Closing the Box

"Closing the box," or finally attaching the top and back to the rib structure, is a sort of milestone for building a violin. Some luthiers even go out and celebrate every time they close a box.

This process first consists of "turning the edges," or creating a smooth, even rounding along the entire edge of both plates. The corners are saved for after the neck is set into the instrument, simply because of their fragility. Turning the edges will not only provide an aesthetically pleasing look, and one that will wear less. It also helps the finish adhere to the edges of the plates.

Next, locating pins are made to fit the holes we drilled when we spot-glued the plates. The picture on the right is kind of fuzzy, but the pin is notched so that the cross-section becomes a semi-circle, allowing it to fit snugly against the purfling. These pins will make sure we get the plates in exactly the same place they were when we spot-glued. They will also keep the plates from slipping while gluing them up.

Speaking of glue, before I even want to think about gluing the plate on, I heat up my hide glue and coat the ends of the blocks with it. Right now I'm only worried about gluing the back on, so I'll get to the top quite a bit later. The glue must be applied to the blocks several times, because it likes to soak in to the exposed end-grain. Then I come around and apply glue to the ribs and linings.

Once the glue is thoroughly dried, I can "dry-clamp" the back on. I stick in the locating pins, put the back on where it belongs, and go around the whole body of the instrument with my spool clamps (pictured on right). Then I peruse the entire edge, looking for gaps. In this case, there were none. But if there were, I'd be going in with a plane or scrapers to remove any high or low spots that might be causing them.

Hide glue only has a working time of about thirty seconds to a minute, depending on temperature and humidity. This means that as soon as you remove it from the warmth of its water bath, you have thirty seconds to apply it to the gluing surface and get it clamped up. That's not enough time to clamp an entire instrument, so the gluing is done in stages. Incidentally, hot hide glue is also a very poor gap-filler, so it encourages (and demands) perfection from the craftsman. So why bother with hide glue? A lot of wood glues boast that the joint is stronger than the wood. That's a great thing most of the time. But for a musical instrument, what it means is that if you bang it on the side of a music stand or if the expansion and contraction of the wood on an aged instrument simply puts too much stress into it, the wood will break instead of the glue. That makes repairs less fun. Titebond is often used on guitars because you can always reach inside through the soundhole. But on a violin, eventually, the top will have to come off for a repair. So they're made to be removable. They have the overhanging margins and they're held together with the more brittle hide glue. It's still a lot of work to remove a top, and should only be done by a professional (unless you want to explain to the professional later exactly how it was that you contributed to the problem). But the hide glue makes things much easier.
Well, that was a long lecture with no pictures, so here's the trick I use. I open three or four spool clamps at a time, use a pallette knife to apply 145 degree fahrenheit water (ideal hide glue temperature) in the joint, and close the clamps back up. This reactivates the hide glue, gives me a slightly weakened bond for when the top does have to come off, and minimizes the amount of cleanup I would have had to do if I'd actually gone in there with hot hide glue on my pallette knife.


I wait overnight for the glue to cure to full strength, then I remove the clamps and I'm ready to extract the mold. I just go in there with a sharpened pallette knife to break the bond between the blocks and the mold. Hide glue bonds should snap, not creak, when they're breaking. So if there's an uncomfortable creaking sound, it's probably the block splitting. That's what you want to avoid.


Just push (gently please) on the block to see whether you successfully broke the bond. If you try to extract the mold and a block is still attached, terrible things could happen...





To demonstrate, here's a picture of the actual mold extraction. Get a firm grip on the mold and lift it up at one of the end blocks. Gently tap on that block with a small hammer. Switch from the end block to the neck block repeatedly to shimmy the mold out of there. It is thick enough that if you attempt to remove it from one end only, you will probably crack the ribs.
So there's the rib structure and the back, with no mold. It always feels surprisingly light at this point. The next thing to do is install the linings for the top and bevel them the way we did for the back. Here's a photo of the bevelling process. I just use a knife to do this, taking care not to dig into the ribs at all. The linings are bevelled so that there won't be a sudden change in the stiffness of the ribs where they begin, which would be prime real estate for a crack.

