[James Lehman]

wrote in message
oups.com...

James Lehman wrote:
If, then, the premise is that the second order distortion
produced by subwoofers is bad because it falls at a
higher frequency that's more easily detected by ears,
than a push-pull arragement, by the same logic, can
be argued as worse, becuase it puts the distortion
products at even higher frequencies.
Use a little imagination here.....

Wiki triangle wave:
...odd harmonics of the fundamental, multiplying every (4n?1)th harmonic
by
?1 (or changing its phase by ?), and rolling off the harmonics by the
inverse square of their relative frequency to the fundamental.

What does that have to do with the nonlinear behavior of
loudspeakers?

Even if you deformed a pure sine wave into a triangle wave; which is a
discontinuous function

Only if the series is carried to an infinite number of terms,
it is.

and has sharp points in the transitions, the first
audible harmonic, the natural third, is only 1/9 th of the fundamental.

That under the rather dubious assumption that the non-
linearities of a speaker approximate those of a triangle
wave. They most assuredly DO NOT.

The point is that if you can eliminate even orders then your first
audible
harmonic is one step farther away from the fundamental and therefore, is
of
much less magnitude.

Based on some rather dubious assumptions, that is.


Would you aggree that natural harmonics tend to diminish in magnitue with
order?

How much second or third can you add to a sine wave before it dominates the
signal and becomes the fundamental?

James. )

[[email protected]]

James Lehman wrote:
wrote in message
oups.com...
The point is that if you can eliminate even orders then your first
audible
harmonic is one step farther away from the fundamental and therefore, is
of
much less magnitude.

Based on some rather dubious assumptions, that is.

Would you aggree that natural harmonics tend to diminish
in magnitue with order?

What one agrees or does not agree is "natural" has nothing
to do with the actual performance of woofers operating non-
linearly. I have seen subwoofers with 25% 3rd harmonic
distortion and only 10% 2nd harmonic distortion. This is
ESPECIALLY true of any symmetrical non-linearity, e.g.,
the type found in push-pull isobarics. It is a feature of
any symmetrical limiting non-linearities, i.e., symmmetrically
exceeding the the linear portion of the motor or the suspension.

Under such conditions, the odd-order products completely
overwhelm the even order products. And depending upon
the order of the non-linearity, there is no assurance that
there is any "natural" diminution of amplitudes with order
as you suggest. Indeed I am sitting here looking at the
FFT analysis of a subwoofer that exhibits a spectrum
(normalized to the fundamental) whose 2nd harmonic
is down 42 dB (about 0.8%), 3rd is -24 dB (6%), 4th is
-39 (1.1%), 5th is -26 (5%) and so on.

But "natural?" Natural has nothing to do with it.

How much second or third can you add to a sine wave
before it dominates the signal and becomes the fundamental?

Well, you want to talk "natural" harmonics, of like, say, a musical
instrument? A 32' principle reed on a baroque style pipe organ
might have all of 10% of it's total energy in the fundamental and
ALL the rest in the harmonics.

To specifically answer your question, exactly 100% harmonic
distortion is needed to "eliminate" the fundamental as you
describe.

[James Lehman]

But "natural?" Natural has nothing to do with it.

What I mean is real objects. After you remove all of the theoretical stuff
like infinite stiffness and zero mass, you are left with real stuff in
dynamic systems.

James. )

[[email protected]]

wrote:
i want to build a box to house (4) 10" subs....should i seperate all
four subs or can i put them in all in one chamber? i will be running
them in mono.
another thought i had is to have them in diamond shape...with opposit
sides facing each other, but would i then have to reverse the polarity
on one the speakers in each pair?
i need some help and or advice

If you have time space and money, then use partitians.
If portability is an issue, I would make separate boxes.
Yes you have to reverse the connections on the opposite in an
isobarik type setup. "diamond".
There is also the bandpass which also reduces orders of
distortion.
Is this for Instruments?

Hello Jb, are you there??

greg

[Phil Allison]

"Arny Krueger"
"James Lehman"

With large cone woofers it is not uncommon that the force that is applied
at the center of the cone causes forward motion that warps the cone for a
brief period of time before that force can propagate out to the edge of
the cone.
A similar, but different distortion of the cone may occur as the cone is
pulled back into the magnet gap. Any distortion of the cone during its
travel will cause some kind of distortion in the sound that it produces.


