EP0318617B1

EP0318617B1 - Apparatus for processing stereo signals and universal am stereo receivers incorporating such apparatus

Info

Publication number: EP0318617B1
Application number: EP19870310677
Authority: EP
Inventors: Bernard D. Loughlin
Original assignee: Hazeltine Corp
Current assignee: BAE Systems Aerospace Inc
Priority date: 1987-12-04
Filing date: 1987-12-04
Publication date: 1993-08-04
Anticipated expiration: 2007-12-04
Also published as: EP0318617A1; DE3786922T2; DE3786922D1

Description

This invention relates to stereo receivers for stereo radio broadcasts, and to apparatus for processing stereo representative signals in general.
On March 4, 1982 the Federal Communications Commission (FCC) adopted a revolutionary Report and Order authorizing the so-called "marketplace approach" to the selection of a national standard for the transmission of stereophonic radio programs in the AM broadcast band (535-1605 kilohertz). Since then, several different AM stereo systems have been competing in the marketplace.
In the various AM stereo systems which have entered the marketplace, the left channel (L) and right channel (R) audio signals are added together to form an L+R signal, which is used to amplitude modulate the radio-frequency (RF) carrier in the usual manner. By subtracting the R from the L signal, an L-R signal is formed which is used to phase modulate (PM) the RF carrier in some AM stereo systems or, to quadrature modulate (QM) the carrier in others.
The transmitted signals from all AM stereo systems which have been and are currently being used in the United States are compatible with monaural AM receivers, but none are compatible with each other.
Because of this mutual incompatibility among the various systems, receiver manufacturers have been faced with the choice of designing AM stereo receivers for one system only (so-called "single-system¨ receivers), or receivers capable of stereophonic reception of signals of more than one of the systems being broadcast (so-called "multi-system" receivers). Both types of receivers are available in the marketplace.
Multi-system receivers have been of two general types. Some include pilot-signal detection circuits which provide automatic switching of those receiver circuits required to properly decode the particular type of AM stereo signal being received. Other multi-system receivers are equipped with manual switches for that purpose.
Among circuits that have been switched in multi-system receivers are: distortion-correction circuits, 90° phase-difference networks used in one of the AM stereo systems (the Kahn/Hazeltine system) limiters, and detectors.
It is, therefore, an object of the invention is to provide a multi-system or "universal" AM stereo receiver which requires no circuit switching, manual or automatic, in the stereo decoder, but which provides stereophonic reception of broadcasts that use any of the different AM stereo systems which remain in the marketplace.
The document "IEEE International Conference on Consumer Electronics, Digest of Technical Papers, 3rd to 6th June, 1986, Session 19 - Automotive Radio, pages 276-277" describes a universal AM stereo receiver capable of demodulating AM stereo signals broadcast according to a plurality of different AM stereo systems, said receiver comprising:
means for receiving radio frequency (RF) AM stereo signals and for converting the received signals to a corresponding intermediate frequency (IF) signal; and
a stereo decoder for processing said IF signal to develop L and R stereo audio output signals, said decoder including detecting means for detecting the (L+R) and the (L-R) components of said IF signal.
The documents US-A-4426728 and US-A-4641341 also describe such receivers. The receivers described in the above three documents all include mode switching by means of pilot signal detectors.
The present invention is characterized in that:
said decoder includes phase means for modifying said detected (L+R) and (L-R) components by imparting a relative phase difference therebetween, said phase difference being in the range of 30° to 60° over a predetermined portion of the ausio frequency spectrum of said signals; and
combining means for combining said modified (L+R) and (L-R) signals to develop said L and R stereo audio output signals.
Thus receivers according to the invention are capable of processing the IF signal through the same decoding circuitry regardless of which of said different AM stereo systems were used to broadcast the signal being received, and therefore do not require special circuitry which is switched into or out of the decoder when any particular AM stereo system signal is being received.
The three different AM stereo systems currently in use in the United States are known to those in the art as the Kahn/Hazeltine system, the Magnavox system and the Motorola system, and certain features are more fully described herein after.
Embodiments of the present invention will now be described, by way of example, with reference to the accompanying drawings, in which:

