Jonathan - 9-5-2008 at 12:25 PM
I am about to start yet another bowl, and am toying with the idea of placing strips of tiled purfling between each stave, such as in the photograph
below.
Any suggestions on how to tackle this? If I glue up the wood that comprises the purfling, I think that the glue will fall apart when I heat it up to
bend it.
Any suggestions would be appreciated.
jdowning - 9-6-2008 at 04:17 AM
I have never tried it but one way might be to glue up each strip of tiled purfling 'on its side' - to the required finished profile using a building
board - i.e. in the same manner as fitting tiles around the edge of a soundboard. Each finished piece of tiled purfling, once released from the
building board, would be quite fragile to handle - but not impossible. It could, also, be considerably reinforced by gluing up each strip on the
building board over a thin veneer of dark wood (or dark paper even) which would then remain as part of the purfling - the dark veneered side being
glued against the light coloured ribs and the un-veneered side being glued against the contrasting dark coloured ribs.
Just a thought.
Jonathan - 9-8-2008 at 10:32 AM
hmmm.
Then what? If I then glued it to one side of the stave, I think I would want to go back later and sand the edge again, at which time I would be
worried about all of the purfling coming off--after all, the glue surface is less than 2mm.
There's got to be a trick to this.
Thanks for your help.
SamirCanada - 9-8-2008 at 11:07 AM
also... I was thinking about this and I think I have a way it could work.
you have your first rib bent and on the mould.
then you could take a strip brown paper bag about the lenght of the rib and about 1 inch wide or so.
then you glue it to the under side of your first rib but leting it overhang by about 6 or 7 mm off the rib.
then you take you could take your little inlay strings and glue it to the side of your first rib and put glue to the underside on that brown paper
strip.
it would have enough suport from that paper strip. Using hideglue would be ideal for this kind of stuff.
then once all the peices are on and glued to the first rib you can join the next rib against the side of your first rib which now has the little tiles
on it the side.
you should still have some paper left after the tiles are in which you could put glue on and it would help with joint on the next rib.
makes sense??
SamirCanada - 9-8-2008 at 11:24 AM
if you look carefully in your picture,
there is a strip of veneer on each side of the tiles.
jdowning - 9-8-2008 at 04:07 PM
Why would you want to sand the purfling (tile) strip once it has been glued to the stave Jonathan? Would you not just shape the matching edge of the
next rib to fit precisely without having to touch the purfling?
Can't determine from the image if there is veneer on one or both sides of the tiles or none at all Samir. The visible lines might just be the glue
joints? Perhaps Jonathan could confirm?
One 'benefit' of contrasting ribs, purfling or lines is that they can effectively visually hide any slight inaccuracies in the rib joints - not that
any self respecting luthier would make anything other than a 'perfect' joint!
carpenter - 9-8-2008 at 04:47 PM
<< not that any self respecting luthier would make anything other than a 'perfect' joint!>>
Not me, surely! Perfection's merely a good place to start from.
I did a bowl-cracks repair of a similar oud; there was a paper-thin piece of veneer on each side of the strip tiles. The tiles went all the way
through (as did the cracks), with no paper backing or anything; 16 - 18 inches worth of cracks. It took forever, made a few new tiles to match,
touched up that finish ... and relaxed later with a little eye surgery, something easy for a change.
My hat's off to whoever does this work on a regular basis! There's a trick in there somewhere, I'm sure. Waiting for a revelation.
Jonathan - 9-8-2008 at 11:31 PM
It looks like there is a veneer on each side of the purfling.
jdowning, if you glue the purfling onto the side of one of the ribs, then you are going to have to go back and sand it down again on the sanding block
for a couple of reasons.
First, in my case at least, there would be no way to get it absolutely flat. Even a half mm discrepency would screw things up when you went to fit
the next rib.
