US 3009991 A
Description (OCR text may contain errors)
Nov. 21, 1961 l. BEKEY souND REPRODUCTION SYSTEM Filed DeC. l, 1955 .N GNN IN V EN TOR.
United States Patent 3,009,991 SOUND REPRODUCTION SYSTEM Ivan Bekey, Santa Monica, Calif. (517 Erskine Drive, Pacific Palisades, Calif.) Filed Dec. 1, 1955, Ser. No. 550,274 4 Claims. (Cl. 179-1) Thisrinvention relates to sound reproduction, and particularly to apparatus for controlling the fidelity of the sound produced.
Particular attention has been given to design of ampliiers and other components of sound reproducing equipment to ensure yfaithful reproduction. Heretofore, the sound produced by the usual form of acoustic system is far from a faithful reproduction of the signal input to the loud speaker. This is due to the frequency response characteristics of the speaker and its cabinet, which is normally very poor in comparison with the frequency response characteristics of high quality amplifiers, and other electrical components. Improvements in loud speakers and cabinet designs are, of course, constantly attempted, and at substantial expense. At best, these attempts are ineffective to overcome any but the more extreme distortions. The principle of degenerative feedback has been successfully used in the design of amplifiers to reduce noise and distortions, and to achieve relative independence between the sound produced and the characteristics of the amplifiers. Speaker systems for use with high fidelity ampliers have been used without the sound produced thereby being included in the feedback loop.
The primary object of this invention is to provide a sound reproduction system in which noises and distortions due to the speaker system and the room in which it is located are largely eliminated, and in which the sound produced is largely independent of the characteristics of the speaker, enclosure or room. This is made possible by utilizing a degenerative feedback arrangement in which the sound produced is the controlled variable and the program material is the reference input.
Use is made of a unidirectional mechano-electrical transducer associated with the source of sound and a feedback circuit connecting the transducer to an intermediate or input stage of the amplier.
This arrangement makes possible a less critical design of speaker systems, reduces the cost of equipment, and materially enhance the faithfulness of the sound produced.
This invention possesses many other advantages, and has other objects which may be made more clearly apparent from a consideration of one embodiment of the invention. For this purpose, there is shown a form in the drawings accompanying and forming part of the present specification. This form will now be described in detail, illustrating the general principles of the invention; but it is to be understood that this detailed description is not to be taken in a limiting sense, since the scope of this invention is best defined by the appended claims.
Referring to the drawings:
FIGURE 1 is a diagrammatic representation of a sound reproducing system incorporating the present invention; and
FIGS. 2 and 3 are diagrammatic representations of alternate pick-up constructions for use in the system.
In the drawings, there is shown a loud speaker accommodated in a suitable cabinet 11. 'Ihe enclosure 11 is located in a room 12 which preferably has highly sound absorbent walls to eliminate distortions due to the selective reflection characteristics. The room 12 and cabinet 11 determine the loading of the speaker 10, and hence influence the sound emitted from the speaker. Connected to the speaker 10 is an amplifier system 13 which is, in turn, connected to an audio signal source 14 through a subtracting or mixing network 27 to be referred to hereinafter.
In order to make the sound produced correspond to the actual signal supplied to the loud speaker, use is made of a degenerative feedback circuit forming, with the ampliier and speaker, a sound servomechanism. Thisv feedback circuit comprises a microphone 15, a feedback shaping circuit 16, an amplifier 17, and a connection 18 to the subtracting or mixing network 27. In this instance, the mixing network is located between the main amplifier 13 and the signal source 14. Optionally, the mixing network 27 could be located at an intermediate stage of the amplifier 13.
The microphone 15 is placed' as close as possible to the cone of the loud speaker and has unidirectional characteristics in order that it be responsive to the sound produced by the loud speaker and not to whatever reflections might be produced by the room 12. In order to prevent mechanical vibrations other than sounds from affecting the microphone 15, it is resiliently mounted in an appropriate manner, as diagrammatically illustrated at 19.
The signal corresponding to the sound at the microphone is fed back degeneratively or at a phase angle of Since the cabinet and room influence the sound emitted, the signal created by the microphone takes into account the characteristics of the cabinet and room. In this arrangement, the sound produced is the variable, controlled by the signal at the input to the amplifier 13. Depending upon the factors in the feedback shaping circuit 16 and amplifier 17, the sound produced is made to be independent, more or less, of the characteristics of the loud speaker enclosure 11 and room 12. Distortions are largely eliminated.
The time required for the sound to reach the microphone 15 causes an additional phase lag in the feedback loop 15, 16, 17, 18. This added phase dilerence varies directly in accordance with the frequency of the signal. The additional phase lag due to this spatial relationship poses a possibility of oscillations at a certain frequency if the gain yfor that frequency is suiciently large. Thus, for example, the spacing between the microphone 15 and the loud speaker 10 will correspond to a half wave length (or a 180 added phase shift) of a certain frequency sound wave. If the spacing is one-half of a foot, a wave having a frequency of about 1000 cycles per second may cause oscillations since the feedback circuit also introduces a 180 lag.
In order to prevent this spatial time lag from creating oscillations, the feedback shaping circuit 16 may be so designed as to attenuate those frequencies that would tend to cause oscillations. It is unnecessary to attenuate those frequencies that have wave lengths longer than twice the spacing between the microphone and the speaker. Since most significant distortions occur in the low frequency range, the attenuation of high frequencies does not detract from the effectiveness of the system.