Not generally true, as there can be either linear or nonlinear
deformations of the cone. Linear deformations don't cause nonlinear
distortion.


** The OP needs to consider how ESL bass panels operate - the edges *never
move* yet the sound generated is remarkably distortion free.

Also, large woofers ( ie 30inch dia ) have been made with NO edge
suspension that work entirely by flexing the cone - again the resulting
sound wave has remarkably low distortion.

The "rigid piston" idea is simply a convenient, mathematical model and
*NOT* the definition of how to generate harmonic free sound waves.

" Linear deformations don't cause nonlinear distortion " is entirely
correct.

Violin and acoustic guitar makers would all be out of business otherwise
!!!





........ Phil

[Mr.T]

wrote in message
oups.com...
Non-ideal materials, in that they do not have infinite stiffness
or propogation velocity, does NOT, a priori, mean that the
resulting flexure and deformation of the cone under stimulus
results in the production of spurious frequencies. This is
ONLY the case if such flexure is non-linear in nature.

Yes, and I readily accept that point. (not that I think linear distortion is
necessarily a good thing either :-)
However the statement I replied to that spurious frequencies *cannot* be
generated just because it's a sub woofer, is still wrong IMO. Fortunately he
has since qualified it to mean only in any sub woofer that does *not*
generate spurious output.

MrT.

[Arny Krueger]

"James Lehman" wrote in message
.. .


Would you aggree that natural harmonics tend to diminish in magnitue with
order?

There's a tendency for that, but there's a stronger tendency for higher
order harmonics to be far more audible.

The operative phenominon is simultaneous masking. Second harmonics tend to
fall under higher ranges of the masking curve.

http://www.umiacs.umd.edu/~desin/Speech1/node10.html

How much second or third can you add to a sine wave before it dominates
the
signal and becomes the fundamental?

Quite a bit, because musical instruments that are common sources of strong
LF signals like pipe organs, generally put out more harmonics than
fundamental in their low ranges. I record a pipe organ quite frequently, and
I observe this result quite consistently, even when using measurement mics
for recording.

[Arny Krueger]

"James Lehman" wrote in message
. ..
Just about any malformation of a sine wave results in harmonics;

Absolutely false as stated.

Prove it.

Simple attenuation causes a change in the form of a sine wave. It makes it
smaller.

[Arny Krueger]

"GregS" wrote in message
...
In article , "Arny Krueger"
wrote:

"James Lehman" wrote in message
m...

"Don Pearce" wrote in message
...
On 29 Jul 2006 01:54:24 -0700, wrote:

But I have to ask - is there any particular reason why you need four
ten inch drivers? In subs, a low resonant frequency is very important,
and you just don't get that from small drivers .....

You can get high excursion drivers with small cones that have very low
resonances.

Within limits. The maximum ratio of excursion to diameter of woofers
tends
to be constant or increase as woofer diameter increases. IOW, smaller
woofers tend to have smaller excursions. This is because maximum excursion
is based on geometry, and as you make something larger, the maximum
excursion tends to incrase accordingly.

I don't believe that there are *any* 4" drivers with 1" Xmax, for example.
But, there are a number of drivers in the 12-18" range that do have 1" or
greater Xmax.

The only problem is that it takes a large surface area to push
enough air at those freqs to produce any appreciable volume.

That's not the only problem, for the reason just given.

That's where using multiple drivers in the same baffle comes in.

There aren't a lot of really good justifications for using multiple
drivers.
One is availability. IOW, if you have 4 10 inchers on hand and can't
easily
liquidate them, you just might want to use them. Another reason would be
a
desire to minimize the depth of the enclosure.


You might also buy 4- 12 or 15 inch drivers that are relatively cheap to
begin with and are on sale. So by using 4 rather than two or one, you can
get more efficiency, more power handling, and less distortion, than the
one
alone.

That's a possibility, but not a given. Furthermore, the multiple drivers
will increase the size and therefore the cost of the enclosure, which is a
major slice of the system cost, particularly if nicely made and finished.

