BINDING GAP REPAIR

After lots of research and speculation I have decided to try and repair a small gap that occurred in the first binding/purfling strip I added. Actually, it's two gaps, each about 0.5 wide, one between the binding and purfling and one around the rosewood veneer line between the purfling and the soundboard.

Had it only been a gap between the binding and purfling, the matter would have been much easier to resolve as I could just have filled the gap with sawdust and glue. But, it's the gap around the rosewood line that bothers me. If I was to fill that I would give a visual appearance of the line growing twice as thick for about 1" and slimming down again to it's normal size. Not good. The question next was how to to fix it.

A LITTLE RESEARCH
I had numerous correspondences with fellow forum members as well as a telephone conversation. Several suggestions as how to fix the problem were mentioned. One was to just add some more glue and clamp the binding in. Another was to make some sort of space at the end of the binding, heat it all up to soften the glue, move the binding into place and reclamping it. Yet another, a variation of the second one, but also adding fresh glue before clamping it.

To figure out how Titebond glue would react to heat I ended up calling Franklin International's, the manufacturer of the Titebond glues, tech support to get the inside scoop. A very friendly tech person informed me that all the Titebond glues does indeed do work as hot melt glues and not only once as some had suggested, but repeatedly so. He recommended heating the glue to a temperature around 240-250º F for best results. I asked if it would be necessary to add more glue to which he responded that as long as there was glue in the joint in the first place adding extra glue shouldn't be needed.

MUCK UPS
I did two little muck ups to run experiments on. One was a 4" long replica of a side with lining and soundboard attached, including binding a purlings with gaps like on the real guitar. The only difference was that I it made from walnut/walnut/spruce rather than bubinga/mahogany/spruce and that it was straight. Otherwise it was dimensioned pretty accurately and it was glued together with the same glue. The other muck up was just some binding I had glued to a piece of bubinga. The idea was to start practicing on the latter just to get a feel for how much heat I need to supply to get the glue soft and then move on to the more elaborate one for a complete trial run.

During the process I tried to reclamp part of the binding just as it was and another part with extra glue added. For some reason I had better luck with getting the binding to stick when I added more glue. I don't know exactly why that was, but I speculated that it may have been because I didn't keep it clamped long enough for the glue to cool down and solidify. Or, maybe I didn't heat it up enough to properly melt and therefore not being able to adhere properly when it got clamped again. Who knows, but because of this I decided that it was probably best to move ahead with the job by trying to open the joint up a little and add more glue.

DOING IT
After some serious procrastination, I took the plunge and got on with it. I opted to drill two small 1/8" holes, one for the purfling and one at the binding centered around 6mm, 1/4", from the side of the neck. This allowed for 1mm for the neck still to have to come off in the final carving process, 1.5mm of slope in the heel from the top to the bottom of the binding, 1.5mm radius of the drill bit and still have at least a 2mm thick wall between the hole and the heel.

I carefully started drilling a shallow hole for the purfling, though it ended up a little deeper from the repeated attempts to get a decent photograph of the process, and a slightly deeper hole for the binding matching it's width. I used my caliper to measure the progress by sticking the depth measuring blade into the hole. I used a hand drill for this which worked really well for this with it's slow operation; No unpleasant surprises of the bit suddenly digging in and coming through the neck.

Next came the softening of the glue. I was very nervous about this. Not that it would be difficult to heat up the glue for the binding, but because I was afraid of heating up the glue around the linings too, weakening the joint between the soundboard and sides.

With the tip of a household iron I began to heat up the binding, being as careful as possible to keep the iron on the binding only and not overheat the area in general. It's a good idea to empty the iron before doing this to prevent the water from the iron to spill all over. Don't ask me how I know this. With a drill bit inserted into the hole I gently tried to pry the binding away from the neck to create a small gap in the joint where I could squeeze in some more glue. But, that didn't do the trick. Instead of the binding on the body opening up, the little bit of binding left on the other side of the hole gave way.

Instead I resorted to just heating up binding to soften the glue around the gap and reclamping the bugger in place.

RECLAMPING
I made as small semi circular cork lined caul out of MDF, that fitted the diameter of the soundhole. I used a flycutter for this. I don't really like using flycutters but the tool worked really well for this application, as the bevelled end of the cutter created a little lip that held the caul in place an prevented it from falling though the soundhole. I did a few dry runs before the real deal, to figure out which clamps to use and where to put them. I found that using three wooden cam clamps covered the troubled area pretty well.

