Sunday, March 15, 2009

FRENCH POLISHING II - SEALING


REFERENCES AND SOME THOUGHTS ON FRENCH POLISHING IN GENERAL
I always find myself procrastinating a bit when I enter unknown territory - should I do this or should I do that? I found this to be particularly true when it came to applying the finish to this, my first guitar. I set out to do some research to prepare myself and as per usual I found a mammoth of conflicting advice, instructions and guidance. Everything from what type of cloth to make your muneca from, how to grain fill and with what to types of oil to used as lubricant, etc. etc. etc. seemed to have endless possible variables and arguments as to what and what not to use and do.

I also found a most of the advice and tutorials to be incomplete or at least leaving a novice like myself with a lot of unanswered questions, wondering why the author recommended doing things a certain way, and most importantly what to be careful not to do and how to rectify it if you did. It also annoyed me a great deal that almost all tutorials only showed pictures of the back being polished, conveniently skipping how to deal with all the tricky areas such as the bridge, the headslots.

That being said I also found some good references and writings on the matter. Many people recommended the Milburn Tutorial to be the most complete practical guide for the beginner and after having read it, I decided to use that as my starting point and main reference too. Actually I just recently discovered that there are now two different tutorials on Milburn's site. I think I may have used the other one for no particular reason other than that was the one I stumbled upon by default.

Some other resources I found to be very helpful companions to Milburn's tutorial was:
1) The Chapter on French polishing in Ervin Somogyi's book: "Making the Responsive Guitar", which is basically a transcript of the notes he wrote while taking lessons from Eugene Clark.
2) Cindy Burton's article "French Polishing with Eugene Clark" printed in GAL #54 (Red Book V).
3) "An introduction to French Polishing" By Cindy Burton, Greg Beyers, Robert Steinegger and Buzz Vineyard From GAL #14 (Red Book II)

Other references I used that should be mentioned here are among others Courtnall's book: "Making Master Guitars", Bogdanovich' book: "Classical Guitar Making" and LMI's article on "French Polish".

MATERIALS
The first thing I had to do was to choose the materials to make up the finish and something to apply it with. I quickly found there is a bewildering variety of shellac to choose from, each type of shape, refinement and origin providing some sort of unique quality. But instead of even making a slight attempt to figure out what it all meant, I decided just to take the easy way out and go with what Milburn recommended: Dewaxed flakes in a color of your choice.

Ervin Somogyi recommends buying shellac in large quantities to have a supply of consistent quality to work with as he claims that the properties of the same type of shellac can vary a great deal from supplier to supplier, sometime even from shipment to shipment from the same supplier. Though I'm sure this is sound advice, I felt it would be a bit of an overkill for me to start out with and so I opted to get a 8oz jar of 'blonde' flakes at my local Rockler store. I went for 'blonde' color as it was the palest they had and I wanted to get a fairly clear, non colored finish.

Secondly, I had to make a choice of what solvent to use. I read great many debates about whether to use denaturated alcohol or pure grain alcohol. Advocates of the denaturated variety seems to like it mostly because it is cheap and readily available. In a GAL article I also read that denaturated alcohol should be of more consistent quality as it is made under much stricter guide lines.

People who toot grain alcohol as the solvent of choice seems to be swayed mostly because of it's benign quality. This argument appealed greatly to me, but I soon found that it wasn't so easy to get hold of. Aparently it is illegal to sell the good stuff, 190 proof Everclear, in many states including California. Only the 151 proof is available here. However, I manage to find a store in New Jersey who were willing to ship 4 bottle to me. When I asked them if I could get in trouble for it they just said: "Ah, forget about it".

For application I went with Milburn's recommendation and cut up some old white t-shirts and underpants in 4x4" squares. I was quite surprise how many 4x4's I could get out of a pair of old undies. Maybe it's because I still wear army sized boxer brief. Anyone who has done their national service in the Danish army will know that their underpants only come in two sizes, too big and much too big. Moving on.....

