Recent work

Well, I've had a bit of a hiatus while I took care of some personal business in the past few weeks, mostly related to a small but significant bit of elective surgery I had over the past weekend. For the record it was not plastic surgery of any kind...

At any rate, over the several weeks since my last post I have accomplished a bit but not had time to post about it. Actually, I also failed to take photos of most of these steps, but fortunately they aren't really visually gripping steps to miss.

What I did over the past few weeks was exactly what I said my next steps would be in the last post. I glued the short braces onto the top and sanded the kerfing to match the radius of curvature of the top. Fortunately the kit builder again thoughtfully configured the sanding stick setup so that the sanding stick, with the additional pad attached, resulted in a radius of the correct curvature. Basically you attach a strip of self-adhesive 80-grit sandpaper to the end of the sanding stick (7" in from the end, actually) and then, keeping the additional pad on the side of the instrument directly across from the area being sanded, you just sand the edges until you have properly contoured and leveled the kerfing. To aid in doing that you scribble with a pencil all over the kerfing. When the pencil lines vanish you have done your job. The key here is to keep moving around the circumference of the edge to ensure that you remove material somewhat evenly.

Here's a photo of the sanding stick setup. See the strip of sandpaper attached to the stick, and the additional pad? This photo was taken after some sanding of the top edge of the sides.



I also glued on the short braces for the top. I can't do any additional gluing of top braces, or really of the bottom brace either, until I get a set of more or less lutherie-dedicated cam clamps, so I'm somewhat stymied at this point. I failed to take a picture of the top with both braces glued-up, but rest assured that the process was identical for both braces and that the top pretty much looks just like this, but with two braces attached. Here is a photo of the brace being glued from the "inside".


And here is a photo of that same gluing operation from the outside, showing a piece cutoff from the top being used as a caul to protect the top. Note that cauls were not used on the braces since they have to be carved and will lose much of the material in the area that the clamp would have been bearing on anyway.


And, finally, here is a photo of the top with one of the two short braces attached.


I also spent some time tuning a bench plane which will be crucial to the shaping of the braces and neck and to the trimming of the top and back to the sides. I never really understood that almost all planes, and other woodworking cutting tools like chisels, require substantial preparation in the shop before they can be used, even if they are purchased brand new. While some of the very high-end planes might come with flat soles and very sharp irons, most do not. On a bench plane you can also spend a considerable amount of time ensuring that the bottom is a true 90 degrees from the side of the plane. Fortunately in my case the sole was very flat when the plane arrived, and the sides were close enough to 90 degrees from the sole to be acceptable for my purposes. I am somewhat surprised by this since because this is just a bench plane and not a high-end plane like a smoothing or jointing plane I bought a Stanley and not a really good plane. But a few minutes of (frankly) significant exertion really got this plane into usable shape. Mostly it was just flattening the sole by polishing it on 80-grit sandpaper adhered to a piece of polished granite I got through my previous employer that is truly flat and very, very stable. I then polished it with ever increasing grits until the sole was flat and mirrorlike. I have yet to sharpen the iron in exactly the same way before being completely pleased with the tool, but it still makes impressive shavings despite the relatively cheap nature of the tool. But this is a bench plane, after all - how accurate are we really expecting it to be?

At some point I will post some details about my sharpening system. Basically it is the Scary Sharp system (link, perhaps not to the originator of the system using that name, to the right), using a Veritas honing guide and several pieces of polished granite with various grits of sandpaper stuck to them. My good friend Joe Barker, who is building a sailboat in a room about 5% larger than the boat and blogs at http://barkerboat.blogspot.com/, uses the same system (I gave him several pieces of the granite, too) and he has more details at his site. Ignore all the Old Style comments over there. He has great taste in beer (and women) - he just wants to feel like a boat builder and you just can't get Narragansett beer in Cincinnati.

Biggest day yet, part II

Blogger kept crashing on me near the end of the last post, so I elected to split them into two posts. If you haven't yet read about shaping the braces, read the next post before this one.

Here you can see the top and back, along with the cutoffs (which the plan suggests keeping for use as cauls later in the project). On the right are the three major components to date – the top, the sides and kerfing, and the back.

And here's an exciting photo – I stacked the parts together on the benchtop just for giggles and took a photo. At some point in the near future the instrument will be glued together like this and then the top and back will be trimmed to final shape. I just thought it would be cool to see it like this.

