Hmmm... 30ppi is fairly non transparent stuff.
Just looked at the pix from the link.
And, expensive apparently if you want a big block from somewhere...
I was thinking, after reading the tail of the Ariel thread earlier that the HOM has to be a subtractive effect. Due to the conservation of energy idea, and adding in the part about the HOM being "another frequency" other than the fundamental, the HOM + remaining response will be = original energy. So, perhaps there is a way to find the HOM via a subtractive means, IF you knew the pre HOM response... just thinking out loud.
Ok, well, one might be able to look at an FFT of the impulse response with and without the foam, and deem the difference due to HOM?? The assumption is that the foam "kills" the HOM significantly, enough to find the differential. Presumably the rolloff from the foam can be reduced to a function that can be corrected for in the software doing the measurement, not in an EQ of the driver?
_-_-bear
Just looked at the pix from the link.
And, expensive apparently if you want a big block from somewhere...
I was thinking, after reading the tail of the Ariel thread earlier that the HOM has to be a subtractive effect. Due to the conservation of energy idea, and adding in the part about the HOM being "another frequency" other than the fundamental, the HOM + remaining response will be = original energy. So, perhaps there is a way to find the HOM via a subtractive means, IF you knew the pre HOM response... just thinking out loud.
Ok, well, one might be able to look at an FFT of the impulse response with and without the foam, and deem the difference due to HOM?? The assumption is that the foam "kills" the HOM significantly, enough to find the differential. Presumably the rolloff from the foam can be reduced to a function that can be corrected for in the software doing the measurement, not in an EQ of the driver?
_-_-bear
bear said:
Hi Bear
Yes, I'd say that is basically correct, give or take a dB or so.
As to UV degeneration, this is not the old days. My speakers have the same foam in them from about 7 years ago and there is no degradation that can be noted. Now in direct sunlight this may be a different thing, of coarse, but foams today aren't the same materials as in the past. These are polyurethanes which are very stable compounds under all kinds of situations.
wxa666 said:
This lowest mode or the principle mode can only be uniquely defined in a few coordinate systems and it does not have a simple description - thats why no one has responded to this question. But it is very precisely defined in those cases where it can be defined (its done in several works under my name). For virtually all coordinate systems (sperical being a notable exception) the principle mode changes shape as it propagates, but it is virtually always close to spherical for large axial distances. You would not be far off to think of the principle mode in an OS waveguide as being a spherical wavefront at the mouth.
FrankWW said:
Hi Frank
The HOM could be measured in any waveguide which has an analytical solution by placing mics along the walls and solving the pertinent equations at those points. But this would only work for those few waveguides which have a closed form solution, otherwise this isn't a possibility.
Some years back at B&C I did something similar to this for compression drivers on a plane wave tube, which would yield the modes in the exit aperature of the driver. But this does not yield the modes in the waveguide of course.
The same technique can be used in reverse to get the impedance of a waveguide, this was an AES paper of mine back in the late 80's. But this still won't get the modes, only the impedance. There is not a one to one relationship between the two things.
You're probably going to beat me to it. CNC shop specialist that I normally work with went on a 2 week vacation.mige0 said:Thanks - yes this stuff seems to be used as filters in Koi ponds or for mufflers and so on.
Just wasn't on the right track as a search for acoustic foam brings up the stuff for heavy (room) dampening in the first place.
Found out that there can be specs for quite different airvolume put through for one and the same PPI (poures per inch) value - so the 30PPI Earl has speced my not be enough to get *exactly* the right one - possibly have to try.
Am working on implementing Soongsc promising contour currently - hope to get some resuts soon (measurements).
Michael
"are you saying that the typical fine weave metal mesh in a typical compression driver causes HF rolloff similar to that of what looks like about 10" of open cell foam??"
"Yes, I'd say that is basically correct, give or take a dB or so."
Assuming the metal screen is reflective, that means that more than half of the sound emitted is reflected back into the CD.
That can't be good and it's no wonder you recommend removing them.
"Yes, I'd say that is basically correct, give or take a dB or so."
Assuming the metal screen is reflective, that means that more than half of the sound emitted is reflected back into the CD.
That can't be good and it's no wonder you recommend removing them.
I would expect it would be necessary to resolve other performance issues before we actually get to HOM effects. Personally I focus on the issues in the following order when working on the waveguide.
1. Good directivity without complicated intersecting of SPL at various angles.
2. Eliminate driver breakup modes as much as possible.
3. Fast CSD decay.
4. Smooth SPL response.
5. HOM investigation.
1. Good directivity without complicated intersecting of SPL at various angles.
2. Eliminate driver breakup modes as much as possible.
3. Fast CSD decay.
4. Smooth SPL response.
5. HOM investigation.
gedlee said:
What You say is, HOM is undefined in any other waveguides than in those to follow a special coordinate tranformation (OS, straight cone etc). On the other hand in those HOM is expected to be insignificant. With the notorious diffraction horns (CD) HOM can not be measured because it lacks definition. In OS, conical devices it can not be measured because it is avoided by first principles.
