Well, how small, and what effect would the microphone actually have on a single impulse WRT altering the soundfield exiting or even inside a plane wave tube, as long as the "microphone" (let's call it a sensor, since we only need to detect the time of arrival of the impulse, so it might be purpose built and very tiny as a result) is a very small fraction of the wavefront's dimension under test??
Intuitively it would seem that with a single pulse traveling down the plane wave tube that merely detecting its time of arrival would be sufficient. I am suggesting that whatever ever effect the sensor has upon the wave, so long as the sensor is physically small the change due to turbulence as the pulse wave meets and passes by is not terribly relevant and will be undetected as well.
Is it correct or incorrect to assume that the wave as it travels down the plane wave tube is essentially unaltered in terms of its wavefront shape?
What am I missing?
_-_-bear
Intuitively it would seem that with a single pulse traveling down the plane wave tube that merely detecting its time of arrival would be sufficient. I am suggesting that whatever ever effect the sensor has upon the wave, so long as the sensor is physically small the change due to turbulence as the pulse wave meets and passes by is not terribly relevant and will be undetected as well.
Is it correct or incorrect to assume that the wave as it travels down the plane wave tube is essentially unaltered in terms of its wavefront shape?
What am I missing?
_-_-bear
bear said:
That is incorrect.
It is also incorrect to assume that the wavefront shape is frequency independent. It will be different at every frequency and it will continue to change shape as it propagates down the tube.
Ultra minuature microphone do exist based on micromachined hot wire principles. As I said early on this is what Behler used at Achen. Its the scanning aperatus that gets complex and costly.
bear said:
I did this about 6 years ago for a horn (duct, really) bend. It worked well. There's info on my website. However, I was doing it at low frequencies. You could make the mechanism to move the mic fairly cheaply - maybe $200 for two small stepper motors, the drivers, and a power supply. The real cost is your time if you haven't done this before and already researched where to get cheap components (speaking from experience, as I just finished building a CNC router...) Add a bit more for some type of linear bearings and framework. I did not do this for my setup - I moved things by hand, but for the low frequencies I was measuring at, the precision I achieved was acceptable to me. The main problem I see in using a PWT is getting the mic in there and moving it around in some sort of practical way.
Also, to clarify, I was using sine waves, not impulses. The behavior of the system definitely changed with frequency.
is <1/5 of 20KHz wavelength small enough?
http://www.knowles.com/search/family.do?family_id=FG/BFG&x_sub_cat_id=8
MEMS microphones can be thinner in one dimension if that helps
http://www.knowles.com/search/family.do?family_id=FG/BFG&x_sub_cat_id=8
MEMS microphones can be thinner in one dimension if that helps
the means & practical wavefronts
Well, "pointless" depends quite a bit on who is speaking!
If I had your Phd. and could solve the problem mathematically, or could hire you or your equivalent to solve the problem then it might be pointless. Otherwise the point becomes more clearly focused.
If you have a means even to do a presumably crude approximation via a means practical to one with less than your obvious expertise, I'm all ears.
However, this then starts to become another discussion based upon your last statement about the wavefront's shape varying with frequency. Based upon some posts in other threads there was some discussion about adusting the exit aperature angle and the way that the phase plug was made so as to make the exit wavefront planar rather than spherical. What you just said makes it unclear that is an optainable goal across a reasonably wide frequency range?
And, are we then saying that the shape of a presumed spherical wavefront produced by a "high quality - well engineered" compression driver changes its apparent radius with frequency? I am presuming it becomes a smaller effective radius as the frequency goes up? Or?
_-_-bear
John, I will take a look at ur site, tnx...
Well, "pointless" depends quite a bit on who is speaking!
If I had your Phd. and could solve the problem mathematically, or could hire you or your equivalent to solve the problem then it might be pointless. Otherwise the point becomes more clearly focused.
If you have a means even to do a presumably crude approximation via a means practical to one with less than your obvious expertise, I'm all ears.
However, this then starts to become another discussion based upon your last statement about the wavefront's shape varying with frequency. Based upon some posts in other threads there was some discussion about adusting the exit aperature angle and the way that the phase plug was made so as to make the exit wavefront planar rather than spherical. What you just said makes it unclear that is an optainable goal across a reasonably wide frequency range?
And, are we then saying that the shape of a presumed spherical wavefront produced by a "high quality - well engineered" compression driver changes its apparent radius with frequency? I am presuming it becomes a smaller effective radius as the frequency goes up? Or?
