JUTE on BESSEL: high quality point source and stereo speakers from multiple cheap drivers

[Andre Jute]

JUTE on BESSEL
Copy to a word processor and set to a monospace font to read the tables
right, or go directly to the source:
http://members.lycos.co.uk/fiultra/JUTE%20on%20BESSEL.htm
at JUTE ON AMPS
http://members.lycos.co.uk/fiultra/JUTE%20ON%20AMPS.htm
All text and illustration is Copyright © Andre Jute 1995, 2000
and may not be reproduced except in the thread KISS xxx on
rec.audio.tubes

XXXXXXX

BESSEL ARRAYS
High quality point source or stereo sound
from an array of cheap drivers

by Andre Jute

WHAT IS THE BESSEL ARRAY GOOD FOR?

The interest of the Bessel array is that a bunch of cheap drivers can
be used to give the same illusion as a high quality point source
speaker (like a Quad ESL-63 electrostat or Lowther concentric driver).
Since only a little power, relative to the single-driver case, is
applied to each driver for the same SPL, the voice coil will always
stay in the magnetic gap and the driver will operate in its most linear
range. SPL is made by the number of the drivers, not by running any one
of them out of band.

WHAT IS THE BESSEL ARRAY?

It is possible to simulate a point source by a line array of drivers,
when the power output and polarity of the drivers represents the values
of the mathematical Bessel row. The simulation will be perfect for an
infinite number of drivers in the line array. But even an array of 5
drivers will result in a good approximation of the point source. The
drivers are arranged in an equally spaced line, or in a square, so that
the array may be 5 or 6 drivers, or 25 or 36. The array of 6 is
actually a Bessel array of 7, with the centre position left blank (put
a tweeter in it for coincident sound!).

Though it is possible with some contortions to use a single amplifier
to drive a Bessel array of any size, two or more amps are more commonly
used to drive a Bessel array the easy way.

SMALL BESSEL ARRAYS

Here is the power allocation from the amplifiers for arrays of 5 and 6
(7 including blank) drivers:

Five elements:

A: 1

B: 2

C: 2

D: -2 (reverse polarity)

E: 1

Six (7) elements:

A: 1

B: 2

C: 2

D: 0 (not needed, leave blank space)

E: -2 (reverse polarity)

F: 2

G: -1 (reverse polarity)

The line array can be constructed as a serial or parallel circuit of
drivers. For a Bessel array with drivers of 8 ohm the aggregate
impedance will thus be 28 Ohm or 2.3 ohm. Spacing, though equal, should
be as close as possible.

BIG BESSEL ARRAYS

Bessel array with 25 elements:

1 2 2 -2 1

2 4 4 -4 2

2 4 4 -4 2

-2 -4 -4 4 -2

1 2 2 -2 1

Bessel array with 36 elements including 13 "dummy" drivers:

1 2 2 0 -2 2 -1

2 4 4 0 -4 4 -2

2 4 4 0 -4 4 -2

0 0 0 0 0 0 0

-2 -4 -4 0 4 -4 2

2 4 4 0 -4 4 -2

-1 -2 -2 0 2 -2 1

The impedance of the array will be the same as the impedance of a
single driver. Spacing should be about 1cm between edges.

STEREO WITH A BESSEL ARRAY

It is possible to make stereo sound using 5 drivers in a Bessel array:

A : 0.5 * (L+R)

B : L-R

C : L+R

D : - (L-R)

E : 0.5 * (L+R)

wherein L and R stands for the left and right channel, and the 0.5
indicates that drivers A and E receive only half the power that the
other drivers receive. The easiest implementation uses four power
amplifiers (one inverted) to drive the array.

This one, having so few drivers, could be an expensive solution to the
wall of sound we all dream of: five Quad ESL-63 in a Bessel array
driven by four 300B or 845 or Svetlana SV572-3 or -10 or -30 SE
amplifiers.

By putting the amplifier distribution into a pre-amp, it is possible to
use only two amplifiers, one for drivers A, C, E, and the other for B,D
-- if proper power distribution to the drivers can be arranged.

CAN YOU BUILD ONE?

For a commercial application, get a license from Philips, who hold the
patent. To build one pair or one stereo "speaker" for your own
enjoyment as a hobbyist, go right ahead.

KEY REFERENCE

"Bessel panels - high-power speaker systems with radial sound
distribution" Philips Technical Publication 091, Eindhoven. Released 15
March 1983.

APPLICATION REFERENCES

"Hi-Fi Performance from Small Speakers", by Charles F. Mahler, Jr.,
Audio, December 1959. He described building up 32 x 6 inch speaks
into an array, equivalent area of five 15 inch speaks, -1dB at 10Hz,
bandwidth to 11kHz.

"Dipole Radiator Systems" by R.J. Newman, JAES Vol 28 #1/2, 1980
January/February. He states, pg. 37: " The basic driver of interest
would then appear to be one with a Qt approaching 1 and with a Vas
being as large as reasonably possible." He builds an evaluation system
with drivers having a Qt of 0.85 at 32 Hz.

GENERAL REFERENCE

Sound System Engineering by Don Davis & Carolyn Davis, Focal Press
(Butterworth-Heinemann), Newton, MA. ISBN 0-240-80305-1

THANKS TO

Bodo Kalthoff, Tom Dunker, Dan "Doc Bottlehead" Schmalle, Jean-Michel
Le Cleac'h, Robert Levrault, Kurt Gabitzsch, Dave Dlugos, and Tony
Weimar.

Copyright © Andre Jute 1995, 2000

[Peter Wieck]

Andrew Jute McCoy exuded:

A badly collated rip-off description of "Bessel Arrays".

Why not go to the source and get a clear understanding, clear
explanation, and no blather?

http://www.prosoundweb.com/install/synaudcon/tt25_4/tt25_4_p1.shtml

http://www.angelfire.com/sd/paulkemble/soundf.html

http://ldsg.snippets.org/boxes.php3#ARRAY

And much more without the brain damage, inaccuracies, false conclusions
and false assumptions... why, some of them even refer to actual
PHYSICS!!

Do a service, Mr. McCoy... just point to your sources, don't interpret
them.

Peter Wieck
Wyncote, PA

[Max]

can you give a reference?
I think 1961's "sweet sixteen" introduced the subject, but those drivers
were all connected to equal power, and in phase, so they suffered from
terrible comb filtering.