Second, and really the key reason to me, is that it would allow you to gradually taper the width of the purfling. If the width of the purfling is 4
mm at the center of the bowl, then it has to narrow down considerably when you get to the tail block and, especially, the neck block. 19 (or 21, 17,
or whatever) lines of purfling at 4mm per line would mean that there wouldn't even be enough room for the purfling by the time you reach the neck, let
alone the ribs themselves. The easiest way to correct this would be to glue the purfling onto the rib, and then treat the new rib/purfling combo as a
new rib. I think. Of course, you couldn't put on that outer strip of veneer, or it would completely disappear.
Even with traditional purfling, you have to make a correction for that. That 1mm purfling adds up by the time you get to the neck, and multiply it by
21. It has screwed me up in the past.
I'll admit I'm a bit stumped on this one.
Jonathan - 9-8-2008 at 11:35 PM
OK, here's another possibility, likely wrong. You mount two ribs, leaving a small gap between them. Then, you go back and put in small (1 inch or
so) segments of purfling that are approximately the right width (just a tiny bit too wide), that taper in at the bottom slightly. Shove them in till
you get a good fit, then go back when you are done and sand them flush with the adjoining ribs.
I've got no idea.
jdowning - 9-9-2008 at 06:05 AM
The attached rough sketch should make what I have in mind a bit clearer.
The tiles - all precisely and equally cut to the required dimensions - are assembled on a flat building board against a removable template cut to the
required profile of the bowl. The template is made either from non stick plastic card or thin card waxed to prevent sticking. On the assumption that
the purfling is to be veneered on both sides - veneer (paper or wood) is placed on the building board temporarily trapped under the template edge to
hold it in place and flat. The tiles are glued end to end (following the template profile) as well as to the veneer. The ends of each tile may require
trimming to a slight angle, as work proceeds, to obtain close fitting glue joints between each tile - the angle being dependant upon the curvature of
the template.
Once the glue has dried, the template is removed and any surplus veneer trimmed away with a sharp knife. The purfling is then flipped over and the
other side glued to another piece of veneer (using a weighted board to hold everything flat). Surplus veneer is again trimmed after the glue has
dried. The preformed 'tiled-purfling' strip may then be lifted from the building board ready for gluing between the ribs of the bowl.
This, of course, will result in parallel sided purfling rather than tapered - both types are viable options - it is all a matter of how you go about
constructing the bowl, how many ribs are required, how accurately you work in jointing the ribs and the allowances/adjustments made for the width of
the purfling as you proceed. 4.0 mm wide purfling seems excessively wide to me (do you mean that?) especially if you are working with 21 ribs.
Alternatively, tapered purfling may be made by carefully sanding each end of the 'open' side of the purfling to the required taper before gluing on
the second layer of veneer.
Either way, making the purfling this way should be pretty straight forward but time consuming and tedious.
If the joint between each rib pair has been accurately cut and prepared then there should be no need to make further adjustments by triming any
material off the purfling.
Johnathan, for my information and interest, you have confirmed that there is veneer on both sides of the purfling under consideration. Is this
purfling parallel sided or tapered? Also what is its width and how many ribs are in the bowl?
Interesting topic
GeorgeK - 9-9-2008 at 01:47 PM
I'll start by saying that I have no oud building experience what-so-ever!! However ideas are ideas so I'm posting mine.
If you look at the attached photo...
First, I'd glue up some strips of wood with the pattern matching what is desired for the purfling. This is seen as the top left diagram. I would then
shape this into the profile of the "purfling rib". This is seen as the top middle diagram. I would then place a laminent sheet on either side, this is
seen on the top right diagram.
Now, this structure can not be bent, at least I think it will come apart if heat is applied...hence I would cut out the rib profile as seen in the
bottom diagram.
This seems easier than playing with hundreds of little pieces, however I think that the tiny squares might appear distorted around the sharp bend in
the ribs.
Andy - 9-9-2008 at 04:40 PM
If I were to use spacers such as the type you show I would score the under side to make the spacer flexable since the glue would break down with heat.
jdowning - 9-10-2008 at 12:42 PM
Another possible solution that comes to mind would be to construct the bowl without any inter-rib purfling. The rib to rib joints could be fairly
inaccurate - within reason. After contouring the bowl by sanding, a groove or channel could be cut (accurately) over each rib joint - to about half
the depth of the ribs, say 1mm deep. Inlay banding of the required width might then be glued into the channel. The banding would only need to be just
over 1mm thick so could then be easily glued into the channel using hide glue and a hot iron.
