JP2726654B2 - Headphone equipment - Google Patents

Description

Description: TECHNICAL FIELD The present invention is used in a relatively compact headphone that can be comfortably worn without excessive pressure on the head, typically by the force of pressing a cup against the head, reducing distortion. A novel device and technique for reducing noise while providing a relatively constant frequency response that is unchanged by the user. The invention more particularly relates to a feedback control system that achieves high gain in a desired frequency range while avoiding oscillation. BACKGROUND OF THE INVENTION A typical prior art method of attenuating noise uses a head horn with a high mass, a large volume and a spring support that exerts a large pressure on the head. The high mass increases the inertia that resists acceleration and contributes to the rigid structure of the head horn wall. High pressure has the effect of increasing the damping of closed low frequencies without air leakage. Large volume compliant air cavities provide high frequency roll-off. However, most of these techniques increase user discomfort. Conventional active noise cancellation techniques include a method of converting external noise using a micro horn outside the head horn. The electrical device then processes the converted noise signal in a manner similar to the attenuation caused by the headphone for the noise signal and provides the headphone driver with an inverted signal to cancel external noise. This method may actually increase the noise level inside the headphone in an open loop system that is not applicable to different users. Other methods use closed loops or servo mechanisms, such as the National Technical Infor
Report AB-A009 274 distributed by mation Service
Patrick Michael Dallosta, April 1975, listed in
"A STUDY OF PROPOSED EAR PROTECTION DEVICES
FOR LOW FREQUENCY NOISE ATTENUATION ". U.S. Pat. No. 3,009,991 discloses a speed sensitive microphone located in the feedback loop immediately adjacent to the loudspeaker diaphragm. U.S. Pat. No. 3,562,429 discloses a movable feedback, a remote acoustic feedback and a feedback around a head horn. OBJECTS It is therefore an object of the present invention to provide an improved feedback control system. According to the present invention, there is provided an amplifying means having an input and an output and amplifying a signal applied to the input, wherein the amplifying means is arranged in a closed loop by feedback means coupling the output to the input. The amplifying means providing a controlled output signal responsive to the input signal, characterized by an open loop gain and phase margin therebetween when the feedback path between the input and the output is interrupted In the feedback control system, means are provided for establishing the open loop gain of the amplifying means to a high level substantially uniformly in a predetermined frequency range substantially bounded at least by one end by the break frequency. The amplifying means includes a circuit configured to establish a frequency characteristic in which a response from a corner frequency to a critical frequency having an open loop gain of substantially 1 is reduced in a region outside the predetermined frequency range, For a significant portion of that area, the slope of the frequency response is greater than 12 dB per octave, while maintaining sufficient phase margin to maintain stability, thereby providing high gain over a given frequency range. Avoid oscillations and minimize the error between the controlled output signal and the desired control output signal indicated by the input signal. The phase margin is desirably at least π / 6.