"Bret Ludwig" > wrote in message
ups.com...
>
> Peter Wieck wrote:
>> Andrew Jute McCoy exuded:
>>
>> A badly collated rip-off description of "Bessel Arrays".
>>
>> Why not go to the source and get a clear understanding, clear
>> explanation, and no blather?
>
>
> Prior Art exists as far back as 1964 because there were published
> designs that purported to emulate the 15" coaxial drivers then de
> rigeur with an array of the oval-cone medium size drivers then common
> in consoles with two way stereo arrays.
>

[Peter Wieck]

Bret Ludwig wrote:

> Prior Art exists as far back as 1964... Rest of crap snipped.

Um... "Bessel Array" is a term-of-art referring to a patent held by
Philips. It is not even the first cousin of multi-driver systems based
on different-sized drivers dedicated to their own piece of the
bandwidth.

Now, if you want to discuss the uncanny superficial resemblance of the
901, that abortion from Bose, to a Bessel Array, that may be open to
interpretation. Which, as it happens, bears an equally uncanny
resemblence to a point-source single driver system inasmuch as it has
neither highs nor lows.

I have actually heard purpose-designed Bessel Arrays in sound
reinforcement applications, each bearing a Davis & Co. Nameplate. From
a considerable distance away, they were quite impressive, a lot from a
little. Too close, and not so much. No danger of "too close" in their
particular habitat, however. Suffice it to say that in most listening
rooms, the term NFW comes to mind. Furthermore, in those applications
where such reinforcement is required, a limited bandwidth is a
reasonable price to pay for the accurate reinforcement of the critical
speech frequencies. I am not stating that such (properly designed)
arrays are necessarily so-restricted, but at the the application I
experienced, voice reinforcement was the primary goal.

Now, some makers have used what they call Bessel Arrays for specific
frequency drivers, mostly tweeters... so the relationship between
necessary distance and speaker size is within most average listening
rooms. What I found surprising is that the particular maker that comes
to mind is known for their extremely high quality monster amps.... so
there would be no lack of power available. I happen to find it
intriguing that they would use the term with specificity, and that they
would even experiment with such a system in a commercial/consumer
product.

But all of this is past Mr. Ludwig's mistaken reference and Mr. McCoy's
fevered, hurried, wildly inaccurate cribbing... OOPS, my most profound
apologies... its "Famous" article from 1995 or whatever.

Peter Wieck
Wyncote, PA

[Ian Iveson]

Peter wrote

> ...I have actually heard purpose-designed Bessel Arrays in sound
> reinforcement applications...

There's a feature on arrays in a current hi-fi comic. Can't remember
which one. Good for churches, is all I can remember, because a
horizontal dispersion at head height can be achieved from vertical
arrays stuck to the pillars.

I had thought the feigning of a point source is a slightly different
matter...but there are so many different ways of using multiple
drivers, in time, space, and frequency distribution. For example:

http://www.sfu.ca/~truax/bourges.html

!

cheers, Ian

[Arny Krueger]

"Max" > wrote in message
news:FPMJg.487751$Mn5.95285@pd7tw3no
> can you give a reference?
> I think 1961's "sweet sixteen" introduced the subject,
> but those drivers were all connected to equal power, and
> in phase, so they suffered from terrible comb filtering.

Agreed.

Unlike the rest of you trolls, I've actually built a Bessel array that is
in use. It is composed of 5 fairly high-quality 4 ohm 6 inch full-range
drivers, along the lines of the parallel array mentioned first in
http://www.angelfire.com/sd/paulkemble/soundf.html . It is currently in use
as a stage monitor.

Doesn't sound half bad with a little eq from the console. No noticable comb
filtering, but not the most efficient stage monitor we have around. It can
take a lot of power without complaining.

[Max]

"Bret Ludwig" > wrote in message
oups.com...
>
> Max wrote:
>> can you give a reference?
>> I think 1961's "sweet sixteen" introduced the subject, but those drivers
>> were all connected to equal power, and in phase, so they suffered from
>> terrible comb filtering.
>>
>>
>
> Unfortunately I was divested of all ny electronics magazines about
> fifteen years ago, but one of the major ones carried a story on it,
> advising the "home recordist" he could "match the performance of the
> Altec" .
>
> It would have been early sixties, very certainly.
>

I've read about a number of systems like this, and not only in American or
British magazines, but European ones also, including several designs which
were published in the Soviet Union, in the 1960s. Some of them are quite
interesting looking, like 8 speakers mounted in a horizontal arc, and put in
the corner of the room, to spread the sound around (mono), or 8 oval
speakers mounted in a circle, in a big bass reflex enclosure.

Incidentally, I have a question about speakers - maybe someone here could
help, or direct me to a place where I could find it - Are there any well
known ways to use a speaker which has a high Q (i.e. 1.2 - 1.7), other than
an open baffle enclosure? Is there some sort of loading I could use, or is
an open baffle the only way to get flat response? Conventional bass reflex,
and also acoustic suspension loading give a big peak when modeled.
Curiously, the famous Diatone P-610 has a Q of 1.2 - 1.5 if the measurements
I've read in other forums are to be believed, and people use them in bass
reflex enclosures, so high Q alone can't be enough to completely damn a
speaker!

The above speaker if built, will be used with a tube amplifier! (trying to
keep it on topic!!)

[Max]

"Arny Krueger" > wrote in message
...
> "Max" > wrote in message
> news:FPMJg.487751$Mn5.95285@pd7tw3no
>> can you give a reference?
>> I think 1961's "sweet sixteen" introduced the subject,
>> but those drivers were all connected to equal power, and
>> in phase, so they suffered from terrible comb filtering.
>
> Agreed.
>
> Unlike the rest of you trolls, I've actually built a Bessel array that is
> in use. It is composed of 5 fairly high-quality 4 ohm 6 inch full-range
> drivers, along the lines of the parallel array mentioned first in
> http://www.angelfire.com/sd/paulkemble/soundf.html . It is currently in
> use as a stage monitor.
>
> Doesn't sound half bad with a little eq from the console. No noticable
> comb filtering, but not the most efficient stage monitor we have around.
> It can take a lot of power without complaining.
>

How does the sensitivity compare to that of a single driver?

[Arny Krueger]

"Max" > wrote in message
news:9s3Kg.491132$Mn5.78568@pd7tw3no
> "Arny Krueger" > wrote in message
> ...
>> "Max" > wrote in message
>> news:FPMJg.487751$Mn5.95285@pd7tw3no
>>> can you give a reference?
>>> I think 1961's "sweet sixteen" introduced the subject,
>>> but those drivers were all connected to equal power, and
>>> in phase, so they suffered from terrible comb filtering.
>>
>> Agreed.
>
>> Unlike the rest of you trolls, I've actually built a
>> Bessel array that is in use. It is composed of 5
>> fairly high-quality 4 ohm 6 inch full-range drivers,
>> along the lines of the parallel array mentioned first in
>> http://www.angelfire.com/sd/paulkemble/soundf.html . It
>> is currently in use as a stage monitor.

