Official Luthiers Forum!

Y'all,

Here's a look at my latest acoustic: The Heresy II








Maple front, walnut back




Neck-to-body joint. Screw-on neck. Mahogany, with a maple strip in the middle.
It eliminates that big hurky, pointy chunk of wood that graces the neck-joint of conventional acoustics.


Rosette, bridge and pickguard made of East Indian Rosewood

Here's a link to a YouTube video. It will give you a pretty good idea of how it sounds.



Like I say in the video, soundwise, I think it's a little "tight". But I think I understand why.

On my next guitar, I'm going to reduce the stiffness of the top - and bottom. I'm thinking of using mahogany for the back. Looking forward to seeing if this softens and sweetens the sound.

Questions, comments, critiques - let em fly!

Cheers,
Jeff
http://www.kozmguitars.com

I'm fascinated by the idea but it's not a little tight....it's a lot tight.

I don't think it's going to ever ring unless the front surface is flat. The curved resonating surface contradicts conventional luthrie for a number of reasons. One reason is runout. The best acoustic facings are lightest, tightest grained, straightest grained, and least runout...which translates to an excellent ratio of lightness to stiffness. Your curves force runout as a matter of design.

The curvature at the edges also blur the distinction between the sides and the resonating face. It intuitively seems to work against the goal of resonance not to have a distinct delineation between the resonant face and the sides. Spreading that delineation over an area, such as you have done with the rounded edges, could only dampen resonance.

I think the concept could work up to a point...maybe even work fairly well. I'm guessing, but I think the grain orientation of the bracing and the sides will probably never allow for the same kind of resonance that you'd hear from a conventionally designed acoustic...but I expect you could get a lot closer to it with some revision.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

I think the Heresy II is very similar in many ways to the early Orville Gibson guitars. Arched surfaces with glued on bridges and integral sides. Eventually Gibson's designs moved away from the glued on bridge and its accompanying physics and on to the floating bridge/tailpiece configuration that led to the modern archtop. The early Gibsons sounded really tight or sometimes downright dead. The glued on "rocker" style bridge wasn't that great at moving the top plate. Later Gibsons with the more violin-like (I like to think of it as more banjo-like) floating bridge/tailpiece arrangements gave the arched top room to sing.

I would counter Zlurgh's thoughts by saying that curved resonating surfaces with runout and all are a part of conventional lutherie, but part of conventional lutherie using a floating bridge and tailpiece with its more piston-like sound board movement moreso than the rocker-like movement of a fixed/pin type bridge. Archtops and flattops are really very different animals. I think the Heresy is more likely to find its place among the archtop animals.

Very cool concept and execution by the way!

Dave

Yes...I was going to mention archtops as a further point but time ran short. Went to see Les Miserables. Don't tell my wife...but...meh.

I'd note that the arch top is created to specifically NOT produce sparking overtones. Even so...the premise is still in line with my comments. With archtops, the edges of the top plate are carved thinly to specifically promote movement of the top plate as a unit. If the edges of an archtop left as thick as the rest of the plate the result would be a boxy, thin sound by comparison to a plate resonating freely. Both the traditional acoustic, with its tapered braces, and the archtop, with its thinned edge use the same premise of a soundboard allowed to resonate by controlling the flexibility of the constrained edge.

Yes...an archtop is curved but its curves move as a unit along with the whole plate. Since Jeff's design constrains movement at the edges, or at least since the design doesn't promote movement, it cannot function as a proper archtop. Since the soundboard is neither flat or super light, it cannot act as traditional acoustic either.

To me....the best way to go forward with this would be to create a very flat front, eliminate to a large extent the fillets along the inside top edge, and instead, experiment with progressively thinner outer edges when cutting the inside front half of the body. Whatever you can do to make the whole front plate move as a unit will do something to increase the bass response of the box. From there I'd whittle away the entire top thickness until it seemed dangerous.