_-_-bear
John, I will take a look at ur site, tnx...
Bear
Any of the methods that you proposed would take more time and cost far more than what I proposed.
The wavefront in the exit aperature of a Compression Driver is designed to be planar - flat not spherical. It will deviate from flat in complex ways and with varying magnitudes across the frequency spectrum. (To what extent this happens is really unknown since no one has actually done a thorough study.) You also have to consider what is meant by flat or planar. The wave can have constant phase across its extent and not constant amplitude and this is "not flat". It can also have constant amplitude but varying phase and this is "not flat" either. Phase plugs based on Smith's design will tend to be constant phase but not constant amplitude.
Any of the methods that you proposed would take more time and cost far more than what I proposed.
The wavefront in the exit aperature of a Compression Driver is designed to be planar - flat not spherical. It will deviate from flat in complex ways and with varying magnitudes across the frequency spectrum. (To what extent this happens is really unknown since no one has actually done a thorough study.) You also have to consider what is meant by flat or planar. The wave can have constant phase across its extent and not constant amplitude and this is "not flat". It can also have constant amplitude but varying phase and this is "not flat" either. Phase plugs based on Smith's design will tend to be constant phase but not constant amplitude.
Excellent!
I feel my brain expanding with good information.
Dr. Geddes, you made reference to the shape of the wave changing as it travels down the planewave tube?
I am wondering on the specifics of that change. Intuitively I would expect the wavefront to "stick" to the sides a bit and create some turbulence as it travels... so does the wave merely bend toward a bulged out spherical shape, or is there another way it can be described?
I think I recall Berenek mentions that the change of angle from an expansion, to a parallel wall (and back out to an expansion) creates (trying to recall now...) well let's say a negative effect? So, I am wondering if this applies to the effect of a plane wave tube placed on the end of the expansion that is usually the throat section of a compression driver as well?? (perhaps it is so small an angle as to have minimal effect?)
_-_-bear
(wishing he took a whole lot more math and acoustics in college now...)
I feel my brain expanding with good information.
Dr. Geddes, you made reference to the shape of the wave changing as it travels down the planewave tube?
I am wondering on the specifics of that change. Intuitively I would expect the wavefront to "stick" to the sides a bit and create some turbulence as it travels... so does the wave merely bend toward a bulged out spherical shape, or is there another way it can be described?
I think I recall Berenek mentions that the change of angle from an expansion, to a parallel wall (and back out to an expansion) creates (trying to recall now...) well let's say a negative effect? So, I am wondering if this applies to the effect of a plane wave tube placed on the end of the expansion that is usually the throat section of a compression driver as well?? (perhaps it is so small an angle as to have minimal effect?)
_-_-bear
(wishing he took a whole lot more math and acoustics in college now...)
bear said:Excellent!
I feel my brain expanding with good information.
Dr. Geddes, you made reference to the shape of the wave changing as it travels down the planewave tube?
I am wondering on the specifics of that change. Intuitively I would expect the wavefront to "stick" to the sides a bit and create some turbulence as it travels... so does the wave merely bend toward a bulged out spherical shape, or is there another way it can be described?
I think I recall Berenek mentions that the change of angle from an expansion, to a parallel wall (and back out to an expansion) creates (trying to recall now...) well let's say a negative effect? So, I am wondering if this applies to the effect of a plane wave tube placed on the end of the expansion that is usually the throat section of a compression driver as well?? (perhaps it is so small an angle as to have minimal effect?)
A perfect plane wave would travel down the frictionless tube unchanged. But since the wavefront form a real compression driver is not perfect, the imperfections will change shape as it travels - those bouncing off of the walls etc. and adding in-phase and out-of-phase at different places as the wave travels. The straight section does force a wavefront change as it enters the plane wave tube, but the caculations take the wavefront back to the point where it just enters the tube so the tube has not had any effect as yet. This is key and why you just can't measure the wavefront down stream, you must calcualte it back to the very end of the driver for it to be correct.
The point about the math being hard seems to me to be irrelavent, because I have already done it. What I have not done is to build a tube to take the data to use in the math. If someone builds the tube and gets me the data I would have no problem running that data and giving back the results. What I will not do is to supply the math calculations to be run elsewhere as that is proprietary and it would make no sense for me to just give it away.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.