Since I still had several hours (at this point I had spent only about 1.5 of my 6 or so) I decided to cut out the soundhole as well. Now, the plans suggest cutting the soundhole patch using a hobby knife. I had been dreading this step because I was frankly worried that I would cut this poorly and damage the top by over cutting. It had to get done sometime, though – so here it goes.

The photo on the left shows the soundhole from the kit manufacturer with the soundhole patch glued over it. The photo on the right shows the results of the rough-in cut out (as suggested by the plans I made a first cut a bit shy of the edges of the soundhole, followed by a final cutout to actual shape). As you can see in the left photo below, I did some minor damage to the top wood surrounding the soundhole when cutting it out just as I had worried. However, the next step in the plan is to smooth and bevel the edges of the soundhole with sandpaper so I decided not to worry too much about minor damage. Fortunately the damage turned out to be very minor indeed – the photo on the right is the soundhole after smoothing and rounding with some 120 grit sandpaper (the MDF of my benchtop is also visible here, and it might make the soundhole hard to make out. Sorry.)

These last two photos are of the top after all the work of today was complete. The one on the left is of the inside of the top (the side that will not be visible after the instrument is complete). You can see the soundhole patch and the markings for all the braces in this photo. The one on the right is of the outside of the instrument.

Next time I will be sanding the kerfing to match the curvature of the top and back, and gluing in some of the braces. Not too long until the body is largely complete!

Biggest day yet

Wow! I got several hours to work in the shop today and the outcome was better than I could have imagined. My wife took the kids and went off the visit her parents, and she instructed me (this is why I know she's the right girl) to work on the mandolin. I got more accomplished today than any other day of work on it yet!

I started off by shaping the braces for the top and back. The kit includes a specially sized “sanding stick” made of ¼” plywood. They even included a little piece of the same plywood for use as a caul for clamping the stick to the benchtop. The kit designers thought this step out well, and planned for it well, too – the caul that was included for the outside of the sides at the neck block it just the right height at the edges to result in the proper radius of curvature when used to elevate the non-clamped part of the stick. The following pictures should give some idea of what the heck I'm talking about:

The photo on the left shows the stick clamped to the benchtop and propped up at the other end using the neck block caul. A clamp holds the neck block caul in place – I found that the two clamps at the bench kept the stick from rotating sufficiently. As directed by the instructions I also supported the stick at two other points along its length; one of the supports is the tail block caul and the other is a piece of scrap left over from the toolbox I built for my son for his birthday. The picture on the right shows the same setup from above just for clarity. You can see the self-adhesive sandpaper (80-grit as recommended) in the photos.

The instructions say to sand one of the braces and compare it to the drawings, adjusting the curve by moving the neck block caul in or out until the curvature is satisfactory. You then sand all of the braces using that same setup so that they are the same. I found that, by luck, my setup worked out very well the first time, so I went about sanding all of the braces to the curvature shown on the plans (which the instructions helpfully tell me is about an 8' radius).

The picture on the left is an action shot of me sanding one of the braces. I used a longitudinal motion to sand them – that is along the length of the stick and not side to side across it. This just seemed easier and the results were good, so I guess it's OK – at any rate it matched the grain direction so it made for clean sanding. The photo on the right shows the completed braces. Note that there are two #3 braces.

The next step is rough-cutting the back and top. I had not originally planned to do this today, but the shaping of the braces went so well and so quickly that I decided to have at it. The instructions say to cut the top and back with a bandsaw if you have one. Since I do I decided to give this a shot.

Up until today I had not had luck with my bandsaw, which I bought used (along with almost all of the rest of my stationary power tools) from a guy who was liquidating his shop. I examined the saw and realized that I had not been setting the damned thing up correctly. Not one of the cuts I had tried with it had been successful until today. Today I adjusted the tensioner more carefully and realized that I had not been properly adjusting the blade guide. After having done this the saw cut like a dream. Lesson learned – adjust your bandsaw carefully for each situation and it will cut much, much better.

I used the bandsaw to cut out the back and sides (the instructions tell you to leave about ¼” proud for later trimming to the actual shape of the instrument). Because I was not being very careful about accuracy since I was leaving the cut proud it went fast. I first had to mark the brace on the back and trace the outline (the left photo) and then cut it out on the bandsaw (the right photo).

Flush with my success I then cut out the top as well.

Soundhole patch

Because of a little glitch in timing today I got to work on the mandolin again. I only did a little bit of work, but it's still nice to get out there!

The task for today was to transfer the bracing pattern to the inside of the top and apply the soundhole patch (the soundhole patch is explained in a post below). In order to do that it was necessary to cut out one of the patterns of the top of the instrument. As an engineer I am loathe to work on an original document. I had therefore prepared for this by making photocopies of the pattern in question. Using the Xacto knife, I cut out the pattern for the top.