You know where I'd start? I'd get an old beater acoustic and apply pressure to the front plate right at the bridge...until it broke. I have load cells and meters for this but a poor man's version of this could be done with a series of rated compression springs. My guess is that Jeff's structure here will be able to take up to several times more pressure than a traditional acoustic. Knowing the failure point of a traditional acoustic would provide a more insightful starting point that you wouldn't have to make a bunch of guitars to arrive at. You could experiment by making successively thinner front surfaces until finally one broke in the same manner as the traditional guitar using the same force. It's just a thought....but you'd only have to make the front plate to do the tests.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

Funny how little comments can make the brain work..... If the model's already in solidworks, it's a small step to simulate the behavior using the simulation plugin. Even the free one that comes with SW might be able to provide a good starting point.

My brain kept wandering and I started investigating and the premium versions of the simulation tools will also do modal analysis up to several hundred Hz which is were all the interesting stuff is.

Unfortunately, the simulation plugins cost more than solidworks itself I believe so maybe I'm just dreaming here.... opens up some fantastic possibilities though...

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Yeah...but the number you really want first is the amount of pressure it takes to break a conventional acoustic...making the assumption that the overall strength of the top face is, more or less, going to be some measure of its ability to resonate and that recreating that strength in Jeff's part is going to get the thing closer to a resonating box than he currently has. This is a bit of a stretch....but you have to start somewhere. I think understanding what is occurring with the geometry of either an acoustic or an archtop guitar to make its unique tone is the first thing to incorporate into this idea. The second thing would be to have some idea of what strength you are hoping to achieve with the surface...thus the destruction of an old beater. The third thing is to test iterations of Jeff's geometry to try to replicate the strength of the old broken beater.

I thought of software analysis too...but in terms of taking the model over to an engineer friend who has Ansys software with which you can simulate stress fields...among other things. I used to get all new propeller designs evaluated with this software. But...this kind of thing is entirely predicated on the mechanical properties of the material inputs being accurate....and wood is the least predictable structural material. I used it more as a comparative tool than any real predictor of mechanical performance. In the case of a guitar, you'd have to prepare the geometry of the acoustic guitar, identify all the woods, and THEN hope the mechanical properties of the materials in the program are valid...or you have to input them manually.

And that's on the Ansys...I trust Solidworks to be good at this FAR less than the $30K Ansys...which I didn't trust all that much.

By the time you're done with all that....breaking an old beater to get a number as a starting point doesn't seem like so much of a hassle.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

I looked a little deeper and the simulation xpress software or whatever comes free with SW standard probably would not be up to the task as it looks like it won't handle orthotropic materials. If it does or if you have the full version, I'd say that you could definitely get a baseline value without destroying anything.

It wouldn't take very long to model a guitar top with all the doming and everything using a typical plan and typical values for plate thickness etc. Heck, just doing it on the design he's got should be pretty easy and you'd at least get a figure for how much the bridge is rotating from string torque - typical values are in the Gore book I think.

A friend of mine works for Altair who also make one of those super high dollar sim packages (HyperWorks) and he said it would be pretty much kid stuff to model a guitar including the air coupling between the top and back etc. Basically once all the parameters are entered, you run it and you get a result that looks like the spectrum analysis people are doing anyway.

Like you say though - interesting stuff but unless you've got a grant or a sponsor or something - rather pricy!

If someone's got an acoustic guitar assembly in SW, I could get have him run a sim on it just to see how things behave. The key would be that it would have to be an assembly with each component a separate entity so that properties could be assigned accordingly.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

I think the SW numbers might be good in relative terms, so I'd be prone to do a simulation comparison between the two with them modeled as orthotropic materials (OK, so already into high end sim software)...and then I'd scale the result to what I got when I smashed in the top of a beater guitar.

