Official Luthiers Forum!

Sorry for a bit of a rabbit trail, but I'm curious as to what adhesive you use for the CF lamination in your bracing? Do you also use that to glue down the bracing to the top and back?

Also, do you factor the use of CF when calculating plate thickness and/or brace width/height?

Thanks!

I use epoxy with anything to do with CF and mostly Titebond otherwise.

Yes.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Can't yo just go down the road and set Maton straight, Trevor?

Ambiguity is good...occasionally!

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Do you mind sharing your preferences for epoxy brands?
Do you use HHG for certain places, and if so, where?

John Hall is getting ready to do a comparison test of how glues might affect tone. Does anyone know of any similar tests already published?

Tim McKnight has some articles on the topic of HHG and affects on tone. I suggest you visit his website and read them. Although I am very interested in hearing form Trevor and Allan more on their opinions and experience with the topic.

Gorjan

one of Tim's articles is on glue hardness. As I recall, his testing on tone - though very well devised - was done on metal plates. It seems to me that the only definitive test would be to compare different glues on a wood sample, and preferably glued down to some sort of wooden rim (if possible, the SAME rim). Comparing different types of wood might also be relevant (hard, soft, etc).

As has been pointed out elsewhere, so little glue is actually used on a guitar, I'd be surprised if it made a discernible difference - but that is what would be nice to determine. If true, then we can focus on the attributes we prefer (set time, repairability, etc), and move on to better understanding the nuances of wood.

I would expect there are people at Martin who have read all of the luthery books and understand these relationships as well as, if not better than, the rest of us. They are in the business of producing thousands of guitars at a price point that makes them competitive for market share. Manufacturing processes will be optimized to the greatest degree possible for low cost and minimal warranty work while providing satisfactory performance.

Part of my job as an R&D engineer is to be aware of what others are doing in my area of specialization and to always consider new methods and technologies for our products (medical imaging equipment). We make prototypes in our business, and some are pretty darn cosmic, but there is no way we could economically manufacture most of them. They are only a first step to something that we can actually put in the market. We are constantly assessing the performance we need to provide so we can maintain or improve our market share; "Extra" performance costs more to produce and does not necessarily result in more profit. I expect there are folks at the major guitar companies doing the same thing.

_________________
Steve Smith
"Music is what feelings sound like"

Trevor Gore wrote:
"Alan's explanation directly above addresses mainly the static issues, because he is measuring and adjusting for stiffness without a specific relationship to mass, so (and I'm sure Al will step in here if I'm selling him short) there's no direct relationship to modal frequencies, which ultimately determine a guitar's sound. Al has other ways of dealing with that."

Right. I take a somewhat different approach than Trevor does, although I strongly suspect that our best efforts would tend to converge simply because 'good' sound is good sound, and that's what we're all trying for (setting aside 'good for what' for the moment...).

If you think about it, there is a whole continuum of ways to make a workable 'flat' guitar top. You could make a solid top that's thick enough to take the load without any bracing. Nobody does it that way that I know of, simply because the top ends up being so heavy. It ought to be possible to make a 'sandwich' structure that would work; essentially a low-density synthetic wood that would require no bracing. You use normal wood, thin it out a bit, and add some bracing to get the stiffness back with less weight, and that's the usual method. You can also make a set of braces that will handle the whole load, with just enough membrane in between to move air; which you could call the 'Smallman' approach. Generally speaking, the lighter you go the more sound you get, and, as John Calkin said in 'American Lutherie': "Give 'em volume, they'll hear tone".

I'm not sure I totally buy into John's assessment, but, of course, that is in large part a matter of taste. The few Smallmans I've heard up close and personal have sounded like tin cans to me, and, although that timbre does even out a lot in a big hall, I'd rather not go there if I could help it. The sound I prefer seems to come about when I get some sort of 'balance' between the top itself and the bracing, which is just about where the best of the 'traditional' designs are. The 'free' plate tuning methods I use are mostly ways to get that balance right, I think, without the necessity of shaving braces through the sound hole. This method does tend to make a somewhat heavier top than you can get with more 'modern' technologies, so the guitars will not be as powerful, but, by being pretty careful about top thickness to begin with, you can at least make a considerable improvement in weight over the factories.