The sound produced by the loud speaker 10 is a function of the character and extent of loading of the loud speaker in the particular room 12. The microphone 15, accordingly, is responsive to the characteristics of the room 12 as well as to the normal innite baille characteristics of the loud speaker 10. The feedback arrangement produces a linear distortionless sound heretofore not achieved by refinements merely in cabinet or loud speaker design.
'Ihe foregoing considerations assume that there are no material resonant characteristics of the acoustic system in the range under consideration. Actually, however, there may be deiinite physical resonances at low frequencies within the range operated upon by the feedback circuit. Since a physical resonance is accompanied by phase shifts which reverse sign through each resonant frequency,
it is possible that this also might produce a condition in which oscillations could occur.
In order to ensure against the possibility of inherent physical resonances creating a condition of possible oscillations and to create a system having greater flexibility of control over the sound produced, use is made of the pick-up structure 20 responsive to a function of axial movement of the loud speaker cone. A degenerative feedback circuit, including a feedback shaping circuit 21, amplifier 22, and connection 23, serves to control the motion of the loud speaker cone and minimize whatever resonant tendencies there may be in the acoustic system. In the present instance, the connection 23 connects with the mixing network 27. It will be understood that a separate mixing network could be used.
The pick-up 20 in this instance is diagrammatically illustrated as a potentiometer type. The potentiometer contact arm 24 is mounted on the cone.
Instead of using a potentiometer type pick-up responsive to the amplitude of motion of the loud speaker cone, various other arrangements could be provided, such as piezoelectric crystals, magnetic pick-up or the like. For example, in FIG. 2 there is shown a coil 30` operatively associated with the voice coil 31 of the loud speaker. The current of the voice coil determines the motion of the speaker cone. The pick-up coil 30 is also responsive to this function.
In FIG. 3 a magnetic element 32 carried by tne loud speaker cone induces a current in a magnetic pick-up 33.
It will be understood that the pick-up 20 or either of the pick-ups shown in FIGS. 2 and 3 itself is insufficient completely to control the sound produced since, unlike the microphone it is responsive to the motion of the speaker and not the sound produced.
Resonant tendencies, such as might upset the balance of the feedback circuit associated with the microphone, are prevented from causing any condition which might tend to sustain oscillations.
The closer the microphone 15 is to the loud speaker cone, the greater the range of frequencies to which the feedback circuit may be made responsive. The closer the microphone 15 is to the ear of the listener, the more the feedback system reects and controls the sound reaching the listeners ear. This, however, may be at a sacrifice of the frequency range of operation.
Usual servomechanism techniques well known in the art may be applied to the feedback circuits to achieve appropriate stability and sufficient range of operation.
Thus, in accordance with the present invention, the acoustical feedback starting with the electrical-sonic transducer` or microphone 15 effects an improvement in the frequency response curve of the speaker system, tending to flatten that curve, particularly by increasing the response at lower frequencies. This improvement, however, tcnds to produce an undesirable secondary effect of introducing undesired `harmonics due to over driving of the speaker. The feedback starting with the motion sensing electrical transducer 20, however, substantially nullifies the aforestated undesirable secondary effect, in that it acts yto` depress or partially suppress the motion of the speaker cone 10. As a result of the combined effects of both feedback systems, a substantial improvement in acoustic output characteristics is thereby obtained.
The inventor claims:
1. A sound servomechanism for cooperation with an audio frequency electrical signal source coupled to a loud speaker output for said source, said servomechanism comprising an electrical-sonic energy transducer spaced from said speaker and responsive to the sonic output of said speaker for producing an electrical signal related to said sonic output, degenerative feedback means coupling an electrical signal derived from the last-mentioned signal to said signal source, a motion sensing electrical transducer positioned to detect physical movements of said speaker and convert the same into a related electrical signal, and degenerative feedback means coupling an electrical signal derived from the lastmentioned signal to said signal source simultaneously with the feedback of the first-mentioned derived electr-ical signal.
2. In combination, an audio frequency electrical signal source, an acoustic output means coupled to said source for converting the electrical energy signal of said source to an acoustic energy signal, an electrical-sonic energy transducer spaced from said output means and responsive to sonic -output of said output means for producing an electrical signal related to said sonic output, degenerative feedback means coupling an electrical signal derived from the last-mentioned signal to said signal source, a motion sensing electrical ltransducer positioned to detect physical movements of said output means and convert the same into a related electrical signal, and degenerative feedback means coupling an electrical signal derived from the lastmentioned signal to said signal source simultaneously with the feedback of the first-mentioned derived electrical signal, whereby the first-mentioned feedback means functions to effect a more uniform frequency response of said output means, and the second mentioned feedback means functions to depress the motion of said output means.
3. In the combination of claim 2, said motion sensing transducer comprising a potentiometer pick up.
4. In the combination of claim 2, said motion sensing transducer comprising a magnetic induction pick up.
References Cited in the tile of this patent UNITED STATES PATENTS 1,822,758 Toulon Sept. 8, 1931 2,194,175 Wilhelm Mar. 19, 1940 2,841,648 Thurston Iuly l, 1958 2,843,671 Wilkins et al. July 15, 1958 2,860,183 Conrad Nov. ll, 1958 2,948,778 Clements Aug. 9, 1960