PREPARATION
It was time to dissolve the shellac. Milburn suggest to make a 2lb cut for sealing and bodying and so I did. The 8oz plastic container from Rockler had conveniently divided the flakes up in four bags of 2oz each. I put the content of one bag in a glass jar and added 8oz of grain alcohol. I didn't grind or crush the flakes before adding the alcohol as some suggest to speed up the dissolving of the flakes but just left them as they were. It took roughly 24 hours for them to dissolve completely and that even despite the fact that the flakes had caked together into one lump.

Once dissolved I poured half the mixture into a 8oz squeeze bottle also bought from Rockler, covered my bench with a large piece of felt and I was ready to go.

APPLICATION
As per Milburn's suggestion I started off by sealing the purflings on the back in order to prevent the color from the rosewood back, bubinga in my case, bleeding onto the paler wood.

I folded a piece of 4x4" cotton in half and then once more into quarters. The corner of the first quarter was then loaded with about 12 drops of shellac which was wiped on the first purfling and binding following the contour of the guitar in one steady swoop without stopping at any time. Even though I don't think the bubinga really bled very much, it still showed some contamination. Whether this was from resins and oils in the wood or just dust, I don't know, but the cloth was not clean anymore.

To keep things clean the cloth was refolded to a fresh, uncontaminated section, reloaded with about 6 drops of shellac and the second back purfling was sealed, also in one steady swoop. Refolding to the third clean quarter, the cloth was reloaded with a few drops of shellac and the center strip was sealed the same way.

The last clean quarter was recharged and used to seal each section of the back with a coat of shellac, however this time the shellac was applied using a circular motion. After the whole back had been sealed the cloth was disposed of.

It is important to note that I didn't use oil for this. In the tutorial I followed, it stated: "use no oil or alcohol" and so I didn't. Apparently, the other Milburn tutorial advices: "then apply a drop of oil". This is not, in any way, meant as a criticism of the Milburns as I found their tutorial exceedingly helpful, but it is a prime example of how easy it is for the novice to get confused.

The sides were sealed in a similar fashion. A fresh cloth was loaded and first the purfling and bindings were done in one single sweep, then the joint between side and the neck, refolding the cloth to a fresh quarter and reloading it after each sweep. I winged it a bit when I got to the inlay at the butt, by using that quarter for the side as well.

I used two cloths for the soundboard as I quickly ran out of fresh quarters, having to deal with both purling/bindings, the rosette the bridge and the fingerboard as well as the soundboard itself. But in essence it was done like the back and sides, in one continuous sweep where dark and pale woods were joining and finished off with circular motion on the spruce itself. The same technique was used on neck/fingerboard joint and around the headplate veneers.

As the application of the shellac is done with some pressure it was sometimes hard to keep the guitar steady. This was particularly true when I did the back, as in this case the guitar was resting on the bridge and neck only. Doing the areas close to the sides did get a little hard to control. I've only just become aware that Milburn suggest using a neck rest while polishing. For some reason I must have skipped that part of the tutorial, hence my troubles. I believe Eugene Clark clamps his guitars by the neck while polishing.

For those of you who wonder if a person from Blue Man Group came to help out during this, I can attest that it was in fact me wearing a protective nitrile glove from Lee Valley. I think it was probably spurred on by the photographs of Bogdanovich from his book, applying shellac sporting a similar outfit. But while they did prevent the shellac from getting onto my hand I don't think they were necessary, apart from the fact that I didn't have to wipe dried shellac off your hands afterward and then even so.

Being a one man band, it was particularly hard to photograph this process. As the shellac dries almost instantly I kinda' had only one shot at each process, hence the somewhat odd placement of items in the shots here and there. It was all a bit nerve wrecking.

SECOND PREPARATION
After the first application the guitar was inspected again only to find there was several areas that needed some further attention before I proceeded.

First of all I noticed a few areas where the bindings hadn't been scraped flush with the sides. Actually you can see one of these in my previous post, French Polish I - Preparation, in the close up photo of the rounded edge. I also noticed some ripples in the surface of the back around the heel cap where I had scraped it flush. I think it must have stemmed from scraping in the same direction making the scraper follow and accentuate the ripples as the work proceeded. I hindsight I think these could have been avoided had I just changed direction scraping diagonally every so often.

I also noted a tiny bit of bleeding from the bridge onto the soundboard, but as it was so only very little I decided to let it slide.