Fig. 1 shows a block diagram of the stereo decoder portion of a universal AM stereo receiver embodying the present invention.
Fig. 2 shows a vector diagram with a vector representing the decoded L+R signal and vectors representing the positive and negative decoded L-R signals at the same phase as L+R.
Fig. 3 shows a vector diagram with positive and negative L-R vectors at an angle of 90° with respect to the L+R vector.
Fig. 4 shows a vector diagram with positive and negative L-R vectors at an angle α with respect to the L+R vector.
Fig. 5 shows a vector diagram of the type shown in Fig. 4, except that L-R has been multiplied by a gain factor, (m).
FIGS. 6A and 6B show two vector diagrams in which L+R is summed with positive and negative values of m(L-R), respectively, the same as was done in Fig. 5. However, in Figs. 6A and 6B the ±m(L-R) vector is represented as having a ± component in phase with L+R and a quadrature component, i.e., the Hilbert transform of the in-phase component. This transformation shifts all Fourier components by 90° [see IEEE Standard Dictionary of Electrical and Electronics Terms.]

The present invention enables construction of an AM stereo receiver which requires no switching in the stereo decoding circuits and which provides universal AM stereo signal decoding for any of the different AM stereo signal types being broadcast. The invention also provides enhanced perceived stereo separation to listeners of such receivers and of other stereo systems.
The 90° phase-difference networks used in receiver decoding circuits for the Kahn/Hazeltine Independent Sideband (ISB) AM Stereo System have, in prior art multiple-system receivers, been switched out of the circuit when signals of other AM stereo systems were being received. (With the present invention, the phase-shift networks need not be switched.) Instead, such networks are designed to provide a phase difference between the L+R and L-R channels of a selected amount less than 90° and preferrably within the range of approximately 30° to 60°, for example 45°. Although this compromises somewhat the mathematically calculated stereo separation for all AM stereo systems being received; nevertheless, the perceived separation is enhanced by another feature of the invention. The gain in the L-R channel is increased above the value normally required to optimize stereo separation, and the increased gain, in conjunction with the selected phase shifting in the aforementioned phase-difference networks, provides a significant increase in the subjectively perceived stereo separation.
Furthermore, AM stereo systems competing in the marketplace require distortion correction in the L-R channel of the stereo receiver. Receivers for one AM stereo system (Magnavox) have employed a limiter in the intermediate-frequency (IF) circuits prior to L-R signal detection. Receivers for other systems have employed inverse modulation with various degrees of correction. In one embodiment of the present invention, a fixed inverse modulation characteristic is used which provides sufficient distortion correction in the L-R channel for all systems.
Finally, since in universal AM stereo receivers which embody the present invention pilot-signal detection is not required for determining which AM stereo system's signal is being received, inclusion of such circuits is at the option of the receiver designer and if included, would serve to activate a stereo indicator light when a stereo broadcast is received.
A block diagram of a "universal" AM stereo receiver embodying the invention is shown in Fig. 1. Antenna 10 and RF/IF unit 11 may be of conventional design and construction, and serve to receive RF signals and to convert the received signals to a corresponding IF signal. The IF signal is coupled to the input terminals of both L+R detector 12 and inverse modulator 13.
L+R detector 12 may be an envelope detector, but other forms of detectors known to persons skilled in the art may be used instead. For example, a synchronous detector having an appropriate reference signal supplied to it could be used. It is to be understood that L+R detector block 12 in Fig. 1 can (if necessary) include means for filtering the detector output to reduce any residual spurious IF signal to an acceptably low level relative to the desired L+R signal output.
The output of L+R detector 12 is coupled to amplifier 14, whose gain is designed to meet the requirements of inverse modulator 13, as will be described shortly. The amplified L+R output of amplifier 14 is fed to inverse modulator 13 where it inversely amplitude modulates the supplied IF signal, effecting a reduction in quadrature distortion in the incoming signal by reducing the amplitude modulation on the signal prior to detection in L-R detector 15.
In the various AM stereo systems competing in the marketplace, at the transmitter the process of combining undistorted L+R envelope modulation with L-R phase, or quadrature, modulation in the transmitted signal results in distortion of the quadrature component of the modulation, which component is a function of the L-R modulation. In some systems correction of this distortion is accomplished partially at the transmitter stereo encoder and partially at the receiver decoder. In another system all correction is performed at the receiver.
For example, consider three different systems. In the Kahn/Hazeltine Independent Sideband (ISB) AM stereo system, approximately half the distortion reduction is effected at the transmitter stereo encoder, and half at the receiver decoder. Thus, an inverse modulation function of

for example, is required in the L-R channel of the receiver. In the denominator of the function, the "1" is a dc component which may correspond to the dc component of the detected L+R signal and "x" is the ac component. Thus, the denominator in this function can vary from 0.5 to 1.5.
In the Magnavox AM stereo system, distortion correction occurs entirely at the receiver decoder by means of an IF signal amplitude limiter. In this case an inverse modulation function of