I then marked a centerline on the inside of the top. Using the centerline, and the soundhole already cut in the top by the kit manufacturer, I centered the cut out pattern on the top and traced around it to mark the approximate shape of the top. I then used the tip of my mechanical pencil (0.5mm!) to press through the pattern at the corners of all the braces, leaving dimples in the wood that I connected with the pencil and the straightedge to complete transferring the brace pattern (this is the method recommended by the instructions).

The final step was to glue the soundhole patch to the top. I did this using the Titebond. The instructions recommend using a weight instead of clamps for this operation - this made sense to me, so you can see the 30 cal ammo box I keep tools from my old Jeep in used as a weight. Because the bottom of the ammo box is not flat I used a cutoff of MDF, originally used as a spacer during glueup on another project (the markings on it have nothing to do with the mandolin) as a caul. A piece of waxed paper under the top, and one between the soundhole patch and the caul, prevent the top from sticking to the table and the caul.

Kit explanation

A good friend of mine (Suave) asked for some detail that separated this kit build from simply building from scratch. After reflection I realized that the aspects of the kit might not be clear to everyone from the descriptions so far, so I sent him this:

I guess I should have spent some time illustrating the kit itself. The kit came with all the wood, the fret wire, the fretboard side and top dots, and the nut and bridge material. The top was bookmatched, properly thicknessed, and had the soundhole cut. The back was also bookmatched. The kerfing was precut and prekerfed, although there were four pieces each of which was far too long and had to be properly trimmed. The braces were provided, properly thicknessed in cross section but requiring shaping both to form the curve of the top and the shape of the side of the braces away from the top or bottom (the manufacturer provided a long narrow piece of plywood that, when clamped to the table at one end and blocked up at the other end, will form the curve necessary for the curvature of the top). I do have to learn to fret an instrument, though - but they provided enough waste fingerboard with fret slots cut into it that you can practice this before going big time, and they also included extra fretwire. Critically, the sides were prebent. This is probably the biggest advantage of the kit since it meant that I would not have to learn to bend sides in concert with everything else (though it should be noted that I have some limited experience with bending wood and it is not that difficult to do - though my experience is not to the level of tolerance required for instrument building). I plan to bend them myself the next time.

Interestingly, not included were a tailpiece (mandolins typically use trapeze-type tailpieces) or tuning machines. I will buy these this spring after completing the instrument. I think the project is going fairly well, although it appears that the mandolin will be slightly nonsymmetrical about the axis of the neck - the sides might not have been exactly the same length and that seems to have resulted in a slight lopsidedness favoring the side that will be "up" when I play it. I don't know that it will be obvious to anyone but me (and, now, you) or that this will be the most out-of-whack thing about it in the long run. So far I am pleased.

I see your point - most of the tough stuff was left for me to do - but just gluing together a pretty much already built instrument would not be all that rewarding, I guess. This has been very rewarding so far. Believe it or not they actually have a kit with less stuff done for you. You have to bend the sides and glue up both the top and the back in that one. I might try that one next. In fact, I almost certainly will.

Kerfing for back

On Monday I took the mandolin out of the clamps (err... clothespins) and examined my work. Here is the result of the glue-up of the kerfing for the top of the mandolin:


Here I'll take a moment to once again try to explain kerfing. I now have a couple of good examples of what it looks like in the field.




See how the kerfs in the wood (the word kerf refers to the bit of wood removed by the passing of a saw blade, so here they are the notches in the wood) allow the wood to bend smoothly while still maintaining a nice bit of glue surface on top?




Here we have two closeup details of what the kerfing looks like. Notice how the kerfs close slightly on inside curves, allowing the wood to bend smoothly.






Finally, I have a closeup of the kerfing in the clamps, showing the small amount I left the kerfing proud of the sides during the glue-up. The instructions say to make this small reveal about 1/32". No idea how close I am to that but it looks fairly good, anyway.





These last two pictures are of the glue-up underway, showing how I laid out the kerfing on the side while clamping. Also shown is a photo of the back with all clamps in place.






After this was done I had some time before my wife and boy came home. The next step in the instructions is to layout the braces on the top and back. However, it recommends cutting up the pattern to make this easier. I'm an engineer and I hate working on an original - it makes it impossible to revert - so I decided to take the pattern/drawing in to the office and make a couple of copies before cutting anything up. The next step after the layout step is to cut the soundhole patch. This step does not rely on the layout on the top so I decided to do this step instead.