Of course the 4 10 inch isobarik can really be small in size, as is
also the case with the 12's or 15's.

I'm not much of a fan of isobarik designs.

Most of the cheap subwoofer drivers I see are down on Xmax - in the 6-10 mm
range. A single mid-priced driver with 20-25 mm Xmax can cost less than 2-3
of the cheapies.

[James Lehman]

"Arny Krueger" wrote in message
...

"James Lehman" wrote in message
. ..
Just about any malformation of a sine wave results in harmonics;

Absolutely false as stated.

Prove it.

Simple attenuation causes a change in the form of a sine wave. It makes it
smaller.



OK fine. But is that distortion? It most definitely isn't a malformation of
the wave. It's just a difference in the efficiency of the driver at that
frequency.

James. )

[[email protected]]

James Lehman wrote:
"Arny Krueger" wrote in message
...
Just about any malformation of a sine wave results in harmonics;
Absolutely false as stated.
Prove it.
Simple attenuation causes a change in the form of a sine wave. It makes it
smaller.
OK fine. But is that distortion? It most definitely isn't a malformation of
the wave. It's just a difference in the efficiency of the driver at that
frequency.

The problem is you set yourself up formulating the statement
as:

"Just about ANY malformation of a sine
wave results in harmonics."

And, strictly speaking, a linear malformation of the wave, i.e.,
attenuation IS a malformation in the strictest technical
sense of the word, and, being linear in the amplitude domain,
most assuredly does NOT result in harmonics, save one:
the first harmonic.

The problem in the statement is manifold. Once of which
is the strict dependence on "sine waves". Consider
generalizing it:

"Just about any malformation of the waveform
results in harmonics."

If that true? Obviously, it is not. Adding, for example,
an all pass function to any complex waveform will
result in the maformation of its shape AND will result
in NO generation of new harmonics.

Take that same generalization, and state the premise
differently:

"Any malformation of the waveform is distortion."

And such a statement is absolutely correct.

Another problem with your statement and, most especially,
your followup question

"But is that distortion?"

Distortion is the result of non-linear behaviro in one or
more of the domains of the signal. By your statement
and your followup question, you limited, implicitly, the
consideration t ONLY the amplitude domain, whereas
it is perfectly reasonable and, indeed, very often used
in the practice, to describe frequency response errors
as frequency-domain distortion, i.e., a non-linear
frequency response. It's a distortion that results in,
to use your term "malformation" of the waveform, and
a "malformation" that DOES NOT lead to the production
of new harmonics. Similarily, a non-linear group delay
is a non-linear distortion in the time domain that will
result in "malformation" of the waveform, but STILL will
not lead to the production of new harmonics not already
present in the signal.

The problem with your use of "sine wave" as the signal
of choice is that the bandwidth of a sine wave is precisely
0. And applying the classic frequency-time incertainty
principle, whe

dT * dF 1/2

that is, the uncertainty in time dT times the uncertainty in
frequency dF can never be less than 1/2, the fact that the
bandwidth of a sine wave (0, or essentially 0 uncertainty
in frequency) leads to essentially an infinite uncertainty in
time (i.e., it's only a sine wave if it lasts forever) results
in two inevitable properties:

1. Because it's bandwidth is 0, a sine wave is utterly
incapable of demonstrating non-linear behavior
in the frequency domain and,

2. Because its uncertainty in time is in essence infinite,
it is utterly incapable of demonstrating non-linear
behavior in the time domain.

That means, then, if you insist on using a sine wave in your
gedanken, you are forced to deal with the fact that the ONLY
nonlinearity a sinewave is capable of revealing is in the
amplitude domain, and that's what leads to the generation
of harmonics. And you are forced to deal with the fact that
it hides all other non-linear behavior.

And that's why sine waves in the traditional sense are about
the poorest choice to use for evaluating the non-linear behavior
of systems, and why they haven't been used for many years
in doing so.

[Arny Krueger]

"James Lehman" wrote in message
. ..

"Arny Krueger" wrote in message
...

"James Lehman" wrote in message
. ..
Just about any malformation of a sine wave results in harmonics;

Absolutely false as stated.