is required in the L-R channel of the receiver.
In the Motorola system, first order distortion correction takes place at the transmitter encoder, whereas correction for higher-order terms requires a

transfer function (where 0 is the instantaneous carrier phase with respect to the unmodulated carrier phase) at the receiver decoder. However, the first-order decoder transfer function is simply 1.
From the above, it may be seen that an inverse modulator transfer function of

for the Kahn/Hazeltine system is conveniently approximately midway between the

and the first-order 1 of the other two systems. Thus, an inverse modulation characteristic of

is a reasonable choice for a universal decoder. Using this characteristic the distortion in the modulation of the L-R component of the transmitted AM stereo signal is essentially corrected in the Kahn/Hazeltine system, and is effectively reduced in the other two systems. While

is a preferred compromise value for the inverse-modulator transfer function, the invention is not restricted to only that form, and receiver designers may employ other transfer functions within the scope of this invention. The gain of amplifier 14 is set to the value required to provide the desired transfer function in inverse modulator 13.
While inverse modulation is desirable in an AM stereo receiver for reduction of distortion in the received L-R component of the AM stereo signal, for low-cost receivers the receiver designer may choose to omit inverse modulator 13 and amplifier 14, but still retain the universal decoding and/or stereo separation enhancement features unique to this invention. Or, the designer may elect to switch among several different inverse-modulator transfer functions to provide more nearly optimum L-R distortion correction for each AM stereo system which may be received. A preferred method of switching this function would be by means of changing the gain of amplifier 14. Switching could be either manual or automatic. Automatic switching could be activated by detection of the pilot signal which is unique to the AM stereo system being received. Circuits for pilot signal detection and automatic switching are known to persons skilled in the art. Omission, or switching, of the inverse-modulator transfer function does not negate applicability of other features of the invention.
In the embodiment of Fig. 1, the inversely modulated IF signal is coupled to L-R detector 15. Detector 15 normally is a synchronous phase, or quadrature, detector requiring a reference signal at the phase necessary to effect the desired phase or quadrature detection. Circuits for deriving such a reference signal are well known to those skilled in the art. For example, the reference signal may be derived from a phase-locked-loop driven by the output of an IF signal amplitude limiter. It will be understood that the L-R detector 15 in Fig. 1 would (if necessary) include means for filtering the detector output to reduce any residual spurious IF signal to an acceptably low level relative to the desired L-R signal output.
The L-R output signal from detector 15 is coupled to amplifier 16, where the signal is amplified to provide a selected level relative to the detected L+R signal, as will be described later.
The detected L+R signal and the amplified L-R signal are coupled to the inputs of phase- shift networks 17 and 18, respectively, as shown in Fig. 1. These circuits can be designed to provide (over a selected frequency range) a reasonably constant phase difference between the two channels. For example, in receivers designed for the Kahn/Hazeltine ISB AM stereo system the phase-shift (φ) in networks 17 and 18, Fig. 1, is nominally ±45° resulting in a 90° phase difference in the transfer functions of the L+R and L-R channels over a specified frequency range. Other AM stereo systems do not require these phase-shift networks. However, it will be shown that in accordance with the invention, a variant of these networks can be used advantageously in a universal AM stereo receiver for all systems.
For a better understanding of the invention, assume that the phase shift (φ) in networks 17 and 18, Fig. 1, is nominally ±45° over a frequency range of 100-6000 Hz, for example. Further, assume that the universal receiver is receiving a Kahn/Hazeltine ISB AM stereo signal modulated with a left-only audio-frequency signal of 1000 Hz, for example. At the transmitter the phases of the L+R and L-R signals are shifted to provide a 90° phase difference between those two signals in order to obtain the essentially single-sideband characteristic of the transmitted ISB signal; however at the transmitter the phase shift (φ) was negative in the L+R channel and positive in the L-R channel. The receiver phase-shift networks essentially restore the original phase relationship of L+R and L-R as it existed prior to the phase-shift networks in the transmitter, and with the gain of amplifier 14 properly adjusted, the vector representation of the L+R and L-R input signals to matrix 19 would be as shown in Fig. 2. In the matrix, L+R vector 20 and L-R vector 21 would add to produce a 2L output, provided L+R and L-R amplitudes were exactly equal. Similarly, if L-R vector 21 were subtracted from L+R vector 20 (or equivalently, if negative L-R vector 22 were added to L+R vector 20) the vectors would cancel, resulting in zero R output. Thus, $\frac{L}{R}$
(stereo separation) can potentially be very large (infinite in the limit).
However, if the receiver with the assumed 90° phase-difference networks (φ=45°) were to receive an AM stereo signal of one of the other competing systems, modulated with a left-only signal, the vector diagram for the input and output signals of matrix 19 would appear as shown in Fig. 3. In this case, the L+R and L-R vectors are in 90° relationship and the sum of L+R vector 30 and L-R vector 31 is vector 33, corresponding to the L output of matrix 17. Likewise, L+R vector 30 and negative L-R vector 32 combine to produce vector 34, corresponding to the R output of matrix 19. In this case, the relationship of the L and R outputs of matrix 19 ( vectors 33 and 34, respectively) is