>> Doesn't sound half bad with a little eq from the
>> console. No noticable comb filtering, but not the most
>> efficient stage monitor we have around. It can take a
>> lot of power without complaining.

> How does the sensitivity compare to that of a single
> driver?

A little worse because of the reverse-polarity driver.

To summarize:

Efficiency - slightly less because of the reverse-polarity driver, and two
drivers in series.
Maximum acoustic output - about doubled
Directivity - about the same once you get a number of driver diameters away
from the array.

[Andre Jute]

[ADDITIONAL INFORMATION ADDED TO THE NETSItE ARTICLE AT
http://members.lycos.co.uk/fiultra/JUTE%20on%20BESSEL.htm ]]]

SHOULD YOU BUILD A BESSEL ARRAY?

As always with speakers, it depends on a number of variables. A Bessel
Array integrates the sound, that is, works as intended, at a distance
of about ten times the length of the array. This natural limitation of
its form doesn't matter overly much if a Bessel array is used for the
purposes for which it was designed (outdoor sound, huge stadia,
architectural applications, large venues like theatres) but in a
domestic environment it means that Bessel arrays of any substantial
size are suitable only for the largest rooms or garden use.


Adventures in Bessel Arrays

One domestic application I have found for this clever design is
multimedia speakers. At first this seems an outrageous idea but there
is no reason computer speakers should be on the desk with the computer.

Specifically, I used five two-inch Coral drivers (too good for Bessels,
really, but that was what I had to experiment with) to build the array,
no-name surplus sealed "sound modules" for amps and a kustom-kludged
box for a distribution "preamp". This "speaker" was useless under my
Mac because it simply sounded like five separate speakers interfering
with each other; I have close relations of the same drivers in coconuts
(made by a Swiss designer and sent to me for an opinion) and close up
and personal the coconuts won hands down. The tiny Bessel array was
then hung in the angle of the ceiling and the wall at the short end of
the room furthest from my desk and angled to beam down the length of
the room over my computer screen. Suddenly it integrated and became a
speaker, a single unit of sound. The sound was of superior quality but
that is no surprise: the Corals are a known-good quantity and priced
accordingly. This tiny Bessel array with the Coral drivers worked so
pleasingly at well beyond the minimum correct distance that I decided
to see how big a Bessel array would work in the same room.

Next I rigged up a standard commercial public address array, five
patently cheap and nasty oval no-name drivers each about 3x5 inches in
an open-backed box about 30 inches long, as a Bessel array by simply
rewiring and adding amps as required. This proved to be the size limit
of a Bessel array in this particular application where the distance
between the far wall and the "listening" position is 24 feet.

An experimental 6ft6in long stereo Bessel array with five 12in
fullrange guitar drivers was too big for a domestic setting but was
stunningly realistic when I took it along to a record club in a public
hall.

Of course, your house might be bigger than mine, or you may want
speakers to hang on the far side of your pool from the barbeque. I have
proved experimentally to my own satisfaction that ten times the length
of the Bessel array is a good room size.

I had several cartons of the two inch Coral drivers, so next I tried a
5x5 stereo square (not an official Bessel configuration, so don't blame
Mr Bessel and Philips if in your hands it turns out an experiment too
far!). Correctly angled near the ceiling so that it beamed over the
computer screen, and with some further in-computer EQ fiddling to sort
out anomalies created by room irregularities, it created a high-quality
soundstage with width as well as depth; it placed the listener in the
centre of the soundstage, a most impressive ability indeed. The problem
was not the sound but that what started as an experiment in el cheapo
mimicry of a Quad ESL63 was becoming expensive; I anyway have real Quad
ESL63 and Lowther horns. Besides, I'm not big on artificially wide
soundstages; in fact, I am often happy with mono, and find that most
people with active imaginations agree with me. The Stereo Square
Mini-Bessel was therefore broken up and the fine drivers reapplied.

However, with cheaper drivers such as the MTM aluminum-coned 4- and
5-inchers, a Stereo Square Midi Bessel Array would be a possibly
economical and certainly space-saving alternative to, for instance, a
pair of my Impresario speakers or other big floorstanders in even
medium-sized living rooms (from about 17ft for the 4-inchers, 21ft for
the 5-inchers). As an experimenter's econonomy, the flat baffle design
for these drivers which I published in Glass Audio (Volume 9, No. 6,
1997) with my review of SEX amp kit can easily be reconfigured as a
Bessel stereo amp.


So, is a Bessel Array any good for hobbyists?

The answer is yes -- and no. As outdoors party motivators, Bessel
Arrays can bring real quality to your barbeques at a surprisingly
modest price. A tiny Bessel array can make a most superior multimedia
speaker. A modest Bessel stereo square can be stunning in a large room.
But anything bigger than a modest Bessel array is suitable only for use
outdoors or in large venues. The key consideration is always the size
of the space. In addition, Bessel Arrays are intended to use many cheap
drivers to give high quality sound at substantial volume. But the
people who live in rooms of the size required to make Bessel Arrays
work inhabit an income bracket where the saving between real Quad ESLxx
and a Bessel Array faux point source speaker looms much less large than
the pleasure of having the real thing. In short, the Bessel Array isn't
really wanted in (nor was ever intended by its inventor for) any
everyday domestic niche.

We have to conclude that the Bessel Array comes into its own as a
public address speaker of quality, and is a curiosity for audiophile
DIYers.