Once that was sorted out, I reheated the binging again as before, added a little bit of fresh glue on the gap and scraped as much glue as I could down into it with an old credit card and added the first clamp furthest away from the neck. This pushed the binding in place where the clamp was and opened up the gap a little more in front of it. This procedure was repeated for the second clamp, though the gap in front of the clamp did not seem to widen further when it went on.

The third clamp went on about less than 1" away from the neck and managed to close the last bit of the gap. I wasn't able to detect whether the end of the binding had moved further in to it's pocket in the neck or not, but speculated that it must have done just that and it only a very small forward movement was needed to press the binding in place. The three clamps were left in place overnight for the repair to dry.

Once the clamps were off, the binding was scraped flush with the sides and the clue was cleaned up on the top. On close inspection of the repair revealed that the binding was as good as new. I was still able to trace an approximately 0.2mm gap on one side of the rosewood veneer line but it was a lot less than before and not something I think anyone will notice unless you know it's there.

All in all I was quite pleased with the end result, particular in the light that this it was a task I dreaded doing and one that had been hanging over me for a some time.

Onto the fingerboard....

BINDING GAP

BINDING GAP CLOSE UP

MUCK UPS

MAKING SPACE FOR THE BINDINGS TO MOVE INTO

HEATING THE BINDING

REGLUING THE BINDING

SCRAPING BINDING AND PURFLING FLUSH

THE FINISHED REPAIR

THE HEEL CAP

DESIGN CONSIDERATIONS

I wanted to make the heel cap into a little appointment that would tie in with the overall color scheme of the wood parts and ornamentation and thus go a little further than just letting it be a little bumper to protect the softer heel/neck.

In this case I chose to echo the look of the binding and purflings, making cap itself the same thickness as the bindings with a maple/bubinga/maple veneer sandwich underneath that matched the purfling. At the very bottom a slice of bubinga veneer was added to create a contrast between the heel itself and the maple/bubinga/maple veneer sandwich. Of course this is not a new concept nor is it an exceptional idea in any way, but there you have it.

Another design element I had to consider was how to deal with the nonparallel planes of the stacked heel and the domed back. In order not to creating a slight change of angle where the cap meets the back, I figured I had two options:

• To have the cap follow the doming of the back by way of sloping the heel step parallel to the tangent of the back and thereby creating a slight downwards angle towards the tip.
  
• To follow the plane stacked heel, but tapering the cap in order for it to continue the tangent of the back.

I decided on the first one as I thought it would be the simples and cleanest looking. Yet another option could also be to make a thicker bottom veneer out of bubinga and then taper that, but I don't know how well that would have worked.

MAKING THE CAP
One good thing about resawing your own wood is that you inevitably end up with lots of scrap wood that is ideal for these purposes and thus I had a nice chunk of 3/4" bubinga that suited the bill perfectly.

I cut off a little piece 1 1/4"x1 3/4" and sliced a 3/16" thick cap off of that. From the leftover I fabricated a 1mm thick veneer that was to go on the bottom. Three similarly small sheets of veneer, two maple and one bubinga, were cut and it was all glued to a sandwich and trued further to a slightly oversized rectangular cap.

FITTING THE CAP
Next up was trimming the heel down to a step that allowed for the veneer sandwich of the heel cap to be aligned with the side purflings. This was done with a 1" chisel first in big chunks and then slowly doing thinner and thinner slivers as the worked progressed to it's final depth.

The binding was trimmed flat over the heel. I didn't take any photographs of this but basically the slight curve in the area over the heel where the bindings meet was pared down to a flat. It wasn't really very much to come off but the curve was pronounced enough to make a nice butt joint with the heel cap impossible.

Using the shooting board, a slight bevel was planed on the back of the heel cap to accommodate for the angled heel that followed the curvature of the back. I didn't use any measuring for this but winged it by resting the end of the heel cap under some scrap binding, to lift it up at an angle. By trial and error I kept moving the binding in until the angle was steep enough to make a nice fit.

