Official Luthiers Forum!

Thanks everyone for your help on this one. I had an idea that the bridge specs were an important part of shaping the response and tone of the instrument but I can see from your responses that it is probably more important than I first imagined.
Hesh thanks for the reassurance , I'v managed to knock another gram off and I'm planning on slotting the bridge so should be around 38g final weight. It is a reasonably large footprint 6.5 inches x 1.55in at the widest point.
Mike the top finished up at just under .125 .Felt reasonably stiff compared to the ADI top on my first instrument. Bracing was all 1/4 sitka . Just under 3/4 high at the X and tapering down to .4 in about 3in from the sides. Tone bars tapered from .450 down to .300. Finger braces were .350. I wanted it to be a bit of an all rounder. Strum and fingerpicking.
Bob is correct in that QLD Maple is not a true maple and to me it has it's own unique sound .Definately closer to a mahogany sound than maple though.
Thanks again folks for your detailed and thoughtful responses. This info will go into my reference folder.

Regards

Craig

Howard Klepper wrote:
"Next, mass has a relatively greater effect on impedance in the treble frequencies, and damping a relatively greater effect on it in the bass. "

It took me a long time to find the handle on impedance, so I may be wrong, but I disagree with you on that one point. One way I found helpful to think about it is to use the electrical analogy.

First: 'impedance' is the AC equivalent of 'resistance' in a DC circuit. DC resistance is a measure of how difficult it is to pass a current through a wire. Impedance measures how hard it is to pass an AC signal through a circuit at a particular frequency. The frequency dependance is important.

Several things can add to the impedance of an electrical circuit. We all know that putting a current through a wire creats a magnetic field: as the current builds up some of the energy is taken out of the current and stored in the magnetic field, to be 'given back' as the current dies down. Thus 'inductance' resists changes in current. Similarly, a capacitor stores some energy in the form of an electrical field ('static electricity') when there's a voltage change, and so a large capacitance opposes changes in voltage within a circuit. Good ol' resistance opposes the current all the time. Generally speaking the effect of a given inductance will be small at low frequencies, and larger as you go up. Capacitance has a large impedance effect at low frequencies, which drops off as you go up.

If you use the analogy that voltage>force and current>velocity, then inductance>mass and capacitance>1/stiffness (a large capacitor is like a soft spring). Resistance>friction, or damping in wood. Using this analogy a heavy bridge should block high frequencies, while a stiff one should block low ones.

The usual measures of damping in wood (or in electrical circuits, for that matter) involve either the 'Q-value' or 'log decrement'. These are actually inverse numbers: Q=1/log dec, iirc(sometimes you see 'pi' in there, depending on whether they're talking about frequency in Hz or radians per second). In log dec you figure out how many cycles it takes for the thing to lose half it's energy at resonance once the power is turned off, in Q value you look at how far off resonance you have to get, driving with a constsnt power, to drop the driven amplitude in half. Q gives you the proportion of energy lost per cycle: a top that has a Q value of 50 loses 1/50th of the energy in it for every cycle of vibration.

If the Q value of a material is inherent, and the same at all frequencies (neither of which is likely to be strictly true, but probably close), then the higher you go in frequency the faster the vibrations die out. Thus a low Q material (high losses) will tend to 'eat' high frequency energy faster. Cardboard is 'deader' than Brazilian rosewood, which is 'deader' than glass.

It is not at all clear to me how the damping factor of a bridge wood would effect the tone of the guitar. Sure it 'ought to' if there's enough of it, and it's bending enough, but the bridge does so little compared with the top, or even the back, that it's hard to assign a definite role, in my mind. I tend to think the mass and stiffness are pretty important, and pay less attention to the damping. I could be wrong: it wouldn't be the first time.

I never thought I'd be disagree with Al on anything of this sort, but my limited experience points towards Howards conclusion. Here's my point of reference which may or may not be valid. About 5 years ago I began experimenting with Steve Klein's Kasha-based design. I followed Steve's body, bridge, and bracing design as faithfully as I could without the benefit of interaction with Steve save one ten minute conversation with Klein and Kaufman at 2003 HGF. I did get the chance to play a couple of Klein's guitars and had a pretty good amount of research data to work with.

I built 4 guitars using the same top material (different B&S materials) and while they were all good sounding guitars, I was not satisified with the bass response in the first guitar. Steve's bridge design has significantly more mass on the bass side than on the treble, and I theorized that the mass of the bridge was having an adverse effect on the bass response. With each subsequent guitar I reduced the bass side mass until on the fourth guitar it was totally symmetrical on the both sides. The bass response improved with each guitar and on the last one, while not outstanding, I felt it had reached an acceptable level, given the Kasha bracing. It was nicely balanced across the entire tonal spectrum. The bridge was the only design deviation, thus my conclusion.

Of course, I have no way of scientifically validating this, but I'm not much of a scientific guy anyway. I've read some threads that want to make my head explode. All perfectly reasoned I'm sure, but I'm still trying to figure out a better way to glue on a fingerboard. It's kind of like the Brazilian vs. Ebony bridge debate. The science says one thing, but many of us hear something else. That's one of the things that makes lutherie so interesting.

