U.S. patent number 3,922,488 [Application Number 05/425,527] was granted by the patent office on 1975-11-25 for feedback-cancelling electro-acoustic transducer apparatus.
This patent grant is currently assigned to A.R.D. Anstalt. Invention is credited to Saad Zaghloul Mohamed Gabr.
United States Patent |
3,922,488 |
Gabr |
November 25, 1975 |
Feedback-cancelling electro-acoustic transducer apparatus
Abstract
Electro acoustic transducer apparatus having a loudspeaker and
one or two microphones includes positional and/or electric
arrangements for reducing or cancelling feedback from the
loudspeaker. Also, a microphone apparatus has positional and
structural features for reducing or cancelling echo.
Inventors: |
Gabr; Saad Zaghloul Mohamed
(Canterbury, EN) |
Assignee: |
A.R.D. Anstalt (Vaduz,
FL)
|
Family
ID: |
26236423 |
Appl.
No.: |
05/425,527 |
Filed: |
December 17, 1973 |
Foreign Application Priority Data
|
|
|
|
|
Dec 15, 1972 [UK] |
|
|
58174/72 |
Jan 9, 1973 [UK] |
|
|
1071/73 |
|
Current U.S.
Class: |
381/93;
379/106.01; 379/388.07 |
Current CPC
Class: |
H04R
3/02 (20130101) |
Current International
Class: |
H04R
3/02 (20060101); H04M 001/60 () |
Field of
Search: |
;179/1HF,1FS,81B |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Brown; Thomas W.
Attorney, Agent or Firm: Young & Thompson
Claims
I claim:
1. An electro-acoustic transducer apparatus comprising a
loudspeaker and microphone means having two closely spaced
transducers, electronic circuit means, said circuit means including
an additive circuit element and a substractive circuit element,
each of said elements receiving output from both of said
transducers, said elements providing two outputs one of which is
wanted sound plus noise and the other of which is only noise, said
circuit means comprising also phase adjustment means to receive
said other output and to produce a noise component 180.degree. out
of phase with said other output, said circuit means comprising also
means to combine said one output with said out-of-phase component
to cancel noise from said one output.
2. Apparatus as claimed in claim 1, and a housing mounting said
loudspeaker and microphone means together.
3. Apparatus as claimed in claim 1, and an amplifier for the
microphone means and an amplifier for the loudspeaker, and
automatic gain control means for controlling the gain of the
amplifiers.
Description
The invention relates to duplex or two-way communication systems,
particularly loudspeaking systems.
All such known systems suffer from the effects of acoustic
feedback, that is, from the effects of sound from the loudspeaker
reaching the microphone, and from the effects of acoustic echoes
and ambient noise. The intelligibility of communication is
diminished by both effects.
Acoustic feedback, that is the undesired transmission of sound from
the loudspeaker to the microphone, whereby the electrical output
signal of the microphone includes an unwanted component
corresponding to and caused by the loudspeaker sound reproduction,
takes place mainly because of the transmission of loudspeaker sound
reproduction to the microphone through one or more of three
channels. One channel is the airborn transmission of sound from the
loudspeaker directly to the microphone diaphragm. The second
channel comprises loudspeaker sound reproduction reaching the
microphone through mechanical vibration of means or media other
than air, for example the loudspeaker and microphone housing means
and associated fixed structures. The third channel is the
transmission of loudspeaker sound reproduction to the microphone by
vibration of the microphone housing due to airborn sound
(microphony).
Echo and ambient noise reach the microphone diaphragm through the
air and by microphony.
It is the object of the invention to eliminate or at least reduce
both acoustic feedback and acoustic echo and ambient noise to a
point such that a satisfactory signal to noise ratio is obtained
whereby the stable sound levels and the intelligibility required
for a duplex, two-way, loudspeaking telephone conversation is
achieved without the need for voice operated switching.
In accordance with the invention, this objective is achieved in
three ways, acoustically, electrically, and electroacoustically. In
all cases the acoustic measures provide for the basic high signal
to noise ratio needed. The microphone means is designed and
positioned in combination with the loudspeaker so as to
discriminate between the loudspeaker sound reproduction and the
wanted speech or other sound input desired for transmission, which
represents information to be communicated. They provide for the
prevention by cancellation or at least the reduction of the
undesired acoustic feedback noise and acoustic echoes, whereby the
electrical output of the apparatus acquires the required signal to
noise ratio which allows satisfactory stable loudspeaking duplex
communication. Further improvement in the overall signal to noise
ratio is achieved if necessary by using non-linear amplifiers and
components and/or automatic gain control operated by and dependent
on the noise level, or arrangements giving similar effects. The
acoustic measures adopted in accordance with the invention consist
in arranging the microphone or microphone means in such a position
with relation to the loudspeaker or loudspeakers that sound from
the latter falls equally on two sides of the diaphragm of the
microphone means. This can be readily effected as far as directly
transmitted sound as well as acoustic echo are concerned, but is
less readily controlled in respect of reflected sound, due to
objects not forming a part of the equipment, the position of which
is variable or unknown. The acoustic measures of the invention
likewise provides constructions for the cabinet or housing means,
including sub-housings for the individual transducers, which
effectively reduce or eliminate microphony, and sound transmission
through media other than air. The effects of sound from
uncontrollable local sources, and of loudspeaker sound reflected
from likewise uncontrollable objects, cannot be effectively dealt
with in either of these ways. To deal with such sources of unwanted
sound components in the microphone output, the invention provides
electrical circuit arrangements, to reduce the effects. The
electro-acoustic meansures of the invention comprise electrical
circuit arrangements in combination with complex electro-acoustic
transducer assemblies. The function of the combination of circuitry
and transducers is firstly to discriminate between the electrical
output signals of the microphone caused by the loudspeaker sound
reproduction and the speech or other signal desired for
transmission. Secondly, the combination functions to deal with each
and to arrange for the cancellation or reduction of the undesired
signals caused by and corresponding to the loudspeaker sound
reproduction and echoes so that the final electrical output signal
of the combined microphone means and electrical circuits acquires
the high, satisfactory ratio of signal to noise, to allow the
stable reproduction of the required sound levels from a duplex
loudspeaking communication or other sound reproduction system.
In addition to the acoustic measures, the cancellation of the
undesired noise can be achieved by simple electrical circuits to
allow for the addition and subtraction of the outputs of the
microphone means, or by complex arrangements for summing the
signals resulting from the addition and subtraction and also by
arranging for phase and amplitude adjustment if necessary to
achieve the required cancellation of noise. It is of course within
the invention to apply one, more than one, or all of these measures
as is deemed appropriate in any particular piece of equipment.
The invention will be more readily understood from the following
illustrative description and accompanying drawings. drawings, in
which:
FIGS. 1A, 1B, 1C and 1D are somewhat schematic showings of various
suitable arrangements of microphones and loudspeakers relative to
each other;
FIGS. 2A-2K are arrangements similar to FIGS. 1A-1D, but in which
the microphones receive sound directly from the loudspeaker;
FIG. 3 is a circuit diagram of apparatus according to the present
invention;
FIGS. 4 and 5 are views similar to FIG. 3 but showing other
embodiments of the invention;
FIGS. 6-11 are views of suitable microphone arrangements for use in
the present invention;
FIG. 12 is a view of a microphone diaphragm and FIG. 13 is a view
of a microphone baffle for use in the present invention; FIG. 14 is
a cross-sectional view of a modified housing for use with the
present invention;
FIGS. 15A and 15B are fragmentary cross-sectional views of tubular
embodiments of housing;
FIG. 16 is a fragmentary cross-sectional view of a diaphragm
mounting; and
FIGS. 17-21 are vector diagrams of signal and noise components
involved in the present invention.
According to a first aspect of the invention illustrated
diagrammatically in FIGS. 1A, 1B, 1C and 1D, a microphone M and
loudspeaker L are so relatively located that both sides of the
microphone diaphragm receive sound directly from the loudspeaker.
The resultant effect of this incident sound will be ideally zero,
so that there is no electrical output from the microphone due to
this sound. The microphone and loudspeaker must be relatively close
together, preferably they should be in close proximity within a
housing means which advantageously comprises sub-housings, one for
each transducer, suitably connected together in a manner or by
means whereby sound transmission between the sub-housings is
eliminated or nearly eliminated. The housing means are shown
schematically in or omitted from the drawings. A variety of
relative positions are possible, of which some only are shown in
the accompanying drawings. In particular, the microphone can be
arranged with its own diaphragm in any plane containing the
loudspeaker axis. The arrangement of FIG. 1A is found to be
particularly advantageous, it is thought because the sound wave
front radiated by the loudspeaker is not broken up or impeded by
the rim of the microphone and associated housing. The rim and
housing should preferably in any event be generally small and thin
in particular to afford better cancellation. Alternatively, the
microphone and loudspeaker diaphragms can be co-planar and in
adjacency. The arrangement must be such that the loudspeaker sound
reaches the two sides of the microphone diaphragm directly or
substantially directly, that is without reflection or undue
obstruction. In particular, the housing means should not cause
undue reflected sound from the loudspeaker to fall on either side
of the diaphragm. One or both sides of the microphone and/or the
loudspeaker can however be covered with a conventional grill or
fabric.
It will be understood that these configurations of loudspeaker and
microphone can be used in any position; these do not in themselves
do anything to counteract acoustic feedback due to reflected sound
from the loudspeaker, and it is accordingly an advantage to have
the loudspeaker diaphragm substantially horizontal to provide for
acoustic echo reduction or cancellation. At least the majority of
any reflected loudspeaker sound which reaches the microphone is
then due to reflection from the ceiling of the room in which the
apparatus is used. The distance of the ceiling will normally result
in considerable atenuation of this reflected sound, which is
moreover a constant factor in any one room, which facilitates
compensation.
Whilst the arrangements shown comprise only a single microphone and
a single loudspeaker, it will of course be appreciated that the
actual tranducer units employed as loudspeaker and microphone can
be duplicated or used in any desired numbers. Thus a loudspeaker
may have one or more additional microphones M' positioned similarly
with respect to the loudspeaker as the microphone M. For example
four microphones can be located in planes containing the
loudspeaker diaphragm axis and angularly spaced at right angles
around the axis. The microphones could thus receive speech from
speakers grouped around the apparatus as at a conference.
It is also within the invention to arrange a pair of microphones in
a position with relation to the loudspeaker such that each
microphone diaphragm receives on one side thereof equal or
substantially equal sound signals directly from the loudspeaker.
Some possible arrangements are diagrammatically shown in FIGS. 2A -
2K. If the arrangement is such that the signals received are in
phase, the outputs of the two microphones must be reversed, so that
the electric signal components due to the loudspeaker sound cancel,
and are thus absent from the common microphone output. If the
arrangement is such that the sound signals are out of phase, the
microphone outputs are simply added to provide the common
electrical output. It will be evident that the or each microphone
of any of the arrangements of FIG. 1 can be replaced by a pair of
microphones arranged with their diaphragms parallel and spaced some
distance apart, and either separated acoustically by a baffle, or
in substantially direct acoustic coupling to realize the present
aspect of the invention. It is however also possible to arrange the
two microphones in other ways. As with the arrangements of FIGS. 1A
- 1D, it is of course readily possible to add one or more
additional pairs of microphones, each pair being appropriately
positioned relative to the loudspeaker. Again, more than one
loudspeaker can be employed, provided the correct relative
positions are preserved.
In order to ensure that the loudspeaker sound signals reaching the
microphones are as nearly as possible identical in all respects,
the distance between the axes of the microphone diaphragms should
be kept small compared to the acoustic wavelength of the signals.
Preferably, the greatest distance involved should not exceed 1/4 of
the wavelength of the highest frequency to be handled by the
apparatus, which may be regarded as 3,000 Hz where only speech is
being transmitted.
It is within the invention to arrange for the operation of the
loudspeaker or loudspeakers and the microphone or microphones
alternately or on a time division basis for example by direct
switching or by switching the energisation of associated
amplifiers. The frequency of alternation should be high enough not
to cause interference with the reproduction. With this method, the
distance between the loudspeaker and microphone diaphragms must be
kept small to avoid an undesirable time lag in transmission.
