Official Luthiers Forum!

What do you like to see for the density of the wood for the sound board of your 000 guitars?

The Lutz spruce tops I have fall into the range of 415 kg/m^3 or 26 lb/cubic feet. I am working through Gore's plate thickness formulas and noticed that my wood densities are higher than his (366 kg.m^3 Average). Therefore my plate mass is higher than he seems to recommend (170 grams area)

Do I need to dump these tops and get new ones?

_________________
_____________

EddieLee

Are your higher density tops also higher stiffness? If you thin them down to have the same stiffness will you be in the ball park for plate mass?

Gores tops seem to be on the extremely low end of the spectrum. I'd be happy with .415 g/cm3, I have made a pile lately from lutz in that range, and it's just fine. I'd you do,decide to dump them, please let me know...

Thanks Meddingfool, That give me a bit of confidence. The Spruce I can find data on is all above what Trevor uses in his book.

Clay, the stiffness numbers on my tops are close to what Trevor has in the book. After thinning to target thickness I am still heavier than his. In the book he rejected tops above 170 grams if I am reading that right.

_________________
_____________

EddieLee

From what I've read in Gore's books, he seems to have a target bracing plan that stays fairly consistent, and targets his material properties to fit those specifications, and therefore has stiffness to weight ratios that "work" with his plan, and selects his materials accordingly.

That doesn't mean you can't use those tops.

Other guys like Somogyi work the materials during assembly by tapping the tops and shaving the braces to target frequencies.

So I would say if you plan on building according to Gore's methods, you should use materials with properties that conform to his plan specifications if you want the correct outcome, and also stick to his plan dimensions.

_________________
Old growth, shmold growth!

That's a good point. Are you going falcate?

No. I am using a standard X brace. I am planing to use some of Trevor's methods, ideas, and calculations along with some of Alan's to quantify my normal tap and listen and flex style. Looks for things to do to get better and/or more consistent.

_________________
_____________

EddieLee

First: keep in mind that Gore's main focus is high 'monopole mobility'. Another term for this might be 'area to mass ratio'. He's looking for the lightest possible top. That might not be the same thing you're after.

I just worked up a spreadsheet of the 38 tops I have in stock, which includes only one Lutz top. That one comes in at a density of 382kg/m^3 (23.875 lb/ft^3). For various reasons, which may or may not be justifiable, I've been using the lengthwise Young's modulus to figure out how thick to make my tops. Using that, and the density, I've worked up a 'mass index number'; not exactly a prediction of what a top would weigh, since I make a lot of different sizes and shapes, but a relative indicator of the expected mass for the various tops when all else is equal. The result was interesting.

Keep in mind that, with a certain amount of normal scatter, including some outliers that are pretty far out there, the density is a reasonable predictor of the lengthwise Young's modulus for softwoods. In this range the values fall pretty close to a linear plot. Denser wood will generally be stiffer at a given thickness. Since the stiffness varies in a non-linear way with thickness, the denser tops will tend to be heavier, but not so much heavier as you might think. Thus, the density values I have run from about 340 to 500, while the mass index numbers are almost all between 160 and 200. If 180 is the norm, then very few of the tops I have would be much more than 20% lighter or heavier.

In his computer modeling study of the pschycoacoustics of guitars, Howard Wright found that people could generally hear a difference in mass of the 'main top' mode of 30%, so 20% would probably be noticeable to at least a lot of people. On the other hand, I had tops over almost the whole range of densities that fell within 5% of the 'norm' of 180, and I'm not sure most people would notice that.

There are some things people like, like 'headroom', that a heavier top seems to provide. I'm not convinced that any single metric is necessarily going to sort out the sheep from the goats. It's more important, IMO, to work on consistency to begin with. When you can do that, you're in a position to move your sound where you want it.

Lots of stuff here!

First, the point of the thicknessing formulae I use is so that you can get to a predictable vibrational performance; i.e. the modal frequencies of the finished guitar are more-or-less where you want them and your guitars sounds more-or-less the same (if that's what you want) from guitar to guitar, eliminating much of the random variation in sound you can get due to the variation in material properties. However, as the material properties vary, some sound boards are going to be lighter than others, even though they vibrate in a similar manner, so there will be variation in monopole mobility and hence responsiveness. Obviously, there is no free lunch in this unless you have "perfect" wood all the time, which none of us do.

In Table 4.5-3 in the book, a sample of my wood measurement results is shown, with Engelmann spruce having an average density of 366kg/m^3. Engelmann, on average, is the lowest density of the commonly used spruces, with the European I've measured coming in at ~430kg/m^3 and the Sitka at ~460kg/m^3. Funny thing is, that my "intuition" lead me to preferring Engelmann, long before I was measuring things!

