Weak trebles

[Mike_Kinny]

Hello all,

I strung my newly built guitar and have a problem. Note that The fingerboard is not attached permanently yet.

The basses sing like a dream. Long sustain, responsive and loud. However the trebles are just the opposite.

Light and stiff back/sides (plywood) reinforced with extra braces, Top lower bout 1 mm and upper bout 8 mm.

Guitar
Top
Back/sides

[Kerbie]

[Mike_Kinny]

Thanks Kerbie.

[charles Tauber]

In classical guitar making, generally, it is easy to create good bass response and more difficult to obtain good treble response.

The response, generally, is a function of mass, stiffness and damping. Some combination of those provides the response the maker wants. Repeating a known successful design increases the likelihood that one will obtain a similar response. The further one deviates from that, the greater the likelihood that the response one gets differs.

In your guitar, based on your description, you have a combination of mass, stiffness and damping that doesn’t give you the response you want. As a gross generalization, Mass favours bass response while stiffness favours treble response. at first glance, it looks to me like you have way over-braced the top back and sides: too much mass, some of which is likely in the wrong places to do much structurally.

Generally, a good sounding classical guitar is lightly constructed with as little mass as possible and with that mass carefully placed to maximize structure. As an example where you have not done that is the upper transverse brace and the braces surrounding the sound hole. In a rectangular cross section, stiffness is proportional to the cube of the height: double its height and it is 8x stiffer while being only twice the mass. You have used very tall sound hole braces, but then scalloped them heavily at each end. That gives you a very stiff brace that is very flexible at its ends. In a short brace, such as you’ve used around the sound hole, the result is a very heavy brace with stiffness little more than the scalloped ends. mass but not much stiffness. Similarly, you’ve heavily scalloped the ends of the transverse brace, reducing its stiffness where it is most needed to resist the torque of the neck. You are left with a brace that is heavy for the amount of stiffness it provides.

Upon further inspection, I see that you have laminated the upper bout of the top to a thickness of 8 mm. Arthur Overholtzer, back in the ‘70’s used a thick upper bout, but used otherwise traditional bracing and top thicknesses. The combination of the 1 mm thick lower bout, massive central longitudinal braces and light lattice braces haven’t given you a combination that yields the sound you want. You’d need to experiment with different combinations until you find the one that gives you the response you want. That’s the downside to deviating from the tried and true.

[Mike_Kinny]

Thanks Charles. The mass favors basses and stiffness favors trebles makes a lot of sense. For the next guitar I think I'm going to use less mass in more critical locations.

Re tried and true I have already made a Torres style guitar and enjoying it too. However I would never build the same guitar again since I already have one. That design like any other design has its limits and you will never build any guitar that's far better than the existing guitars of that style. For example IMHO both double top guitars and Smalman guitars are head and shoulder above traditional style classical guitars.

It's the search for the holly grail that keeps me building.

[charles Tauber]

There is nothing wrong with experimenting. The interesting question is how to go about it.

One approach is trial and error. You try something, it doesn’t work, you try something else you hope might produce a more desirable result. The downside is that you end up with a lot of failure.

Another approach is to try to reduce the amount of failed attempts by better understanding the desired outcome and the variables that produce it. It is more difficult in a complex system such as a guitar, but it provides potential direction for one’s experiments. Regardless, there is still trial and error, just maybe less of it, more of a more educated guess.

[Mike_Kinny]

I think I know why this happened. The two longitudinal braces neutralize most of the string pull. Here I'm not sure, does string pull add stiffness to the soundboard? If so then it explains it all. Basses excel and trebles suffer.

What do you think?

[charles Tauber]

"Stiffness is the extent to which an object resists deformation in response to an applied force." https://en.wikipedia.org/wiki/Stiffness

Stiffness can be characterized based on how forces are applied: bending stiffness, rotational stiffness, axial stiffness.

Stiffness is an "extensive property" that is a function of the material used and its geometry. Stiffness is not a function of the forces applied.


The strings exert a force on the guitar’s structure. For practical purposes, that force neither increases nor decreases that structure’s ability to resist those forces unless the forces sufficiently deform the structure into shapes that are less capable of resisting the applied forces, at which point the structure has essentially failed. The stiffness discussed in guitar construction is largely bending stiffness.

As previously noted, bending stiffness is a function of the geometry and the inherent properties of the material used. For a rectangular cross section, stiffness is proportional to the cube of the height: a small change in height produces a large change in stiffness, but proportionally much less change in mass.

Guitar response, like that of any mechanical system, is a function of mass, stiffness and damping. One produces a desirable sound as a result of finding the combination of those factors - mass, stiffness and damping - that gives the desired response. There are many possible practical implementations that achieve a desired combination of those factors, hence many different variations on guitar design. The practical goal of the guitar maker is to use some combination of the many variables that go into a guitar so as to provide the maker's desired response.

