Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Art Welter]

Bob,

All right, the 66th post here for 6o6 , and now he’s SuperpositionMan  

Whether called  superposition, wave addition, or mutual coupling,  you are of course correct in that length vs frequency is a continuum, though below 1/4 wave length the continuum does not change much, and above a quarter wave length the changes happen more rapidly.
Not that 1/4 + 1/16th of a wavelength would be a big change, just saying 1/2 wavelength to 1/4 is a bigger change than a 1/4 to 1/8th.

At 16 kHz, where a 1/4 wave length is about the size of the type on the page (+/- 3 dB depending on your eye condition), it is difficult to physically do anything about the differences, we can’t get loudspeakers to dance on the head of a pin.

Down in the <100 Hz range,  wave lengths are long enough that several speakers can be co-located in a near enough configuration to tend to act as a single speaker in the area of interest.  A single EV MTL-4 would be an old example of four 18” drivers with an exit less than 1/4 wavelength in width.

I have not used an FFT analyzer, but using a dB meter or Smaart, I have consistently measured a 6 dB increase doubling cones and power when the sources are within 1/4 wavelength.

That said, large sub arrays are generally well beyond 1/4 wavelength in width, so the “tendency to act as a single speaker” no longer applies.

Art Welter

[Tom Danley]

Hi Bob, Nick

“So much is made of 1/4 wavelength. But as I see it, it is just a milestone in our brains.”

Well it like this I think, we have some definitions which have a strict meaning, for example, a square wave has harmonics which extend from DC to Microwaves. Or, in acoustics a point source is infinitely small or a line source is infinitely large in one dimension.

In real life on the other hand, there are practical definitions too, for example, on an oscilloscope, it takes a bandwidth about 10 X and 1/10 the fundamental to produce a square wave that looks good perfect to the eye.
A line source gradually stops changing acoustically when it is about 40 wavelengths long and  point sources can combine into a single point source of greater efficiency, if they are less than about ¼ wl apart

Trust me, I didn’t make radiation resistance up and neither did John Eargle or the numerous other acoustic references one can find. Mutual coupling, radiator wave number, radiation resistance curves  etc are real.


“But if somebody wants to give me a procedure that I can use on my FFT analyzer to directly observe "tendencies to act like a single speaker" I will be happy to try them out.”

Procedure?
Sure, I guess so, here are a couple.
Take a number of small identical subwoofers and scatter them around your parking lot.
These are small enough to be omni directional point sources in sub range.
Put your mic somewhere “out there” with a red cone so it won’t get squashed.

Do a narrow band energy vs time measurement with all the subs on.
Now gradually assemble them all in one location, measuring the narrow band ETC each time.
What you will see is they gradually look more like and then combine into one radiation in time.  Once you have all the radiations in less than ¼ wl spacing at the highest frequency in question, there is little further change moving them closer.

If you can do this in a large enough reverberant room with pink noise, you can also monitor the sound power by averaging the spl from a number of far locations, what you will see then is the reverberant sound level will climb from the initial level as the woofers begin to have mutual coupling.
This way you can see the effect of the mutual radiation as it raises the total acoustic power relative to the power of the individual sources up to the frequency where the drivers are too far apart..

If you to try and integrate that subwoofer system with a full range speaker above, you will also find your task becomes vastly easier once the individual subwoofers are all less than a quarter wavelength apart at xover.  
I guess those should do.

How this applies is in what Ivan measured, that the close coupled array of four woofers, produced more acoustic power than any of the “array” configurations, even the on axis SPL was less for the directive array, not just the total radiated power.
All that is good when you sell speakers and amplifiers though haha.

When you consider the essentially hemispherical pattern of the close-coupled sources and the power being more or less “area under the graph” it is clear that is so.
There are times when it is desirable to shape the dispersion pattern, the original issue was how best to construct arrays and how do they actually measure.
Best,
Tom Danley

[Anthony Risi]

Hello Tom,

I am a little confused about something here.  I have 4 of the TH-115s, and they are absolutely amazing.  I have always either stacked them on either side of the stage ( mainly because my tops need something to sit on) or have center clustered them.  I have always been amazed at the power increase by clustering them all together.  I understand by reading this post how that is happening, so here is my question.  If I got 4 more TH-115s would it be better to put 4 on each side of the stage, or center cluster a block of 8 of them.  I ask because would a block of 8 of them put the drivers further apart then a 1/4 WL at the highest frequency?  I am no expert on this and didn't know if I was understanding what was being said correctly.  

Thanks
Anthony

[Ivan Beaver]

Tom Danley wrote on Tue, 20 April 2010 16:12

How this applies is in what Ivan measured,

Is is just me or am I the only one who is actually interested in MEASURING something?