The cheapest way I've found to get relatively well-gauged high loads is to get a budget air cylinder with a 3-4" bore. Calibrate your load per PSI with a bathroom scale and then go to town. Good within a few percentage points, which is all we need here. I pressured up the cylinder with a digital tire gauge/inflater which read down to 1PSI (unlike an analog pressure gauge, which is inaccurate for the first 20% or so of its travel) which then fed into a 'booster' tank with a lot more volume than the cylinder (to avoid the motion of the cylinder dramatically changing the pressure reading). It worked well enough for my masters thesis

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

Ya...good thinking. Once you have the model responding to the simulation exactly as your real world test indicated, you're in like flint on subsequent designs that don't stray too far beyond the basic geometry. What I was thinking was that you'd model the old beater perfectly, break it, then alter the mechanical properties of the materials until the model simulation broke at the real world test values. Knowing the material values are accurate brings the analysis software into play as a true design tool. We are talking about two different ways to achieve the same thing I think.

But I'm talking as if I can actually use the Ansys system and I want to be clear I can't. I needed to work with an engineer trained in that software. I just have vague notion of its full capabilities. I'd confer with him to design testing fixtures. I'd go home and make them and then we'd bend a propeller blade til it broke. We'd yank on 'em til they broke. We'd impact the leading edge. Then all these stresses were simulated with Ansys. After a while it got to the point where I had a kind of "gestalt" understanding of the fiberlay that would work with new design and things were relatively trouble free after that. That was the basic goal.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

Haha...that's why I tried to invent a methodology that Jeff could actually apply and get somewhere with.

Breaking the old beater and having a number as a reference, making maybe four or five front plates at different thickness, and breaking all those might....might give him the design he can make into a whole guitar and get a better tonal result without have to make five whole guitars.

I do love a good discussion on design though!

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

Ya but I think it would be a relatively low load. I'm guessing between 50-100 pounds of point load right on the bridge is going to break an acoustic. What do you think?

For a point load that small it might be easier to just create a simple jig and stack it with 5 lb. sandbags until the old beater broke. You'd have your number.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

As I was discussing this idea with my buddy at Altair and I asked about the air coupling thing he said something along the lines of - "how do you think cars got so quite in the last 10 years?". These sims are used heavily in automotive and other applications by the NVH teams to design out noise and stuff before a part is ever made.

I was talking to my bud at a solidworks re-seller and they had actually done a demo for a guitar maker and had a video of the solidworks tools doing modal analysis. Here's a screen shot. I'm working on getting permission to post a link to the video. Now, granted SW simulation is only like a $8k (only!) add on and I suspect it's a far cry from what the things like Ansys and Altair can do.

Super cool stuff I think. Stuart, with your experience in composites, you should work up CF acoustic guitar with all the right mode shapes to make it sound super nice then stamp em out by the thousands. Imagine, a student guitar that sounds fantastic and, you can use it as a sled in the winter time.

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com

Tell me Andrew....did you see the movie "Adaptation"?

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

Using spruce instead of maple for the top would make a much lighter top and go a very long way towards better tone IMO.

Y’all,

Thanks for the great comments and insights about my post.
A couple of thoughts from my side:

Regarding the idea of doing FEA on an acoustic guitar:
I agree that this would be a tremendously cool tool to have. Not so much as a tool to predict how a specifically designed guitar might sound - but as a means of doing “what if” experiments. Because of the grain-related complexities and “non-isotropicities” (now there’s a cool name for a band!) of wood, it would be a fool’s errand to try to meticulously match the actual FEA with the specific structural qualities and idiosyncrasies of each piece of wood used in a specific guitar… if your objective was to create a perfect model of a specific guitar. On the other hand, having a GENERIC model to play with, you could then do what-if experiments - investigating, e.g., the effects of thinning out the top, or a specific part of the top, or fattening up a particular rib, or changing the density and speed-of-sound qualities of one of the elements, etc. You could then feed this info into your next design.

Regarding modal-resonance patterns, chladni patterns, and the cult of the wood-tappers: I’ve been thinking a lot about this, and my thoughts are still a little half-baked, but I throw them out on the table here, for entertainment value, if nothing else.



