Classical Guitar 253(B)

This guitar was started in 2019. The intent was initially to build a guitar that sounded as good as #251 which I built several years ago in Fort Collins. That guitar initially didn’t speak to me at all. It was a Cedar double top with Cocobolo back and sides and a 5 piece laminated neck. Now I tend to impatiently wait to listen to an instrument. I will often string a guitar up long before it is ready. This was the case with 251. I’ve been using water borne finishes for several years now and one of the things about it is that it takes up to a month for it to fully cure. Until it does, the guitar sounds dull. So the finish wasn’t cured when I first strung 251 up and I set it aside for several years while I built a few guitars that I french polished. No long waiting time for those.

At any rate, this summer I dragged 251 out of the case and started playing it and realized that it was one of the best sounding instruments I’d built. It was loud, had a great bass and the treble was as loud as the bass without being harsh. I impressed a few local players with it and decided to do it again, maybe trying a solid top this time. The photo below is of #251 before attaching the back.

Top bracing and configuration of #251

The main thing missing from the photo is the set screw pivot that acts to support the small transverse brace below the soundhole from the carbon fiber rod. The other thing it doesn’t show is the support rods that replace the upper transverse bar.

The bracing pattern is my own creation, which I have dubbed the diamond spider. Diamond because of the centrally located diamond, and spider because it has 8 ancillary braces, or legs like a spider. All the angles except for the center brace adjoin at 90 degrees. The center brace adjoins the others at 45 degrees. I’d say this qualifies as a radial bracing pattern, of which there are many. But it started as a large, rather open lattice. Most lattice bracing is joined together at the crossings, often with carbon fiber threads and epoxy. My first try at this (a year or so earlier) sounded really tight and I tore the back off and reworked the bracing to roughly what the photo shows before it started singing well. So far, this bracing yields the best sound from a double top for me.

Now I can seldom leave well enough alone, so in addition to doing this bracing with a solid cedar top I decided to change the bracing a bit.

The bracing above is what I tried. I extended the top legs up into the upper bout and shaved the small transverse bar down a bit. The cedar top was thinned down to 2mm around the edge with almost 3mm in the center. This is a bit thicker than I’d normally have done it in the past, but I was going more by weight than thickness. Since I was in experimental mode, I strung this up with the back clamped on. I played it for several days and didn’t like it at all. There was no bass response and the treble was thin. I proceeded to drop the bracing down to 4mm tall and glued on the back. Oh, did I mention that I was using a wood rod instead of the carbon fiber ? I thinned that rod down too. Comparing it to #251 side by side was sad. I got a bit more bass out of it, but still not good.

The last thing I did, with the back still glued on, was to sand down those two extending braces below the soundhole almost level with the surface of the top. There was marginal improvement, but it wasn’t balanced. The treble was too shrill and the bass without depth. It has now been a couple months or more. I sent 251 off to it’s new owner and built another copy, #257 using Brazilian Rosewood. The next entry will be about that guitar. Against that guitar 253 sounded just as weak… sigh… I should mention that when I talk about the sound of this guitar it is in comparison to the other guitars it’s being compared to. In reality 253A wasn’t that bad a guitar and would have held it’s own against a good many guitars in most stores. But the bar is high here.

So top A came off and a new top B was prepared, along with some other changes…

This version (253B) is much more like the original 251 (and 257) and is a cedar double top. The transverse bar isn’t shown here. In addition, I made a few changes to the body as well…

Here the transverse support pivot can be seen with the set screw in place. I also added a carbon fiber support in the lower bout to sturdy up the lower bout, which felt a little too loose to me. That was probably because I had just finished the sides for #258, which are extremely thick and stiff. That brace across the lower bout would be a really bad idea if I didn’t have a vacuum press for gluing on the bridge. I’ll also mention that 251 and 257 don’t have the longitudinal back bracing, instead they have 3 transverse aromatic cedar braces. At this point I added the cedar (pink) transverse brace on the back. Then it was attach the new top, re bind and finish the top.

So to summarize all that. I started with a top that was too stiff, or such was my surmise having very little low end and a shrill high end. All my efforts didn’t get me to where I was satisfied, so I replaced the top and stiffened the back and sides. Having sent #251 away I wasn’t really able to compare the same way I had before the switch, however I can say with certainty that there is more depth in the bass response and that the treble is not shrill at all. Did I get close to #251 ? I don’t know for sure since 251 is gone, but it’s certainly much closer. It’s quite loud, which is a nice change. When I play it, I am pulled into a more romantic style, pulling for deep emotion. It has grown in the month that it has been strung up. It started sounding a bit mushy (uncured finish ?) but has opened up nicely in the past week or so. Another month or so and polishing up the finish on the new binding around the top and it should be ready for someone. I expect that I’ll be using it to judge 258…

Building a highly responsive guitar: Tops – stiffness vs weight

There are two dominant factors that must be considered when building a highly responsive guitar. The weight and stiffness of the soundboard. These two factors determine the volume and tonal properties of the instrument and while bracing is very important, it can only do so much. It won’t make a poor soundboard equal a great one.

The dance between weight and stiffness is a hard one to master. The best soundboards are lightweight and relatively stiff. The problem is that stiffness and weight are inextricably linked. Reducing the weight of a solid soundboard is normally only possible by making it thinner which makes it less stiff. If it isn’t stiff enough, the tone suffers and consistency suffers more. The result is erratic instruments with loud and quiet notes causing problems with consistent performance. A thicker soundboard will be stiffer and thus less erratic, but won’t have the volume required by today’s demanding musicians. This is the base dilemma of building stringed instruments from solid wood.