I'm prepared to be totally wrong, so I'd like to hear differing opinions. Just my $0.02.

_________________
Jimmy Caldwell
http://www.caldwellguitars.com

This has been an interesting thread to read through. Lots of opinions and ideas.

I've found that the weight of the bridge can present a certain level of damping if
excessive,no matter what wood it's made from. A heavy piece of Brazilian Rosewood
is likely to present more damping than a lighter piece of Ebony.

I messed around with bridge shape and the area covered by the footprint of the
back in the mid 90s and found that these actually had much more bearing on tone
than weight in every case.

The best results of all the tests done seemed to be achieved with straight bridges
and, surprisingly to me, those with the pyramid style wings. This was just my experience
and just shaped my opinions concerning the issue.

I still love Ebony as a bridge material and the very best of the best sounding guitars
that I've ever played have had had Ebony bridges on them so I'll look to them and their
builders as my reference base.

I do, however, get very picky when choosing my bridge blanks. I listen to them for
both resonance volume and resonant frequency. There are widely varying characteristics
that are very obvious when listening to, say, 500 Ebony bridge blanks. Some are extremely
lively and have a very high frequency to the resonance and seem to ring like a piece of
glass while others have a dead thud to them and have very low frequency to what little
resonance they exhibit....and almost every variable between the two can be heard.

Because of this, it's obvious that the bridges made from the blanks will have their own
set of effects on tone, response and sustain of any guitar they're used on so the material
needs to be given a bit more consideration than color or grain orientation.

On top of their tonal characteristics, I keep all of my bridge blanks to about 40 grams
so that I'll know that they will finish up after shaping to my goal weight of between 25 and
30 grams.

Just my opinions gleaned from some fun observations and experiments over the years.

Regards,
Kevin Gallagher/Omega Guitars

Right, opinion time!

I've got my latest build nearing finishing stage, which means I'll be making the bridge soon (generally while the finish cures). Specs: Walnut/Cedar Jumbo (16.7" lower bout), deep body (4.75-4.8" tail block depth, plus wedge, so almost 5" treble side, 4" bass side). Tuned the top and back using chladni patterns, came out to around the 245 Hz mark for both, reasonably closed ring shapes (aside: impressive how I thought the top was 'good', then saw shapes, removed more mass, shapes and tap both opened up a bit). Helmholtz for the box is around G on low E (by ear), raised it to G# by drilling that 1.25" sound port in the upper bout (other aside: very easy to hear the difference when tapping the bridge area with hole covered and uncovered).

The top's nice and stiff WRC, about .140 thick at the center, tapering to a .110 at the lower bout edges (will likely loosen that up a little, take it down a little bit more, although the top's already very resonant), Adi bracing (with minimal CF in the X, Mario-style, so not adding that much stiffness), good ring. Back's Wanut, two X-braces (one lower, one upper bout), all X's about 5/8" tall and capped, with CF, tapered to almost nothing at the rims (0.4mm for the back braces).

The question becomes which bridge to use: I've got similar sized blanks of dark EIR (heavy, nice ring), black ebony (not a bad ring for ebony, not great) and macassar ebony (much nicer ring, between the two in weight). Given the helmholtz (fairly low) and the lively coupling of top and back (damping the back significantly damps the box's overall resonance), I'm not too worried about bass response, but I do want to ensure I get some nice high-end sparkle. At least, I *think* I don't need to worry too much about bass response - this is the first guitar I make that's this big, in all directions. It's almost too big to be comfortable, but the friend I'm making it for wanted a 'big country Jumbo', so that's what he's getting.

Thoughts? I've got madagascan rosewood as well which is a complete mismatch for the rest of the guitar, aesthetically, and some brazillian I want to resaw into back/side sets for a parlour sized guitar that can yield some bridge blank sized 'leftover' pieces, but that still needs cutting. I'm hoping for a guitar that'll respond reasonably well to fingerpicking (think Ryan Grand Cathedral or Lowden O series), within the limitations of the body size. I did try to talk the guy into an OM/OOO....

Jimmy Caldwell wrote:
"Steve's bridge design has significantly more mass on the bass side than on the treble, and I theorized that the mass of the bridge was having an adverse effect on the bass response. With each subsequent guitar I reduced the bass side mass until on the fourth guitar it was totally symmetrical on the both sides. The bass response improved with each guitar and on the last one, while not outstanding, I felt it had reached an acceptable level, given the Kasha bracing. It was nicely balanced across the entire tonal spectrum. The bridge was the only design deviation, thus my conclusion."

I have to wonder if the problem might have been the asymmetry.

It's commonly thought that guitar tops have 'bass' and 'treble' sides in a functional sense: that the low notes come off one side and the high notes off the other. This is not the case: the lowest notes involve the whole lower bout vibrating, and as you go up it breaks up into smaller and smaller areas over the entire top.

One theory is that if the higher order modes are too symmetric the different areas will simply cancel each other out, and you won't get any sound to speak of. The Kasha design is a somewhat extreme response to this.