GLUING THE CAP
Once everything as aligned and fitted the heel cap was cut on the bandsaw to a slighly oversized cap that was ready to be glued on. A la Cumpiano I used two cam-clamps for this. The clamps were positioned at a slight angle, so any sliding that could occur in the wet glue would be towards the body.

A note on the bandsaw operation. Getting my fingers this close to a running blade was not my favorite way to spend an evening and I only did it because the friendly photographer was very insistent. Next time I think I'll be using a coping saw or just chisel away at it once it is glued on.

CARVING AND TRIMMING THE CAP
After the glued had dried I pared the heel cap down, close to a final shape. The last bit will be taken off once I get to the final carving of the neck and heel.

For now the cap was shaped in facets as I did with the initial heel carving. First the sides were taken down to fit the width of the heel. Then I carved the tip to a flat. After these two reference planes were done a primary facet were cut, followed by two secondary facets. Last the facets were cleaned up to a smooth round curve that extended the shape of the heel.

The cap also had to be trimmed flush with the backplate. For this I used a block plane to get rid of the bulk of it and finished it off by scraping.

THE HEEL CAP COMPONENTS

GLUING THE HEEL CAP TOGETHER

PARING THE HEEL STEP TO PROPER HEIGHT

MEASURING HEEL STEP DEPTH

TRIMMING THE BACK OF THE CAP TO AN ANGLE

PRELIMINARY SHAPING

GLUING THE CAP ON

CARVING HEEL CAP WIDTH

FACETING THE CAP

PLANING HEEL CAP FLUSH TO BACK

THE FINISHED HEEL CAP

THE FINISHED HEEL CAP

THE FINISHED HEEL CAP

THE FINISHED HEEL CAP

BINDING THE BODY

THE END GRAFT

The end graft was made from leftovers of the back strip. To cut the channel for it I clamped the body of the guitar to the bench with two 8" handscrews, neck pointing down. The handscrews themselves were clamped to the bench with two bar clamps and had their jaws padded with cork.

This was a trick I picked up from Cumpiano's book, though he used four cam clamps instead, two above the bench and two below. That would probably provide a little more stability, however the jaws of the handscrews seemed to be wide enough to give adequate support.

With a small dovetail saw butted up against a square two cuts were made. The square was held up against the soundboard so the cuts would be just on the inside of the back. I had made a stop for the saw so it wouldn't cut any deeper than the sides themselves, about 2.0mm. After the cuts were done the waste was cleared and the channel was carefully widened with a 1/4" chisel until the end graft went in with a snug fit. Well, I didn't end up with quite as snug a fit as I would have liked, but close enough.

The endgraft was clamped in with two long cam clamps, a corked MDF caul and some waxpaper. Once the glue had dried I first planed, then scraped the graft flush with the sides.

CUTTING THE BINDING CHANNELS
For this I had bought a BOSCH COLT laminate trimmer fitted with a 1/4" spiral downcut bit and a binding router base from Luthiertool. I was quite nervous about using the tool as I'm not very familiar with using powertools. Actually I don't really like using them. I always feel the you just cough and whoops a big chuck of wood has gone.

I did a lot of test cutting in various wood, some soundboard scrap in particular. I had heard that it was easy to get tearout when doing this and I thought it best to give it a spin, so to speak, to see for myself. To my surprise I didn't encounter any problem at all. I routed two circles, one in each direction, and both came out as clean as a baby's bottom.

Encourage by this discovery I felt I was ready to go at it on the guitar. I clamped the guitar by the neck shaft into the patternmakers vise and this proved to be a good way to hold the body while routing the channels.

First up was cutting the end graft to size. As I intended to miter the purfling lines The center of the end graft has to protrude up from the binding channel. I carefully adjusted the depth and width of the cut and carefully routed little steps at each ends of the end graft. So far so good.

Next up was the binding channel. I started on the back and moved confidently along about 2" when I discovered the channel looked a little wide, and sure enough somehow I managed to add a 1/16" to the width. I don't know how it happened, but I have a feeling it was the locking screw that hadn't been tightened enough and had come loose. Regardless, I had to reset the width and try again. Second time went better. I flipped the guitar over and started the soundboard side.

This did not go so well, I started to experiment with holding the router in different ways and learned that it is best controlled when holding the fence with one hand while the other hand holds the base itself. I tried to have the other hand hold the router too and this appeared more comfortable to me and provided easy access to the on/off switch. However, I also learned that holding the router this way gave less stability to the hole unit, resulting in tilting the router towards the guitar and thereby widening the cut.