Last I found a few area around the heel where traces of chisel marks were still visible, as were some planemarks on the back. These were all carefully sanded back again using StewMac's Stikit120 grit followed by 220 grit. After this the whole guitar got another two seal coats for a total of three.


MY FRENCH POLISHING MATERIALS

SEALING BACK STRIP

SEAL COATING BACK

SEALING SIDES BINDINGS AND PURFLINGS

SEALING ROSETTE

SEALING EDGE OF BRIDGE

SEAL COATING SOUNDBOARD

SEALING FINGERBOARD EDGE

SEALING NECK

SEALING HEAD

CONTAMINATED WIPING CLOTHS

BLEEDING FROM BRIDGE

BINDING DETAIL

ROSETTE DETAIL

HEEL DETAIL

THROAT DETAIL

HEAD DETAIL

GUITAR FINISHED WITH THREE SEAL COATS


Sunday, March 08, 2009

FRENCH POLISHING I - PREPARATION


FILLING GAPS AND FIXING DENTS
The first thing I did was to fill the gaps at the fret ends. This was done by melting a stick of dark brown shellac into the gaps with a soldering iron. I don't know if a soldering iron was the best tool for this as I got the feeling the iron was too hot and burned the shellac as oppose to just melting it. But it was what I had and so it was what I used. Maybe a little alcohol lamp and a pallet knife would have been more suitable. Next time...

Then the whole guitar was inspected from A-Z in an attempt to locate all the dents, gaps, scratches and other imperfections. Inevitably there was a few around, but generally speaking it wasn't too bad. The two worst cases was a fairly deep ding in the soundboard and a tear out from the spokeshave on the side of the fingerboard. That occurred when it was trimmed flush with the neck.

I was considering using the shellac stick to fill the tear out on the fingerboard, but after my experience with the soldering iron I decided just to fill it with some 5-min epoxy, which seemed to work very well. I thought of tinting the epoxy with saw dust but decided not to and just use it clear. It somehow seemed to make a more invisible repair showing the grain of the wood as opposed to having a homogeneously colored blob of filler covering it up.

I made an attempt to steam out the ding in the soundboard using the soldering iron and a damps cloth the create steam. I went at it for quite some time but even though it did lift the indentation, I wasn't able to raise it completely. Eventually I gave up and decided to leave it till later and then decide how to proceed to fix it, if at all.

Other stuff I noticed included small gaps around the bindings but these seemed insignificant enough to be left alone.

SANDING, SANDING AND MORE SANDING
One of the more hidden flaws I detected were tiny ridges left from the plane blade on the back and sides. At this point I had not done any sanding on the surface of the body and these ridges became apparent once the body was held up against a low light. This took me by surprise as I had be very careful to avoid this very thing. But there they were laughing me straight in my face.

I also noticed some waviness running across the grain of the sides together with a few places where the bindings hadn't been trimmed completely flush. I carefully used a scraper to level all these blemishes the best I could.

Next up was rounding the sides. Some people like to create a very soft rounded edge around the body, but personally I prefer a sharper, more defined edge.

In his book Ervin Somogyi states that Eugene Clark recommends using a smooth file for this, working with deliberate stokes in one direction only. I tried this, but I never really got to grips with it. I couldn't really see the benefit of this approach and I had a hard time working in a smooth edge around the inside curve of the waist area. Soon I switched to a cork backed sanding block with 120 grit garnet paper which I had an easier time with. First the edges was treated to a small single bevel that then was rounded over. Having the bevel made it easier to control the homogeneity and evenness of the edge all the way around the body.

The head needed some attention too. The inside of the string slots showed traces of chisel marks from when they were done initially. Likewise a few marks from the drill bit were traceable a the end of the slots.

The inside of the slots were sanded with a flat sanding stick with some of StewMac's Stickit paper, first with 120 grit followed by 220 grit. The rounded ends got a similar treatment though this time a piece of dowel was use as a sanding stick. Like the body the edge of the head and string slots were softened a little, though not quite as much.

WIPING, SANDING, WIPING, SANDING AND WIPING
Now the whole guitar was wiped with Naphtha and inspected once more to see if any other blemishes were showing up. As it looked fine to me I proceeded to sand the entire guitar with 120 grit garnet paper. I really don't like sanding but I guess it has to be done. All that dust everywhere, yuck!