Now I get to shape the blocks. To do this, I use a chisel for the end blocks and an incannel gouge for the corner blocks. The main purpose of this process is to remove weight from inside the instrument. How much weight? I was curious, so I measured before and after. The fiddle lost 12 grams. It doesn't sound like a lot, but for a fiddle, that's not too shabby. That's like going down two dress sizes. Or something like that.


The pictures look scary, but I'm using a very controlled chopping motion with my chisel and gouge. The idea is basically to split the wood and clean it up afterwards. So make sure when buying blockwood to get the straightest-grained stuff you can find.


Now that looks much better.
Now I punch in my serial number on the neck block, where it will be visible to anyone looking in through the hole for the endbutton. I'll save time in the blogging process and explain that the top goes on pretty much exactly with the back, starting with coating the ends of those freshly-shaped blocks with hide glue...
And voila! My box is closed.
Mmmm... So tasty...
Next, I'll be blogging about the beginning process for two separate fiddle necks. But for now, I'm going to take a break, and give a break to my readers after that oh-so-long-winded blog. Now you see why I can't do Twitter.

The f-holes and the bass bar

Here's a picture of the f-holes all cut out. Unfortunately, I didn't take any of the actual process, but here's how it went. I started out by 1/8" drilling holes in the center of each lobe, and in the middle of the f. Then I used a coping saw with the smallest blade I had to remove the wood along the center strip of the f. I then used a jeweler's saw to remove the wood from the lobes and wings. I cleaned it up with my knives, then files.
Once I cut the f-holes, I was ready to install the bass bar. After preparing my stock so that it was flat, straight, and of even thickness, I determined its final position, then set it on the top and traced along the edge with a pencil so that I'd have a rough guide for the underside of the arch.

I shaped the bar and set it into its proper position, then installed cleats all around it, tacking them down with hide glue. This way I can remove the bar over and over again as I'm fitting it to the inside of the top, and every time I put it back, it'll be in the same exact location.

Now here's why I'll need to remove it so often. I spread chalk along the inside of the top, where the bar goes.

Then I set the bar down, shift it forward and back ever-so-slightly ( I left a 1/2 mm gap at the end of the bar when I set down the lowest cleat). The chalk adheres to the underside of the bar and lets me know where I need to scrape off material. Don't let the pictures fool you. In person, the chalk can be very difficult to see on the light wood. But colored chalk stains the wood, so I just squint around and constantly shift the lighting.




Once the entire surface is chalk-covered, I know I'll have a perfect fit with no gaps. Then I glue in the bar and pop out the cleats. But I can't leave it like that!



I take the bar down to very specific heights along at five points along its length, creating smooth transitions between those points. Then I use fingerplanes, chisels, knives, and finally, sandpaper, to create a bullet-shaped cross-section.
Isn't it lovely? And for my next update, I get to close the box!

And Now, For Something Completely Different

Between commissions for instruments and various repairs, I always have several projects going on in my shop at once. In this case, this lovely cigar box is about to become my next fiddle.

The person who commissioned me to build this also requested photographs of the process. No problem, says I. So here they are in my blog.

You're likely to notice right off the bat how much faster this project moves along than my other fiddle. Until I get to the neck, anyways.


So I'll start out with the blocks to support the neck and tailpiece. This one chunk of spruce is about to be turned into both blocks, with some scrap left over.






But notice when I go to put the blocks in, they overlap with this bit of paper on the edge. This is a common problem in cigar-box instruments. So common, in fact, that it doesn't phase me anymore. I just cut away the paper as shown, to make room.



The next step is to cut out the f-holes.

I lay these out so that the notches will indicate the bridge position and the upper lobes are 42mm away from one another (the width of the bridge feet). There's still something missing, though. Because the top lacks an arch, it doesn't have the inherent strength of an actual violin. So in order to reinforce it, I glue in some cross-braces, along with a shortened (though probably unnecessary) bass bar. Here's what my box looks like with the braces glued in, after I've shaped them.

Now I have to remove lots of the paper from the insides of the sides so that I can glue kerfing on. This will give me a larger gluing surface on which to glue the back.


Speaking of the wily back, there's not only paper to remove from it, but lacquer as well. I scrape off the paper, then sand away the lacquer, being careful to keep the surface level and without divots so that I'll have a consistent gluing surface. I also use a flat file to level the kerfing and blocks to a flat, consistent surface.