I've found that the more symmetric I can make my tops the better I like them. There are always some sources of asymmetry in the natural variation of the wood, if nothing else, and I don't have much trouble with high end if I can get the low end pretty symmetric. I do find that pronounced asymmetry seems to be detrimental, at least in my guitars. Too much mass or stiffness on one side does seem to 'deaden' the response, possibly by making the lower modes less efficient. That would be a hard thing to measure, so I'm forced to go on gut feelings here.

At any rate, at low frequencies the bridge itself is pretty much one lump that all moves together, so you have to consider the impedance of the whole thing rather than the 'bass' and 'treble' ends (unless you're using a split bridge, of course, and even then....). Are the tapered bridges the same overall weight and stiffness as the straight ones?

I've found that the more symmetric I can make my tops the better I like them.

I don't suppose that includes making a bridge that is symmetric in thickness from bass to treble side? I've pondered trying to change the neck/body geometry to be able to achieve that but I can't see a way of doing it without making the guitar's inherent ergonomic problems worse. Kind of like a wedge with the thinner side down. I suppose a person could add some hidden mass to the treble side to balance the weight if not the stiffness...

_________________
http://www.chassonguitars.com

Kent Chasson asked:
"I don't suppose that includes making a bridge that is symmetric in thickness from bass to treble side? "

I should have written: "...more symmetric _within reason_..." The relatively small difference in bridge height between bass and treble shouldn't matter in terms of sound. Those Kasha 'trumpet' bridges do, I think.

Al, I enjoyed your analogy, could you expand on it to include the pluck of the string, which must be like a voltage (or is it current source?) and the string etc.

Is there a way to model the impedance of the string? Is there a way to model the string as a “voltage source equivalent? I think I get what the pluck does, but the relationship between the impedance of the string, now the “voltage source” for the guitar and how it drives it is not clear to me.

wow.

_________________
"Preoccupation with an effect gives it power and enhances the error"
from "Your Owner's Manual" by Burt Hotchkiss.

This seems to clear up my questions about string impedance.

http://www.physics.uiowa.edu/~fskiff/Physics_044/Impedance.pdf

It seems that it's a constant for a given string and just a function of its tension and mass.

Iirc, the 'characteristic impedance' of a string is a constant, proportional to the sqrt(T*m) where T is the tension and m is the mass per unit length. This is the impedance on an *infinitely long* string, and is the same at any frequency.

Fletcher and Rossing discuss the transverse impedance of a real string on an instrument, in chapter 2, section 11 of 'The Physics of Musical Instruments'. The math is too involved to put into a post here, and anyway, I'm not comfortable at that level. Suffice to say that the impedance is very nearly zero at the resonant pitches of the string and quite high in between. I strongly suspect that's an 'ideal' string on a 'rigid' mounting, but real strings probably aren't far off most of the time.

It used to be thought that a 'perfect' set of strings would all have the same tension, and thus the same 'feel'. In the real world most string sets are not like that, but depart a bit from equal tension in favor of a more equalized impedance. Since the impedance is a measure of how hard the string can push on the bridge (voltage, it seems) it stands to reason that you'd want all of the strings to push the same at the same amplitude. The match is not exact in the sets I've checked; another example of the sort of compromise that we see all the time in instrument making.

It's really handy to calculate the characteristic impedance of the strings when you're designing a harp or a hammered dulcimer, to make sure that ity varies smoothly. You don't want one to stand out, orit would be hard to play the instrument well.

I'm not following all of this. Here's what I got.

We can have a string all by itself. In the ideal model for that it doesn't favor any frequency so impedance is a tension- mass kind of thing.

Then constrain the string with two endpoints. Now it wants to swing at certain frequencies. So, impedance is low for the frequencies that is likes and high for all others.

Now if we have two different strings constrained by the same fixed points and one string is heavier than the other, and just to keep things simple, the tension is adjusted so that they swing at the same frequencies, then how do their impedances compare?
They're the same, right? Because the ratio of tension/mass is the same.

So what are you really doing when you calculate what you're calling the "characteristic impedance" for harp strings?

As always, your thoughtful clarifications are appreciated.

John

john Platko wrote:
"Now if we have two different strings constrained by the same fixed points and one string is heavier than the other, and just to keep things simple, the tension is adjusted so that they swing at the same frequencies, then how do their impedances compare?
They're the same, right? Because the ratio of tension/mass is the same."

No, they're different. Remember, the frequency of a string is proportional to sqrt(t/m), and the impedance is proportional to sqrt(t*m).

Suppose you have two plain steel strings, .010" and .014" in diameter. since the cross sectional area goes as the square of the radius, at the same pitch the .014" will be carrying about twice the tension as the .010" one(sqrt(2/2)=1). however, the impedance of the thicker string will be twice as high, since both the mass and tension are twice as great, and sqrt(2*2)=2 . Is trick, no?

Are wood density and dampening correlated? Why would Padouk be used for marimbas? Seems the denser the wood the better the sound transmission.???

_________________
A higher purpose for wood.
Rich Smith
Issaquah, WA

I suppose I should have done the math ...

Sooo. now we have two strings with the same modes but one has, (in your example) twice the impedance. How does that relate to: