Subject: Re: [JN] Unity horn and Reinkus-Heinz Coentrant horn
Date: Fri, 22 Feb 2002 10:16:35 -0600
From: Thomas Danley <Tom@ppci.com>
To: "Le Cleac'h Jean-Michel" <lecleach@paris.ensmp.fr>
CC:"Joenet (E-Mail)" <sound@lists.io.com>

Le Cleac'h Jean-Michel wrote:

> Hello, I cannot remember that any parallel was ever been done between
> the Unity horn and the Reinkus-Heinz "coentrant" horn. Did this was the
> subject of any discussion on Joenet? How do they compare? Best
> regards, Jean-Michel Le Cleac'h, Paris, France

Hi

Sorry for the delay
The RH and Unity are visibly similar and both designs intend to produce
the radiation pattern and acoustic phase of a single source while being
comprised of multiple drivers.
The RH co-entrant design was about the first acoustically arrayable
speaker (while for years trapezoid shaped cabinets were sold as
array-able, they are only in the physical sense, polar plots of multiple
boxes show what really happens).

While I do not feel it is appropriate to discuss the RH design, I can
explain the unity.
First of all, Lambda is selling one of our horns under license as a DIY
kit.
My background is the DIY area so a number of times I have made our stuff
available to the DIY'ers, with the Unity it was through Nick at Lambda.
In fact right now on the live audio board I am leading a bass horn DIY
project (although a tad big for the home)
The Unity idea is simple.

A horn can be thought of as a high pass filter, a high pass filter who's
corner frequency is set by the expansion rate of the horn passage. If
the area doubles every 24 inches or so, that corresponds to about a 30
Hz exponential flare rate, doubling every 2.4 inches ~ = 300 Hz and so
on.
Small variations in the rate such as imposed by a Hyperbolic flare rate
effect the horns acoustic impedance and reactance as one approaches the
low cutoff, often the choice of the right hyperbolic expansion can
significantly extend (downward) the usable portion of the horn / drivers
response by nulling out the drivers reactance at / near the low cutoff.

All curved wall horns have another rarely discussed property, the
dispersion angle changes radically with frequency.
This does several things.
#1 The reverberant sound field in the room is proportional to the
systems acoustic power and room absorption and in the case of the curved
horn, the reverberant field spectrum is very different than the on axis
spectrum (un-desirable).
#2 The other side of the same coin, the sweet spot (the listening area
where the response is flat) gets smaller and smaller with increasing
frequency.

All compression drivers have an acoustic power response which rolls off
starting between about 2-2.5 KHz typically (the Tad 2001 is at about 4
KHz).
On a plane wave tube measurement which reflects the real acoustic power,
this roll off is clearly visible, it is a fact of life.
A horn which has changing directivity, can, by focussing the sound into
a decreasing angle, act as an EQ, flattening the response (on axis).
Many horns exploit this effect to make the acoustic power fall off "go
away" (again, only on axis) and one could figure any curved wall horn
used for HF is doing this.
The Altec 811 / 511 / 311's are examples of horns where the falling
acoustic power was matched pretty closely to the the collapsing
dispersion pattern resulting in a much closer to flat response (on
axis) than the drivers p/o.
In this case, it is the vertical pattern which shrinks most.

Constant directivity horns can be constructed in several different ways
but the goal there is to make the frequency response the same everywhere
in the coverage pattern, in other words the highest frequencies are not
focussed to the center, all frequencies are dispersed equally.
Assuming the horn was built properly, it does what is intended and now
the frequency response anywhere in front looks much more like the plane
wave tube measurement, that is the response represents the actual
acoustic power of the driver. To compensate for the falling acoustic
power above mid range, a "CD" filter in used to flatten the response.
In real life this filter and driver roll off are minimum phase and not
part of a time delay so they are conjugated, in other words the
magnitude and phases of each cancel out leaving a straight line (when
properly done).
Even perfect CD horns require this correction or they sound horrid
(typically measuring a big peak in the 2-2.5 K area)
Also, one finds that CD horns do not go as low as expo variants of the
same size.
This "problem" was one of the clues that lead me to the Unity.

When one examines the expansion rate of a CD or other conic flare, one
finds that at the apex where the driver is connected, the expansion rate
is very rapid (A HIGH FREQUENCY FLARE) so low frequency energy is not
coupled to the mouth.
A few inches further down the horn, one finds a slower expansion rate in
fact the closer to the mouth one looks, the slower the rate becomes.
With the Unity, one finds where the expansion rate is suitable for mid
range, drivers are made suitable for horn loading in that frequency
range, for that size horn and mounted on the outside of the horn.
Small holes pass the sound from the drivers into the horn, the size of
the holes and air volume under the cone forms an acoustic low pass
filter, set just above the crossover which blocks out of band distortion
components from entering the horn and the small hole size minimizes the
acoustic discontinuity in the horn itself.
In the Unity Nick sells, the mid drivers are actually about 3 dB more
efficient than the compression driver and handle 3 times the power so in
operation they are loafing along and have a distortion spectra typically
1/10 to 1/50 or less than any other speaker I have measured at the same
SPL.
The mid drivers are forward of the compression driver, by an amount
which is off set by the phase shift difference between the mid and high
sections, they are automatically "time correct".
Since the high / mid energies combine in a part of the horn where the
dimensions are less than 1/4 wavelength across, the truly add into a
single source.
While the horn Nick sells only has 2 horn loaded ranges, we sell several
products which have a 3 way horn and there is no limit to the number of
stages one could build.

What do they sound like?
Having invented it and developed all of our products, I have heard lots
of different kinds of Unity's.
I can say that when everything is "right" they are the best speaker I
have ever heard, no trace of horn sound, effortless dynamics, a huge
reduction in room interaction due to the directivity AND lack of
interference with multiple ranges and drivers. Imagine a single, really
really clean and dynamic horn that covers the most of the audible range
from a single exit and doesn't sound like a horn, it doesn't sound
"like" anything..
Hope this helps

Feel free to visit our somewhat lame pro sound web site

Servodrive.com

Tom Danley
Servodrive / Sound Physics Labs