Otherwise, it is advantageous for this distance to be relatively
large, to allow the wave front of loudspeaker sound to be in phase
across the microphone diaphragms.
It will be understood that the arrangements of FIGS. 2A - 2K will
include housing means preferably of the kind described with
reference to FIGS. 1A - 1D. The measures so far described provide a
substantial improvement over known arrangements, which have dealt
only with acoustic feedback due to transmitted sounds reaching the
microphone diaphragm or diaphragms via the air directly, that is,
direct air-born sounds from the loudspeaker.
In any of the two-microphone arrangements of the invention, the
microphones can be constituted by the so-called button or midget
microphones, the inlet nozzles or tubular apertures of which will
as regards position replace the diaphragms illustrated. The use of
these button microphones provides the closest possible spacing
between the microphone sound input positions so that the phase
difference between the microphone outputs is minimised.
FIG. 3 shows an acustic feedback-free transducer unit F providing a
high signal to noise ratio and including two microphones M. Within
a phase adjustment unit U incorporating electric circuit
arrangements for further improving the signal to noise ratio the
outputs of the two microphones M are applied both to an additive
circuit element A and to a subtractive circuit element S, for
example a transformer or a differential amplifier. The additive and
subtractive outputs are then taken through phase and amplitude
control means P and summed either simply or repeatedly, preferably
in a further amplifier O, to provide the electrical output signal
to be transmitted. The treatment of the microphone outputs is
described in detail below with reference to FIGS. 17 - 21. The
circuitry can include the circuit elements necessary to effect
appropriate phase and/or amplitude adjustment. Each of the various
circuit elements is preferably adjustable and may be arranged
either for presetting or for control from outside the unit. The
unit F includes a loudspeaker L, the relative positioning of the
microphones M and the loudspeaker shown diagrammatically being of
course only one of various possible arrangements.
It will be appreciated that this aspect of the invention consists
in isolating an electric signal produced by unwanted sound falling
by any channel on a microphone or microphones, and subtracting this
electric signal from a microphone output containing both an
unwanted sound component and also a wanted sound component. The
output of the unit U and the input to the loudspeaker L and its
associated amplifier are connected through a hybrid H if required,
to a line and thereby to like apparatus at a remote station. A
conventional automatic gain control circuit G, e.g. in the form of
two conventional units, may be included as shown to control a
loudspeaker amplifier K and the output amplifier O to improve
efficiency, but this can often be omitted if desired.
Two other circuit arrangements for the electrical cancellation of
feedback are illustrated respectively in FIGS. 4 and 5. These can
also include automatic gain control arrangements as described with
reference to FIG. 3 if desired. In each arrangement, an acoustic
feedback free transducer unit F of any desired kind comprises
housing means mounting loudspeaker means represented by a single
loudspeaker L and microphone means represented by a pair of
microphones M.
Referring now to FIG. 4 the loudspeaker L has two speech coils 12,
14. The microphone outputs and the loudspeaker reproduction coil 12
are connected through respective amplifiers 15, 16 to a hybrid 18
and thus to a line terminal 19. The hybrid 18 enables the station
to be connected to a two-wire transmission system but may be
omitted if a three or four line transmission system is
employed.
In FIG. 4, part of the output of the transmission amplifier 15 is
taken on a line 20 through an adjustable resistor 21 to the coil
14, the transmission signal cancellation coil, of the loudspeaker.
The line 20 also extends through a phase adjuster 22 which may
instead be in either of the alternative positions indicated at 22'
and 22". When there is a sound input to the microphone, part of the
resultant signal reaches each of the coils 12, 14 through the
hybrid or through coupling between the lines caused by the other
station or stations. In order to remove this a part of the signal
is fed through the line 20 and the circuit components 21 and 22
such that these signals are equal and 180.degree. out of phase.
In FIG. 5, an amplifier 24, preferably of variable gain, is
included in the line 20 together with an adjustable phase shifter
25 which can instead be placed at the position indicated by 25'.
The operation of the circuit is essentially the same as that of
FIG. 4 except that the amplifier 24 is provided to ensure that the
signal applied to the coil 14 is of adequate strength. Also, the
loudspeaker 11 has only a single speech coil 26 connected with the
secondary winding 28 of a transformer having primary windings 29,
30 performing the functions of the coils 12, 14 of the loudspeaker
of FIG. 4.
It will be evident that the circuits of FIGS. 4 and 5, operating
between the hybrid and the transducer unit, can be employed in
loudspeaking telephones and can be adapted to systems using two
pairs of wires for transmission as for intercommunication or public
address systems. Although cancellation takes place at the
loudspeaker in both circuits, cancellation in a like manner at the
microphone, additionally or instead, can be provided. Moreover, the
circuit arrangement of FIG. 3 can be employed in combination with
the arrangements of FIGS. 4 and 5 whenever the transducer unit F
includes two microphones.
Whilst the feedback free transducer unit of the invention can
include loudspeakers and microphones of any suitable construction,
microphones of preferred construction are shown in FIG. 6 and
7.
Referring to FIG. 6, an electro-acoustic transducer for use as a
microphone or loudspeaker consists of a diaphragm 51, a speech coil
53 and a frame 54 for rigidly supporting a magnet 52. The frame 54
exposes a major area of the adjacent side of the diaphragm. The
diaphragm 1 therefore may be exposed to sound waves on both sides.
Such a microphone has bi-directional characteristics and therefore
is particularly useful for acoustic suppression of the feedback.
Ideally, the microphone should have symmetrical bi-directional
characteristics and to this end the planar diaphragm 51 is used as
distinct from the conventional conical diaphragm. As the magnet,
coil and frame structure are on one side only of the diaphragm, the
microphone is mounted, in accordance with the invention, in a
housing providing dummy, elements (not shown) on the other side the
dummy elements being of at least approximately the same shape and
location. The apparatys comprisisng the microphone and the housing
thus has the desired symmetry. The diaphragm 51 can if required be
corrugated whilst retaining its generally planar form.
The operative sensitivity curves for such a transducer depends on
the mechanical suspension or electrical sensitivity control of the
diaphragm. As the former cannot be readily adjusted, it is
convenient to provide for electrical sensitivity control by biasing
to obtain the required performance. The microphone can thus be
provided with biasing coils in addition to the normal speech
coil.
An electro-acoustic transducer apparatus which is itself physically
and electrically symmetrical will avoid the need to provide dummy
elements in its housing, and is shown in FIG. 7. This transducer
corresponds in its general form on one side of the diaphragm to the
transducer of FIG. 6. However the speech coil, magnet and frame
structure is repeated symmetrically on the other side, the parts on
this side being marked with primed reference numerals corresponding
to those on the one side. A biasing coil can be provided to adjust
the suspension of the diaphragm, as previously described.
FIG. 8 shows a microphone unit having a pair of like
electro-acoustic transducers M mounted one on either side of a
baffle board 66. The transducers can be of conventional moving coil
construction, with the axes aligned.
The conical diaphragms of the transducers face one another and are
spaced apart at their edges only by the edge mounting boards 70
(which may be omitted) and the thickness of the board 66. The
transducers are each mounted in an open ended loading chamber or
baffle box 72 having one wall constituted by the board 66 and
adjacent four sides 74 at right angles to the board and surrounding
the transducer. Slits 76 can be provided in the walls 74 to ensure
a communication between the interior of the box and outside. The
walls 74 should not extend so far axially of the transducers as to
impose directional characteristics on these, unless such
characteristics are desired. The mounting arrangements thus
comprise a housing providing for each transducer something
approaching an infinite baffle.
Circuit arrangements associated with the transducers can comprise
means for adjusting the phase and/or amplitude of the output of the
transducers as previously explained. As shown, a potentiometer is
connected across the coil of each, the tapping contacts of the
potentiometers providing the output of the unit across terminals
75.
In use, the microphone unit will be effectively echo-cancelling
because reflected sound will in general fall in substantially equal
intensities on the two transducers and will consequently cancel out
in the electrical output signal from the unit. Desired sound, for
example, from a speaker holding the unit so that he directly faces
one of the transducers will be fully reproduced in the electrical
output, as also will speech from a second speaker opposite the
first. The potentiometers permit adjustment of the unit to
compensate for departures thereof from exact symmetry, electrical
or structural, about the central plane of the baffle board 66.