Copyright © Andre Jute 2006


Andre Jute wrote:
> JUTE on BESSEL
> Copy to a word processor and set to a monospace font to read the tables
> right, or go directly to the source:
> http://members.lycos.co.uk/fiultra/JUTE%20on%20BESSEL.htm
> at JUTE ON AMPS
> http://members.lycos.co.uk/fiultra/JUTE%20ON%20AMPS.htm
> All text and illustration is Copyright © Andre Jute 1995, 2000
> and may not be reproduced except in the thread KISS xxx on
> rec.audio.tubes
>
> XXXXXXX
>
> BESSEL ARRAYS
> High quality point source or stereo sound
> from an array of cheap drivers
>
> by Andre Jute
>
> WHAT IS THE BESSEL ARRAY GOOD FOR?
>
> The interest of the Bessel array is that a bunch of cheap drivers can
> be used to give the same illusion as a high quality point source
> speaker (like a Quad ESL-63 electrostat or Lowther concentric driver).
> Since only a little power, relative to the single-driver case, is
> applied to each driver for the same SPL, the voice coil will always
> stay in the magnetic gap and the driver will operate in its most linear
> range. SPL is made by the number of the drivers, not by running any one
> of them out of band.
>
> WHAT IS THE BESSEL ARRAY?
>
> It is possible to simulate a point source by a line array of drivers,
> when the power output and polarity of the drivers represents the values
> of the mathematical Bessel row. The simulation will be perfect for an
> infinite number of drivers in the line array. But even an array of 5
> drivers will result in a good approximation of the point source. The
> drivers are arranged in an equally spaced line, or in a square, so that
> the array may be 5 or 6 drivers, or 25 or 36. The array of 6 is
> actually a Bessel array of 7, with the centre position left blank (put
> a tweeter in it for coincident sound!).
>
> Though it is possible with some contortions to use a single amplifier
> to drive a Bessel array of any size, two or more amps are more commonly
> used to drive a Bessel array the easy way.
>
> SMALL BESSEL ARRAYS
>
> Here is the power allocation from the amplifiers for arrays of 5 and 6
> (7 including blank) drivers:
>
> Five elements:
>
> A: 1
>
> B: 2
>
> C: 2
>
> D: -2 (reverse polarity)
>
> E: 1
>
> Six (7) elements:
>
> A: 1
>
> B: 2
>
> C: 2
>
> D: 0 (not needed, leave blank space)
>
> E: -2 (reverse polarity)
>
> F: 2
>
> G: -1 (reverse polarity)
>
> The line array can be constructed as a serial or parallel circuit of
> drivers. For a Bessel array with drivers of 8 ohm the aggregate
> impedance will thus be 28 Ohm or 2.3 ohm. Spacing, though equal, should
> be as close as possible.
>
> BIG BESSEL ARRAYS
>
> Bessel array with 25 elements:
>
> 1 2 2 -2 1
>
> 2 4 4 -4 2
>
> 2 4 4 -4 2
>
> -2 -4 -4 4 -2
>
> 1 2 2 -2 1
>
> Bessel array with 36 elements including 13 "dummy" drivers:
>
> 1 2 2 0 -2 2 -1
>
> 2 4 4 0 -4 4 -2
>
> 2 4 4 0 -4 4 -2
>
> 0 0 0 0 0 0 0
>
> -2 -4 -4 0 4 -4 2
>
> 2 4 4 0 -4 4 -2
>
> -1 -2 -2 0 2 -2 1
>
> The impedance of the array will be the same as the impedance of a
> single driver. Spacing should be about 1cm between edges.
>
> STEREO WITH A BESSEL ARRAY
>
> It is possible to make stereo sound using 5 drivers in a Bessel array:
>
> A : 0.5 * (L+R)
>
> B : L-R
>
> C : L+R
>
> D : - (L-R)
>
> E : 0.5 * (L+R)
>
> wherein L and R stands for the left and right channel, and the 0.5
> indicates that drivers A and E receive only half the power that the
> other drivers receive. The easiest implementation uses four power
> amplifiers (one inverted) to drive the array.
>
> This one, having so few drivers, could be an expensive solution to the
> wall of sound we all dream of: five Quad ESL-63 in a Bessel array
> driven by four 300B or 845 or Svetlana SV572-3 or -10 or -30 SE
> amplifiers.
>
> By putting the amplifier distribution into a pre-amp, it is possible to
> use only two amplifiers, one for drivers A, C, E, and the other for B,D
> -- if proper power distribution to the drivers can be arranged.
>
> CAN YOU BUILD ONE?
>
> For a commercial application, get a license from Philips, who hold the
> patent. To build one pair or one stereo "speaker" for your own
> enjoyment as a hobbyist, go right ahead.
>
> KEY REFERENCE
>
> "Bessel panels - high-power speaker systems with radial sound
> distribution" Philips Technical Publication 091, Eindhoven. Released 15
> March 1983.
>
> APPLICATION REFERENCES
>
> "Hi-Fi Performance from Small Speakers", by Charles F. Mahler, Jr.,
> Audio, December 1959. He described building up 32 x 6 inch speaks
> into an array, equivalent area of five 15 inch speaks, -1dB at 10Hz,
> bandwidth to 11kHz.
>
> "Dipole Radiator Systems" by R.J. Newman, JAES Vol 28 #1/2, 1980
> January/February. He states, pg. 37: " The basic driver of interest
> would then appear to be one with a Qt approaching 1 and with a Vas
> being as large as reasonably possible." He builds an evaluation system
> with drivers having a Qt of 0.85 at 32 Hz.
>
> GENERAL REFERENCE
>
> Sound System Engineering by Don Davis & Carolyn Davis, Focal Press
> (Butterworth-Heinemann), Newton, MA. ISBN 0-240-80305-1
>
> THANKS TO
>
> Bodo Kalthoff, Tom Dunker, Dan "Doc Bottlehead" Schmalle, Jean-Michel
> Le Cleac'h, Robert Levrault, Kurt Gabitzsch, Dave Dlugos, and Tony
> Weimar.
>
> Copyright © Andre Jute 1995, 2000

[Chris Hornbeck]

On Fri, 01 Sep 2006 23:33:48 GMT, "Max" > wrote:

>Incidentally, I have a question about speakers - maybe someone here could
>help, or direct me to a place where I could find it - Are there any well
>known ways to use a speaker which has a high Q (i.e. 1.2 - 1.7), other than
>an open baffle enclosure? Is there some sort of loading I could use, or is
>an open baffle the only way to get flat response?

There are both acoustical and electronic approaches. The classic
acoustical approach is to load the backwave into a damped shorted
(open-ended) quarter wavelength line. The resistively lossy mass
of air in the line couples tightly to moving mass of the driver,
reducing Q, F-sub-c (to use the term over-broadly) and sensitivity.

In the bad old days this was a way of getting good sounding bass
that didn't involve horns or open baffles.

An electronic approach is possible for sealed box drivers. A single
inverting op-amp stage can convert any F-sub-c and Q-sub-tc to any
other F-sub-c and Q-sub-tc. See Siegfried Linkwitz's seminal paper
in the very early Speaker Builder for a well worked out example;
(important additions to his earlier AES article; very highest
recommendation; must read).