I think it’s great to be able to vibrate a sound-board with a speaker-coil-like transducer, and sweep the frequency, looking for distinct modes of resonance… and then visually analyze it with the use of interferometers and strobe-lights and such. It can give us an idea of how a classically-braced soundboard behaves, and what resonant frequencies seem to be subjectively-optimal and culturally expected.

On the other hand, methinks this might all be a red herring. For 2 reasons:
1) The idea of “an ideal” sound board having a harmonic resonance at some specific frequency is, if nothing else, interesting. But it’s also kind of “so what”. I mean, if your objective was to design an instrument that had specific “howl points”, where instead of being a uniform radiator of acoustic energy, it had discrete peaks in its output spectrum… then you’d want to design the soundboard in such a classic manner.
However, if we didn’t have guitars or violins today, and someone suddenly came up with the idea to build such a device, I think the argument could be made that you’d want to design the box so that it DIDN’T have such acoustically-prominent “obnoxicities” . I.e., using the optical analogy of a diffraction grating, or prism, you wouldn’t want to shine uniform white light into one end, and have a weird signature with dramatic spikes in the orange and blue areas, and virtually nothing in the greens and violets, coming out the other side. Or maybe you would! IF (and only if) you’re ultimate objective was to duplicate the most popularly-received guitars of yore, despite their um, acoustic shortcomings.

2) The second reason I think there’s too much significance placed on resonant frequencies of soundboards, is the dynamic nature of music. It would be one thing if typical music consisted of looooong, single notes… plucked, and allowed to die out… and if the guitar was designed to resonate at those specific notes of the music. However, most music, is pretty dynamic. Notes are played rather quickly, sometimes 2, 3 or 6 notes at a time. Strings are bent and vibrato-ed. Further, to get a specific mode of a guitar to actually resonate at it’s fundamental pitch, it takes a finite length of time and excitation to actually get it to resonate. Again, unless you’re playing extremely slow musical pieces, with looong notes, I’d guess that they typically don’t really come into play.
Again; interesting information. Data points on the curve. But so what.

Regarding flatness vs curvature, and the topic of “runout”: I certainly wouldn’t be surprised if my non-flat guitar bodies never sound exactly like today’s revered and expected conventional acoustic guitars. As a species of finicky music lovers, we’ve come to know what we like, and expect. Most of us, anyways.
On the other hand, I’m hoping that I can come pretty close. Ideally, it would be cool to build “a lot” of different Heresy 2’s, using different types of woods, different wall thicknesses, different bracing patterns, etc., etc… and fully characterize how each input affects the overall acoustic signature. Maybe I’d eventually stumble upon the specific combination of qualities that would mimic today’s flat designs.

However, I’m hoping that I’ll only have to build a “small number” of guitars, to determine what the most-significant “knobs and levers” are. And then I’ll be able to get “close enough” to the sound of a conventional guitar.
However however; from the perspective of a hypothetical creative musician - playing an instrument with a different sound can be inspiring. It can also force you to drive it in a way you never driven a conventional guitar, and maybe cause you discover some terra nuevo, some new mojo, or some cool voodoo funk.

Lastly, regarding the lust for flatness: From the perspective of nature, “flat” is very UN-natural. Other than frozen ponds, I’m hard-pressed to think of many natural examples of perfectly flat objects. I would bet that the human larynx, and the equivalent devices found in the sound-paths of song birds aren’t flat either… and yet they don’t seem to suffer. Also, bells, organ pipes, flutes, clarinets and cymbals don’t seem to object to being flat.
I’m hoping that, in taking a more natural, organic, non-flat approach to guitar building… I’ll discover that it really doesn’t make that much of a difference. The mechanical energy injected into a plucked string will find its way… through the saddles, into the bridge…into the neck, into the body… back into the strings, back into the body… and eventually, it’s going to come out somewhere… over some period of time.

OK, that’s way more than I was planning on saying. Thanks for bearing with me.
Looking forward to your comments!

Jeff

http://www.kozmguitars.com