The amplifying means may have a slope of the open loop gain from the corner frequency to a first frequency between the corner frequency and the critical frequency that is substantially greater than 6 dB per octave and higher than the first frequency to the critical frequency. The slope to the second frequency is less than 12 dB per octave,
Means for setting the change in frequency response such that the slope of the frequency above the frequency is substantially zero decibels per octave. And the slope between the corner frequency and the first frequency is at least 12 dB per octave;
The slope between the first and second frequencies is substantially 6 dB per octave. (Example) The present invention will be described in detail with reference to examples. First, with reference to FIG. 1, a headphone that can effectively use the feedback control system according to the present invention will be described. A microphone 11 has a headphone housing 13, a driver 17 and a driver
The area between the outer ear 16 and the cushion support 21 of the head horn is arranged coaxially with the diaphragm 14 and the cushion 15
Seals. The microphone 11 is arranged near the entrance of the ear canal 18 so that the amplitude of the pressure wave in the microphone 11 is substantially the same as the entrance of the ear canal 18. The cavity 12 is reduced so that the pressure is substantially constant throughout the cavity. To this end, the cushion 15 has a high mechanical compliance, a high flow resistance, a high density, an axial cross-sectional area approximately equal to the diaphragm 14 and an axial ring of the cushion 15 around the cavity 12. Smaller than the cross-sectional area. The head horn housing 13 is coupled to a resilient head band 23 in a known manner by a friction ball joint 22, as shown in FIG. A typical material of the headhorn cushion 15 is a continuous foam urethane foam that is slow to recover. The cushion 15 holds the outer ear 16 over a relatively large area, disperses the force necessary to maintain a good seal over a sufficiently large area, effectively seals, and reduces the pressure on the ear sufficiently. To make the user uncomfortable. Continuous bubble high flow resistance materials are independent in that they exhibit the mechanical advantages of a continuous bubble material that conforms to the irregular shape of the ear, while significantly attenuating spectral components above a certain frequency in the middle range, e.g., 2 KHz. It has the acoustic advantages of bubble material. It also keeps the pressure in the cavity substantially constant in this frequency range. Fluid filled cushions also have these properties. Such an arrangement of the present invention provides for the overhead generated in response to the force required to maintain a closed or continuous bubble low flow resistance cushion around the ear that negligibly attenuates low frequency signals as in the prior art. And a large cavity with high pressure. Such conventional cavities are characterized by divergence in large pressure fields, requiring a larger diaphragm deflection to produce a constant sound pressure level than the small cavities according to the present invention. Thus, the present invention can achieve better acoustic efficiency with a smaller set that gives the user more comfort. Referring to FIG. 2, there is shown a block diagram showing the logical configuration of the device according to the present invention. Signal combiner 30 is input terminal 24
The signal desired to be reproduced by the head horn is algebraically combined with the feedback signal provided by the microphone preamplifier 35. The signal combiner 30 supplies the combined signal to a compressor 31, which limits the level of the high level signal. Then, the compressor sends the compressed signal to the compensator 31A. Compensation circuit 31A adapts the oven loop gain to the Nyquist stability criterion to prevent oscillation when the loop is closed.
The illustrated device is similarly made for the left and right ears. Power amplifier 32 energizes head horn driver 17 to generate an acoustic signal in cavity 12, which signal is
An external noise signal entering the cavity from the region 25 is combined with a portion indicated by a circle 36, and the combined sound pressure signal is input to the microphone 11
And converted there. The microphone preamplifier 35 amplifies the converted signal and transmits it to the signal combiner 30. The compensation circuit 31A has a loop gain T _(s) corresponding to the curve 2 in FIG.
It is designed to be maximum (open loop) between 40 and 2000 Hz as indicated by 0. This loop gain is P _(s) =
(CBDEMA), where DM is the transfer function of the electrical signal output of microphone 11 to the electrical input to driver 17, and A, B, C, E, D and M are microphone preamplifier 35, power 9 shows transfer characteristics of the amplifier 32, the compressor circuit 31, the compensation circuit 31A, the driver 17, and the microphone 11. This loop gain is maximized in consideration of the phase margin and amplitude margin sufficient to ensure stability under different conditions, including differences in the user's head and removal from the head. The closed loop transfer function P ₀ from electrical input to pressure output is P ₀ / V _I = T _u = CBDE / (1 + CBDEMA) at the entrance to the ear canal. The amplitude of this closed loop transfer function as a function of frequency corresponds to curve 21 in FIG. The active noise reduction amount when P _I corresponds to the acoustic noise input is (P ₀ / P _I ) = N _R = 1 + CBDEMA = 1 + T _(s) . Referring to FIG. 3, a curve 23 representing the actual noise reduction measured by the microphone simulated in the middle ear.
