As some here may know, I've been doing measurements on acoustic guitar bodies. My preferred way is with pink noise as excitation, though I can and have done it with Swept Sine. I'm using a Dayton DAEX25FHE exciter, weighing in at about 3.5 oz. I simply rest the exciter output on the guitar's bridge/saddle and use a rubber band and clip to keep it from "walking". It doesnt take much amplitude to get a clear response spectrum; the amp I use (for the pink noise anyway) runs off a +5V USB port, with its volume setting nowhere near full.
I presented this in another forum, with spectrums before and after the application of a "Tornavoz"- guitar speak for a ducted port - invented in the late 17th century. In the spectrums, both the cavity / port and first mode of the top of the guitar resonance was evident. One astute reader challenged the validity, saying the exciter mass weighing down the top - as now part of the system - would change its resonant frequency appreciably.
Made sense to me.
Of course, the going from a plain hole to a ducted port only changes the air cavity resonance. Adding a weight to the bridge area wont change that. So the effect of the Tornavoz was valid frequency wise (it brought resonance down, as expected)
So I tried an experiment; one, where the 3.5 oz weight of the exciter was applied as a downward force; held by clip and band, the other, the guitar was held vertically as if played, with the exciter held by hand, pressed against the saddle / bridge by hand. So the exciter weight was hand held, versus guitar held.
No difference in top resonance, in fact it went slightly down being held by hand - the opposite of what one would think. How can this be?
Something to do with the compliance of the exciter at resonance makes its virtual weight go to zero - I just cant wrap my head around it. Any help with thinking this through would be appreciated! Thanks.
I presented this in another forum, with spectrums before and after the application of a "Tornavoz"- guitar speak for a ducted port - invented in the late 17th century. In the spectrums, both the cavity / port and first mode of the top of the guitar resonance was evident. One astute reader challenged the validity, saying the exciter mass weighing down the top - as now part of the system - would change its resonant frequency appreciably.
Made sense to me.
Of course, the going from a plain hole to a ducted port only changes the air cavity resonance. Adding a weight to the bridge area wont change that. So the effect of the Tornavoz was valid frequency wise (it brought resonance down, as expected)
So I tried an experiment; one, where the 3.5 oz weight of the exciter was applied as a downward force; held by clip and band, the other, the guitar was held vertically as if played, with the exciter held by hand, pressed against the saddle / bridge by hand. So the exciter weight was hand held, versus guitar held.
No difference in top resonance, in fact it went slightly down being held by hand - the opposite of what one would think. How can this be?
Something to do with the compliance of the exciter at resonance makes its virtual weight go to zero - I just cant wrap my head around it. Any help with thinking this through would be appreciated! Thanks.
Interesting thought.
The weight is a constant, it is a 'DC' / constant load; but what you are interested in, is the AC response to excitation of the surface / cavity resonance
Presumably then the body responance is fairly high , what say 80Hz+; so if your measurement window is an effective high-pass filter, the the mass of the exciter I think does disappear. it is a pure forcing function of AC excitation. Even if the Dayton unit were heavy enough to deflect the top - the effect of the mass/top stiffness : air - which is effectively a Helmholz resonance - is not going to be changed.
Experiment with adding some known dead weight(s) in addition to the top of the dayton unit, and try again; I bet the resonance you are interested in, barely changes.
The weight is a constant, it is a 'DC' / constant load; but what you are interested in, is the AC response to excitation of the surface / cavity resonance
Presumably then the body responance is fairly high , what say 80Hz+; so if your measurement window is an effective high-pass filter, the the mass of the exciter I think does disappear. it is a pure forcing function of AC excitation. Even if the Dayton unit were heavy enough to deflect the top - the effect of the mass/top stiffness : air - which is effectively a Helmholz resonance - is not going to be changed.
Experiment with adding some known dead weight(s) in addition to the top of the dayton unit, and try again; I bet the resonance you are interested in, barely changes.
Understandably, the Helmholz resonance (typically 110 Hz) doesnt change, as adding weight to the guitar top changes neither the box dimensions nor the size of the port. But the next resonance up (typically 240 Hz) is due to the top's 1st mode motion, like that of a trampoline.
If what you're saying is the top @ resonance is isolated from the mass of the exciter by its suspension compliance - which is perhaps compressed just a tiny amount by the 3.5 oz weight. I could add weight to see the resonance change, but what I still dont understand is why adding the force of weight through the suspension compliance - in motion, in phase with the response its driving - has no effect.
All I can think of is it must be the driver suspension that isolates the driver's mass from the resonant system. Why it does that in a "How It Works" style explanation, is still beyond me.
If what you're saying is the top @ resonance is isolated from the mass of the exciter by its suspension compliance - which is perhaps compressed just a tiny amount by the 3.5 oz weight. I could add weight to see the resonance change, but what I still dont understand is why adding the force of weight through the suspension compliance - in motion, in phase with the response its driving - has no effect.
All I can think of is it must be the driver suspension that isolates the driver's mass from the resonant system. Why it does that in a "How It Works" style explanation, is still beyond me.
A thought:
What does this guitar under test bracing structure look like?
A sketch or diagram of the bracing and the exciter location may be useful.
As far as the vertical test resulting in slightly lower, perhaps that is from slightly reduced mass on the soundboard (gravity) allowing it to be less stiff.
What does this guitar under test bracing structure look like?
A sketch or diagram of the bracing and the exciter location may be useful.
As far as the vertical test resulting in slightly lower, perhaps that is from slightly reduced mass on the soundboard (gravity) allowing it to be less stiff.
There is a device called Kundt's Tube, studied in high school physics.
https://en.wikipedia.org/wiki/Kundt's_tube
Some other devices are also mentioned at the bottom of the article.