I bit flustered by this setback, I decided to sleep on it and continue the day after with, hopefully, fresh eyes. Before, I left it I had the 'ingenious' idea to maybe I could build up the troubled spot with some veneer and so I did. The next day I went over the trouble spot with the router and it seemed to do the trick.

Next up was the purfling channels. These were done in the same manner as the binding channels though the router had to be adjusted for the wider more shallow cut. Again testing was done on scrap. I first cut the back and then the soundboard side. The cut on the soundboard side had to be adjusted to an even wider channel as an extra rosewood veneer band was going to be added to contrast the maple purfling line with the soundboard spruce.

Last the end of purling and binding channels were completed with a 1/4" freshly sharpened chisel by slowly paring the waste away. At the end where the purfling channels met the end strip and back strip, particular care had to be taken not to let the chisel slip and cut into the purflings surrounding them. At the neck shaft a razor saw was used to cut the binding channel across and a 1mm chisel to clear the waste. At the soundboard the web from the neck was left intact and the channels were stopped at about 3/4" into the neck. On the back the binding channel was continued through the heel and the purfling channel stopped at the back strip.

Cutting the last bit of the channels with just a chisel, ensuring continuation of the smooth, flowing curves while maintaining accuracy was quite challenging. A purfling cutter would have been very helpful, but patience, diligence and a fair amount of trial and error did go along way.

Last the all the ends of the purflings on the back and end strip was trimmed to a 45° miter. I used the reflection in the polished back of the chisel as a guide for this.

FITTING THE BINDING AND PURFLING STRIPS -

THE FIRST SOUNDBOARD BINDING STRIP

Following the books I have it appears that each author have their own way of going about this. Cumpiano glues all the elements on at the same time but a few inches at the time. Bogdanovich and Courtnall and others do the purflings first then the binding.

Since both my purflings and bindings were a little on the high side I though it would be best to glue them all on in one go. This way the purflings would sit snugly in the ledge between the soundboard and the binding with little possibility of tilting when taped on. Other than that I wanted to try out different scenarios, both adding glue to the whole set-up at once as well as in stages per Cumpiano.

First to go on was the top right binding (looking at the top with the neck pointing upwards). I had thought of bending the extra line of rosewood veneer that but it was quite bendable in itself and so I didn't really see the need for it.

Both the binding and purlings were taped on dry and the ends were marked with pencil and cut to length at the neck and slightly long and the tail. Furthermore the side purflings that had already been glued to the sides had to be mitered. This was done right on the mark and the waste cleared with a chisel. Like on the back and end strip I used the polished back of the chisel as a visual aid to get a perfect 45º angle.

A final dry fit was performed and a few adjustments were made to get a snug fit and everything was ready to be glued on.

On the first strip I wanted to try and glue it all in at once and so I added glue, Titebond II slow set, to the binding channel, the outside of the purfling and the rosewood veneer strip and spread it as quickly as I could. All the strips were held in place and the first tape strip was added about 1" away from the tail end. This gave a clear view of the the mitered joint between the side and end strip purflings and a chance to correct errors before the very end was taped on, a tip I lifted from Cumpiano's book.

In order avoid having the strips move about too much smearing glue onto the soundboard I loosely put tape strips on about 4" apart until the whole binding/purflings were secured. Then the binding was firmly taped on adding strips 1 - 1½" apart starting from the tail all the way up to the neck.

However, the next day I realized that was not a good way to go about things as the binding and purflings had buckled a bit and left gaps of about  0.5mm, 1/64", one between the binding and the purfling and one around the rosewood line around the upper bout. I think that the glue might had set as I it was loosely taped on before I got it taped on properly and there was no way for the strip to move into it's pocket between the body and the neck. That, or the binding and purflings were simply cut too long. Well, you live and learn.

I removed the taped and started to trim the excess binding and purflings off. On the top I did this with a spokeshave set with the blade set at an angle, meaning retrieved fully on the side that was over the soundboard and set to a moderate cut on the other side, giving the center a very fine cut. This way I could monitor how thick the cuts would be depending on how the spokeshave was held and it eliminated the chance of the blade suddenly digging into the soundboard.