Your guitar has used a very thin lower bout - 1 mm thick. That makes for a very low stiffness and low mass. You have then used wide and very tall longitudinal braces near the centre that add tremendous stiffness in the direction of the string pull - though very little across the width of the instrument - and a lot of mass. That, in combination with the other variables you have used, produce the response your instrument has. It isn't, generally, one variable, one factor, but the combination of many of the variables that produces the response one gets. A guitar maker tries to find the right combination for the values of those many variables that produce the tone he or she wants. With the right values of other variables, the tall longitudinal braces can probably be made to produce the response you want. To find the right values of those variables has traditionally been accomplished through extensive trial and error, one result of which has been the standard designs most often used.

[redir]

Otherwise it's a cool looking guitar and you might find that the tone changes in the coming weeks if it's the first time you strung it up. You could also reach into the sound hole and modify those big bar braces too. Don't forget that generally speaking more mass in braces usually means bigger braces which are stiffer too.

[printer2]

What he said, pretty guitar, some of the mass in those braces have to go. Start at the ends and tapper to the bridge area.

[Alan Carruth]

Listen to Charles.

Most of the differences between classical and steel string guitars are aimed at addressing the differences between the strings. The rate at which energy is dissipated in particular is different. Steel as a material has low losses as compared with nylon: if you tap on a piece of steel it tends to ring for a while, where nylon 'thuds'. In addition, nylon strings are fatter than steel, so that have to move more air as they vibrate. This doesn't produce sound: it's more like trying to run in knee deep water; you don't don't make a current but just waste energy. According to one source these mechanisms are about equal in a nylon string, while a steel string, which has low internal loss, has even lower losses to viscosity. These sorts of mechanisms tend to dissipate a certain proportion of energy per cycle of vibration, so they usually cut down on power at high frequencies much faster.

If you pluck a steel string and a nylon string on a 'rigid' rig (a Les Paul...) in the same place and measure the signal produced at the top of the saddle, they start out exactly the same, with the same amount of energy in each partial. After a second or two there will be no energy left in the nylon string at high frequencies, but the steel string will still be going strong all the way up.

So, the main problem in making a good steel string guitar is to get enough bass to balance out all of the high end sound in the strings. This is generally pretty simple: make the box bigger (up to a point...). On a nylon string guitar the problem is to get as much use as you can out of the small amount of high frequency there is in the string signal. This is harder.

Most of the energy driving the top will come from the vertical component of the transverse string force. As the string moves 'up and down' relative to the soundboard it pulls the soundboard along with it, moving the lower bout like a loudspeaker. This is an effective way to produce sound. The top only needs to be stiff enough to keep from folding up under the static string load. This varies with the height of the strings off the top. Your longitudinal braces are 'way heavier than they need to be to take up that load. As a rule of thumb, Trevor Gore says that the bridge should rotate forward under string load by about two degrees; more than that and the soundboard is too loose, and much less and it's too stiff. I don't think I've ever seen a guitar top fail from being too weak; that is, fold up and pull apart as soon as it's strung up. They fail when they're not stiff enough, and creep into too much of a belly, at which point the bridge peels off and takes the top with it.

Most of the sound is produced by the lower bout of the top anyway, so I'm not too concerned with the beefed up upper bout. I do like to see some sort of 'balance' in the structure of the lower bout, though.

If you look at the classical guitars that use thin tops and take up the string tension with bracing, the braces tend to be distributed. Smallman uses a lattice of balsa and CF to take the load, with the 'soundboard' being more a membrane in between to move air. The 'sandwich' tops use what could be called a 'distributed I-beam' construction, where the lattice is simply a spacer ('web') to separate the 'flanges' of the upper and lower skins. The stiffness comes from the height of the web, and the flanges only need to be thick enough to avoid buckling under the load. The lattice of braces you have is, in itself, very nearly stiff enough (compare it with the fans on your Torres pattern), and with the elements so close together you could reduce the thickness of the top accordingly.

I suspect you reduced it too much, though. You say that the bass is strong, but you lack treble. That suggests to me that the lattice areas outside of the main longitudinal braces are vibrating at lower pitches than the structure of the bridge and main braces. This is part of what I mean by 'balance'; you've got spots that are much harder or softer than others, rather than a more 'graded' structure. It would be hard to add much to the stiffness of the top in this one, but you could certainly reduce the stiffness of the longitudinal braces to a more suitable value. I suspect you'll still end up with a 'bassy' instrument. I have seen classical guitars with lower bouts that were really soft and they still had decent treble response if the bracing and top were properly balanced. The only issue then is how long they hold up....