That's a lot of pictures, but the process doesn't take nearly as long as, say a violin. This point in the construction process, when the top and back are both glued on to the rib assembly, is called "closing the box," and is a sort of milestone towards the completion of most instruments. As of today, I have closed the box on both the cigar box fiddle and the violin. The cigar box fiddle took me a grand total of five working hours to close. The violin took fifty-one and a half. Which is actually not bad. On Tuesday, I'll be starting the necks for both fiddles.

The Graduations

Well, it's graduation time again. I'll be carving the insides of the violin plates and I'll need a few guides so that I don't go too far. First of all, I don't want to remove any wood from the edges, where I'll need to glue the plates on to the ribs. I went through all that trouble flattening them, and I want to keep them flat. I use a compass to mark an even distance from the edges, a distance of 4.5mm, if I remember correctly. Then I "bridge" the dips in the corners to allow surface for gluing the the blocks, and draw straight lines across the bottom and top, 18mm from the scribed line on the plates, to allow for the neck and end blocks. Whatever else happens, I don't want to go past these lines.

But how do I keep from cutting too deep and tearing through my plates?
This handy contraption serves as a depth guide. It's just a dowel glued into a board, with a rounded, leather-padded tip. This simple jig is arranged so that, by setting the depth stop on the drill press and setting the plate so that its face touches the padded dowel, I can drill a guide hole into the plate, without fear of going too deep.

It's not quite so easy as it sounds, because the depth stop has an inaccurate gauge, and I have different depths that I have to drill to in different ares of the plates. Consequently, I have to set and reset the depth stop, drill into a test piece, measure that, and perhaps reset the depth stop again. Then I drill all the holes that require that depth, and repeat the process. When that's all done, I go in with a gouge and carve away until I get to the bottoms of the holes.

I use a "cradle" to support the arch, and to keep the plates from rocking back and forth, as they would clearly do if I attempted to carve their insides without one. Once I get down to the bottoms of the holes with my gouge (a time-consuming process) I switch to fingerplanes. This concludes the rough graduation.

Now I have to come back to this whole thing, drill more holes, and repeat the process. But before the final graduations can be done on the top, the f-holes require a little TLC. First, they need "fluting." Fluting the f-holes is a process by which the lower lobes are scooped out so that they don't stick up, get caught on things, and break off. Fluting also has an acoustical effect on the final instrument, since it changes the resonant frequencies of these lower arches.

This isn't a very good picture, but you'll get the idea. Next I finalize the arches of the plates, then I'm ready to cut out the f-holes. In order to get a good idea of when to stop once I've gotten through the second set of drill markers, I flex the back and top (gently), then I decide whether it feels right to me. If there are any areas stiffer than others, they need to be treated; if the whole plate is still what I would deem to be too stiff, the whole plate must me thinned.









I will not be blogging the f-holes or bass bar on my next post. They will have to wait for two posts from now. Something different has come up and, at the request of the customer, it will appear in my blog as well. But it's probably perhaps a little more interesting.

The Final Arch

Now that the purfling is installed, I can focus on finalizing the arches of the top and back.

For the first step, I return to the purfling. Using a 12mm gouge with a number five sweep, I cut a nice, even trough all around the edges of the plates, bringing them down to near-final thickness.


Then I'm ready to get down to business. Using my arching gouge whenever it will save me time and my fingerplanes when it won't, I carefully sculpt the outsides of the plates. As guides, I use "third generation" aluminum templates. That is to say, they are a copy of a copy of a set of arching templates, in this case derived from a 1715 (golden period) Stradivari Violin.

They are reasonable accurate, but it's important to keep in mind that each piece of wood is different, and that each golden-period Strad is different. Consequently, as my instructor John Reed once reminded me, they are guides, not gods.

I start out using standard convex blades in the arching gouges, but as I get down to finer and finer work, I switch the blades out for toothed blades, which tear out less.

Finally, I finish the surface with my scrapers, negating the necessity for tedious sanding. Scrapers are wonderful when properly used, and actually give you a finer, more open surface than sandpaper of any grit could manage.

I always start with the back, and do the top next. But the process is essentially the same.

Stay tuned for my next post, when I'll begin the "graduations," or the carving of the inside of the plates. I took some good pictures, and things went pretty swimmingly, so it should be exciting. That is, if you like that sort of thing.