In the embodiment of the invention, shown in FIG. 9, one or more
features of which may be interchanged with those of the unit shown
in FIG. 8 as appropriate, two like transducers M are again
employed. These transducers have substantially planar diaphragms
are previously described. The transducers M are mounted coaxially
on a baffle board 76 which is apertured either over substantially
the whole area of the diaphragms, or over part of this area. In
this arrngement, the microphone orientation can be reversed, to
expose the full area of the diaphragms to incoming sound. The
transducers M are connected into appropriate circuitry, for
example, as shown in FIG. 8 or as previously explained with
reference to FIG. 3.
In use the microphone unit of FIG. 9 as so far described will be
seen to have the same echo-cancelling properties as that of FIG. 8.
The acoustic coupling of the two transducers however causes both to
respond to an incoming desired sound along the common axis from one
side of the unit, so that the output and thus the ratio of this to
any uncancelled noise is substantially doubled.
The microphone unit of FIG. 9 additionally incorporates two further
electro-acoustic transducers M' for use as loudspeakers, so that
the unit becomes a device for two-way communications, usable for
example as a conventional or loudspeaking telephone. The
transducers M' are mounted so as to preserve the symmetry of the
device and with their axes at right angles or nearly at right
angles to the common axis of the transducer M. Such an arrangement
provides for acoustic cancellation of feedback without loss of the
echo-cancelling properties of the microphone unit comprising the
transducers M.
When a baffle is provided between the two diaphragms, the other
diaphragm of course remains unaffected by this sound and the output
from the diaphragm receiving this wanted signal is substantially
unchanged by the sound travelling parallel to the diaphragm. If
there is no baffle between the two diaphragms, the second
diaphragm, which does not directly receive the incoming wanted
sound, moves in sympathy with the first mentioned diaphragm because
of the vibrations of the first mentioned diaphragm conveyed through
the air between them.
The microphone unit shown in FIGS. 10 and 11 has a single
transducer having a generally planar diaphragm 80 and a pair of
speech coils, one on either side, co-operating with separate
magnets. It is however to be understood that a pair of microphones
can be used instead, and that appropriate features of the units of
FIGS. 8 and 9 will then be included in the present unit. In the
present embodiment, apertures, which may be selectively adjustable
in effective area, are provided between the loading chambers
associated with the respective sides of the transducer.
As shown in FIG. 10, one form of such an aperture comprises an
annular aperture 34 around the transducer and a baffle 35, which
functions as a by-pass or short circuiting aperture. Thus the
diaphragm suspension can be clamped at its outer edge between a
pair of rings 86 from each of which radial arms 88 extend to
support one of the magnets. Two bolts or other fasteners extend
outwardly from the rings 86 at diametrically opposed points and are
the sole means of connection with the baffle 85, which has an
aperture of greater diameter than the rings. The resulting annular
aperture 84 effects a degree of equalization of sound vibrations on
opposed sides of the baffle. The bolts can extend through rubber or
plastics grommets 91 so that the transducer is effectively isolated
from sound vibrations transmitted through the housing.
Where the baffle is a square as in the unit of FIGS. 10 and 11 or
rectangular, an aperture 92 in the baffle can be provided at each
corner to minimise the effect of sound reflections at the corners
between the walls of the loading chambers extending from the baffle
85.
Instead of, or as well as, providing such apertures in the baffle,
it is within the invention to provide them in the actual transducer
diaphragm or diaphragms.
The effective cross-sectional area of all the various kinds of
apertures described is preferable made selectively variable, and
even reducible to zero, by any suitable form of closure or shutter
to permit either pre-set adjustment or adjustment in use or both.
As shown in FIG. 12, apertures can be arranged for example as a
series of holes 100 along a radius or a diameter of a diaphragm 101
and their area rendered adjustable by a blocking means (not shown)
rotatable about the axis of the transducer. The blocking means can
be rotatable disc but need comprise only a radial or diametral arm
to function as required and can be constituted by a radial arm of
the transducer structure by which the magnet is connected to the
outer ring from which the diaphragm is suspended.