Of course, accurately cutting the channel in the first place would be a challenge but some suitable tool might be designed for this duty with a little
thought.
Just an idea.
Jameel - 9-10-2008 at 02:19 PM
Interesting topic. I've thought about this many a time. I asked Nazih about this once, and he said that he places the spacers between the ribs as he
builds the bowl. Never asked him about how he actually makes the tiled spacers though. I think John's building board would be a precise way to do it.
Jonathan - 9-10-2008 at 03:54 PM
Wow.
Thanks, guys, for some great tips.
John--I don't know why I didn't think of something like that. Now that you have pointed it out, it seems sort of basic, which is always a good sign.
I should have taken a better look at the inside of that bowl before I sent it for repairs, to see if the tiles went all the way thru.
My concern, I guess, is that the glue joint is so amazingly thin, that I think I would have a good chance of going all the way thru when I routed the
channel. On the other hand, this might not matter that much, because another glue joint will take its place.
I really like that idea a lot.
Jameel--I don't know what he means. Could he place two ribs in position, leaving a gap between, and then, when they are completely dried at the neck
and tail block, add in the tiles?
George, I love the logic behind your plan.
Time to start making some tests, I guess.
Here's another pic of the instrument. I like how he switched over to thin maple purfling strips as he neared the tail block. It gives it a cool,
almost sunburst effect.
Jameel - 9-10-2008 at 04:34 PM
No, he doesn't leave a gap. That could never work since the gap would have to be dead parallel. I suppose you could leave a gap by using a spacer of
teflon or plastic---something that would hold the ribs apart but be removable afterward. I think he probably has the strips ready to go, just like
veneer spacers, and then when he has the rib fit, he just slips it in.
If you wanted to plow a groove along the joint line I wouldn't use a router. I'd make a curved fence (think a thick rib) that I could clamp in place
along the joint and then run a purfling cutter (two knives spaced apart--like for cutting violin purfling grooves) along the fence to establish the
width. You could then remove the waste between the lines with a chisel, or better yet, a small router plane like a vintage Stanley or Lee
Valley/Lie-Nielsen. That would provide you with good control and a relatively flat bottom.
Jonathan - 9-10-2008 at 04:41 PM
Gotcha. Thanks!
More skills I need to learn.
jdowning - 9-11-2008 at 04:19 AM
Jameel's concept for cutting the inlay channel over the glue joint should work quite well. Holding the fence firmly in correct registration while
using the purfling cutter would need to be figured out - possibly a frame that might retain the fence at both ends or supermagnets or both.
The little Stanley router planes might be a bit difficult to manage (working over a surface that curves both laterally and longitudinally). Hand
chiselling the waste might give better control to get the depth of the channel close to what is wanted and then to finish to uniform depth with the
kind of tool used by gunsmiths to cut channels for inletting gunstocks. This tool is essential a short length of file (smooth sides with cutting teeth
on the bottom) attached to a handle. Such a tool can be easily made so will post an sketch of what I have in mind later today when I get a chance.
The bottom of the channel might not need to be perfect as it will be filled with glue after the inlay banding has been glued in place. Early guitar
makers would sometimes set inlay banding and decorative 'lines' on a 'bed' of black mastic (instead of hide glue) which would be squeezed out along
the sides to make a 'perfect' glue line in appearance. Or they might just fill the grooves with mastic alone if the decoration was fine enough.
The channel cut by this method would, of course be parallel sided.
It looks as though the purfling on the oud is parallel sided as well?
jdowning - 9-11-2008 at 09:00 AM
The attached image gives a rough idea about the 'inlay plane' that I had in mind. A very simple affair - essentially a wooden block with the cutter
glued in place with epoxy resin. The underside of the block is a shallow 'V' section so that it will 'self centre' on the curved surface of the bowl.
The blade height might be adjustedto the required depth of cut for the inlay channel by gluing card or veneer of the required thickness on the surface
of the 'V'.