Prove it.

Simple attenuation causes a change in the form of a sine wave. It makes
it
smaller.


It is a form of linear distortion.

It most definitely isn't a malformation of
OK fine. But is that distortion?


the wave.

Sure it is.

It's just a difference in the efficiency of the driver at that frequency.

just one of many possible explanations.

[James Lehman]

Well this was an active thread!

I just want you all to know that I am currently (not) working on a set of
subs that are the same dual isobaric woofers as in the "ES Twin Iso 8";
except this time they are opposing each other with the cones showing on the
outside of the box. They are 6.4 cu.ft. each and will be tuned to 22.7 Hz
for a -3dB of 19Hz. The max dispersion freq is 23Hz. There are 4 cabinets, 2
per side, for a total of 16 driven 8 inch woofers for the stereo set; 8
woofer cones facing into the room. I have measured the woofers grouped in
iso sets and the figures that I get would indicate that they should be able
to put out a max SPL of about 117dB at 1 meter, per side. In a room, at sub
freqs, 1 meter is irrelevant. These subs are to go behind a large panel line
source array; which is 36 full-range drivers per side, wired together as
one, 72 drivers per stereo pair. The system is, of course, bi-amped with the
Behringer 2 way stereo active crossover.

The reason I say (not) is because I've been on crutches since April 11th! So
all of this, along with 2 other speaker projects, is sitting around my
basement and garage unfinished. I should be getting back to it fairly soon.
I will have pictures.

James. )

[Phil Allison]

"James Lehman"

The max dispersion freq is 23Hz.



** Ok - WTF is this " max dispersion freq " crapology ?


Or did you just make its up like all your other factoids ??






........ Phil

[[email protected]]

James Lehman wrote:
They are 6.4 cu.ft. each and will be tuned to 22.7 Hz
for a -3dB of 19Hz. The max dispersion freq is 23Hz.

What on earth is "dispersion frequency?"

[[email protected]]

James Lehman wrote:
I have measured the woofers grouped in
iso sets

What is an "iso set?"

and the figures that I get would indicate that they should be able
to put out a max SPL of about 117dB at 1 meter,

At what frequency? Since these things are most assuredly
excursion limited, as the vast majority of woofers are at
low frequencies, and since excursion goes essentially
as the inverse square root of frequency, a max SPL with
a frequency limit is pretty meaningless.

[James Lehman]

wrote in message
oups.com...

James Lehman wrote:
I have measured the woofers grouped in
iso sets

What is an "iso set?"

4 woofers combined into a set of 2 isobaric coupled pairs, with no cabinet
behind them.


and the figures that I get would indicate that they should be able
to put out a max SPL of about 117dB at 1 meter,

At what frequency? Since these things are most assuredly
excursion limited, as the vast majority of woofers are at
low frequencies, and since excursion goes essentially
as the inverse square root of frequency, a max SPL with
a frequency limit is pretty meaningless.


Max dispersion freq is where the cone moves the easiest to produces the
greatest volume in the room. At 23Hz, at Xmax, for all the woofers together,
you can expect to use the least power to produce the most sound. It has to
do with the output of the port combined with the output of the cone.

James. )

[Phil Allison]

"James Lehman ****wit "

What is an "iso set?"

4 woofers combined into a set of 2 isobaric coupled pairs, with no cabinet
behind them.


** This pig ignorant ASSHOLE is inventing his own technical terms.

Wot a jerk off.



Max dispersion freq is where the cone moves the easiest to produces the
greatest volume in the room.


** This pig ignorant ASSHOLE is inventing his own technical terms.

Wot a jerk off.


At 23Hz, at Xmax, for all the woofers together,
you can expect to use the least power to produce the most sound. It has to
do with the output of the port combined with the output of the cone.


** Absolute drivel.




....... Phil

[James Lehman]

Crawl back into the slime, you pathetic lump.

[Phil Allison]

"James Lehman ****wit "


What is an "iso set?"

4 woofers combined into a set of 2 isobaric coupled pairs, with no cabinet
behind them.


** This pig ignorant ASSHOLE is inventing his own technical terms.

Wot a jerk off.