What's interesting to me is that my experience pretty well confirms what researchers like Wright have found: that the actual resonant pitches of the completed box don't make a lot of difference in the overall assessment of sound quality in listening tests. The resonance tend to mostly change the timbre of notes that are close to their pitches, and, as Trevor points out, you do want to avoid having strong resonances line up on played pitches if you can, particularly on lightly built and responsive instruments. And, of course, there is no simply predictable relationship between 'tap tone' or 'free' plate Chladni frequencies and the assembled modes of the box.

What does seem to make some difference in the completed guitar is the relative strengths of the low order resonances, their relative frequency placements, the form of the power spectrum in the low range, and the relative abundance and 'peakiness' of the high range peaks. Again, much of this is a matter of taste: I've got some ideas about how I like the spectra of my guitars to look, but you might not like what I do.

At any rate, whatever sound you like you probably would like to get more of it if you can do so and still end up with the tone you (or, more importantly, your customers) want. Since the bracing usually weighs less than the top plate it follows that the place to save weight is the top if you can figure out how. If you're counting on the top to provide any stiffness at all, you might find some system like my 'index numbers' to be useful, with the understanding that the numbers I gave are the ones that work for my system, and might not be right for you. If the top on your guitars is not a structural member, then maybe you need to figure out ways to 'tune' the bracing.

I'll note that mass in the top does not always seem to be a bad thing. More mass seems to correlate pretty well with more 'headroom', all else being equal. If you're going to make an all-out flatpicking Dread with scalloped braces you will probably find you'll get more satisfactory results with a nice dense piece of Red or Sitka, rather than a light WRC or Engelmann top. There's a reason behind tradition...

If you're going to make an all-out flatpicking Dread with scalloped braces you will probably find you'll get more satisfactory results with a nice dense piece of Red or Sitka, rather than a light WRC or Engelmann top.

There are, as always, exceptions to every rule... I've built absolute canons with WRC and Engelmann tops... In fact, it was Engelmann-topped dreads that started my career. Not because I was a good salesman, or did any advertising. Those who heard the first few guitars I built(all with Engelmann tops) absolutely had to have one. By my 5th guitar, I was selling to professional musicians, and by my 7th, had a 2+ year waiting list(now running at about 10 years..., and I've yet to spend a nickel on advertising). All because of the incredible volume and headroom my Englemann-topped dreads had, while not suffering from the "thin tone" malady that overly-light guitars suffer in exchange for volume.

This thread has digressed way too far from the original question; we(I'm guilty, too) should make an effort to stay on topic here.

True, but it's too good of a discussion for me to let it be, even if it doesn't match the thread title.



I'd be very interested in giving the approach in your books a try. But I think I'll wait until my second attempt to do that. Too many basic things to worry about in just getting the first one built!



That is probably to be expected. In most of the musical range, and certainly in the range where human hearing is more sensitive, the usual sizes of guitar will have high modal overlap. So the sound quality will be governed more by the modal density, the "peakiness" of the response, and any discernible shapes or patterns in the frequency response.



Interesting. It seems to me that there should be such a relationship, at least a loose one. If you are building a dread, for instance, there will certainly be differences between tops (thickness, bracing, etc.) that will affect the modal frequencies. But the low order free plate mode shapes will be basically the same. If you hold the air volume of the box and the modes of the back plate constant, you should get relatively the same changes when assembling the box and going to basically clamped plate modes, with the biggest variable being how closely the top couples to the other elements of the system due to frequency differences. But of course, this is lutherie, where the only thing we can hold constant is the speed of sound in air (and even that will be different in Denver than in Miami!).

If one had access to the proper software (I do) and enough time (I do not) a big, automated finite element Monte Carlo study could shed some light on the subject.



Is your system similar to the traditional x-bracing of Martin and Gibson? I am planning on a J-45 inspired guitar for my first build, but I would like very much to go beyond just building straight according to the plans. Would your tap tuning and stiffness index methods work within that framework?