Alternatively, as shown in FIG. 13, the apertures may comprise
holes 102 towards the edge of a circular baffle 104. The effective
area of these holes can be adjusted by rotating of a ring 105,
mounted concentrically with the baffle, which has holes
corresponding in shape and position to the holes 102.
According to a further feature of the present invention, the
transducer or transducers of the microphone unit are provided with
a sound diffusing cover to minimise the directional effects of
noise and the effects of reflected noise. Such a cover can be
constituted by a piece of acoustic transparent fabric secured over
each side of the transducer frame. The diffusion effect can be
controlled by controlling the acoustic transparency of the
fabric.
The microphone units so far described are inherently
bi-directional, being completely symmetrical about the plane of the
baffle. The units can however be rendered uni-directional, as
regards the incoming sound vibrations which are to be transmitted
electrically, without loss of their noise and echo free properties.
This can be achieved as shown in FIG. 14 by modifying the housing
means to provide at the outer end of one of the loading chambers
106 passage means 107 communicating with the chamber at one end and
with the desired direction of sound reception at the other, entry
to this chamber of sound from the other direction being blocked. As
the microphone unit is constructed as an elongated tube, the
passage forming means 107 is constituted simply by a tube of larger
diameter with one end closed, the bi-directional unit being mounted
with one end extending concentrically into this tube. The other end
of the unit functions normally but sound can reach the side of the
diaphragm towards the tube only through the annular passage between
the tube and the outside of the housing of the unit. As the force
acting on the diaphragm is inversely proportional to the square of
the distance between the source and the diaphragm and since ambient
noise and echo can be regarded as from a distant source, the extra
path length represented by the annular passage has no appreciable
effect on the ambient noise and echo cancellation properties of the
unit, whilst having appreciable effect on sound from a relatively
near source. The annular passage will of course reduce the noise
level as well as the microphony effect of the microphone housing
but will also reduce the sensitivity, as it provides for the wanted
sound to reach both sides of the diaphragm so that a degree of
cancellation of this sound also occurs. however, the wanted sound
will be from a source which is relatively close to the unit so that
extra path length of the part of this sound reaching the remoter
side of the diaphragm will cause a lowering of the intensity of
this part as compared with the part reaching the near side of the
diaphragm directly, so that an adequate net output is obtained. It
will be evident that the means forming the passage which converts
the bi-directional unit to a uni-directional unit can be
constructed in various way other than as described.
Although the invention is not limited to the use of any particular
material for the housing means, sub-housings, housing, baffle,
etc., these are preferably built up in a manner similar to that
illustrated in FIG. 15 which shows a tubular housing made from a
plurality of rings of like inner and outer diameter arranged in
co-axial alignment to form a tube in the manner shown in FIG. 15A.
The rings are made of different materials a, b, c, each having
different sound conductive properties. Thus one ring may be formed
of metal, the next made of glass fibre and the third of rubber or a
hard or soft plastics material. Sound transmitted through any one
ring to the next is very substantially attenuated at the interface
and the whole structure is effectively acoustically "dead" or
non-microphonous. Instead of repeating the sequence of, say, metal,
glass fibre and plastics rings, it is preferred instead to use for
the next three, rings of the same three materials but in a
different order, again as shown in FIG. 15A, for example, rings of
glass fibre, metal then plastics. The next three rings can then be
arranged in the order metal, plastics and glass fibre. In this way
a sequence of rings of different materials is obtained, in which
each ring has on either side of it a ring of different material,
and the number of instances in which a sequence of adjacent rings
is kept to a minimum. Obviously this aspect of the invention is not
confined to the use of three materials only, to the materials
mentioned above, or to any particular sequence in which the rings,
or layers of other shapes, are arranged.
It will be appreciated that a tubular microphone housing built up
as described will have almost infinite resistance to sound
conduction in the axial direction. Sound conductivity within each
ring radially of the housing will however be no more reduced than
if the housing were constructed of an integral piece of the
material of the ring. In order to obtain the required high sound
attenuation characteristics in the radial direction also, the
housing can be formed not simply of axially adjacent single rings
but of axially adjacent concentric rings, preferably at least three
in number and again of different materials, as shown in FIG. 15B.