All good fortune,

Chris Hornbeck
"History consists of truths which in the end turn into lies,
while myth consists of lies which finally turn into truths."
- Jean Cocteau

[maxhifi]

"Chris Hornbeck" > wrote in message
...
> On Fri, 01 Sep 2006 23:33:48 GMT, "Max" > wrote:
>
>>Incidentally, I have a question about speakers - maybe someone here could
>>help, or direct me to a place where I could find it - Are there any well
>>known ways to use a speaker which has a high Q (i.e. 1.2 - 1.7), other
>>than
>>an open baffle enclosure? Is there some sort of loading I could use, or is
>>an open baffle the only way to get flat response?
>
> There are both acoustical and electronic approaches. The classic
> acoustical approach is to load the backwave into a damped shorted
> (open-ended) quarter wavelength line. The resistively lossy mass
> of air in the line couples tightly to moving mass of the driver,
> reducing Q, F-sub-c (to use the term over-broadly) and sensitivity.
>
> In the bad old days this was a way of getting good sounding bass
> that didn't involve horns or open baffles.
>
> An electronic approach is possible for sealed box drivers. A single
> inverting op-amp stage can convert any F-sub-c and Q-sub-tc to any
> other F-sub-c and Q-sub-tc. See Siegfried Linkwitz's seminal paper
> in the very early Speaker Builder for a well worked out example;
> (important additions to his earlier AES article; very highest
> recommendation; must read).
>
> All good fortune,
>
> Chris Hornbeck
> "History consists of truths which in the end turn into lies,
> while myth consists of lies which finally turn into truths."
> - Jean Cocteau

Thanks Chris!

This gives me somewhere to start!

I have done a bit of research, and found out that the classic Diatone P-610
has a Q of 1.5, and is used in a bass reflex enclosure!!! anyone here ever
experience it??

[Andre Jute]

maxhifi wrote:
> "Chris Hornbeck" > wrote in message
> ...
> > On Fri, 01 Sep 2006 23:33:48 GMT, "Max" > wrote:
> >
> >>Incidentally, I have a question about speakers - maybe someone here could
> >>help, or direct me to a place where I could find it - Are there any well
> >>known ways to use a speaker which has a high Q (i.e. 1.2 - 1.7), other
> >>than
> >>an open baffle enclosure? Is there some sort of loading I could use, or is
> >>an open baffle the only way to get flat response?
> >
> > There are both acoustical and electronic approaches. The classic
> > acoustical approach is to load the backwave into a damped shorted
> > (open-ended) quarter wavelength line. The resistively lossy mass
> > of air in the line couples tightly to moving mass of the driver,
> > reducing Q, F-sub-c (to use the term over-broadly) and sensitivity.
> >
> > In the bad old days this was a way of getting good sounding bass
> > that didn't involve horns or open baffles.
> >
> > An electronic approach is possible for sealed box drivers. A single
> > inverting op-amp stage can convert any F-sub-c and Q-sub-tc to any
> > other F-sub-c and Q-sub-tc. See Siegfried Linkwitz's seminal paper
> > in the very early Speaker Builder for a well worked out example;
> > (important additions to his earlier AES article; very highest
> > recommendation; must read).
> >
> > All good fortune,
> >
> > Chris Hornbeck
> > "History consists of truths which in the end turn into lies,
> > while myth consists of lies which finally turn into truths."
> > - Jean Cocteau
>
> Thanks Chris!
>
> This gives me somewhere to start!
>
> I have done a bit of research, and found out that the classic Diatone P-610
> has a Q of 1.5, and is used in a bass reflex enclosure!!! anyone here ever
> experience it??

Look up Gordon Rankin's netsite (he's an American niche market amp
manufacturer called Wavelength IIRC). He used to build speakers with
Diatone drivers and there might still be something on his site. Also
look up Diatone on the Joenet (officially the Sound List) which still
has an archive on a server at Harvard. There was a discussion there of
Diatones in which Grank gave the proper box size and rolloff details,
and Lynn Olson, a super speaker designer, and other knowledgeable
parties were probably also in the discussion.

Koji at EIFL in Japan was the last Diatone dealer. He had some specs
and recommended or possibly even OEM enclosures on his site, I seem to
remember.

Chris has already given you the quarter-wave gospel. The bad news is
that some quarterwave designs can sound a bit screechy compared to say
electrostats; tapered pipes have a poor rep in this regard but
sometimes the case is overstated by people who already have good
speakers experimenting with QWTP, putting Lowther drivers into what is
supposed to be a cheap makeshift, and then being loudly disappointed
out of all proportion to what they have a right to expect from a
speaker with 7inch front baffle, with no internal woodwork -- usually
chosen by them precisely because it is easy to build! My experience is
that when audiophiles who don't already have expensive speakers (Quad,
Lowther, something equally upmarket) hear a QWTP, with say a modest
Fostex driver, they are stunned by the quality. (As an aside, even a
small, inexpensive transmission line, if well designed, can have the
same effect; I heartily recommend Vivian Capel's tiny Kapellmeister to
those who have space only for a bookshelf speaker. Reflex boxes, while
convenient for manufacturers, are really vastly overrated as hi-fi
devices.)

I really don't see the problem with a flat baffle. To me it looks like
a spacesaver! Even a small horn, like my Fidelio-type bicor HWAF
http://members.lycos.co.uk/fiultra/KISS%20194%20T91HWAF3.jpg has a
footprint of only a bit over 11x17 inches, which compares favourably
with more standard type of point source speaker as in my Impresario
http://members.lycos.co.uk/fiultra/KISS%20195%20The%20Impresario.jpg --
a flat baffle, bent or otherwise, will take up exactly the same space.

A baffle, incidentally, needn't be flat to work like a flat baffle. It
can be an open-backed box; the only important point is the distance
from the cone edge to the edge of the panel where the front and rear
wavefronts meet and cause a frequency cancellation, a dropout; the
corner will not matter hugely to the quality of the sound.

Nor need a horn or any dipole speaker have only its front width and
sides for bandwidth projection, deliberate rolloff or dropout to tame
peaks, or whatever parameters concern you. You can easily fix folding
wings to the sides of the speaker, fold them open for serious
listening, close them to conserve space. Thomas Dunker, a Norwegian
audiophile with an English netsite (you'll find him too on the Joenet,
the Sound List), used piano hinges to put wings on the sides of a horn.
IIRC, his horn looked like an antique, dark with age, but that was just
the way he painted or varnished it; it was a new construction.

Incidentally, all houses have baffles built in. They're called walls.
Do you really need so many windows in your listening room? Perhaps you
could build a speaker into the cutout for an unnecessary window. Or cut
a hole in a wall between two rooms and put a driver in it. Of course,
some people will always go too far. One fellow, while his wife was away
staying with her mother, dug up her mature garden, built a 16Hz (and a
few feet more just to be certain!) unfolded tapered horn underground to
exit in the floor of his listening room, and was surprised when he
landed in the divorce court...