Is shown in comparison to a theoretical curve 24 obtained by adding one to the open loop gain. Referring to FIG. 5, there is shown a block diagram illustrating the logic configuration of the preferred embodiment of the present invention. For convenience, it is represented by six blocks indicated by numerals 1 to 6, the compensating circuit block 2 is further divided into sub-blocks 2a, 2b and 2c, the power amplifier block 3 is divided into sub-blocks 3a and 3b, and the compressor block 5 is further divided into 5 blocks. Into five sub-blocks 5a, 5b, 5c, 5d and 5e.
The circuit sections forming the block of FIG. 5 are indicated by the dashed boundaries in FIG. One of ordinary skill in the art will be able to practice the present invention by assembling the circuit of FIG. 6 and will not require too much detail. Next, an embodiment of the present invention will be described with reference to FIGS. The adder / multiplier 1 has an adder 30 which receives an input audio signal from an input terminal 24 and a feedback signal supplied from an amplifier 35. The adder 30 is realized by an operational amplifier connected in the form of a usual inverting and adding amplifier, as shown in FIG. A capacitor connected between the connection point of a pair of similar input resistors and system ground shunts high frequency signal components to system ground. Half of the analog switch U304 transmits the remaining low frequency components to a virtual ground at the input of the integrated circuit U102. The modulation signal on the MOD line from the compressor block 5 turns the switch on and off at a rate of 50 KHz. Multiplication is performed by controlling the switch closing time in each cycle. For signals characterized by spectral components of bandwidth much less than 50 KHz, the switch can be thought of as an impedance whose magnitude is proportional to the duty cycle of the modulated signal waveform. In normal operation, the switch is closed most of the time. When the compressor 5 detects a very large input amplitude, the duty cycle is reduced, attenuating the low frequency spectral components of the input signal. The series connected resistors and capacitors transmit the high frequency signal components directly to the input of U102 so that they are not affected by the multiplication operation. The output of the adder / multiplier block 1 is sent to a compensation block 2, which is characterized by amplitude and phase characteristics which ensure the stability of the feedback loop without significantly reducing the overall loop gain. Active filters. Section 2c provides gain with adequate roll-off at high frequencies. Sections 2a and 2b compensate for the phase response of the loop gain at the low and high frequency crossover points, respectively (specifically, advance or delay the phase). The principle of this preferred form of compensation will be described later, while maintaining stability to prevent oscillations while providing high gain in the frequency band containing most voice spectral components. Power amplifier block 3 receives the output signal from compensation block 2. Section 3b is a well-known non-inverting amplifier with a separate output current buffer. Section 3a is a simple diode limiter that protects the catastrophic power consumption level from being tripped. The light emitting diode emits light when a limiter operation is occurring. Preferably the input is AC coupled to remove the DC offset from the previous stage. The driver / microphone / ear system block 6 is not shown in FIG. Driver 17 receives the amplified signal from power amplifier block 3 and generates an acoustic signal that is sensed by ear 16 and converted by microphone 11. Incomplete sealing of cushion 15 causes a low frequency roll-off. A composite structure that produces multiple resonances at frequencies above a few KHz is also a feature of block 6. In addition, excessive phase shift occurs due to the propagation delay from driver 17 to microphone 11 and the distributed source characteristics of driver 17. However, the components of the system work together to compensate for these non-uniform characteristics, resulting in a substantially uniform closed loop frequency response between input 24 and ear canal 18. The microphone preamplifier block 4 is a low noise operational amplifier that receives the converted signal from the microphone 11 and is connected to have a non-inverting gain. The amplifier and gain are chosen so that the microphone's self-noise predominates, thereby minimizing the effect of the system electronics on the noise level at the ear 16.