In the end, you have to make a decision about plate thickness by one method or another

-Most factories are going to assume a fairly low stiffness to allow for the variability of the material and avoid excess warranty problems.
-you can just use the figure on the plan you have
-you can base your target on average figures for that species (Dangerous assumptiion)
-Use Alan's density method
-use various deflection testing methods either empirically or to determine a modulus
-use the gore tap frequency method and plate thickness calculations
-use your fingers and other senses, perhaps in conjuction with a standard panel for comparison

any of the last 4 methods will probably give you better consistency than a factory guitar

I think in the end Jeff sums it up pretty well. For me the testing is about consistency and sound board selection.
I found that when I tried Trevors thickness formula I got a number that jives with what I would have done using my data from previously builds, his number was maybe a touch thicker. Also it's worth noting that Trevors formula has a "knob" in it you can adjust if you don't agree with his setting. It'll bring you consistency if you're interested in that.

_________________
Jim Watts
http://jameswattsguitars.com

grumpy wrote:
"There are, as always, exceptions to every rule... I've built absolute canons with WRC and Engelmann tops... "

Yup; it's entirely possible to make cannons with WRC and Engelmann, and even to get decent headroom out of them. And, although WRC and Engelmann are usually low density woods, some of the densest spruce I've got is Engelmann, and I have a WRC top that exactly matches a Red spruce top in my stash in everything but damping. "The race is not always to the swift, nor the battle to the strong, but that's where the smart money goes".

ScooberJake wrote:
"If one had access to the proper software (I do) and enough time (I do not) a big, automated finite element Monte Carlo study could shed some light on the subject."

I'm sure there must be some relationship between 'free' top mode frequencies and the modes of the completed guitar, it's just that there are so many other variables involved. I can say on the basis of some 'matched pair' experiments that if you keep everything 'the same', the modes do track pretty well, but you don't have to change anything much to throw them out. For example, it's tempting to draw a parallel between the free top 'ring+' mode and the 'monopole' on the assembled top. You might, for example, assume that otherwise identical tops with the same ring+ frequency would yield guitars with the same monopole pitch, assuming everything else, such as shape, material and so on, is the same. In fact, if the ring+ shapes differ, you could end up with different monopole pitches, and the guitars will probably sound different even if the monopole pitches are the same. It's been a while since I tried to make a truly 'matched' pair, and at some point I need to do that again. Maybe next time I'll actually end up with guitars that sound 'the same'!

"Is your system similar to the traditional x-bracing of Martin and Gibson? I am planning on a J-45 inspired guitar for my first build, but I would like very much to go beyond just building straight according to the plans. Would your tap tuning and stiffness index methods work within that framework?"

I've used a variant of Martin's X bracing, as most of us have, but in the last few years I've gone to a symmetric 'double-X' pattern. My last 'almost matched pair' had top bracing as the variable, and blind trials at an ASIA meeting showed that two out of three players preferred the double-X by a small margin. The interesting thing was that many people commented that the 'standard' braced guitar had a more 'traditional' sound, while the 'double-X' one was more 'modern'. It's naturally easier to get symmetrical patterns on a symmetrically braced top.

I don't make many acoustic instruments these days so take this from an interested observer. I am of the opinion that the best top wood for acoustic instruments is determined by measuring its radiation ratio. It can be done in a wedge or a rectangle. In fact if the wood has a regular shape it can be averaged. I am not alone in this belief.
edit: just to head off the "what is best" questions, rightly so of course, and my answer is in every case from mandolin to double bass. The timbre can be altered with any method you chose but starting with high RR wood is a recipe for success.

_________________
Ken

grumpy wrote:
"There are, as always, exceptions to every rule... I've built absolute canons with WRC and Engelmann tops... "

Yup; it's entirely possible to make cannons with WRC and Engelmann, and even to get decent headroom out of them. And, although WRC and Engelmann are usually low density woods, some of the densest spruce I've got is Engelmann, and I have a WRC top that exactly matches a Red spruce top in my stash in everything but damping.

Al, I'm not talking of eng. and wrc that were exceptional and stiff and dense. I'm talking of run of the mill eng. and wrc..... The rule I was making exception of was that we need stiff, dense top wood in order to have great volume and headroom. I am also not talking about 'decent' headroom, but pretty much unlimited headroom.

To this day, I still don,t seek-out the stiffest tops.....

There's much more to all of this than the obvious. Unfortunately, what I found can't easily be described, much less taught. But I can at least put it "out there", and have for years, to hopefully make others realize and look into other methods.

Maybe some day I'll find a way to put it into words, and write book titled "Right Brain Lutherie"......

Ken McKay wrote:
" I am of the opinion that the best top wood for acoustic instruments is determined by measuring its radiation ratio."

