response - but no
bounce either.
jdowning - 4-24-2015 at 04:03 AM
Jody - I forgot to mention that I have been in contact with Oliver a couple of times by email about the functionality of his program StringCalc32 so
his email address on his website is OK.
Note that I have now also tested the program on Linux Ubuntu 14 32 bit using the Wine application that allows some software designed for Windows to
run on Linux. I can confirm that it runs as designed (except for the tone generator facility that is non functional).
Oddly I have not been able to get StringCalc32 to work properly in either Windows XP Home or XP pro.
I have also just tested Paul Beier's string and fret calculator application version 2. This is shareware and can be downloaded for free testing (30
tests) before registering (cost 10 Euros). It covers both monofilament strings and wound strings by the well known makers and includes a tone
generator. It works as designed in Windows 7 pro 64 bit but has limited functionality in Linux 32bit via Wine.
| Pages:
1
2
3
4 |