The Zener diode is an electret microphone 11
_Is supplied with a bias voltage V _CC . Amplifier 35 of adder / multiplier block 1 receives the amplified signal provided by microphone preamplifier 4. Compressor block 5 monitors both the signal at input terminal 24 and the feedback signal at the output of microphone preamplifier 4, and provides a modulating signal on the MOD line which modulates the low frequency gain of adder / multiplier block 1. . Section 5a sums the input signal with both the left and right channel feedback signals. A low-pass filter having a corner frequency (typically 400 Hz) selectively transmits the combined signal for full-wave rectification. Category 5b
Averages the rectified signal with a fast attack time and a slow fall time and provides an output signal proportional to the low frequency spectral energy of both the left and right loops. Section 5c converts the latter signal into a proportional current having an offset, the gain and offset of which are controlled by potentiometers. Section 5d receives the output current signal from section 5c and provides a 50 KHz modulation signal on the MOD line. The sixth integrated circuit U305 includes:
It consists of a 50 KHz clock pulse source that triggers the integrated circuit U306 and resets its output to ground every 20 microseconds. The capacitor voltage on pin 2 of integrated circuit U306 drops linearly until a threshold level is reached at a rate proportional to the output current provided by section 5c, switching the output switch of integrated circuit U306 high and the capacitor at terminal 2 Reset voltage to positive supply voltage until triggered again. Since the analog switch of adder / multiplier block 1 is closed to the ground potential of control pins 1 and 8, the adder / multiplier gain for low frequencies is inversely proportional to the current level provided by section 5c. The large current causes the capacitor potential on pin 2 of integrated circuit U306 to quickly reach the threshold level, and analog switch U304 is correspondingly closed in a short time. Section 5e drives an LED bar graph display that indicates the amount of compression.
Because the low frequency spectral components carry most of the typical input signal, it is sufficient to sense the low frequencies and act accordingly. In summary, the system can be considered as a servo system with two input signals. The first input is the audio electrical signal to be reproduced. The second input is the acoustic noise signal around the ear. The output of the system is an acoustic signal generated in the ear. The feedback signal is a voltage proportional to the instantaneous sound pressure at the entrance to the ear canal. This sound pressure is a combination of the sound provided by the driver and the surrounding acoustic noise. A small electret microphone 11 converts this signal, a preamplifier block 1 amplifies it, and an adder / multiplier block 1 adds this feedback signal to the input audio signal to provide the actual sound pressure to the ear. And an error signal representing the difference between the signal and the desired sound pressure. The desired sound pressure is proportional to the input audio signal. Compensation block 2 selectively transmits the spectral components of the error signal to ensure loop stability. Amplifier block 3 amplifies the compensated signal and sends the signal to driver 17 to generate a sound pressure at the ear corresponding to the desired audio input signal. Thus, over the frequency range in which the feedback loop is active, the loop corrects for the spectral distribution of the driver / microphone / ear system and cancels out ambient noise. The amount of correction is related to the amount of stable gain that the loop can provide. Compressor block 5 cooperates with the multiplier portion of adder / multiplier block 1 to prevent the input audio or audio signal from driving too much of the loop and clipping. Having described the entire system, the subsystem and its features will now be described. Compressor block 5 is implemented in a specially advantageous manner to reduce artifacts during compression and to avoid non-linear oscillations. Conventional compressors are typically categorized into basic types, n-to-1 compression and threshold compression. The n-to-1 compressor produces a 1 dB change in output level for each ndB input level. Threshold compression is linear for input signals below a certain threshold, and above this threshold results in an n-to-1 compression ratio, which may be infinite above the threshold level, limiting the average output level. The N-to-1 compressor compresses a signal transmitted through a noisy communication channel or a signal to be recorded on a noisy medium, then expands the compressed signal after detection to reduce the recording or communication channel noise after expansion. It is widely used in companders that return to the first dynamic range with significant attenuation. The threshold compressor is n to 1 if properly designed
It is commonly used in systems where the signal is not later decompressed because there are less unwanted products in the output signal than the compressor. Threshold compressors having an infinite compression ratio above the threshold are typically created with a feedback loop. If the compressor gain and compressor attack and decay time constants are not carefully selected, significantly unwanted audible sounds can be generated, especially for input levels just above the threshold, and the system can oscillate, This produces an unpleasant audible sound. The present invention improves upon a conventional threshold compressor that has an infinite compression ratio above the threshold. FIG. 7 is a block diagram showing the logical configuration of the device implemented in block 5. This device is responsive to an input signal X at terminal 51 and provides a compressed signal Y at terminal 52. The dividend input of the divider 53 is connected to the input terminal 51. Input terminal 51 is also connected to the input of full wave rectifier 54. The output of full wave rectifier 54 is connected to the input of averaging low pass filter 55, which is characterized by a drop response time constant that is much larger than the attack response time constant. The output (X) of the averaging low-pass filter 55 is an amplifier 56 having a compressor gain K.