Engelmann and WRC produce tops of around the same mass, on average, (trading density and stiffness) which, on average, are about 15% or so lighter than the other spruces, but because of the square root in the monopole mobility formula you only see about half of that (IIRC) ~7% improvement in the monopole mobility.

Whether to chose Sitka, Engelmann, European or Lutz depends on what you're aiming for. Lutz has a nice combination of properties, relatively low density, relatively high stiffness whilst being relatively hard (unlike WRC) and certainly produces nice guitars.

The main idea of my methods is to get the same result (given my initial explanation) regardless of the wood you use. You can "tune" the formulae to produce your own preferred result if you're not keen on mine. The test as to whether this stuff works or not is whether it produces a decent guitar from "strange" wood. I think "The Shed" guitar demonstrated that, build thread here, sound sample here. What I hear in this sample is that the radiata sounds somewhat more damped than say Engelmann or WRC (on average the top woods with the lowest damping), but you'd possibly not pick that unless you had "live" examples to compare it with.

That's perfectly justifiable! It's where I started until I figured out how to incorporate the cross grain stiffness etc. and I didn't know its relative importance. As it turns out, because the long grain stiffness is so much higher than the cross grain stiffness, the cross grain stiffness has only a small bearing on the vibrational performance of the top. A 50% variation in cross grain stiffness only results in a 3% variation in the vibrational frequency of the panel, so panel vibration is almost independent of cross grain stiffness.

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

http://www.goreguitars.com.au

Thanks Trevor: I wondered when you'd check in!

My feeling, based on experience and some of the things that came up in Dave Hurd's 'Left Brain Lutherie' book, is that cross grain stiffness becomes less useful in controlling top distortion over time. It probably still contributes to vibration stiffness, though.

Basically, the cells in softwoods have a more or less square cross section in the end grain view. One of the reasons that perfectly quartered softwood has such high cross grain stiffness is that the squares can take compression and tension forces along the faces. When the wood is cut off quarter the squares distort into diamond shapes at much lower force levels. As the top bends the squares can be distorted into parallelograms fairly easily, and will 'take a set' in the new shape over time. The top bellies and dips, and as time goes on the distortion becomes greater. However, the parallelograms are not so far out of square that they don't act like a flat piece for purposes of vibration transmission.

Again, this is a mental model that I have not had the time to test out. It does seem to fit the data I have, though. If it's correct then relying on cross grain stiffness to prevent long term top distortion is not a good idea. I used to think of high cross grain stiffness as a 'free ride' of sorts: it enabled you to reduce the thickness of the top a bit, and make it lighter. Now I'm less sure. I'd prefer to be conservative in thicknessing tops: a little extra thickness probably won't hurt the sound too much, and if it contributes to long-term stability and road-worthiness, I'm for it.

I referring to if he stuck strictly to the result targeted by the "build" book.

You have a part with a formula to find the Young's Modulous for particular piece of wood chosen, which determines the final thickness for that piece of wood, but don't you reject tops that fall out of a certain range, which is based on "your" target sound?

The design book outlines the process for determining what a particular builder likes, and describes how to determine the properties of the materials that produce a sound he likes, so that he can develop his own formula for repeatability, which will then allow him to select suitable materials which may have different qualities than your sound.

It doesn't look to me like you spend an inordinate amount of time shaping and chiseling braces and top thicknesses to achieve your sound, it seems you work from a plan based on material properties with some slight fine tuning after (as opposed to a set plan no matter what materials are used).

Or am I missing the point (again)?

_________________
Old growth, shmold growth!

Am I wrong in thinking that cross grain stiffness is irrelevant to long term structural stability? It seems to me that the only time there is any cross grain stressing is when the strings are activated causing the bridge to move. The rest of the time it is all long grain in the direction of string pull. If I'm wrong, please tell me.

I end up "rejecting" very few tops. I have good relationships with the suppliers I use and take a lot of trouble explaining what I want and generally that's what I get (and I thank them for taking the trouble!). That doesn't mean all the wood is the same of course, but the vast majority is usable for one type of guitar or another. For example, the tops that are going to turn out on the heavy side are going to have a lower monopole mobility, so that wood gets reserved for stage guitars with pick-ups, because the last thing you want on stage is an over-sensitive guitar that acts like a large microphone. Putting the modal frequencies where you want them gives you the fundamentals of the sound you want. Monopole mobility, simple put, just governs how much of that sound you get.

Yes, that's close to right. I would say "which will then allow him to use the materials he has to produce his sound which may have different qualities to your sound".

The vast majority of the time I use the top thickness predicted by the formulae in the book, just as written. With falcate bracing, the shaping of the braces is pretty straight forward. Just tapering. No scalloping or anything. About the only thing I vary is the starting height of the braces; higher when I want less sensitivity. After the CF is on, there's little you can do after the fact, without destroying the structural integrity, hence all the techniques on working on the coupled resonators (back, sides etc.) to do final trimming once the box is complete. Before completion, there's edge thinning, bridge mass etc. that can be used to trim things to intermediate targets.