My experience has shown that a truly responsive guitar (nylon/carbon strung), in terms of volume requires that a soundboard, thicknessed, cut to shape and ready for bracing should weigh no more than 120 grams for spruce. This will differ depending on the body shape, but it’s a good rough number for this article. Depending on the variety (Sitka, Englemann, European etc) this will be very thin.  As mentioned before, thinner means less stiff and less stiff means erratic results.

The answer to this dilemma for me is composite soundboards. I can meet weight targets while increasing stiffness by building soundboards that are mostly air. More often referred to as “double tops”, the construction is two thin sheets of wood encasing an open Aramid® honeycomb core. The vast majority of weight is in the wood sheets, which must be very thin. Spruce and cedar are used for the outer sheets, which can be as thin as .5mm. Care must be taken when building these tops and proper construction can nearly double the time it takes to build the instrument, but it is worth it. In all cases I’ve built, the finished double-top weighs less and is stiffer than solid tops from the same woods.

Of course it is quite easy to build a bad double top as well. The devil is in the details that take considerable experience to master. I have replaced or completely reconfigured many double-tops in the past decade. Each time the resulting instrument vastly outperformed it’s original setup.

Building a highly responsive guitar: back and sides

Way too much attention has been given to the back and sides in discussions of guitar responsiveness. There are no magic woods that if used for a back and/or sides have any guarantee of tonal quality, volume or responsiveness. For the soundboard you can make a case, but not the back and sides. The reason being that the back is not the primary resonator on a guitar.

But let’s look at the back anyway. There are a few primary functions for the back:

  • The back defines the volume of the body, i.e. air cavity. This results in a major resonance known as the body resonance or the air resonance, depending on who you are talking to.
    • The placement and size of the soundhole is also a factor in determining this resonance.
    • The flexibility of the back also contributes to determining this resonance. A stiffer back will push the air resonance up somewhat and a more flexible back will lower it somewhat.
  • The back itself has it’s own set of resonances. The main back resonance is the least important of the three main resonances; top, air/body and back. However, if it is in the wrong relationship it can cause problems with comparatively loud or soft notes or even causing a specific note to waver.
  • Other than the effect on the resonances, a stiffer back is more likely to act as a reflector. A more flexible back is more likely to act as a resonator. Any back will act as both to some extent, but to what degree depends on the stiffness (and weight).
  • Some woods are highly resonant in the sense that they resonate strongly within a confined frequency range. These woods tend to be very dense and heavy woods, and true rosewoods are in this category. But density alone doesn’t assure a highly resonant wood.
    • In “tap” tests, the unassembled back is tapped and analyzed by how clear the resulting sound is. When viewed on a spectrogram, this is a narrow, sharp peak for the highly resonant pieces.
    • The assumption has always been that the highly resonant back woods are superior, but guitars built with them can be noticeably erratic in the response, volume and sustain of individual notes.
  • Other woods are much less resonant, and will resonate more evenly across a broader spectrum of frequencies. Woods like maple are in this category. Lower density does not guarantee this kind of  response characteristic.
    • This actually makes for a more balanced guitar, with fewer hot notes and fewer dead notes overall. It’s still possible to have these problems if the overall frequencies are in conflict.

I categorize back/side woods according to density, response type and appearance. Appearance being unimportant to sound.

Building highly responsive guitars: Bracing

The road to building exceptionally responsive classical guitars is littered with failed attempts and discarded theories. As I experiment a lot I’ve had my fair share of both. But scattered among my failures are diamonds that have pointed the way to designs, materials and processes that consistently result in instruments deemed extraordinary by experienced professionals.

The bracing pattern outlined above is the basis for the most recent, and most highly acclaimed series of guitars I’ve built. I refer to it as the Spyder pattern. I seldom build two identical instruments, but all the Spyder braced guitars have this common conceptual beginning. This design is extremely flexible and seemingly minor variations, like the size of the center diamond, can have very discernible effects on the sound of the instrument. 

When paired with a composite soundboard, and properly constructed, this pattern is a great platform for a highly responsive instrument.

Building a highly responsive guitar: The bridge

The stiffest and heaviest brace on a guitar is the bridge. Changing the mass and/or the design of this important element will have discernible effects on the sound of the strings.

  • A relatively heavy bridge will cause the attack (response) to be slow when compared to a relatively light bridge.
  • A relatively heavy bridge will also reduce the overall volume as it takes more energy to move a heavier object than a light one. Thus a heavier bridge makes for a less responsive guitar.
  • However the law of inertia implies that it will also take longer for the heavier bridge to stop vibrating (An object at rest stays at rest and an object in motion stays in motion with the same speed and in the same direction unless acted upon by an unbalanced force.) So the heavier bridge augments sustain. How much that matters with the weights involved here is debatable.

Taking all this into account, my approach is to go for maximum attack and volume by using relatively light weight woods for bridges. My target weight is 20 grams for the bridge, not including the saddle. Using a lighter wood like sapele, I can get a bridge as light as 14 grams, which includes a tie block overlay of a denser wood like ebony or gidgee. I also save a few grams by making the saddle from a dense wood like african blackwood or snakewood. The end result, when used in conjunction with properly built composite soundboard with diamond bracing is a very loud and responsive guitar. No one has complained of lacking sustain on these instruments.