I've made a few arch top Classical guitars; trying for a tone that was usable for the normal repertoire, but with more power. It took a few tries to get it reasonably close, and I'm hoping to make another at some point to improve it further. Here's what I think I've learned so far.
It's normal to think that you'd scale the arch height to the length of the box, but that doesn't work. You scale the height to the thickness of the top. So what you want to do is figure out how thick you want to make the top, and go from there. Most archtops use a somewhat thicker top than flat tops, and thus a higher arch.
On my arch Classicals I started out with thin tops, to keep the weight down, and a high arch to get the stiffness The first two ended up sounding 'tinny'. When I read about that scaling rule I decided to use a 3mm thick top, and since that's about what a violin uses, I went to the same overall arch height of 15mm above the rib edge. Those worked much better.
The first unsuccessful arch top Classical that I made used a padauk back, which is a decent BRW substitute (except for the way it cracks!): it was too 'lively', with a 'peaky' spectrum, and that contributed to the tinny sound IMO. With Thomastic 'S' series steel wire core Classical strings it made a pretty good Blues box. Stick with the usual arch top woods.
All the arch tops I've made to date, both 'normal' steel string and nylon string, have used X bracing. I'm now working on one (for steel strings) that has 'parallel' bracing. That's supposed to give a punchier sound. We'll see.
Use a round or oval hole. The F-holes move the 'main air' resonance up too much, even if you keep them very narrow. A round hole arch top tends to sound somewhere between an arch top and a flat top. There's also a benefit to having the main hole in the 'normal' spot: again, it's complicated, but it's part of what makes things sound like a guitar instead of a cittern.
From what I can find in the experiments I've done, side ports are best looked at as 'monitors'. They can direct a useful amount of high frequency sound out toward the player if you can look into the port as you play. Normally the player only gets most of the high from room reflections, so that in a large, dead, or noisy room (the 'restaurant gig') it can be hard to hear a guitar acoustically without a port. Some folks with high frequency hearing loss report that they're beneficial in normal playing, which fits. You don't need a big one: a 1" or 1-1/2" hole in the upper bout near the wide point on the right side works well for most people. If you want to use only a port and do away with the normal hole then the port should go on the side at the same height as the hole would have; just above the waist.
Another instrument with a 'tween sound is a flat top with an arched back. I made one of those years ago and a student made one some time after that. The arching makes a wood like maple or walnut act more like a rosewood on the guitar: it has higher mass and lower damping than a flat braced back of the same material.
Flat top guitars with the strings tied or pinned to the bridge get a small amount of energy from the twice-per-cycle tension change as the string vibrates, which rocks the bridge fore and aft. This is not the major driver on a flat top, contrary to an opinion that is widely held on these groups, but nowhere within the guitar acoustics research community. Bridge rocking contributes just enough to the signal to color the sound a bit, so that a fairly large change in string height off the top with that system makes an audible difference. I've done that experiment.
At any rate, that small bit of signal is missing on arch top guitars with tail pieces. You'll note that they don't sound an octave lower than flat tops with the same strings.
Changing the string height off the top of an arch top also changes the break angle , and thus the down bearing force on the top. Again, from experiments I've done, more break angle does not equate with more sound. In fact, too much break angle kills it. I don't know what the limits are for this. Benedetto seems to recommend about 5-6 degrees of break angle, which is barely enough to stop the string vibration at the saddle top, assuming there are notches in it to stop sideways motion. I tend to keep it about there on the theory that I know that too much angle hurts the tone, and since I don't know how to say that 'too much' is, I'll go with less. It's possible that there is some 'ideal' amount of down bearing, but short of a lot of experimenting with each guitar I don't know how to find it.
The tailpiece can have quite an effect. It's got at least one resonant frequency that can push on the bridge, and tuning that relative to the other resonant modes of the top and air can effect those. This gets complicated, and a lot depends on what you want to do with the sound. The actual string tension used is one variable, along with the mass and length of the tail piece, so it can be hard to know exactly until you get the thing strung up. In a general way, all else equal, the longer the tail piece (shorter back strings) the more effect it will have. I saw Jim D'Aquisto in one talk where he said that he could move the guitar sound around a lot with a different tailpiece.
That's a start. As always, these things get complicated. It would help if I knew just exactly what sound you're after.
|