I mean it is not that hard-Heck for the point in this discussion you don't even need to do polars-just a simple on axis measurement-a decent distance away from the array-whatever type it is.  The argument is not about what is happening at 180

[Adam Schaible]

Valid point, but from my perspective it's nice to read a lot of the theory, comprehend, etc - so I think it's great.

[Ivan Beaver]

Anthony Risi wrote on Tue, 20 April 2010 17:50

Hello Tom, I am a little confused about something here.  I have 4 of the TH-115s, and they are absolutely amazing.  I have always either stacked them on either side of the stage ( mainly because my tops need something to sit on) or have center clustered them.  I have always been amazed at the power increase by clustering them all together.  I understand by reading this post how that is happening, so here is my question.  If I got 4 more TH-115s would it be better to put 4 on each side of the stage, or center cluster a block of 8 of them.  I ask because would a block of 8 of them put the drivers further apart then a 1/4 WL at the highest frequency?  I am no expert on this and didn't know if I was understanding what was being said correctly.   Thanks Anthony

As usual-it depends.  There are two things at play here-center cluster vs side stacks.  There are advantages and disadvantages to each method and it depends on the particular situation and what works well in one does not work so well in another.

The best way to stack multiple TH115's is to lay 2 of them on their sides with the exits in the middle, and go as high as you can or have cabinets.  In the case of 8, you would go 4 high-but in a center cluster that would be to high for most stages.

If you just stand them up, then 8 would start to form a sideways "line array" and you would start to pull in the coverage on the sides.  This may or may not be a good thing-depending on what area you are trying to cover.

In most cases I would put 4 per side. 2 over 2.

You can help to reduce the power alley by turning down 1 side a little bit.

Everything regarding loudspeakers and how they are deplyed is a compromise.  It is all about finding the best compromise for a particular situation.

[Bob McCarthy]

Shazzam!!!! You mean there's analyzers out there! I always wanted to try one.

Ivan,

I have an opportunity next week during a seminar I am teaching to set up 4 subs in end-fire and monoblock config. It will be indoors in a small room - but since I am not a manufacturer, rental house or holding any research grants, it is the best I can do. My plan is to compare the forward direction SPL of these array configs to that of a single unit. Working title: "The Search for the Lost dB"

Will keep you posted but it will be some time before it leaves post-production.

6

[Ivan Beaver]

Adam Schaible wrote on Tue, 20 April 2010 18:33

Valid point, but from my perspective it's nice to read a lot of the theory, comprehend, etc - so I think it's great.

And the theory is great-but it is also a very good idea to see how the real world stands up to the theory and vice versa.

[Ivan Beaver]

Bob McCarthy wrote on Tue, 20 April 2010 19:53

Shazzam!!!! You mean there's analyzers out there! I always wanted to try one.

That is exactly my point.  I first brought up the point of the directional arrays not having the output of the same number of stacked boxes almost 2 years ago.

And to date NOBDOY (that I am aware of) has tried to repeat even the on axis measurements.

Surely there are people on here who have access to all the needed gear.

Lots of people have disagreed-but none have even taken a measurement to prove their point.  I am not sure what that is saying.

And remember that my measurements were taken outdoor in a free field.  (OK-a parking lot).  I have not done any indoor measurements to compare the different types-so I have no idea what the results would be.  I would suspect the room will have lots to do with the final outcome.

[Bob McCarthy]

Tom Danley wrote:

Quote:

Trust me, I didn’t make radiation resistance up and neither did John Eargle or the numerous other acoustic references one can find. Mutual coupling, radiator wave number, radiation resistance curves etc are real.

Understood that you did not make these terms up and no skepticism as to their having demonstrable effects. Hats off to you for working them into a sentence.

Understood also that as sources get closer together they more closely resemble a single source.  Here is where (I believe) the expression/wording etc. got fuzzy: Way up this thread it was stated that the 4 closely coupled boxes would have the same polar pattern a single speaker- I think it was in your first one of this thread.  This is what I disputed but wait.....

If you meant that it will have the polar of A single speaker. Fine.  
If you meant (and this is what I took it to mean) that the combined will have the same pattern as the 4 individual elements that comprise it - that is where I diverge - and everything I have posted on this thread is prediciated on THAT premise.

Combined to create a NEW and UNIQUE quasi-"single" speaker, yes I am with you, regardless of whether it is the result of resistance, radiators or air conditioners.  the lack of nulls that comes with 1/4 wl spacing makes it look similar to a single speaker.  

The combined "speaker" is narrower than the unit component by SOME amount (as previously stated). Kind of looks like we built some horn steering for the speaker, although it is a funny horn - leaks out the back!

So, yes I completely agree that we can fuse the speaker together, but the displacement, however small will reshape it into a new polar - albeit - a polar recognizable as a combined entity.

Cheers

6o6

ps: If there is "mutual coupling" then there must be "non-mutual coupling" eh?  

[ProSoundWeb Community]