No, I think you have a pretty good understanding of what I'm up to.

From the statics point of view (withstanding string tension) it is mainly the long grain stiffness that counts for long term stability. But cross grain stiffness does allow load shedding across the grain which couldn't happen if there was no cross grain stiffness. For traditional X braced guitars, the X dumps a lot of the load across the grain. For fan braced classical guitars the story is very different, because the braces are more-or-less grain aligned. So the bridge alone has the job of spreading the load laterally, which it does to a degree, but on lightly built fan braced classical guitars of any age you will almost always see significant brace print-through. On X braced SS guitars, the bracing (including bridge plate and bridge) will provide most of the cross grain stiffness, rather than the top wood itself. For fan braced classicals, the cross grain stiffness has a greater role, and will impact more on the frequencies of the higher transverse modes of vibration.

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

http://www.goreguitars.com.au

As Trevor says, almost any top will work on something; the tricks are:
1) to match the top with the guitar you want to build, and
2) to not end up with a lot of tops that won't work for the stuff you like/want to make.

If you're making Classical guitars you'll probably want to be looking for low density tops with at least moderate cross grain stiffness. If the objective is Dreads for flat picking, maybe you want more density, and if Jumbos are your thing, you're looking for higher cross grain stiffness along with that, IMO.

What I was talking about in my last post was what Trevor refers to as 'load shedding'. It's my impression that cross grain stiffness might help with that at first, but becomes less of a factor over time as 'cold creep' sets in.

If you're using 'tap tones' or 'Chladni tuning' you might find that it pays to look at the cross grain stiffness, though. Mark Blanchard gave a talk at a Healdsburg Festival a few years ago about this. He's been looking at the Chladni patterns on 'bare' unbraced tops for some time. What's interesting about them is that you see many of the same patterns on bare tops that you see on braced ones; in particular, the 'ring+' pattern that we often use as a guide shows up on most bare tops. Mark originally tried to work with the thickness of the bare tops to get the 'ring+' to show up clearly there, with the notion that this would allow him to standardize on his bracing and spend less time tuning at the end. It didn't work out that way. The stiffness ratio of the top wood itself is a pretty stubborn thing. What he did find was that, if the pattern showed up clearly on the bare plate, it would be easy to tune once it was braced, whereas if it didn't show up on the bare plate, it would be harder to tune. Ultimately he decided that "the sound is in the top", with the bracing being mostly for structure, and fine-tuning. You can certainly mess up a good top with bad bracing design or execution, but you can't make a poor top sound good simply by getting the bracing right.

Ultimately, he started sorting tops this way. He joints them all, takes them to a generous thickness, and cuts them to the shape of his largest box. Then he checks the modes. If they're not right, he cuts the shape down to the next smaller size, and tries again. Since, in general, guitar bodies are much the same length, but vary a lot in width, this is a process of matching the 'aspect ratio' of the body with the stiffness ratio of the wood. Tops with high cross grain stiffness work more naturally on Jumbo patterns, and wood with lower cross grain stiffness ends up on 00s. This is a way of working with Mother Nature, which is always the path of least resistance. You very seldom get her to do what you want if she doesn't, and the cost is always high.

Mark doesn't measure the actual Young's moduli, so far as I know. I've added his mode shape tracking to me own stiffness ratio measurements for several years. In the long run, with enough data, it should be possible to say what the workable range of stiffness ratios is for any given outline. Those of you who don't buy into Chladni testing need not pay any attention.

[quote= Funny thing is, that my "intuition" lead me to preferring Engelmann, long before I was measuring things![/quote]

I like to measure things. I get pretty obsessive about set-up measurements. I have notebooks full of displacements, densities, and eigenmodes. I don't pretend to know anywhere near as much about these as Trevor, or Alan, Carleen, Arthur, Ervin, or...

But, is it possible that we're talking about which tastes better, Scotch or Bourbon? I've never heard an Engelman topped guitar that I liked. I've never heard an Engelman top that didn't have a lot of damping. Still, some of my customers love the sharp attack, and quick decay that comes with a low density, high damping top. The classical guitars that appeal to me have a sweeter, ringing tone. European spruce at its best is my favorite. Sitka can be a close second. WRC is the best bargain on the market. The Lutz I have is somewhere between Sitka and Engelman.

This is my "intuition", but it's informed by customers' responses, including prominent dealers who mistook Sitka for WRC based on sound and appearance.

A guitar that has really turned my thinking on its ear is a current restoration project: a ladder-braced Jose Ramirez 1911 Tablao model. Rift-sawn pine top, crude workmanship, a single massive cross-brace in the lower bout, and a great sound. I will be building a copy of this guitar for myself, and I have no illusion that it will sound like the original.