I continued this way until the binding and purflings were almost flush with the soundboard and switched to a scraper that had the side that went over the soundboard protected with tape.

The sides of the binding were almost flush as it was and it was finished with the scraper almost exclusively. However, I did leave the section at the upper bout that had the gaps as I want to try and fix that later.

The very end of the binding was cut right at the center of the template with a razor saw held at about an 45º, taking care not to cut into the soundboard and the end strip. The waste was cleaned up with a 1/4" chisel, and the second strip was ready to go on.

THE SECOND SOUNDBOARD BINDING STRIP

This was done pretty much the same way as the first one. However, in order not to get a gap in the binding on the second strip I decided to try and go with Cumpiano's suggestion of gluing the binding on in increments. This worked fairly well, if a bit fiddly having to add glue to the rosewood strip, purfling and binding channel. But armed with the experience from the round I took care making sure everything went in nice and snug.

Another thing I did differently was to use the shooting board to true the ends of the purflings and binding. This worked really well and ensured a nice fit where they butted up against each other at the center.

THE FIRST AND SECOND BACK BINDING STRIP

For these two I followed most of the procedures of the first ones, but this time the purflings had to be mitered at both ends and the bindings had to be cut to exact length as they were butted up against each other both at the tail and neck end. The rosewood line was omitted.

This time around I didn't glue the first binding strip in extending over the middle of the template, but cut it to exact length. I thought this made more sense since the mitered side purflings functioned as a stop that determined the exact position of the binding in the channel.

This was altogether a little more challenging to do, but a lot of dry runs eventually yielded a nice fit. Also the purflings and bindings were glued on together in one go. I thought this was doable since it was only two components to deal with. Had all the purfling lines not been glued together before hand, I think Cumpiano's method of gluing in 1-2" increments had made more sense.

A SHORT NOTE ON BINDING TAPE

I have read several accounts on how easy it is to tear wood fibers from the soundboard when using tape to clamp the binding in. In order to figure out what I was up against I taped three different kinds of tape the some soundboard scrap - regular masking tape, blue 'light tack' masking tape, and StewMac brown binding tape.

I cut a 4" strip of each of them and pressed them to the soundboard scrap as firmly as I could and left them overnight. The next day I started by peeling off the first half slowly and then ripping the rest off with considerate force. To my surprise none of them teared the slightest.

I started to wonder why this was and came to the conclusion that maybe it had something to do with the fact that the soundboard still had a planed surface on it that had never been sanded. The fact that a sanded surface, even a finely sanded one, leaves a surface of torn wood fibers and a planed one doesn't may have been the reason. I don't know if this is true but it was the only logical conclusion I could think of.

StewMac brown binding tape was very nice to work with, though it is a bit pricier than regular masking tape. It seemed very strong and was very resistant to tearing.

CUTTING THE END GRAFT CHANNEL

PARING END GRAFT CHANNEL WASTE

GLUING TH END GRAFT

SCRAPING THE END GRAFT FLUSH

THE END GRAFT CUT

CUTTING THE BINDING REBATE

BINDING CHANNEL ROUTER DUST

CUTTING THE PURFLING REBATE

TRIMMING THE BINDING CHANNEL TO THE END GRAFT

CUTTING THE CONTINUING BINDING CHANNEL AT THE HEEL

EXCAVATING BINDING CHANNEL AT THE HEEL

PARING THE CONTINUING PURFLING CHANNEL AT THE HEEL

MITERING THE PURFLINGS

TRIMMING BINDING TO LENGTH

MITERING SIDE PURFLINGS

TRIMMING FIRST BINDING AND PURFLINGS TO LENGTH

PARING FIRST BINDING AND PURFLING WASTE

MARKING SECOND BINDING AND PURFLINGS FOR LENGTH

CUTTING BINDING TO LENGTH

TRUING BINDING END

BUILT UP BINDING CHANNEL

MITERING BACK PURFLINGS

GLUING ON BINDING AND PURFLINGS FOR THE BACK

THE BACK WITH BINDING AND PURFLING TAPED ON

THE BACK WITH BINDING AND PURFLING TAPED ON

SHAVING THE TOP OF THE BINDING FLUSH

SCRAPING THE SIDE OF THE BINDING FLUSH

BINDING SHAVINGS

BINDING DETAIL

BINDING DETAIL

BINDING DETAIL