The blade is made from a piece of steel of slightly less width than the inlay channel with saw teeth cut either with a needle file or with a steel
cutting chisel. The teeth are cut straight across the blade and have no 'set' as a normal saw blade would have. This is in essence a miniature plane
maker's "float" or file which produces a very smooth cut. For occasional work, the teeth likely would not require hardening.
I make miniature chisels - suitable for cutting out the waste material from the banding channel - from carbon steel rod, heated to red heat with a
propane torch and hammered and bent to the required shape followed by oil quenching and tempering.
I shall make a prototype of this tool to see how well it works and will post the results later as a separate thread. However, if anyone wants to go
ahead and try their hand at making one feel free to do so. We can compare notes later!
To make the tiled inlay banding, the attached image illustrates the method I would use. The tiles start life as wooden sticks of the required section
glued together side by side and then glued as a 'sandwich' between two sheets of veneer. After the glue has dried, the 'sandwich' is then cut into
slices of the required thickness (about 1mm +) on a bandsaw using a fine toothed blade. The saw marks may then be removed (if necessary) using a
simple purfling thicknesser of the type that I am using in the thread " Old Project - New Lute" in the Projects Forum. The saw marks could also be
left - if fine enough - as they will be hidden in the inlay channel on one side and will be removed by final sanding on the outer face.
SamirCanada - 9-11-2008 at 09:44 AM
BRAVO!
that is really really clever!
Jameel - 9-11-2008 at 11:01 AM
Nice idea incorporating the float instead of the chisel. You probably wouldn't have more than a tooth or two engaged in the cut at any one time
because of the curve of the bowl, but the high angle of the float would provide a more controlled cut. I like your drawings almost as much as your
ideas, John. Good stuff!
SamirCanada - 9-11-2008 at 11:41 AM
I was thinking would it be viable to sharpen the blade and put a hook on it like on a scraper blade? that would avoid making all this precision filing
jdowning - 9-11-2008 at 12:29 PM
I am not sure that a single 'scraper' style edge would work Samir - as Jameel notes, only a single tooth would be cutting at any one time but you
would need two or three teeth to ensure that at least one tooth would be always cutting due to the longitudinal curvature of the bowl. Besides - due
to the small size of the blade - filing the saw teeth should not be a great problem. The best shape of tooth would be one cut with a chisel - like a
hand made file - but this might require a bit of skill and practice to 'get it right'!
I should say that I do not think that this proposal to cut a uniform, shallow channel for a banding inlay will be quite as easy in practice as it
might appear on paper - which is why it likely is not a traditional method used on oud bowls (is it?) But it would be a nice option if it could be
made to work easily, accurately and efficiently.
More than likely, between rib purfling has always been made to the full depth or thickness of a rib - probably because it is less time consuming in
the long run for an experienced luthier to do it this way.
Incidentally, for information, I have checked the old oud that I have and the purfling (conventional white-black-white) is about 3mm wide overall,
parallel along its length (more or less!) and is full depth.
That is one nice looking oud.
jdowning - 11-21-2008 at 07:24 AM
I have not yet explored further the possibility of cutting half depth banding or purfling grooves over the rib joints of an oud but came across this
bit of information that may be of interest.
Re-examining a museum drawing of the 16th C Dias guitar in the Royal College of Music, London recently, the drawing shows that the (ivory?) fillets
between the ribs are full rib depth but that there is also decorative purfling set into the centre rib that is only about half rib depth. The depth
(or thickness) of the ribs is about 1.2 mm and the purfling about 1.5 mm wide by 0.5 mm thick, a 'sandwich of kingwood between ivory - pretty delicate
stuff. The rib surface below the purfling is reinforced with a strip of parchment. The sides of the guitar are also decorated with the same half depth
purfling.
The guitar is a fluted rib vaulted back design. The longitudinal curve of the ribs is much less than that of an oud but the figure '8' sides have
tighter radius curves than the bowl of an oud and the surfaces are concave as well as convex. The question is how was the channel for the purfling
originally cut and with what kind of tool? Was the channel cut after the guitar body had been fully assembled or were the channels cut before the ribs
and sides were bent and the purfling glued in place after the body had been fully assembled?