i.e., there is essentially no stereo separation. It must be understood, that in Fig. 3 the positive detected L-R signal may be at an angle of either + or - 90° with respect to the detected L+R signal, depending on the system being received, but this does not alter the conclusion that

and there is essentially no stereo separation.
In accordance with the present invention, choice of the amount of phase-shift (φ) provided in each of the networks 17 and 18 of Fig. 1 is not critical and can be in the range of 15° to 30°, for example, (giving a corresponding phase difference between the L+R and L-R channels in the range of 30° to 60°), and still provide performance within the scope of the invention. However, for purposes of explaining the operation and benefits of the phase-shift networks in the invention, a median value of φ = 22.5° (providing 45° phase difference in the transfer characteristics through phase-shift networks 17 and 18) will be used.
Assuming again, reception in the universal decoder of a Kahn/Hazeltine ISB signal modulated with L-only audio-frequency information, and with the L-R gain of amplifier 16 adjusted to make the amplitude of the L-R input to matrix 19 equal to the amplitude of the L+R input, the vector diagram for the input and ouput signals of matrix 19 would be as shown in Fig. 4. For the example chosen, the angle in Fig. 4 is equal to 45°. Thus, L+R vector 40 and L-R vector 41 (equal in length) add in matrix 17 to form an output at the L terminal represented by vector 43. Similarly, L+R vector 40 and negative L-R vector 42 combine in the matrix to form an output at the R terminal corresponding to vector 44. With the parameters chosen for this example the ratio of the amplitudes of the L and R outputs (vectors 43 and 44 respectively) is $\frac{L}{R} = 2.414$

, corresponding to a stereo separation of 7.66dB. Similarly, if signals of other AM stereo systems were received, the angle α in the diagram of Fig. 4 may be positive or negative, depending on the system, but the basic operation in providing some separation for all systems would be the same.
While the stereo separation for circuit conditions corresponding to Fig. 4 is calculated to be 7.66dB, as described above, another phenomenon which enhances the perceived separation is present which is better explained in the following example.
In this example, assume, as in the previous example that φ (in Fig. 1) = 22.5° providing a phase difference of 45° between the transfer characteristics of the L+R and L-R channels and that the gain of amplifier 16 in the L-R channel is adjusted to make $(L-R) = √2(L+R)$
. Also assume that the signal being received is a Kahn/Hazeltine ISB AM stereo signal modulated with a left-only audio-frequency signal. For these assumed conditions, the vector diagram for the audio-frequency signals at the inputs and outputs of matrix 19 is as shown in Fig. 5. For the example cited, m in Fig. 5 equals √2 and α = 45°.
As shown in Fig. 5, the L+R vector 50 and positive m(L-R) vector 51 are summed in matrix 19 to produce an output at the L terminal represented by vector 53. Similarly, L+R vector 50 and negative m(L-R) vector 52 are combined to produce an output at the R terminal represented by vector 54. The ratio of the L and R output amplitudes ( vectors 53 and 54, respectively) is $\frac{L}{R} = 2.24$
corresponding to a calculated stereo separation of 7.0dB. However,the separation perceived by listeners in this case corresponds more nearly to perfect separation, for reasons to be described below. Again, if the signals of other AM stereo systems are received, the vector diagrams will be essentially equivalent to the diagram in Fig. 5, except that the angle α may be positive or negative, depending on the system.
To further assist in understanding the manner in which the perception of stereo separation is enhanced using the invention, the vector diagram of Fig. 5 may be replaced by its equivalent in Figs. 6A and 6B. In Fig. 6A, L+R vector 60 is the same as vector 50 in Fig. 5. However, the +m(L-R) vector 51 in Fig. 5 is represented by component 61 in Fig. 6A, in-phase with the axis of L+R vector 51, and the Hilbert transform of 61, which is quadrature vector 62. Thus, the vector sum of vectors 61 and 62 is equal to the +m(L-R) vector 51 of Fig. 5. Similarly, in Fig. 6B L+R vector 60 is the same as vector 50 in Fig. 5, and vectors 64 and 65 are the equivalent in-phase component and its Hilbert transform, respectively, of the -m(L-R) vector 52 in Fig. 5.
Therefore, in Fig. 6A, it can be seen that the left output of matrix 19, represented by vector 63 (equal to vector 53), has a component 66 at the phase of the L+R axis. This in-phase component 66 is the sum of vectors 60 and 61, and represents the L output intended in the transmitted signal. Also present in the L output of matrix 19 is the Hilbert transform component, vector 62, representing L-R in quadrature with the axis of the intended L-output signal. The effect of this quadrature L-R component will be described following the discussion of Fig. 6B below.
In Fig. 6B, the L+R vector 60 is the same as vector 50 in Fig. 5, and the -m(L-R) vector 52 in Fig. 5 is represented in Fig. 6B by its equivalent two vector components, vector 64 (at the same phase as the L+R axis), and its Hilbert transform, vector 65. Vectors 60 and 64 are combined in matrix 19 and effectively cancel at the R output of the matrix, as intended in the transmitted signal. However, there is also present in the R output the Hilbert transform, vector 65, representing L-R in quadrature with the axis of the desired signal output. Thus, considering for the moment only the in-phase components of Figs. 6A and 6B, the potential for perfect, or near perfect, stereo separation exists. The effect of the quadrature components of the L-R signal in the ouput is explained as follows.
In the special case just cited in relation to Figs. 5 and 6, L-R and L+R are in the relationship