The pitot tubes and angle of attack sensors on aircraft are all based on air pressure sensing, so the answer to your query may be found there, sometimes you have to start at the basics.
That happens sometimes, people assume you have the basic knowledge, some people don't, some have forgotten.
Your query was how the resonance pattern changes when the guitar is vertical and horizontal, was it not?
Can you do that experiment with another shaped string instrument?
I mean a different style guitar, mandolin, violin, or similar.
Maybe the smaller port location changes from vertical to horizontal, and the interference pattern changes?
I am assuming two ports, a picture wold be nice.
Sometimes the strings are supported on a tube rather than solid wood? That also would be considered a tube?
I am not a guitar expert, forgive me for not having much knowledge...
https://en.wikipedia.org/wiki/Kundt's_tube
Some other devices are also mentioned at the bottom of the article.
The pitot tubes and angle of attack sensors on aircraft are all based on air pressure sensing, so the answer to your query may be found there, sometimes you have to start at the basics.
That happens sometimes, people assume you have the basic knowledge, some people don't, some have forgotten.
Your query was how the resonance pattern changes when the guitar is vertical and horizontal, was it not?
Can you do that experiment with another shaped string instrument?
I mean a different style guitar, mandolin, violin, or similar.
Maybe the smaller port location changes from vertical to horizontal, and the interference pattern changes?
I am assuming two ports, a picture wold be nice.
Sometimes the strings are supported on a tube rather than solid wood? That also would be considered a tube?
I am not a guitar expert, forgive me for not having much knowledge...
I read the post again, your query was hand held vs. rubber band?
Noise cancellation at work?
The trial can be made with a different exciter for further research....if it is a different design, that difference may show up as well.
Noise cancellation at work?
The trial can be made with a different exciter for further research....if it is a different design, that difference may show up as well.
Could you 'tap' the bridge area (with finger or some other contraption) to make an impulse measurement, and then compare that to the exciter measurement? My gut feeling too is that the weight of the exciter would affect the result.
Thanks, guys. I'm planning on doing that, in fact I've collected data on a number of different instruments, which needs to be verified.
Verified in the context of what interests me, is when the guitar is strung and played, does the top resonate at the same frequency as revealed by the test process? The standing objection is that no, it wont, because of the added 3.5 oz weight of the transducer. Here's a picture of how I've attached it to drive sound into the guitar bridge / saddle;
The little 1/2 size guitar is flat on its back, with the top horizontal. The strings are tuned, but mechanically muted to eliminate their sympathetic vibration. A measurement mic is placed above the guitar, pointing down at it, with a typical measurement looking something like this;
The first peak at ~150 Hz is the Helmholtz resonance, a product of the air chamber and sound hole diameter. I introduced a round duct, which lowered that peak in frequency - as expected, no challenge. The second peak, at ~300Hz, is the 1st mode of the wood top and its bracing, with a trampoline like motion. It is this value that's contended, due to the weight of the exciter. And I assume everything north of it as well.
Repeating the experiment, taking the weight of the exciter off the guitar, by holding it vertically and with exciter in hand, touching it to the saddle as shown, sans band and clip. I get pretty much the same value, in fact it goes a bit lower without the added weight. Why? It could simply be whatever contact pressure - as the tranducer has to contact the bridge somehow to effect resonance - I applied by hand is the same as, or pretty close to, the 3.5 oz weight.
I can forget trying to understand why and just go after the validity of the measurement technique. Which is what I want to ultimately establish. Guitars are easy to tune. I could simply find a note close to the "310" value (as indicated above), tune that string so the note is exactly 310Hz, right at the resonance point. I should clearly hear it. Then, sound the notes at either side, # or b, and the response would soften, which should be as well perceptible by ear.
I just thought why would be an interesting Lounge topic. As any acoustical experiment might be, where what you're getting is a little surprising and unexpected. Applicable to anything DIYAudio? I dunno; has anyone ever made a speaker cabinet with thin walls, tuned for a desired response via a bracing pattern - as the master Luthiers of old did? No. All of them are either you get what you get, or some herculean effort is made to make box walls acoustically dead. It would be interesting to create a valid measurement method applicable for this sort of thing also.
Last edited:
Would this be better than the one you have?
https://www.digikey.com/en/products/detail/seeed-technology-co.,-ltd/101020031/5488092?utm_adgroup=Seeed Technology CO., LTD.&utm_source=google&utm_medium=cpc&utm_campaign=Shopping_DK+Supplier_Tier 1 - Block 3&utm_term=&utm_content=Seeed Technology CO., LTD.&gclid=EAIaIQobChMIramy57Cx_QIVQh6tBh16owhnEAQYASABEgLOlPD_BwE
Maybe play with layers of lacquer,bracing, etc., until it sounds good, then measure that and document it for future reference.
https://www.digikey.com/en/products/detail/seeed-technology-co.,-ltd/101020031/5488092?utm_adgroup=Seeed Technology CO., LTD.&utm_source=google&utm_medium=cpc&utm_campaign=Shopping_DK+Supplier_Tier 1 - Block 3&utm_term=&utm_content=Seeed Technology CO., LTD.&gclid=EAIaIQobChMIramy57Cx_QIVQh6tBh16owhnEAQYASABEgLOlPD_BwE
Maybe play with layers of lacquer,bracing, etc., until it sounds good, then measure that and document it for future reference.
Try a sound deadening material used in cars?
The ones used here resemble coal tar felt, which could be a roofing / water proofing material as well.
An old phono pick up would work as a sound level pick up as well, as would many industrial stethoscopes, or as above, a vibration sensor.
The ones used here resemble coal tar felt, which could be a roofing / water proofing material as well.
An old phono pick up would work as a sound level pick up as well, as would many industrial stethoscopes, or as above, a vibration sensor.