Official Luthiers Forum!

I just finished assembling a soundbox on an OM'ish steel string....no binding yet...and not finish sanded yet. It doesn't have glue all over it either. The top is Western Red Cedar. Back and sides are East Indian Rosewood. Bracewood is spruce for everything. The depth is just shy of 4.250 at the tail block and 3.562 at the neck block. Soundhole is just shy of 4 inches...3.90.

Currently the main air resonance of the box is 25 cent to the flat side of A.

I am attempting to bring the main air resonance down. My perception, and bear in mind that I have absolutely NOTHING to base this on, is that it's too high. My last 2 guitars were between G and G#. They had spruce tops though.

So....any ideas?

_________________
There is no difference between the man that thinks he can....and the man that thinks he cannot.

If your referring to the main air resonance (not the top resonance), then try binding the soundhole making it smaller.

_________________
Formerly known as Adaboy.......

Lowering the main air resonance can also be done (ON purpose... or inadvertantly) by the tightness/looseness of the top and back bracing scheme, as well as the flexiness of the top and back..

The more flexy the top and back -- the lower the main air.

Now... I am not the one to ask about how to achieve PREDICTABLE results on this.... as I managed to re-top a cheapo Esteban dread and get an air resonance at 83 hz (Low E)... My goal was to loosen it up a *Bit*... Now, I gotta go pull the back and redo the bracing to try to get the main air back up into a desirable range (like main air ~90+ hz-ish ) and the main Top back up to a desirable range (190+ hz-ish)

I can tell you that if you make itty bitty tone bars and finger bars..... and space them about 1" off of the X brace.... it will produce a very low main air and a loose top.... BUT... you probably don't want to go this extreme for other reasons.

Good luck

John

I don't have anything to contribute to this thread, except that I'd like to nominate 'Flexiness' for OLF Word of the Week, and submit it for acceptance into the Official Luthiers Glossary.

_________________
Jim Hansen

I think it would be a good idea to put up some pics and add some data for some of the internal components.

Here's the top.
Tonebars - lowest valley is .360 (.258 wide when installed)
Lower X Brace arms - lowest valley is .470 (.235 wide when installed)
Top thickness is .115 -.119 depending on where you measure
Bridge Plate is EIR @ 90 thou



I don't have a photo of the bracing of the back bracing but there are 3 cross grain supports where joints meet - It's a 6 piece back - Here's a photo of the outside.
The two lower braces of the back started at .252 wide x .705 tall
I forgot to record the upper braces but I can tell you they are wider but not as tall.


Previous to posting this I worked the top bracing down a tad farther with an Ibex finger plane. Where the main air resonance was EXACTLY A before.....the main air resonance dropped almost 30 cent (headed towards G#) after the carving. A good estimate for the amount I took off the braces for the top would about 75 thou off the height off all the valleys and the peaks of the finger braces. I don't think I want to go too much farther because the x brace was installed at less than 250 thou wide.

_________________
There is no difference between the man that thinks he can....and the man that thinks he cannot.

As an inexperienced noob, it looks to me like you could move the "tapered part" or the peaks of the X-braces a bit more to the center, perhaps 1/4 to 1/2". That should loosen the top a bit and drop the top mode some. Not sure how the interactions with the air mode are so obviously YMMV.

I base this thought on looking through some pictures of bracing of my own guitar and others I've seen here and elsewhere (just googled some and checked). Perhaps a thread where people post their bracing would be nice to have eh?

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

My back braces are never taller than .590, maybe something to chop down on yours.

_________________
Tony Karol
www.karol-guitars.com
"let my passion .. fulfill yours"

Would it be possible for someone to do an explanation on what you are using to measure the resonance and why you would want to tune it to a particular note? I know this is a different way of voicing than many people use, but I am very curious about it. What would one call it and where would I find more information about it? Who uses it? Thanks in advance!!

It would be great if there were some simple way to predict what all the resonant frequencies of the comppleted box would be before you glue it up. It's SO much easier to get at stuff when it's on the bench, rather than having to reach in through the hole! Sadly, the very complexity of response that makes the guitar interesting to play also makes it very hard to predict what's going to happen. I have seen computer models that can give you a good idea of what the modes will be, but they're complicated to use, and only put the problem back a step or two. To get useful results out of them you have to be able to measure the wood properties very exactly, and that would take a lot of effort in itself.

In the low range the guitar acts as a 'bass reflex enclosure' with flexible walls. If you look this up in a book on speaker cabinets, you'll see that the ideal BRE has rigid walls, high damping for both the speaker cone and the air resonance, and the cone and the cabinet pitch are the same. Physically, this is a mechanical analog to the 'Butterworth' filter circuit, in which two L-C networks with the same properties are linked, but the speaker setup uses high damping where the goal in the electronic case is usually low damping.

With two resonant elements tuned to the same pitch you'd expect a single tall, narrow peak, but that's not what happens. Instead since it's impossible to ever have two things exactly alike, the resonant pitches of the two elements will always be slightly different, and that's enough of an opening for one of them to change phase relative to the other and start soaking up power (in a sense) rather than putting it out. The output of a Butterworth filter thus actually will have two peaks, with a dip in between, and it will fall off very quickly on either side. With care, you can have something approaching a 'square' response curve, with little output below or above selected values, and a fairly high, and nearly uniform output between. This makes it a good narrow-band filter.

In the case of the speaker cab, the high losses make the output less 'peaky', and the dip in between the two peaks you do have is shallow. The output rises pretty fast as you approach the resonant band, stays within 3dB of the peak height over a wide frequency range, and then falls off fairly slowly, so that it takes it a long time to drop 3dB below the peaks. Since a 3dB change in loudness is barely audible, the response curve of the speaker is fairly 'flat' over a wide band. It's also much more efficient than an 'infinite baffle': the box and port capture some of the energy from the back of the speaker and put it out in front.

On the guitar the 'real' Helmholtz pitch, and the 'real main top' pitch are not the same: often they would be about 6-8 semitones apart in isolation. They are, however, strongly coupled by pressure changes inside the box, and one outcome of that is to shift the pitches. The stronger the coupling, the more the pitch shift, so box depth and hole size get into the equation: a larger hole or a deeper box gives less pressure change for a given top amplitude, and thus less coupling. Thus when the 'Helmholtz' mode starts out around, say, 125 Hz, it can end up at more like 98 (G). I've seen it anywhere from 87 (F) to 233 (A#) on 'standard' gutiars with 'normal' soundholes. Similarly, the 'main top' mode, which might be around 180 Hz without the air coupling, can end up at more like 196 (G). Note that the air mode, I think because it involves moving less mass, shifts more in pitch.

There are various ideas about what the 'right' pitch is for these modes on the completed guitar, and why a particular pitch is 'right'. In a broad sense, I like to have the 'main air' pitch within a third or so of the lowest note on the guitar in normal tuning (whatever is normal for that guitar). The 'main air' mode fills in a lot of the fundamental of the lower notes. If it's too low, then you don't get the benefit very far up the string, and if it's too high, the lowest notes sound 'thin'. I like it between F# and G# for the most part.

It's even harder to give a fast rule about the main top pitch. In part, of course, you're sort of stuck with what you get when you make the top stiff enough not to fold up. Many people do caution against having the 'main top' pitch right on a played note, say the open G string. This is because top motion at this frequency (and the 'main air' pitch as well) can react back on the string and shift the pitch, or suck the energy out too fast, giving a 'wolf' note.

It gets more complicated when you factor in the back. In a way you can think of the back in the low ferequency range as an extension of the top area: if the 'main back' tap tone is anywhere near that 'mian top' pitch the two will couple. This has the effect of dropping the 'main air' pitch (the air has to move the back as well as the top, so the pitch is dragged down) and making the air resonance stronger. Even the sides can get into this act, so you can see why it's hard to calculate.

Anything that makes the top or back more flexible will tend to drop the 'main air' pitch. It turns out that loosening the top in the middle, by using scalloped braces, is more effective than loosening the edge. Apparently the top is a larger 'equivalent piston' when the center is flexing. Scalloped bracing also drops the 'real main top' pitch, so that it's closer to the 'real Helmholtz' pitch, and this makes the coupling stronger, which, in turn, causes a larger pitch shift (did I mention that the total pitch shift is a measure of the coupling strength?).

Making the body deeper drops the 'real Helmholtz' pitch, but it also decreases the coupling strength all else equal. The upshot is that the 'main air' frequency stays about the same (lower to begin with but less shifted down), and the 'main top' pitch can drop (same pitch to start with, but less shift upward). This seems counterintuitive until you understand the physics.

A larger soundhole in the same place raises the 'real Helmholtz' pitch, and decreases the coupling, so that the 'main air' mode can end up higher in pitch, and the 'main top' lower. Usually the top shifts less, again, because it's heavier.

And so on.....

Alan, thanks for taking so much time for such a well thought out response. I, as so many here, listen up when you speak. I am currently trying to find my own "voice" in the voicing process. I am pulled toward resonant tuning. I have a lot of research to do before I can understand more of what you have to offer. Do you have anything published? Have you considered it?

Alan thanks for the response. I've read it 4 times. Talking with you on the phone was a pleasure.

So here are the options I have available to me:
1.) Work the braces on the back - the lower 2 braces are currently around 680 thou tall.
2.) Work the top bracing - It's quite possible that I have room here especially when one considers the original dimensions. Plus - I'm fearless and will try anything once.
3.) Change the diameter of the soundhole.

Question - How does one go about making a soundhole smaller? I know binding it is the answer but some ideas regarding how to do it on an assembled body would be welcome.

Question - How does one determine how far to go in reducing the size of the soundhole?

I think I got a handle on working the braces. I did it on the last guitar with really good results.

Once again thanks to all that chime in. Many ideas in one place usually ends up forming answers....

Chris

_________________
There is no difference between the man that thinks he can....and the man that thinks he cannot.

Thanks Alan! Jody

From Al's post:

Al, sounds like teh back should not be tuned with the top to minimize the coupling affect. From your experience should the back be tuned lower or higher than the top to produce a good tone? I've no idea why but I've always thought of the back being tuned higher. Tuned higher would mean stiffer, and if you want a responsive back seems you would want to minimize the stiffness (though structurally it would be nice). Leaves me scratching my head <smile>.

_________________
Formerly known as Adaboy.......

Chris:
One way to drop the 'main air' pitch in a hurry is to use some sort of sleeve in the soundhole. If you look up Helmholtz resonators on line or wherever, you'll see that the classic model is a bottle with a neck. The air in the bottle acts as a spring, and the air in the neck is a piston with some mass. The longer the neck, the greater the mass, and the lower the resonant pitch for a given size of spring. On a guitar it doesn't look as though you've got much of a neck; just the thickness of the top. Actually, though, in any Helmholtz resonator, you have to throw in an 'end correction' to take account of the air near the hole that has to get pushed out of the way. This usually amounts to at least one hole diameter of added tube length, iirc. So the 'neck' on a normal guitar is something like 3-4" long. A sleeve will make it that much longer, of course.

You can make a sleeve out of almost anything: paper will do. Make it pretty deep (say, a couple of inches) to begin with, and cut it down until you like the result. Then you can make a more permanent one out of almost anything. Clear plastic makes a nice 'stealth' sleeve. Note that the sleeve reduces the sterngth of the Halmholtz resonance, since it adds drag to the air flow. That's one reason to prefer other methods, such as shaving back braces, but any port in a storm.

BTW, if you test the air mode pitch exactly, and then put your hand above the hole and check it again, you'll find that the pitch does, indeed, change. If you like to find that air pitch by singing into the hole, make sure and keep your face well back.

Daryl:
In terms of coupling strength, you get the greatest effect (the lowest air resonant frequency) when the 'main top' and 'main back' frequencies match exactly. There are risks with this, though.

For one thing, remember that the back is, in some sense, a parasite off the top. All of the energy that drives the guitar gets in through the bridge and top, and everything else steals energy from it. Since the top is the most effective sound producer on the guitar, you don't want to make anything else too good at stealing from it. It turns out that, as far as I can tell, the 'main back' resonant mode is the only back mode that can give a direct gain in overall output, and that's because of the way it couples with the 'main air' mode. So, you do want the back to be working at the low frequency range, but maybe not too well.

If the top and back are tuned exactly the same it's likely that the back would take a lot of energy from the top, and you could end up with a 'wolf' at that pitch. The energy would be extracted from the string so fast that it would lack sustain.

A bigger problem is when the back and top ALMOST match. I had a young lady bring in a guitar just before her senior recital that the Conservatory, to take care of the buzzes. Over a couple of hours we found and corrected all of the usual suspects: nut buzz, back buzz, fret buzz, and so on. This left a stubborn buzz on the low G# that nothing would fix. There were no loose braces or other such problems. When I checked the top and back modes, I found they were within 7 Hz of each other, with the top just at G#(about 208 Hz) an octave higher than the effected G#. The back was at 215 or so.

Apparently what was happening was that the back was close enough in pitch to the top to be able to effectively steal energy from it. This made the whole guitar body swell and shrink, like a big balloon that's having the air sucked in and out, with large pressure changes. Since the back (like most rosewood ones) had low damping and high mass it could store that energy like a flywheel. Eventually, the back would build up enough amplitude to start fighting with the top, and I think it actually got to the point of stopping the top from moving. At this point, there is nothing driving the back, so it switches over to it's preffered frequency (a little higher than the G#), and 'dumps' all of the energy it's storing out through the soundhole. Once the back stopped moving, the top could get going again, and the whole cycle built up to another 'dump'. This happened at the diffrerence frequency between the two plates, 7Hz, and sounded for all the world like a buzzing fret. A quick check, putting a small C-clamp on the saddle to drop the top pitch, eliminated the buzz.

Once we knew what the problem was the remedy was obvious: either shave the fan braces a little or add some weight to the top. Since it was getting late, and it's usually a good idea to do things reverseably, I elected to add weight. A few trials with various things turned up a square 1/4-20 nut that just did the trick. I super glued that to a little piece of neoprene, and stuck it inside the top under the bridge with double-sided tape, so it could be removed later if she wanted. The 'main top' pitch was now 11 Hz below the 'main back' and this was enough to reduce the coupling strength to harmlessness. The tone change was minimal, and there was no reduction in top stiffness, which is good.

I'll note that the 'main top' pitch on a new guitar typically drops as much as 1/4 semitone in the first month of playing. I also normally see the 'main top' pitch drop by about 1/2 semitone when I glue on the bridge. Thus, if I assemble a guitar and find that the 'main top' and 'main back' tones match before I put the bridge on, I can be pretty confident that they will end up around a semitone apart; close enough to be useful and far enough apart not to be a problem (usually). Sometimes there will ba a bit of 'chuff' on the attack of the notes near the 'main air' pitch, but that plays out fairly quickly.

Keep in mind that tops and backs can change pitch differently when the humidity changes, so a little distance is useful.