I suppose that the easiest solution would be to cut the purfling groove prior to bending - although this might increase the risk of the ribs splitting
under the stress of bending.
Cutting the groove after bending might still be possible but guiding the tool accurately and, at the same time, maintaining a consistent depth of
groove, would be a very skilled operation.
Of course, to pre-cut a half depth channel over a rib joint would not be a viable option for an oud bowl.
jdowning - 11-21-2008 at 01:44 PM
I revisited Steven Barber's web site today to look for images of the fluted back Dias guitar. Under "Footnotes", Barber makes some additional
observations about the rib purfling - comments that were otherwise omitted on his museum drawing and notes. He notes that the rib purfling in the
centre of the rib is wavey (wanders from side to side) when viewed along its length and is not straight like a planed rib joint would be. Also, he
notes that, although the purfling is cut to half the depth of the rib, it is cut to full depth at the end of the ribs - giving the impression that the
purfling is full depth. Barber concludes that the purfling is cut to half depth because there is no evidence to the contrary - that would be indicated
by 'breakthrough' of purfling on the interior surface of the rib. He reckons that the linen (not parchment) strip reinforcement under the purfling was
applied before the purfling channel was cut - to provide necessary support.
As Steven Barber has examined the instrument at first hand, these observations are significant and indicate that the purfling channel in the rib
centre was likely cut by hand after the body of the guitar had been fully assembled - not before the rib was bent.
Incredible craftsmanship - the purfling channel, cut freehand by Dias (so it is not perfectly straight). But what tool did he use? My guess would,
again, be a type of gunsmiths in-letting tool - with teeth cut like a planemaker's float. The longitudinal grain of the ribs would need to be as
straight as possible to minimise tendency of the cutting tool to wander from a straight line.
Jameel - 11-21-2008 at 07:22 PM
Good points John. I would like to have seen him cut those channels. But you really think it was freehand? The waviness may be from age. Then again a
double-blade purfling cutter, with several light cuts, might ride the ridge nicely. You've obviously put a lot of thought into this, and I really
enjoy reading about your efforts.
jdowning - 11-22-2008 at 07:06 AM
Thanks Jameel. What I mean by freehand is to cut the channel to a scribed line by eye rather than by using some kind of straight edge as a guide. The
waveyness observed in the Dias guitar may be due to the cutting tool tending to follow the grain of the wood. Actually, if the degree of crookedness
is small this likely would not be noticeable in a finished instrument.
A readily available tool that caught my attention, as one that might be adapted for the purpose of cutting purfling channels, are jig saw blades - not
the cheap variety that have their teeth set like an ordinary saw - but the high quality ones that have no set. Instead the blades are taper ground to
the back of the blade to provide clearance for the saw dust.
The attached image shows an example of this kind of blade. The teeth are ground at alternating angles - like a chainsaw blade - the sharp points
initially making a scribing cut. This blade tapers from 0.92 mm at the teeth to 0.73 mm at the back of the blade with blade width of 7 mm. (0.036" to
0.029" , 1/4 inch wide). The blade has 10 teeth per inch so cuts very aggressively but I have made a few test cuts to see if there might be
possibilities. I was not able find any finer toothed blades of this type in the local stores so am not sure if they are made. However, the blade could
be easily reground with much finer teeth (using a Dremel tool) - cut straight across like a planemakers float so would have less tendency to wander
off course and follow the wood grain. A short length of the reground blade could then be fixed in a handle and used by pulling the blade for optimum
control (like Japanese saws). The length of blade projecting from the handle would provide a 'line of sight' for guidance along a scribed line (or the
rib glue joints in the case of an oud).
jdowning - 11-22-2008 at 07:11 AM
...... and here are a couple of trial cuts made with the saw in a piece of scrap figured maple. Straight grained wood would be easier to cut cleanly
and straight.
If a wider channel than 1 mm is required, two or more blades might be epoxy glued side by side together.