for universal compatibility. In this case, L-R was shown to have a component of the intended value, i.e. the L-R component was of the amplitude and phase required to properly matrix with L+R to reproduce the originally intended L and R components in the transmitted signals. However, the additional L-R component, which is the Hilbert transform of the intended signal, is at an angle of 90° (which will be + or - depending on the system being received) with respect to the intended signal.
This Hilbert transform (quadrature) component of the L-R signal has the characteristics of a reverberant signal. For reverberant signals the L+R and L-R components of the stereo signal are essentially uncorrelated signals from all directions, and, in the example cited, another uncorrelated L-R component (the Hilbert transform, above) is present. Because of the non-directional nature of these signals by themselves, they have a reverberant characteristic. However, the matrixing of the in-phase components of L-R and L+R produces the originally intended directionally- correlated direct path signals with the proper time and intensity difference to provide localization to the listener. On the other hand, the remaining Hilbert transform component of the L-R signal, being at an angle of 90° with respect to the intended direct-signal, is uncorrelated and does not upset the intended time and intensity differences of the direct signal, and is present equally in the L and R outputs. Therefore, the criteria for properly localizing the intended signal have been maintained, and there has been a slight increase in reverberant power, calculated to be 1.76dB.
In the discussion related to Fig. 5, it was shown that conventionally calculated stereo separation for the case cited is 7dB. However, using the analysis just described the perceived stereo separation, or localization, can be perfect, or near perfect, with some added reverberation effect.
In the July/August, 1973 issue of Technology Review (published by MIT) an article by Amar G. Bose, "Sound Recording and Reproduction, Part Two, Spatial and Temporal Dimensions", pp. 25-33, dealt with some aspects of the above analysis. On p. 28 Bose states, "... virtually all of the audience in a concert hall is seated in the region where the reverberant field is dominant¨. He also said, "The reverberant field, while contributing nothing to our ability to localize, plays a very important role in our perception of the timbre of music". On p. 29 he goes on to say that a "small amount of direct sound is all that is required for localization". In his summarizations on p. 29, Bose stated, "From our studies of the spatial characteristics of sound fields, we learned that we should design loudspeakers so that they place the listener in a predominantly reverberant field through the use of the correct proportions of direct and reflected sound". In general the article substantiates the concept that a listener can localize a sound source in the presence of reverberation, even where the reverberant sound field is large in comparison to the direct sound field. It also confirms the desirability of providing reverberation effects for the listener.
Actual tests proved that the seemingly poor calculated stereo separation of 7dB in the example cited above was judged to sound better than (and in most cases was preferred to) normal stereo with much greater stereo separation, again confirming the analysis. Therefore, with a slight increase in reverberant energy this special case for universal stereo reception could be properly termed, "full stereo".
It should be noted that while the above example was based on reception of a Kahn/Hazeltine AM stereo system signal, the analysis is applicable to other systems as well. Further, the general principle of the L-R signal having a component at the same phase as L+R and a reverberant Hilbert transform component, still applies where the angle between the transfer characteristics of the L+R and L-R channels is other than 45° and where the L-R gain relative to L+R is other than the square root of two. Tests have shown that L-R gain enhancement up to at least 6dB has produced what has been judged to be more pleasing sound than conventional prior art AM stereo reception, in most cases. Gain enhancement in the L-R channel greater than 6dB, while producing otherwise pleasing effects, may result in deterioration of the signal-to-noise ratio which may not be deemed acceptable, depending upon the amount of enhancement.
Even where there is no gain enhancement, as in the conditions stated in the example represented by Fig. 4, and where the conventionally calculated stereo separation is only 7.66dB, using the analytic approach just described of separating L-R into an in-phase (with L+R) component and a quadrature, or reverberant, component, the true perceived separation is more accurately represented by the amount 15.3dB. This latter amount is the ratio of the components of the L and R outputs of matrix 19 corresponding to the sum and difference, respectively, of the in-phase components of the L-R and L+R inputs to the matrix.
The concept of producing a fuller, more pleasing sound through use of the above-described quadrature component of the intended L-R signal is not restricted to AM stereo systems only. Introduction of a relatively constant phase difference between L+R and L-R signals over a selected frequency range in other types of stereo systems, e.g. frequency-modulation (FM) stereo systems, possibly with L-R gain enhancement, will produce the above described pleasing reverberant effect.
Those skilled in the art will readily recognize that the apparatus comprising units 16-18, inclusive, in Fig. 1 can be incorporated in an FM stereo receiver, for example, between the outputs of the L+R and L-R signal detectors and the inputs of the matrix (corresponding to unit 19 in Fig. 1). Referring to Fig. 5, if