What I learn from it is that "HOM" is by no means a wording applicable to an ubiquitous issue as the so called horn sound. Or, more precisely some subtle annoyance with higher levels when listend to for longer periods, hmm? The constructive superiority of OS waveguides due to missing HOM is evident. But this special superiority can not be compared to any other system. These other systems show no HOM either, because its simply not defined. OS features somthing peerless. It is outstanding by definition. Regarding itself it defines - with scientific rigor - the absence of something that for other devices is meaningless.
The answer to the originating question is a loud and clear "Maybe". What can be done is to check an OS whether or not it works as predicted. Any other would be silly.
ZilchLab said:
Hi Zilch,
I'm afraid we are thru. For me it was nothing new in the end. The concept of HOM simply doesn't apply until You buy the highly acclaimed stuff from the OEM. Then You are asked to forget about it.
Til that I would like to experiment with the ol JBL 2352 a bit. I'm curious whether a fin (courtesy of EV) within the throat would straighten out the issues @ ~8kHz that all biradials show to some extent. It is said to be a non axial mode, that makes it.
so long
Dr. Geddes explained the technique of microphones along the walls of a plane wave tube, and presumably this would work as well if you knew the mathematics pertaining to a horn shape that was definable per Dr. Geddes' explanation.
Otoh, the paper cited earlier gave a very specific method of using FEM and a mechanical X/Y scanning microphone device for sampling...
But I suggested a practical comparison of foam/no foam. This may not yield only HOM information, but it will certainly give one back usable data that pertains to the specific horn/wavguide being tested...
It seems that HOMs are not that simple to measure empirically... just as 15 years ago and more, FFT measurements were beyond the ability of almost all hobbyists...
Beyond that, dunno. But based upon what Dr. Geddes has said - which is a reasonable amount - I think I know what sort of experiments I am going to try.
Fwiw, a "fin" in the path of the wavefront of your biradial would almost certainly create more problems than it could solve...
_-_-bear
Otoh, the paper cited earlier gave a very specific method of using FEM and a mechanical X/Y scanning microphone device for sampling...
But I suggested a practical comparison of foam/no foam. This may not yield only HOM information, but it will certainly give one back usable data that pertains to the specific horn/wavguide being tested...
It seems that HOMs are not that simple to measure empirically... just as 15 years ago and more, FFT measurements were beyond the ability of almost all hobbyists...
Beyond that, dunno. But based upon what Dr. Geddes has said - which is a reasonable amount - I think I know what sort of experiments I am going to try.
Fwiw, a "fin" in the path of the wavefront of your biradial would almost certainly create more problems than it could solve...
_-_-bear
gedlee said:
Really?
I think something was terribly wrong. I use the same drivers you do - the DE250 - and I've measured with and without the bug screen. Almost no difference in amplitude response anywhere in the passband, including 10kHz +
Maybe they changed bug screens at some point. The ones I measured were made about a year ago. I have 6 in stock now - I'll do a sweep with LMS and see what current production models look like, with and without the screen.
One thing that puzzles me - If the screen removes as much HF as the foam plug, why is any EQ necessary after the screen is removed and replaced by the foam plug?
soongsc said:
You're probably going to beat me to it. CNC shop specialist that I normally work with went on a 2 week vacation.
I apologise
– but I simply couldn't resist to give this approach a try – as, for me too – it seemed a logically thing to be done (I just didn't have BEM – and possibly not enough time to learn and use it if I had)!Anyway - the outcome of your approach is excellent IMO,.
Allow me to sum up what can be gained of correlating simus and measurements with respect to HOM
In the context of this thread - looking at your sim's / measurements back at
http://www.diyaudio.com/forums/showthread.php?postid=1772631#post1772631
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I would say the HF peak you didn't get captured by your BEM simu is something similar like shown earlier in my "bunch of pretty pictures" (
though Earl possibly wouldn't agree) – keep in mind that my simu is for a very simple modelled cone not for the complex shapes of your contours - the correlation is pretty good nevertheless.
This behaviour possibly applies equally for my measurements shown for the LeCleach contour - just that I have equalised for flat on-axis response, making for a dip at all other angels.
I originally have thought the HF dip developing towards off-axis comes from the cable holes in the contour I didn't seal :
An externally hosted image should be here but it was not working when we last tested it.
In contrary this are my 0deg referenced measurements of your new contour you showed us a week ago at:
http://www.diyaudio.com/forums/showthread.php?postid=1851334#post1851334
which may closely relate to a sound field pattern like this (again, equalising for flat on-axis response, making for a peak here at all other angels):
Funny that CARA seems to capture the behaviour we are interested in, whereas BEM does not.
Michael
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.