The really strange thing is that I've heard a few people say that about Engelmann, until they've heard one of mine. The Engelmann I've measured (and subsequently built with) has always come out with really low damping numbers. The only top wood with generically lower damping (that I've measured) is WRC, and not by much. The Engelmann I've had has been more akin to a white WRC (soft, low density, low Young's modulus, low damping). But again, most people classify WRC as "dark" as the sound goes. I use WRC when I want a bright guitar! (that means lots of high frequencies, not to be confused with lots of "punch", which is what you get if you string a decent Adi Dred with PB 13s)

The Engelmann I've used is generally of lower Young's modulus than the other spruces, so I leave it a bit thicker, but its lower density means it ends up as a lower mass top. Certainly, if it gets thicknessed the same as the "hard" spruces (red, Sitka) it will produce a guitar with lower frequency resonant modes, which can verge on the muddy sounding if taken too far.

Not even Ervin? . Eric, you sell your self short... Ask Ervin what an Eigenmode is and he'll call you a "techno-weenie". (second photo down, top right corner; 'Techno-weenies only' sign given to Brian by Ervin).

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

http://www.goreguitars.com.au

Skimming through the numbers I've got, the Engelmann I've tested might run a little lower in Q value than Euro on the average, but there's so much variation that it's hard to say for sure. Some of it is about as low in damping as some Cedar or Redwood I've seen. Engelmann is also all over the place in density: some of the lowest, and highest, density spruce tops I've got are Engelmann. Go figure.

Yes, it's a crap shoot if you just pick wood off a pile. All the more reason for measuring the properties of the pieces you're about to turn into a guitar!

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

http://www.goreguitars.com.au

Thanks all for a really stimulating discussion. Lots of good stuff here.

Alan - Thanks for sharing that info about Mark Blanchard's process. I have heard of someone doing about the same thing with deflection measurements (I think). I went back and did a more precise long grain stiffness measurement and my mass to stiffness was in line with your measurements. I have your CD on free plate tuning. I will also try to incorporate some of those techniques into my next build. You wrote that Trevor's main focus is high 'monopole mobility' and that may not be what I am after. What are the trade off in other areas to get high 'monopole mobility'.

In my last guitar I did focus on trying to create a low mass top and selected a low mass bridge. I was curious if I really needed to worry about reducing the bridge mass by 5g to 10g. So I attached a 5g mass to the bridge. Just the 5 grams made a very noticeable difference in the tone and volume. So at this point I am leaning hard toward low mass. At least for finger style guitars. Although the mass of that guitar top may still be high by Trevor's standards.

This may not be best approach but I am picking a choosing techniques and formulas from Trevor's book to work into my next build. I view the thickness formula (4.5-7) as analogous to deflection testing with changing the f term equivalent to changing the amount of deflection. I hope that approach is justified. I also plan on using the formulas that relate the main monopole resonates to the fully coupled ones as a guide for voicing the top along with some of Alan's free plate tuning techniques. I will see if these clash with what I would build without these techniques. I am hoping to find thing that are measurable to back up what I am hearing and hopefully make a better guitar.

_________________
_____________

EddieLee

Eddie Lee wrote:
"I am hoping to find thing that are measurable to back up what I am hearing and hopefully make a better guitar."
Me too.

I don't have as much experience with really low mass tops as Trevor does, so I'm not as qualified to speak to the consequences of going very far one way or the other. In broad terms, it always seems to come down to a matter of 'balance' in some sense: not simply 'treble vs bass' or 'attack vs sustain' but a bunch of other things. For example, as you move more toward a 'responsive' instrument, you tend to get less 'headroom', and run more of a risk of 'wolf notes'. There are costs and benefits to everything, and you need to think about what you want, and what you're willing to pay for it.

As you've found, in some cases small changes in things like bridge mass can have fairly large consequences. In other cases you have to make pretty big changes before you notice anything at all. Sometimes you can give a little with one hand and take back with the other: I find that low-damping tops combined with higher damping B&S can work well in some cases, for example.

For me, I tend not to prefer extremes. The few really low-mass tops I've heard have had a timbre that I find unpleasant, and we all know what 'tanks' tend to sound like. Some of the numbers that Trevor gives seem 'way out of line from my experience: I'm not as happy with bridges as light as he uses, for example. There's no 'right' or 'wrong' in this, since the judgement of sound is subjective, after all. Everybody comes up with a system that works for them, and makes the sound they want to make, and that's fine, but my system, or Trevor's, may not be exactly what you're after.

Great comments! I have been an EE for a long time and know that Mother Nature seldom lets you have everything you what in a design

Thanks again for all the comments. All have been eye openers.

_________________
_____________

EddieLee

If it's any consolation I would have a better chance with 415 Kg than with 366 Kg. I too find the stuff at the light end of the scale difficult but that's probably because I'm more accustomed to stuff in the 400+ range.