For those of you not familiar with it, the 'radiation ratio' of a piece of wood is equal to the speed of sound over the density, so:
R=c/rho

Density is easy. Speed of sound refers to the speed of a compression wave in the wood, and is proportional to the square root of Young's modulus over density. So:
R= sqrt(E/rho)/rho

Hence the old ditty: "Rho, rho, rho your boat" referring to light, fast wooden racing shells... Sorry...

E along the grain varies linearly with the density in soft woods, or close enough, while across the grain it's difficult to find anything that correlates strongly except ring angle. Generally it works out that low density pieces of soft wood will have the highest radiation ratios.

R was a measure that John Schelling came up with, iirc, back in the early days of the Catgut Society and their research on fiddle acoustics. Basically, the higher the R rating, the lighter the top can be and still be stiff enough to work. The lighter top is easier to move, and makes more sound, all else equal. Note that 'more sound' might not equate with 'louder'.

Mario wrote:
"Al, I'm not talking of eng. and wrc that were exceptional and stiff and dense. I'm talking of run of the mill eng. and wrc..... The rule I was making exception of was that we need stiff, dense top wood in order to have great volume and headroom. I am also not talking about 'decent' headroom, but pretty much unlimited headroom."

... and so have I. One of my customers told about another builder trying to over drive one of my guitars at a gathering, and giving up after 45 minutes or so of really hard playing. Again, I was making a general rule: in my experience, it's easier to get a lot of headroom from denser tops, but there are always exceptions.

So, to get this back to the subject in the header: how do you measure wood stiffness, Mario?

My approach is more related to the practical selection of top which might be best in another thred, so my apologies up front.

The higher the radiation ratio the stiffer the plate is per weight.
Speed of sound can be easily measured by taping the billet of wood. Half a plate is good.
Get the first longitudinal frequency by holding the plate at one quarter its length and tapping it in the center or the end. Either record it into audacity or hum the tone and get its frequency.

To get the speed of sound of your wood multiply the above frequency by the a correction factor of .98 times the length in cm squared and divide that by the average thickness. For meters per second divide by 100.
Speed of sound divided by density is your radiation number.

Mario, I would be curious where your good top wood comes in? I have to notice that from what you are saying, high RR wood might not be your preference. Humm interesting, you should write a book.

_________________
Ken

how do you measure wood stiffness, Mario?

I simply flex it, with my hands, but possibly the thing I pay most attention to is the wood's "springback". I hold the plate in my palms(IE: no grip), and move my arms forward then pull-back quickly. This sends the plate into a "whip" of sorts. What I can't describe is what I'm seeking, and when to stop thinning. It's all intuitive, but I can state that it appears that the wood qualities that lead to what I want appear to be somewhat independent of stiffness and/or density. Yes, there is a direct correlation, but much less than I'd expect, and there are enough outright exceptions that I wouldn't trust my choices simply to structural numbers, no matter how they were arrived at.

And of course, I tap and listen, but not to the pitch. I listen to the tone, and how simple or complex it may be. And always, listen for any internal or invisible problems which will often manifest themselves as strange pitches(wolfs?) or barely discernible buzzing.

Also, keep in mind that I'm not seeking absolute consistency; my goal, and what I possibly do best, is to tailor the tone to the owner's needs and/or wishes. I don't rely on mere wood choices to "predict" the final tone of a guitar or mandolin. More often than not, I'll tell my clients to choose the back and sides with their eyes, then tell me how they want it to sound, and more importantly, what they need from the instrument and where they'll play it most. I take it from there and choose everything else accordingly.

Wouldn't this radiation thing be pretty much what we do when we tap and listen, and compare that to the stiffness? Speed of sound is directly related to damping, no?

I think I get the gist of what you are describing Mario...and your testing for springback on a top is certainly very logical as the results of said testing are an indication of the amount/type of energy a top would be capable of producing...unlike metals, woods have much more complex (unpredictable?) characteristics that won't be explained by a simple Young's modulus test...

what I wonder is if part of your thought process is bringing out the best of that piece of wood as opposed to tailoring it to a certain customer...seems to me that you would be forced at times to discard a top because it would inherently be incapable of giving the response you envision for a particular client...

seems to me that you would be forced at times to discard a top because it would inherently be incapable of giving the response you envision for a particular client...