The soundhole patch is a very thin piece of wood that reinforces the top at the soundhole. The top is made of bookmatched spruce - which essentially means that it is made of two identical pieces with a joint running from the tailpiece to the neck - and the soundhole patch strengthens the top in the area where the material was removed to make the soundhole. The wood is very thin - almost a veneer, really - and the intention is for the soundhole patch to have its grain run perpendicular to the grain of the top. It's trapezoidal and because the wood is very thin I cut it with my Exacto knife and a straightedge. A few moments of layout and the cuts were made.


The photo on the left shows the layout of the soundhole patch cuts, including the patch itself, the framing square I am using as a straightedge, and the drawing/pattern. The one to the right shows the soundhole patch after trimming with the straightedge and the Exacto knife. I should also take a moment to apologize for the incredibly ugly cutting mat. Darn you, Wal Mart!!!



The final image of the day shows the drawing/pattern, the soundhole patch, and the top blank. The next step, probably on Wednesday, will be to transfer the brace locations and commence to shaping the braces, and to glue the soundhole patch onto the top.

Return to the shop

I finally got to return to the shop today. As promised, today's steps were exciting ones that really moved things in the right direction. I prepared the "mold" for the mandolin, and also glued up the kerfing. These are the biggest steps I've taken yet towards actually finishing the mandolin (which is still a LONG way off).

Along the way, a few explanations.

First, the mold. Stringed instruments are constructed of wood (most of the time, and this one is in that class). Wood has a tendency to deflect under pressure - that is, it bends when you pick it up or work on it. Since the sides of the instrument are made of very thin wood, they will tend to deflect quite a bit while you are attaching the kerfing and the top and bottom. Of course, once the top or bottom have been added to the instrument the mandolin will be fairly stiff, but until then you want to keep the instrument in the same shape as the final product as far as possible because every part added will tend to stiffen it, and if parts are added while the sides are deformed due to handling they will tend to cause a system-wide deformation in the final product. So typically builders use a form to ensure that the instrument keeps true to its final shape wile being worked upon. These molds take many shapes and forms - molds for guitars are frequently made of wood and form the outside of the instrument. My kit calls for a cardboard mold that fits inside the instrument as shown here.

The process started with selecting a piece of cardboard that the pattern would fit onto (the kit says to use the "included" cardboard but my kit did not include a piece - fortunately the lid of the shipping box for the kit was large enough to cut a piece out of). I then used a glue stick to adhere the paper pattern to the cardboard and, using an Exacto knife with a fine-tipped blade and a straightedge, I cut the mold out by following the lines (the yellow-green thing in the photos is the cutting mat). First I made a series of straight cuts to remove the bulk of the material, staying just outside of the lines and cutting the waste away. Having done that I cut the pattern out by carefully following the lines with the Exacto knife. I did this in small sections by making cuts perpendicular to the cut line out to the edge of the waste area and then removing the waste by cutting along the cut lines. Here's the final product of this process:

You can see some of the waste pieces lying off to the left and the Exacto knife I used to the right in this photo.

Once the mold was in place it was time to add the kerfing. Kerfing is actually a process by which a series of cuts are made parallel (usually) to one another in a piece of wood to make it easier to bend the wood. In stringed instruments the work also refers to the small pieces of wood, often triangular in section, which are used to glue the top and back onto an instrument. These pieces are kerfed to make them easier to bend into shape. The purpose of kerfing is simple: the sides of the instrument are far too thin to offer enough glue surface to make a secure joint between the top and side. The kerfing is securely glued to the sides and then glued to the top, offering enough glue surface to make a strong joint. My instructions say to use clothespins for clamps since they can be placed very closely together along the sides to ensure even clamping pressure. They also recommend wrapping the clothespins with rubberbands if they have weak clamping force. My clothespins, bought at Wal Mart, had terrible clamping force, so I did this. I started by dry fitting the kerfing to the mandolin to size it and to determine how many clothespins I would need. I fit 23 pins in the dry fit, so I took a few minutes and wrapped 23 pins with rubber bands. I then spread glue onto the first piece of kerfing and clamped it in place using the prepared pins. It turns out that because the pins are significantly wider when wrapped with rubber bands I only used 20 of them. To the left is the first piece of kerfing glued in place. You then repeat the process with the second side (shown on the right).

One final note: the instructions tell you to leave the kerfing about 1/32" above the sides. This will be sanded down at a later moment in the proper shape for the slight bow curve of the top. I tried to do this, and the detail below is the result.