One final remark, probably not necessary for you, but of interest to
someone less well read. It is counterproductive to put the driver in
the centre of the baffle. That merely makes the frequency dropout at a
particular frequency 3dB larger. That could be the difference between a
dropout audible only to instruments and one that disturbs the ear.
Instead divide the baffle up into an odd number of equal parts and put
the driver or drivers off-centre with reference to each driver's
frequency range. That creates two dropouts or a longer, flatter
dropout, per driver, but at least the dropout will not be so obvious. A
scheme that found favour when I published it in Glass Audio in 1997 was
a baffle division in 11 parts with (two similar fullrange) drivers 4
and 7 units from the edges. I later built one with four drivers on that
pattern, with each driver rolling off differently in two directions,
which had a particularly smooth sound; I made that one with hinged
wings, cut away at an angle at the bottom to lean the speaker back when
open so that some of the drivers are nearer the back wall, which also
helped to smooth the sound out and, of course, by opening the wings
more or less, more or less lean could be used to aim the focal point of
the sound at the listening position. Details like these matter in the
musical experience quite out of proportion to the extra effort involved
in thought and construction.

HTH.

Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

[maxhifi]

Hi Andre!

Thanks for the informative post! All of it is excellent info, however one
passage is of particular interest to me. My (perhaps misguided) reason to
shy away from open baffles, is quite simply, to extend low frequency
response below the 100Hz resonant frequency of some specific drivers. I
could always use my subwoofer with them though, if need be. (or, maybe not -
maybe I won't miss it too badly for a while)

"A
scheme that found favour when I published it in Glass Audio in 1997 was
a baffle division in 11 parts with (two similar fullrange) drivers 4
and 7 units from the edges. I later built one with four drivers on that
pattern, with each driver rolling off differently in two directions,
which had a particularly smooth sound; I made that one with hinged
wings, cut away at an angle at the bottom to lean the speaker back when
open so that some of the drivers are nearer the back wall, which also
helped to smooth the sound out and, of course, by opening the wings
more or less, more or less lean could be used to aim the focal point of
the sound at the listening position."

Can you give a few more details about the 4 driver arrangement?? I have 10
160cm x 100cm full range drivers coming to me, and I was thinking about
using them (4 per side, two spares) in some sort of ported line arrays, but
your idea sounds more interesting. The true Qts of the drivers is as yet
unknown to me, but I expect it to be in the 1.2 - 1.6 range.

I can tolerate a few dB of upper bass lift - it might help compensate for
the big drop under 100Hz, but I don't want something overly boomy - so 6dB
boost would be out of the question. I need to think of a way to employ these
speakers to good effect, because they are very sweet sounding, with good
treble and decent dispersion characteristics. I really like the hinged idea,
so that they will not take a lot of storage space, and be relatively easy to
transport. Also, this will be easier to construct, and my free time is
fairly limited these days, so something like a horn would end up like
certain other half-completed projects, collecting dust for a living in my
garage ;)

[Peter Wieck]

Andrew Jute McCoy exuded:

>More incomprehensible crap misinterpreted and cribbed outright from others.

One day, Mr. McCoy will actually attempt to follow its own advice. It
may live through the experience.

Peter Wieck
Wyncote, PA

[Andre Jute]

maxhifi > wrote:

> Hi Andre!
>
> Thanks for the informative post! All of it is excellent info, however one
> passage is of particular interest to me. My (perhaps misguided) reason to
> shy away from open baffles, is quite simply, to extend low frequency
> response below the 100Hz resonant frequency of some specific drivers. I
> could always use my subwoofer with them though, if need be. (or, maybe not -
> maybe I won't miss it too badly for a while)
>
> "A
> scheme that found favour when I published it in Glass Audio in 1997 was
> a baffle division in 11 parts with (two similar fullrange) drivers 4
> and 7 units from the edges. I later built one with four drivers on that
> pattern, with each driver rolling off differently in two directions,
> which had a particularly smooth sound; I made that one with hinged
> wings, cut away at an angle at the bottom to lean the speaker back when
> open so that some of the drivers are nearer the back wall, which also
> helped to smooth the sound out and, of course, by opening the wings
> more or less, more or less lean could be used to aim the focal point of
> the sound at the listening position."
>
> Can you give a few more details about the 4 driver arrangement?? I have 10
> 160cm x 100cm full range drivers coming to me, and I was thinking about
> using them (4 per side, two spares) in some sort of ported line arrays, but
> your idea sounds more interesting. The true Qts of the drivers is as yet
> unknown to me, but I expect it to be in the 1.2 - 1.6 range.
>
> I can tolerate a few dB of upper bass lift - it might help compensate for
> the big drop under 100Hz, but I don't want something overly boomy - so 6dB
> boost would be out of the question. I need to think of a way to employ these
> speakers to good effect, because they are very sweet sounding, with good
> treble and decent dispersion characteristics. I really like the hinged idea,
> so that they will not take a lot of storage space, and be relatively easy to
> transport. Also, this will be easier to construct, and my free time is
> fairly limited these days, so something like a horn would end up like
> certain other half-completed projects, collecting dust for a living in my
> garage ;)

Yo, Max:

First of all, you got the idea of multiple drivers on the baffle by the
tail end. You're thinking like an engineer, who asks, How can I get the
most volume out of these drivers? A music lover asks, What should I
sacrifice in terms of efficiency to get the best sound? Before the
music lover can answer that, he must ask another question, What is the
best sound? It isn't the most bass, or the widest bandwidth. It is the
smoothest midrange and the cleanest bass and reasonable high frequency
extension.

There will be no bass bump such as you fear; you will roll off the
presence frequencies and the HF by the placement of drivers on the
baffle. Each doubling of the number of drivers gives you somewhat less
than 3dB of volume extra which, if it is clean, carries the overtones
so that your ear constructs the bass fundamental from the harmonics.
However, on a baffle, unless it is very large, you will almost always
get *a little* less bass extension on a meter than in one of the other
methods Chris offered you; the upside is that clean bass to experienced
listeners sounds like more than it really is in the same way that to
another class of listener the huge LF distortions of boomboxes sounds
like "bass".

The idea is further to use the wall and the floor and the ceiling as
extensions to your baffle and its wings in the bass, and to use the
varying distances between each driver and the wall consequent upon
angling of the baffle in relation to the wall to smooth the sound out
even more so that there is no perceptible lift anywhere.

The drivers I worked with were cheap ali-cone jobs, simply because I
had them. Fs = 115Hz, so you're starting with superior drivers to mine.
With four drivers up, I managed to make the baffle sound pretty good
next to a Lowther horn I used for a reference. It couldn't match the
very clean Quad ESL63 down to 40Hz but my rule of thumb is that if a
speaker that would retail for less than two grand makes a clean 60 or
70Hz sound, you're doing pretty good; yeah, I know what the ponyracers
claim but they're lying.

HF is never a problem with baffles of any size but you must get the
midrange right before you do anything else.