Then it was time to gently wipe the guitar with a very tightly wrung damp cloth. This raised the grain a little and got rid of most of the dust left on the surface from the sanding. The raised grain was leveled by a final sanding using 180 grit garnet paper.

Some people continue sanding ending up with much finer grits, but Eugene Clark recommends stopping a 180 or 220, in order to avoid clogging up the pores with dust, or chocking it as he puts it. I believe Cumpiano recommends stopping at 220 too. Either way it suited me fine keeping sanding to a minimum.

When all was done the entire instrument was dusted and wiped with Naphtha again to get rid of the last bit of dust allowing for one final inspection.


FILLING FRET SLOT GAPS


FILLING TEAR OUT WITH EPOXY

STEAMING OUT DENT

SCRAPING SIDES

SANDING BINDING EDGES

ROUNDED BINDINGS

SANDING FLAT HEAD SLOT SIDES 120 GRIT

SANDING ROUND HEAD SLOT ENDS 220 GRIT

WIPING GUITAR WITH NAPHTHA

FIRST SANDING 120 GRIT

WIPING GUITAR WITH DAMP CLOTH

FINAL SANDING 180 GRIT

WIPING GUITAR DOWN WITH NAPHTHA

THE DETAILED GUITAR

Monday, March 02, 2009

THE BRIGDE

INITIAL THOUGHTS ON BRIDGE DESIGN
In my early twenties I built windsurf boards with a friend of mine, Ole. Ole is a structural engineer and his background had a big impact on the way we built. Apart from shape, the basic mantra was as light and stiff as possible. Surfboards are typically built around a foam core onto which several layers of glassfiber is laminated. What we ended up doing was, to cut the core straight down the middle, hollow out each half and then glue it back together again. On the out side it would look like any other board, but on the inside holes would run across the entire board from the nose to the tail. This was a fairly labor intensive process but it paid off. We managed to make them about 8 or 9% lighter than had we not hollowed them out. It produced incredibly responsive boards; quick planing, fast and easy to maneuver.

A while back I read an article by Ervin Somogyi where he explained that a heavier bridge would increase sustain but have a slower response and lower volume, while a lighter bridge would have the opposite characteristics. I have also read about builders that experimented by adding weights to the bridge and various places of the soundboard in order to bring out a certain sound. One could argue that the 'optimal' classical guitar sounds a certain way exactly because of the traditional bridge design is the way it is, and monkeying with that will produce a less desirable sound quality even though it might be louder and more responsive.

While that certainly may be the case, I couldn't help but think that maybe, like our surfboards, the guitar could benefit from having a light bridge regardless. After all, much effort goes into maximizing the stiffness to weight ratio of the soundboard as it is, so why not extend that to include the bridge. Off I went.

SHAPING THE BRIDGE BLANK
Like the fingerboard the bridge blank was first cut out to a slightly oversized blank on the bandsaw. However, this time both edges and ends were trued to exact measurements, 25mmx165mm, with a low angle block plane. At this stage I weighed the blank and it came out to be 37.2 grams. As with the fingerboard I had to be very careful to avoid tear-out.

Next a slight concave was scraped into the bottom of the blank, matching the dome of the soundboard. This was done by 'eye'. I had taped a piece of 150 grit sand paper to the soundboard and used that to scratch the gluing surface of the blank, as a guide to where more scraping was needed. This was done until the whole gluing surface of the blank got evenly covered with scratch marks. These scratch marks were cleaned up with a few light strokes by the scraper.

Some people advocate roughing up the gluing surface with coarse sandpaper, to create better bonding gluing surface. I've even seen people using a Stanley knife to cut marks in the bottom. I don't really agree with that practice. In my book the cleaner and smoother the gluing surface the better. Each to their own.

STRING HOLES
Next up were the sting holes. The string spacing at the bridge for this small instrument is only 10mm, making them span 50mm altogether. It's not much, but should fit a child's hand well. As there was going to be a little 2mm high string shelf built into the bridge to protect the soundboard, the holes were drilled slightly higher on the side to end up in the centered between the shelf and the top. To counter this the holes were angle downwards too to increase the string break angle.