the matrixing of L+R and the component of m(L-R) which is in-phase with the L+R axis will provide perfect, or near perfect stereo separation of the intended signal, and the quadrature component of m(L-R) will provide the reverberant effect. However, even if this relationship of m and cos α is not maintained, the reverberant effect may be obtained with some compromise of separation, within the scope of the invention.
It will also be recognized that although the preferred universal case was described above as having 45° phase difference between the L+R and L-R channels, [45° + (N x 90°)] is the general universal case. Thus, where N = 0, the phase difference is 45°; where N = 1 the phase difference is 135°; and so on. Therefore, although the 45° phase difference networks could be replaced with 135° networks, the 45° case was used for simplicity and to facilitate a better understanding of the present invention.

Claims

A universal AM stereo receiver capable of demodulating AM stereo signals broadcast according to a plurality of different AM stereo systems, said receiver comprising:
means (10,11) for receiving radio frequency (RF) AM stereo signals and for converting the received signals to a corresponding intermediate frequency (IF) signal; and
a stereo decoder (12,15,17,18,19) for processing said IF signal to develop L and R stereo audio output signals, said decoder including detecting means (12,15) for detecting the L+R and the L-R components of said IF signal;
characterized in that:
said decoder includes phase means (17,18) for modifying said detected L+R and L-R components by imparting a relative phase difference therebetween, said phase difference being in the range of 30° to 60° over a predetermined portion of the audio frequency spectrum of said signals; and
combining means (19) for combining said modified L+R and L-R signals to develop said L and R stereo audio output signals.
A receiver according to claim 1 characterized in that said decoder (12,15,17,18,19) includes amplifier means (16) to impart a predetermined gain to the L-R component relative to the L+R component to enhance the perceived stereo separation between the eventual L and R stereo audio signals.
A receiver according to claim 1 or claim 2 characterized in that said decoder (12,15,17,18,19) includes an inverse modulator (13) responsive to said IF signal and to the detected L+R component to reduce inherent L-R modulation distortion in said IF signal prior to said L-R detection of said IF signal, the inverse modulator having an inverse modulation function selected to reduce said distortion in IF signals originating from each of said different AM stereo systems.
A receiver according to any one of claims 1 to 3 characterized in that said phase difference is substantially equal to 45° + (N x 90°), where N is an integer.
A receiver according to claim 4 characterized in that N = 0 so that said phase difference is substantially 45°.