The procedure is to study the response plot of your driver and decide
where you want to lop off peaks, fill in troughs. Calculate the
wavelengths for those frequencies. What you're making here is very
smooth midrange. Remember that the sound coming off the wall behind
your dipole adds to the front projection. At this stage you can
calculate the angle of lean towards the wall (the distance to the wall
is simple trig, diagonal as the square root of the sum of the squares
on the square sides) or preferably the corner as another smoothing
tool: clearly you must adjust placement on the baffle with lean towards
the wall to end up with the best compromise. I found it useful to make
a proto that was round so I could in effect roll it in front of the
wall at an angle and study the effects but it was too big to be
practical as an everyday speak. Also, if you have four speakers on the
baffle, you can arrange them so that each one has a different edge
rolloff in two directions, eight different rolloffs altogether, so that
everything overlaps. Notice that the distance between speakers create
more opportunities for manipulation, and also that your oval speakers
radiate differently across the long axis to the small axis. If you're
desperate to manipulate a troublesome frequency or to manipulate it by
only a very small bit, or to smooth off a peak gradually, there is no
law that says the top edge of your baffle must be straight or parallel
to the bottom, or that the baffle must be whole where it isn't pierced
by speakers.

Now you can attend to the bass. Never mind, even in the planning stage,
how big it will be -- with four drivers it will be much, much better
than with one. What you want is to integrate the bass and to smooth it
out. Consider that the wings can hinge forward as well as backwards, so
that they can join the wall wherever you want to behind or before the
plane of the drivers; a 20mm cutaway on the outside of each wing at the
bottom to the height of the skirting board is good to make the join
reasonably smooth. I have never bothered with tapered edges, rubber
fills and all that anal retentive stuff; a small discontinuity might
matter in the treble but in the bass you really have to leave a gap as
wide as a door to hear the difference, and more often than not the
difference caused by the discontinuity is pleasing. The smoothing
process in the bass consists of building the main body of the speak and
the wings as big as your space allows, then trying varying wing and
wall angling arrangements in situ.

You need have no fear that the bass will be lumpy. The problem is the
other way, that you need to put at least two of your drivers pretty
near the centre (or towards the bottom if it will stand flush with the
floor and be angled) of the baffle vertically to stop rolloff at the
top edge of the baffle (the lean helps but not perfectly), and any
drivers you put nearer the edges to smooth out frequency peaks will add
less to the bass. I imagine you know that I like a reasonable amount of
clean smooth bass a lot better than lots of one-note boom.

A few notes that may be more obvious to you than to me when I started
out. Off the bottom of a dipole standing flush with the floor, the
wavelength is all the way to your listening chair. It may be smart to
leave one side of the baffle, usually the side along the back wall,
open i.e. not touching the wall, especially if you're leaning the
baffle backwards; you don't want to enclose the air behind a dipole
because it will start working unpredictably. The angle between the
walls and the ceiling, especially in corners, is a good place to put
speakers; baffles work everywhere. Build an Excel spreadsheet to
calculate all distances and frequencies; it saves a lot of bother
later.

One final point. It makes constructing, wiring and tuning the speaks
in-room easier if you put all the drivers on the main panel, but if
you're good on a computer or with mental geometry (or if you don't mind
doing the woodwork twice), you can put drivers on the wings as well, or
you can make a four-panel baffle with drivers on the centre two panels,
which makes aiming them a little simpler. I built a four-panel
three-hinge job a long time ago (well, actually, I bought a thick ply
hinged divider screen at a shopfitter's sale and sawed holes in it for
drivers) and it ended up being permanently folded into an open-back box
with a shallow W profile and a heavy piece of marble that was the base
for another speaker sitting on top closing the top end and weighing it
down in place. The guy who owns it now has a thick stack of oversize
black roof slates on it...

Good luck.

Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

[maxhifi]

Hi Andre,

I have considered putting the speakers up by the ceiling, in either corner
of the short end of a room which is about twice as long as it is wide, and
angling them down, because I already have a speaker system system set up in
this room (actually, two - floor standers, and also, a full 5.1 setup!) I
was always under the impression that open baffle speakers needed a lot of
space around them, so I hadn't considered it would be possible to put them
too close to the back or side walls, let alone touch them! What thickness of
plywood have you used to make your baffles?? I was thinking of using maybe
5/8, with a couple coats of shellac to protect it.

Have you ever tried this piece of software? http://www.tolvan.com/edge/ I
have been fooling around with it for a while, and seems quite useful for
planning an open baffle. I also found an excel spreadsheet someone prepared,
which I don't fully understand how to use (it only mentions width, but not
height of the baffle for its calculations)

I never really thought about using the above program to find a response
which mirrors that of the driver, but it seems an excellent idea! Response
graphs are not available for the drivers I am using, and all I've got on
hand is a Radio Shack Spl meter (although, if I need to, I could borrow a
B&K).... it would make more work, but I could take a baffle outside and
measure the driver.

My one concern here, is the effect of comb filtering produced by having the
drivers located on the same baffle, and not vertically in line - I can't see
how such a system will be immune to it, but it seems to not subjectively
matter in other speakers which have been produced.