I clamped a 1x2" piece of oak to my drill press as a fence, and added a 1mm spacer to the bottom of it to achieve the desired angle. In order to keep the bit straight as possible while going through the work, I did the drilling in many steps, going a little bit deeper each time. This worked really well and kept the bit from getting clogged up and wandering. A 5/64" brad point bit was used for this.


SADDLE SLOT, STRING TROUGH AND STRING SHELF
This procedure was started by marking up the tie-block and making two preliminary cuts across the blank with a dovetail saw, indicating the location of the tie-block. I had attached a thin piece of maple with some spring clamps that functioned as a simple depth stop, keeping the cut from going deeper than the thickness of the wings.

To route the slots for the saddle and the string trough, I made a jig based on the one shown in Roy Courtnall's book. I modified Courtnall's version by making a kind of shelf onto which the bridge rested upon. The shelf slit into the main body of the jig and could be tightened and locked in place by four hex screws. This modification will enabled me to use it for a wider range of bridge sizes and shapes in future.

Both the saddle slot and the string trough were routed in three goes to a total depth of about 5.5mm, 7/32", leaving about 3mm or 1/8" of solid wood under the saddles. I had to go a little deeper on the string trough to expose the string holes fully. This was done with the width of the tie block as it would cut into the top of the wings otherwise. I used a 1/8" downcut spiral bit for both slots.

The little string shelf was done by planing a small rebate with a shoulder plane. By placing a 2mm thick piece of scrap from one of the sides under the plane I just went at it until the the desired depth of 2mm was achieved. It was slow work but doable. If I'd had a router table at my disposal that would definitely have been an easier and faster way to do it, but I didn't. Next time around!

Then the waste above the wings were trimmed on the bandsaw and last the bevels at the end of the wings were cut as a miter using my shooting board and the miter jig I made for the rosette log. The shooting was done with my low-angle jack plane.

CHAMBERING THE TIE BLOCK
As I mentioned above, I wanted to try and make my bridge lighter. One obvious place to start was by hollowing out the tie block and so I did. I figured that with a 10mm string spacing and 2mm string holes I had 8mm to play with, but as this was my first go at it ,I decided stay on the conservative side of thing. I calculated that using a 3/16" drill bit would give me a wall of roughly 1/16" thickness between the chamber and the string hole, This I reckoned would leave amble strength in the bridge. Ole later confirmed this and recommended I could go thinner next time with out problems if I wanted to.

First the tie-block height was reduced about 1mm to accommodate for the cap that eventually as going to cover the holes. Then the holes were marked slightly offset towards the back as I wanted a bit of strength to the front of the tie block where the strings are pulling upwards as they bend over the saddle. The holes were drilled through the tie block leaving only about 1mm of wood at the bottom.

I weighed the bridge before and after the drilling and found that I had managed to reduce the weight only a mere 0.7 grams from 21 to 20.3 grams. I was a little disappointed with this, though it turned out that the final bridge weight of 18.7 grams had been reduced by almost 4% which was not really that bad. Had I gone with a 1/4" bit instead I'd be able to get that number up to over 7% instead.

THE TIE BLOCK CAP
The tie block cap was made from a 2mm thick piece of scrap I had left over from the bridge stock. As the total width of the tie-block was 10mm, I calculated the center wood strip would be around 4.5mm or 3/16", Leaving space for the two bone pieces each 1.6mm or 1/16", and 4 contrasting veneer lines of a total of 2.4 mm, roughly 3/32". The wood center was trimmed to size with a block plane.

Then the cap was glued together with the veneer lines and bone pieces. I didn't take any photos of this, but basically it was done in a similar fashion to my backs and soundboards; by clamping them from the side and top at the same time. From the side to glue them together and from the top to keep it all flat. Once it was all dry I thinned cap to about 1mm and trimmed the width and length to fit the tie block accurately.

I prepared a bone piece to fit snug into the string trough, it will later become the saddle, and used it as a stop to fit the cap up against. Then glue was added to the tie-block and the cap was carefully clamped on with three 1" spring clamps. The clamps were positioned so if any movement was to happen it would be towards the bone piece. The bone piece was covered in wax paper to protect it against squeezed out glue.