Of course! But I have enough wood on hand to just keep searching for the right one. The trick, if there is one, is to not get 'married' to any one set of wood.....

I like it! Good old lutherie! That is the good stuff.

As for damping, Alan can describe it better but damping is related to how much a certain resonance in the spectrum is present. When you tap you can hear the damping characteristics.
As an example, Balsa might have a very high RR but would obviously not be right for an acousitic guitar due to its damping characteristics. So basicaly there is RR and then there is damping. Separate properties.

_________________
Ken

Ah, that's another great antipodean way of testing wood, originally pioneered by Rolf Harris. And for the completely addicted, here's more and yet more...

Joking aside, I can imagine that that is as good a way as any of the tactile methods, given some experience (which I'm sure Mario doesn't lack).


There has to be; but when you consider the number of variables involved and how many of them couple, it's just not a practical way of working.

Radiation ratio, sometimes called sound radiation coefficient (depending on where you come from) can be a useful comparative parameter; but remember that the waves in guitar tops that produce the sound we hear are bending waves, not longitudinal compression waves. But bending waves are dispersive, which is probably why the speed of the longitudinal wave is usually used in the radiation ratio formula.

No. Speed of sound is just the wave velocity. Damping is about the rate at which the amplitude of the waves decay. Put another way, it's about how long the tap tone lasts for. The sound radiation coefficient (radiation ratio) if you go back to basic formula, combines a factor indicating how easy a material is to deflect (Young's modulus) with a factor indicating how easy it is to accelerate (density) and so is an index of the acoustical power that can be produced by a driven vibrating plate before damping losses are taken into account.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Grumpy:
I do all that stuff too, along with the objective tests. You can't have too much information.

Some people are really good at finding wood stiffness 'by hand', others are not. A report I got of some testing at a violin society meeting suggests that most people are not as good as they think they are. OTOH, if it's working for you, that's great. I do find it a little amusing that you insist on having your humidifier properly calibrated, and work top thickness by feel.

Trevor Gore wrote:
".....when you consider the number of variables involved and how many of them couple, it's just not a practical way of working."

Not if what you're trying to do is get exact reproducible assembled mode frequencies. I find I get good sounding guitars this way, and the assembled modes are never very far out of line: I seldom need to do much 'fine tuning' after assembling the box due to 'wolf' note issues, for example.

"...remember that the waves in guitar tops that produce the sound we hear are bending waves, not longitudinal compression waves. But bending waves are dispersive, which is probably why the speed of the longitudinal wave is usually used in the radiation ratio formula."

Right. I'm sure Schelling was primarily looking for a fairly simple useful way to characterize the difference between wood samples. We all know that wood properties are only a place to start, and not an iron law. The man was an electrical engineer and not a luthier, but he had some very useful insights.

All of this really makes the point that there are lots of opinions as to the 'best' way to work, and lots of good ways to make guitars. It always seems to me that, to the extent we're trying for the same things, all the workable methods have to converge. If we could really get together, and talk about this stuff in a context that would allow us to cross check each other's methods, we'd find we're not that far apart, even though we may seem to be sometimes. I'll always remember 'tracking' Dana Bourgeois through 'tap tuning' a guitar top at a GAL convention. I'd never been able to make heads or tails of it from his writing, but once he was able to show me what he meant by terms like 'kick', and I was able to see how his actions changed the way the guitar modes worked, it all made sense. Alas, the guitar top cracked when he took it home, so we were unable to do the further tests we wanted to on the completed instrument. If not for that we'd have published something in American Lutherie.

I do find it a little amusing that you insist on having your humidifier properly calibrated, and work top thickness by feel

Well, it's pretty simple..... I can't tell what the RH is with my hands or ears, so I rely on a precision instrument to do so for me.... But ya know, with my hair reaching down to my belt these days, I could probably gauge the RH just by how frizzy(or not) my hair is...

For sure, there is no -one- "best" way to do anything in lutherie. I spoke-up here to offer a counterpoint to the left brain methods in case we were to scare away a newbie, or even discourage a seasoned amateur or two into thinking that because they can't quite "grasp" what y'all are saying that it's pointless to even try to pursue building better and better instruments. In other words, it can be as simple as choosing and/or working the wood until it simply "feels right". After all, a lot of excellent instruments, some worth millions, were created by men who did it strictly by intuition....