Max dispersion freq is where the cone moves the easiest to produces the
greatest volume in the room.


** This pig ignorant ASSHOLE is inventing his own technical terms.

Wot a jerk off.


At 23Hz, at Xmax, for all the woofers together,
you can expect to use the least power to produce the most sound. It has to
do with the output of the port combined with the output of the cone.


** Absolute drivel.




....... Phil

[GregS]

In article , "Phil Allison" wrote:


"James Lehman ****wit "


What is an "iso set?"

4 woofers combined into a set of 2 isobaric coupled pairs, with no cabinet
behind them.


** This pig ignorant ASSHOLE is inventing his own technical terms.

Wot a jerk off.



Max dispersion freq is where the cone moves the easiest to produces the
greatest volume in the room.


** This pig ignorant ASSHOLE is inventing his own technical terms.

Wot a jerk off.


At 23Hz, at Xmax, for all the woofers together,
you can expect to use the least power to produce the most sound. It has to
do with the output of the port combined with the output of the cone.


** Absolute drivel.




....... Phil

After 60 follow-ups the poor scoundral has not showed back up after his
first post, I mean, there is a lot of good info here.

greg

[[email protected]]

James Lehman wrote:
wrote in message
oups.com...

James Lehman wrote:
I have measured the woofers grouped in
iso sets

What is an "iso set?"

4 woofers combined into a set of 2 isobaric coupled pairs,
with no cabinet behind them.

So, why not call them that, instead of inventing some nonsense
term?

and the figures that I get would indicate that they should be able
to put out a max SPL of about 117dB at 1 meter,

At what frequency? Since these things are most assuredly
excursion limited, as the vast majority of woofers are at
low frequencies, and since excursion goes essentially
as the inverse square root of frequency, a max SPL with
a frequency limit is pretty meaningless.

Max dispersion freq is where the cone moves the easiest
to produces the greatest volume in the room. At 23Hz, at
Xmax, for all the woofers together, you can expect to use
the least power to produce the most sound. It has to
do with the output of the port combined with the output of
the cone.

Several points:

First, it is noted how you have failed to answer the question.

Second, the term "dispersion frequency", given your
explanation, is simply nonsense. "Dispersion" in this
realm, already has well understood meanings, and your
misuse of the term is also noted.

In looking through a wide variety of references to the
term, there is not a single instance found in ANY of
them that uses the term "dispersion" in the manner
you do. There are simply none that even come close.
In the context of loudspeakers, the term has two
meanings:

1. The separation of a complex signal into its constituent
frequencies or wavelengths, e.g., dispersion due to
non-uniform group delay or,

2. The distribution of an acoustic signal across an angle
or an area, e.g., the measure of the directivity of an
acoustic source.

Third, there is already a measure for "the least power to
produce the most sound." It's called efficiency, and it is
the ratio of the acoustic power output to the electrical
power input. It has nothing to do with XMax. do you
know what Xmax means and why it is independent of
frequency?

Fourth, Your description of what is happening at the
box tuning frequency is basically at odds with the well-
known and well-understood behavior of vented loud-
speaker systems at the enclosure resonance. Contrary
to your statement:

"It has to do with the output of the port combined
with the output of the cone."

the VAST majority of the total system volume velocity
at Fb (the enclosure resonant frequency) comes from
the port, and very little comes from the cone. To quote
Small:

"Over the frequency range near Fb where the
passive radiator (or vent) contronutes most
usefully to the system output, it does so
through reducing and replacing, rather than
supplementing (as so often implied) the motion
of the driver."
JAES, vol 22, no 11, 1974 Nov.

Examine also the expression for displacement vs
frequency Small presents (Eq. 71 and 73, JAES,
vol 20, no 10, 1973 Oct), which clearly illustrates
Xd(w) showing a minimum at Fb. That also
corresponds to a minimum in the electrical
impedance at the same frequency.

Given these facts, it's thus obvious that two of your
assertions:

1. That the output of the port is combined with the
output of the woofers at Fb to increase the total
output of the system, and

2. The system efficiency, what you call "the least
power to produce the most sound," is maximum
at Fb,

are wrong.