SHAPING THE WINGS
The rough shape of the wings were done with a file. Again, here I used facets to control the progress and symmetry, by first working in a primary bevel then two secondary bevels. Once I was getting close to the final shape I switched to sanding. I made a radiused sanding block for this.

By knowing that the wings were 25mm wide and 4mm high minus a 1mm edge, making t
he peak of the arch 3mm high, I calculated the radius of the arch to be just over 27.5mm making the diameter somewhere between 2 1/8" and 2 1/4". With a 2 1/4" forstner bit I drilled out a piece of MDF and cut it into three 120˚ segments, glued them together creating an arched sanding block that was covered with sandpaper. I figured that the thickness of the sandpaper would make up for the oversized, probably not to complete accuracy but close enough.

With some 120 grit 3M Stikit paper on the sanding block I made a few preliminary strokes on each wing to see where the high spots were and the proceeded to work those back with a file
, while monitoring the progress carefully. Once the wings were down to the final height I finished it off with witch 220 grit.

BEVELLING THE TIE-BLOCK, STRING TROUGH- AND THE SADDLE WALLS
To finish off the bridge I beveled the two walls that holds the saddle and the tie-block. The outside wall was done first. I Stuck in a piece off wood into the string trough, turned the bridge upside down, which held the bridge at an angle. Then the bevel was trimmed back on the shooting board with a block plane.

The wall adjacent to the string trough was pared back with a chisel, as was the tiny bevels on the tie-block ends. The bone segments on the side of the tie-block was beveled with the block plane too.

GLUING THE BRIDGE ON
For this I made a jig out of plexiglass to locate the bridge placement. I made it 581.5mm long, 580mm scale plus 1.5mm compensation, and 66mm wide, the width at 580mm when following the fingerboard taper. Then a center line was scribed down the length, as was the 12th fret across. Measuring from the center the width of the nut, 46mm, was marked at the nut end and two lines following the taper of the fingerboard was scribes on each side of the jig.

I put rough cut nut in place and butted the jig up against it and clamped it to the fingerboard with two cam-clamps, one above the 1st fret and one above the 9th. The jig was carefully aligned with the fingerboard having the two scribed outer lines following its taper.

With a rough cut saddle in the saddle slot, I was now able to place the bridge in the correct location by butting it up against the jig. That is, I skewed it 1mm toward the bass side in order to have more 'shelf' on the fingerboard on the treble side.

To prevent it from sliding out of place during gluing, three layers of blue masking tape was placed up against and around the periphery of the bridge. Then the notched gluing caul with a 25' dome was placed inside the guitar over the fan bracing and the bridge was ready to be glued on.

For the actual gluing Titebond was used again. I like Titebond! I made three cauls
out of 1/2 plywood to go on top of the bridge. One for middle and one for each of the wings. The two that went on top of the wings had thin lines of cork on the edges to locate the clamping pressure at the edges of the wings. I had bought four fancy bridge clamps from StewMac, but while they seemed to do the job OK, I was somewhat disappointed with them. They seemed very flimsy and wobbly and I would have preferred them to be a little more sturdy. The assembly was left to dry over night.

And that is all I have to say about that!



CUTTING UP THE BLANK

TRUING THE SIDE OF THE BRIDGE BLANK

TRUING THE END OF THE BRIDGE BLANK

CONCAVING THE GLUING SURFACE

SANDING THE GLUING SURFACE

MARKING STRING HOLES

DRILLING STRING HOLES

CUTTING WING REBATES

ROUTING SADDLE SLOT

PLANING THE TIE BLOCK REBATE

TRIMMING OFF THE WINGS WASTE

BEVELING THE WING ENDS

DRILLING TIE BLOCK CHAMBERS

CHAMBERED BRIDGE AND TIE BLOCK CAP


GLUING THE TIE BLOCK INLAY

FACETING THE WINGS, PRIMARY FACET

ROUNDING THE WINGS

BRIDGE WING SANDING BLOCK

FACETING SADDLE SLOT WALL
FACETING TROUGH WALL
FACETING TIE BLOCK

THE FINISHED BRIDGE

THE FINISHED BRIDGE WITH CAUL

LOCATING BRIDGE POSITION

GLUING BRIDGE

GLUING BRIDGE