Connected to the input of The adder 57 outputs a signal to one of its inputs.
Upon receiving K ₀ , a divisor signal is given to the divisor input of the divider 53. The divider 53 is an output amplifier that supplies a compressed output signal Y.
Give the quotient signal x / a to 58 inputs. It can be seen that the input-output gain is Y / X = 1 / (K ₀ + K (X)) for a static signal, for example, a sine wave. FIG. 8 is a graph showing this compression characteristic. This characteristic is that Y / X = 1 / K ₀ = constant for small signals and Y / X = 1 / KX or Y = 1 / K for large signals. Is similar to the characteristics of a threshold compressor having However, the compressor according to the invention has at least two advantages over conventional compressors. Since the transition from non-compression to complete compression is performed smoothly, there is little audible compression generated sound for a complex input signal, for example, music. Further, since there is no feedback, non-linear oscillation cannot occur. Here, a preferred compression mode will be described. Gain A
_If the attenuation of _(ω) is known for the entire frequency range, the phase φ _(ω) for the minimum phase network is uniquely determined, and similarly, φ _(ω) is known for the entire frequency range. When A _(ω) is present, A _(ω) is uniquely determined for a “minimum phase” function with no poles and zeros in the right half of the s or p plane. This property, already known, is
ITRADIATION LABORATORY SERIES vol.25 `` Theory of
Servomechanisms § 4.9 ATTENUATION-PHASE RELAT
IONSHIPS FOR SERVO TRANSFER FUNCTIONS ". This relationship was initially based on the Journal of Mathematics a
nd Physics, reported in YWLee's dissertation in June 1932, using Bode's "NETWORK ANALYSIS AND FEEDBACK AMPL
IFIER DESIGN "(D. Van Nostr and Co. New York, 19
45) “Relations between Real and Imaginar” in Chapter XIV
y Components of Network Functions ". §4.8 of the above “Theory of Servome chanisms”
States that an analysis of the "attenuation-phase" type is described as the most satisfactory approach to the servo design problem, and that the criterion for phase margin at the feedback cutoff frequency is at least 30 for a good practical criterion for system stability. ° desirably should be at least 45 °. Relative octave per 6db at least cut-off, and the this section, the frequency gain A is 1 (log A = 0) should take a sufficient phase margin in the least 2 _^1/2-octave than cut-off frequency Described. The purpose of providing a phase margin is to avoid situations where undesired oscillation can continue, and to eliminate peaking that amplifies external noise. The disadvantage of providing such a large area between the frequency _c with unity gain and the corner frequency is that the desired effect of negative feedback on spectral components within that frequency range is significantly reduced. The present invention
The drawback is to combine the network with a high open loop gain while maintaining stability in the relevant frequency band to provide the proper phase margin at unity gain frequency _c . Establish both attenuation or gain characteristics and phase characteristics. The present invention establishes a stable phase margin at frequency _c , while simultaneously providing an open loop gain or damped frequency response with any slope at the end or ends of the passband where the gain falls to zero. Are characterized. These principles are better understood with the examples given below. Referring to FIGS. 9A and 9B, there are shown graphs of gain or attenuation and phase characteristics of an operational amplifier having normal primary characteristics. The gain is constant up to half-power or breakpoint frequency ω ₀ , and then drops linearly at 6 db per octave. Referring to FIGS.10A and 10B, the attenuation and phase characteristics shown in FIGS.9A and 9B are modified by applying the principles of the present invention to provide a larger loop loop at or above _zero while maintaining approximately the same phase margin. A graph for achieving gain is shown.
