Neither do I.
I can say with a degree of certainty that all the compression drivers I currently have tested exhibit rather gross distortion after reaching around 150 dB in the throat, but have no way to measure the level at the diaphragm, which would be higher by many dB.
As the Wikki entry says:
"The equations that govern nonlinear acoustics are quite complicated and unfortunately they do not have general analytical solutions."
I'm an empiricist, if someone wants to pay for more experimentation and derive general analytical solutions from the data, by all means contact me
.Art
The Wiki article talks about a -60 dB conversion factor, which is just another way of saying the distortion components are 60 dB down or about 0.1% distortion (at a typical drive level). Nonlinearity is always progressive, that is, the percentage goes up as the stimulus level rises. This is also in proportion to distortion order, rising faster for 3rd or 4th order than for second.
Aren't we just talking about square law air compression nolinearity?
David
The thing about nonlinearity in the air is that it happens continuosly. Its not a one pass thing like in an amplifier of a voice coil. The air gets distortorted as it travels and the longer it travels at high SPL the greater the distortion. Only after the wave have been attenuated by spreading does the nonlinearity fall to insignificant levels. If the HF are very directional then this spreading does not occur very rapidly. This is part of the reason that the HF will be much more distorted than the lows, but the other part is that the wavelengths are approaching the molecular distances and things go bad fast as that occurs.
The distortion over a distance sounds plausible. I hadn't thought of that.
I'm not following your comment on high directivity. Once out of the nearfield doesn't a highly directional source spread the same as any other? A 3 degree beamwidth would still quadruple its area for every doubling of distance? Still drop 6dB for every distance doubling?
Are you thinking nearfield?
David S.
I'm not following your comment on high directivity. Once out of the nearfield doesn't a highly directional source spread the same as any other? A 3 degree beamwidth would still quadruple its area for every doubling of distance? Still drop 6dB for every distance doubling?
Are you thinking nearfield?
David S.
Yes, this all goes on in the near field. After that spreading ceases the process.
But this is what's odd about Art's experiment. He has two horns, two drivers and the signal mixing some distance in front of the horns. It isn't under extreme pressure conditions in the throats, its at 100dB or less, some distance out from the horns.
I'm not sure why his mixing products are as loud as they are.
David
I'm not sure why his mixing products are as loud as they are.
David
I read these later posts briefly and if I thought it had any relavence to audio I might have read it more carefully and even commented. But, to me, its all just an academic discussion with no outcome that has any real significance.
Nonlinear acoustics is very interesting, I'll grant you that, I took a one semester course on it back in school, I probably have the notes somewhere (there wasn't even a text at that time). Underwater, where the nonlinearity of the medium is more than 10 times greater than in air, the applications are significant (much of modern sonar use these techniques), but in air the nonlinearity is limited to very HFs or very high SPLs, niether of which is very interesting for audio.
Nonlinear acoustics is very interesting, I'll grant you that, I took a one semester course on it back in school, I probably have the notes somewhere (there wasn't even a text at that time). Underwater, where the nonlinearity of the medium is more than 10 times greater than in air, the applications are significant (much of modern sonar use these techniques), but in air the nonlinearity is limited to very HFs or very high SPLs, niether of which is very interesting for audio.
Last edited:
I guess they are not interesting if you limit your playback levels to low levels with little HF
.For those of us that for various reasons require high SPL at high frequencies, the subject has more than a passing interest.
I stand corrected regarding my previous statement that the SPL at a compression driver’s diaphragm level would be higher than the throat.
The highest level would be at the narrowest point of the throat.
In the driver I measured recently, the screen (as close as the mic can get) is not at the apex of the phase plug, so there still would be some level increase closer to the diaphragm than at what I referred to as the throat.
The throat screen SPL in my 104 dB test would have been very close to 129.7 dB, using the test attached below as a benchmark.
I used a 1.4" exit driver for the test to lessen the effect the microphone displacement in the throat would have, the smaller diaphragm 1" exit drivers at 104 dB may be about 3dB louder at their throats to achieve similar level.
In a shallow horn as in your Avatar, similar, or higher, throat peak SPL would occur at 3 meters even at what some might call "moderate" listening levels.
Beranek's book “Acoustics” describes the nonlinear distortion mechanism:
"If the horn were simply a long cylindrical pipe, the distortion would increase the farther the wave progressed..."
Doubling the distance would double second harmonic distortion in a pipe.
The nonlinear distortion increases linearly with distance traveled in the tube, or the length of the horn.
"in the case of an exponential horn, the amplitude of the fundamental decreases as the wave travels away from the throat, so that the second-harmonic distortion does not increase linearly with distance."
If my (copy of John Meyer’s) interpretation of Beranek is correct, in a long, narrow horn like the Maltese, even though it has a more rapid expansion than an exponential horn of the same length, while SPL is falling at roughly the inverse distance law, the nonlinear (air) distortion increases with distance.
The distance from the horn throat to the microphone in my VHF test was 48.5 inches (26 inches inside the horn, 22.5” from horn mouth to microphone), many times the distance in Pano’s test, allowing much more distance for air-nonlinearity to take place, thus increasing the relative level of the audible IM tone.
Art
Attachments
If air nonlinearity is a factor then we are in different worlds because in a home setting this could never be a factor. Now in a large venue or concert, yes this may be a factor, but there are lots of much more important ones.
I'm sorry, I just get fixated on those things that really matter. Its just a querk of mine.
I have no doubt that the pressure in the throat would be much higher than external to the horn, but didn't you say that the to-be-mixed HF signals came from two separate drivers on two separate horns? This would mean that you might get harmonic distortion of the individual tones (high pressure in the throat), but the IM could only occur in the lower SPL environment outside the horn.
I can't say whether these things matter or not, but I am technically curious and eager to learn.
David S.
I can't say whether these things matter or not, but I am technically curious and eager to learn.
David S.
Oh, beautiful dismissal!
(Just FWIW, some people do need high SPLs. Some people do need substantial HF - E.G. for long throws in dry air. Yeah, that's audio, too.)
What Dave, no blinkers?
It is a very curious fact that the IM generated in a single horn with twice the power used in the stereo horns produces no audible IM component, while the stereo horn set up does.
My explanation in #251 is as good as I can get with my present knowledge, which at this point in time is empirical data from my own tests, and some limited information on the subject gained from the internet and various audio texts and articles read over several decades.
There seems to be no lack of AES papers containing contradictory complicated formulas explaining the phenomenon, but little data backing up how those formulas were derived.
At any rate, I satisfied my initial goal of testing whether beat frequency IM tones from stereo high frequency drivers are audible, as it turns out, they are.
My goal did not include the "why", if anyone reading can provide a better explanation than in #251, I'm all ears
.Art
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.