I'll let you know how it works out

Max


"Andre Jute" > wrote in message
ups.com...
> maxhifi > wrote:
>
>> Hi Andre!
>>
>> Thanks for the informative post! All of it is excellent info, however one
>> passage is of particular interest to me. My (perhaps misguided) reason to
>> shy away from open baffles, is quite simply, to extend low frequency
>> response below the 100Hz resonant frequency of some specific drivers. I
>> could always use my subwoofer with them though, if need be. (or, maybe
>> not -
>> maybe I won't miss it too badly for a while)
>>
>> "A
>> scheme that found favour when I published it in Glass Audio in 1997 was
>> a baffle division in 11 parts with (two similar fullrange) drivers 4
>> and 7 units from the edges. I later built one with four drivers on that
>> pattern, with each driver rolling off differently in two directions,
>> which had a particularly smooth sound; I made that one with hinged
>> wings, cut away at an angle at the bottom to lean the speaker back when
>> open so that some of the drivers are nearer the back wall, which also
>> helped to smooth the sound out and, of course, by opening the wings
>> more or less, more or less lean could be used to aim the focal point of
>> the sound at the listening position."
>>
>> Can you give a few more details about the 4 driver arrangement?? I have
>> 10
>> 160cm x 100cm full range drivers coming to me, and I was thinking about
>> using them (4 per side, two spares) in some sort of ported line arrays,
>> but
>> your idea sounds more interesting. The true Qts of the drivers is as yet
>> unknown to me, but I expect it to be in the 1.2 - 1.6 range.
>>
>> I can tolerate a few dB of upper bass lift - it might help compensate for
>> the big drop under 100Hz, but I don't want something overly boomy - so
>> 6dB
>> boost would be out of the question. I need to think of a way to employ
>> these
>> speakers to good effect, because they are very sweet sounding, with good
>> treble and decent dispersion characteristics. I really like the hinged
>> idea,
>> so that they will not take a lot of storage space, and be relatively easy
>> to
>> transport. Also, this will be easier to construct, and my free time is
>> fairly limited these days, so something like a horn would end up like
>> certain other half-completed projects, collecting dust for a living in my
>> garage ;)
>
> Yo, Max:
>
> First of all, you got the idea of multiple drivers on the baffle by the
> tail end. You're thinking like an engineer, who asks, How can I get the
> most volume out of these drivers? A music lover asks, What should I
> sacrifice in terms of efficiency to get the best sound? Before the
> music lover can answer that, he must ask another question, What is the
> best sound? It isn't the most bass, or the widest bandwidth. It is the
> smoothest midrange and the cleanest bass and reasonable high frequency
> extension.
>
> There will be no bass bump such as you fear; you will roll off the
> presence frequencies and the HF by the placement of drivers on the
> baffle. Each doubling of the number of drivers gives you somewhat less
> than 3dB of volume extra which, if it is clean, carries the overtones
> so that your ear constructs the bass fundamental from the harmonics.
> However, on a baffle, unless it is very large, you will almost always
> get *a little* less bass extension on a meter than in one of the other
> methods Chris offered you; the upside is that clean bass to experienced
> listeners sounds like more than it really is in the same way that to
> another class of listener the huge LF distortions of boomboxes sounds
> like "bass".
>
> The idea is further to use the wall and the floor and the ceiling as
> extensions to your baffle and its wings in the bass, and to use the
> varying distances between each driver and the wall consequent upon
> angling of the baffle in relation to the wall to smooth the sound out
> even more so that there is no perceptible lift anywhere.
>
> The drivers I worked with were cheap ali-cone jobs, simply because I
> had them. Fs = 115Hz, so you're starting with superior drivers to mine.
> With four drivers up, I managed to make the baffle sound pretty good
> next to a Lowther horn I used for a reference. It couldn't match the
> very clean Quad ESL63 down to 40Hz but my rule of thumb is that if a
> speaker that would retail for less than two grand makes a clean 60 or
> 70Hz sound, you're doing pretty good; yeah, I know what the ponyracers
> claim but they're lying.
>
> HF is never a problem with baffles of any size but you must get the
> midrange right before you do anything else.
>
> The procedure is to study the response plot of your driver and decide
> where you want to lop off peaks, fill in troughs. Calculate the
> wavelengths for those frequencies. What you're making here is very
> smooth midrange. Remember that the sound coming off the wall behind
> your dipole adds to the front projection. At this stage you can
> calculate the angle of lean towards the wall (the distance to the wall
> is simple trig, diagonal as the square root of the sum of the squares
> on the square sides) or preferably the corner as another smoothing
> tool: clearly you must adjust placement on the baffle with lean towards
> the wall to end up with the best compromise. I found it useful to make
> a proto that was round so I could in effect roll it in front of the
> wall at an angle and study the effects but it was too big to be
> practical as an everyday speak. Also, if you have four speakers on the
> baffle, you can arrange them so that each one has a different edge
> rolloff in two directions, eight different rolloffs altogether, so that
> everything overlaps. Notice that the distance between speakers create
> more opportunities for manipulation, and also that your oval speakers
> radiate differently across the long axis to the small axis. If you're
> desperate to manipulate a troublesome frequency or to manipulate it by
> only a very small bit, or to smooth off a peak gradually, there is no
> law that says the top edge of your baffle must be straight or parallel
> to the bottom, or that the baffle must be whole where it isn't pierced
> by speakers.
>
> Now you can attend to the bass. Never mind, even in the planning stage,
> how big it will be -- with four drivers it will be much, much better
> than with one. What you want is to integrate the bass and to smooth it
> out. Consider that the wings can hinge forward as well as backwards, so
> that they can join the wall wherever you want to behind or before the
> plane of the drivers; a 20mm cutaway on the outside of each wing at the
> bottom to the height of the skirting board is good to make the join
> reasonably smooth. I have never bothered with tapered edges, rubber
> fills and all that anal retentive stuff; a small discontinuity might
> matter in the treble but in the bass you really have to leave a gap as
> wide as a door to hear the difference, and more often than not the
> difference caused by the discontinuity is pleasing. The smoothing
> process in the bass consists of building the main body of the speak and
> the wings as big as your space allows, then trying varying wing and
> wall angling arrangements in situ.
>
> You need have no fear that the bass will be lumpy. The problem is the
> other way, that you need to put at least two of your drivers pretty
> near the centre (or towards the bottom if it will stand flush with the
> floor and be angled) of the baffle vertically to stop rolloff at the
> top edge of the baffle (the lean helps but not perfectly), and any
> drivers you put nearer the edges to smooth out frequency peaks will add
> less to the bass. I imagine you know that I like a reasonable amount of
> clean smooth bass a lot better than lots of one-note boom.
>
> A few notes that may be more obvious to you than to me when I started
> out. Off the bottom of a dipole standing flush with the floor, the
> wavelength is all the way to your listening chair. It may be smart to
> leave one side of the baffle, usually the side along the back wall,
> open i.e. not touching the wall, especially if you're leaning the
> baffle backwards; you don't want to enclose the air behind a dipole
> because it will start working unpredictably. The angle between the
> walls and the ceiling, especially in corners, is a good place to put
> speakers; baffles work everywhere. Build an Excel spreadsheet to
> calculate all distances and frequencies; it saves a lot of bother
> later.
>
> One final point. It makes constructing, wiring and tuning the speaks
> in-room easier if you put all the drivers on the main panel, but if
> you're good on a computer or with mental geometry (or if you don't mind
> doing the woodwork twice), you can put drivers on the wings as well, or
> you can make a four-panel baffle with drivers on the centre two panels,
> which makes aiming them a little simpler. I built a four-panel
> three-hinge job a long time ago (well, actually, I bought a thick ply
> hinged divider screen at a shopfitter's sale and sawed holes in it for
> drivers) and it ended up being permanently folded into an open-back box
> with a shallow W profile and a heavy piece of marble that was the base
> for another speaker sitting on top closing the top end and weighing it
> down in place. The guy who owns it now has a thick stack of oversize
> black roof slates on it...
>
> Good luck.
>
> Andre Jute
> Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
> "wonderfully well written and reasoned information
> for the tube audio constructor"
> John Broskie TubeCAD & GlassWare
> "an unbelievably comprehensive web site
> containing vital gems of wisdom"
> Stuart Perry Hi-Fi News & Record Review
>