Thus the concentric ring set just described may have on one side
three similarly shaped rings of which the inner one is of fibre
glass, the middle of metal and the outer of rubber, whilst on the
other side the glass fibre is on the outside, the metal ring being
the inner ring and with rubber inbetween.
It will be evident that the structure described for a microphone
housing can be adapted readily to housings of any shape, and to
housings and sub-housings for other apparatus than the
echo-cancelling microphone units and transducer apparatus
described. Thus flat panels can be formed from several layers of
different material the layers themselves being made up of part of
different material.
The various components of the housing described can be connected
together in any suitable way, as by adhesive layers, or by
snap-fitting inter-connections making use of resilience inherently
possessed by some of the materials used.
The structure described can be applied, as well as to the housing,
also to the mounting and actual structure of the electro-acoustic
transducers. Each transducer as shown in FIG. 16 thus may have the
planar diaphragm constituted by a circular metal disc 120, a pair
of speech coils one on each side, and a frame supporting magnets
for co-operation with the coils. The frame consists of two like
parts 121 which clamp the diaphragm suspension 122 between them.
Each frame portion is made of metal or rigid plastics material but
the two portions are at no point directly in contact. Instead, each
portion is given a coating of a softer, resilient material 124 such
as rubber or a plastics material. The symmetry of the arrangement
is rigidly preserved and sound reflecting obstructions such as
boards mounting connecting tags are eliminated as far as possible.
The frame is thus a structure of three different materials, that of
the frame portions, the coating, and the suspension.
In a modified form of the transducer, one of the speech coils and
the associated magnet are omitted, the other features described
being however retained.
The microphone can be mounted within the tubular housing described
above by providing the latter with a pair of inwardly projecting
flanges, advantageously of the same laminated construction as the
housing, the outer edge of the frame of the microphone being
clamped or otherwise secured between them.
Throughout the unit, internal sound reflections are reduced or
eliminated by the provision of rough, sound absorbent, surfaces.
Any metal surfaces can be embossed or etched and all surfaces may
be covered by flocking or by securement of a suede or other
non-reflective fabric. A rubber compound can be applied by dipping
or spraying to the various parts of the unit.
The invention can also be embodied in a telephone hand-set. The
hand-set embodying the invention comprises three effectively
acoustically separated portions, the walls of which may be built up
from laminations in a manner similar to that already described in
connection with the echo-cancelling microphone housing. These are a
stem, an ear-piece at one end of the stem, and a microphone and its
associated housing at the other.
The functioning of the apparatus of FIG. 3 is explained in FIGS. 17
to 21. Each of the microphones M produces an output including a
wanted component or signal S due to sound generated remote from the
apparatus, typically the speech of someone using the apparatus. The
signals S in the microphone outputs can be regarded as being
effectively equal in all respects. In addition, each output
includes a component due to sound from the loudspeaker, which is an
unwanted component or noise. The noise components N.sub.1, N.sub.2
will differ in phases and amplitude from each other and from the
signals S, as shown in FIGS. 17, 18 which represent the separate
microphone outputs.
By the circuits A, S the two microphone outputs are respectively
added and subtracted. The output of circuit A contains the signal S
effectively doubled in amplitude, but the signal S is substantially
lacking from the output of circuit S. As shown in FIGS. 19, 20, the
circuit outputs contain respective noise resultants N.sub.1.sub.+2,
N.sub.1.sub.-2. The microphone outputs have thus been analysed and
a component N.sub.1.sub.-2 has been separated out which represents
substantially pure noise. Such a pure noise component can be
regarded in practice as having a substantially fixed relationship
with the noise components N.sub.1, N.sub.2 and with the additive
noise component N.sub.1.sub.+2. It can accordingly be produced and
treated to be substantially equal in amplitude and opposite in
phase, for the purpose of cancellation, to any one of N.sub.1,
N.sub.2 and N.sub.1.sub.+2. In the illustrated apparatus, the noise
resultant N.sub.1.sub.-2 is amplified and shifted in phase through
180.degree. in a circuit P and added to or combined with the output
of circuit A as shown in FIG. 21, as in the amplifier O. The output
of this amplifier which constitutes a common microphone output is
thus free or substantially free of components due to loudspeaker
noise.
* * * * *