This is accomplished by adding a dashed line to the break at ₁ above ₀ and providing a 12db slope per octave to frequency ₂ where a gentle slope of 6db per octave begins. The modified phase characteristic shown by the dashed line in FIG. 10B has a phase margin of approximately ^π / ₂ . Referring to FIG. 11, there is shown a modified bandpass response with a dashed line representing a gain or attenuation modification for a more conventional approach having a slope of 6 db per octave on either side of the transmission band. These compensation circuits have an appropriate phase margin and a critical frequency with a gain of 1.
It has a slope closer to the corner frequency with a larger amplitude than the slope at a frequency close to _c . The compensation circuit of block 2 in FIGS. 5 and 6 implements these principles. It can be seen that the loop compensation follows the guidelines from the open loop gain curve. In FIG. 4, the slope after the corner frequency (500 Hz) is 18 db / octave. The network of section 2a has a zero at 80 Hz, a pole at 92 Hz, and damping constants of 0.56 and 1.1, respectively. The high frequency circuit of category 2b has a zero point at 3.1KHz.
It has a pole at 7.3 KHz with damping constants of 0.49 and 1, respectively. The circuit and low-pass filter of Category 2c are 1.6KHz and
It has zeros at 3.4 KHz with poles at 160 Hz, 320 Hz, 800 Hz and 34 KHz. Although the present invention has been described with reference to the embodiments, it will be apparent to those skilled in the art that many changes and modifications can be made within the scope of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a cross-sectional view taken along line 1-1 of FIG. 1A with a headphone attached to the ear, which allows the feedback control system according to the present invention to be used effectively. FIG. 1A is a plan view of the head horn viewed from the ear. FIG. 2 is a block diagram showing a logical configuration of the servo system according to the present invention. FIG. 3 is a graph showing the noise reduction measurements achieved by the present invention as compared to theoretical noise reduction. FIG. 4 is a graph showing the open loop gain and the closed loop gain of the servo system on a common frequency scale. FIG. 5 is a block diagram showing the logical configuration of the preferred embodiment of the present invention. FIG. 6 is a circuit diagram of an electric circuit for executing the block diagram of FIG. FIG. 7 is a block diagram showing a preferred logical configuration of the compressor. FIG. 8 shows the compression characteristics of the compressor shown in FIG. 9A and 9B show gain and phase characteristics of a normal operational amplifier having a first-order characteristic. 10A and 10B show the characteristics shown in FIGS. 9A and 9B modified according to the present invention. FIG. 11 shows a bandpass response modified in accordance with the present invention. (Description of symbols) 11: microphone, 12: cavity 13: housing, 14: diaphragm 15: cushion, 16: outer ear 17: driver, 30: signal coupler 31: compressor, 31A: compensation circuit 32: power amplifier, 35: Microphone preamplifier

   ────────────────────────────────────────────────── ─── Continuation of front page    (56) References JP-A-55-39460 (JP, A)                 Actual opening 1979-109513 (JP, U)                 Jikken 44-31054 (JP, Y1)

Claims (1)

(57) [Claims] Amplifying means having an input and an output for amplifying an input signal applied to the input, wherein the amplifying means is arranged in a closed loop by feedback means coupling the output to the input, wherein the amplifying means comprises When the feedback path between the
A feedback control system for providing a controlled output signal in response to an input signal, characterized by loop gain and phase margin, wherein a passband of a predetermined low frequency range substantially bounded at least at one end by a corner frequency. And compensating means for establishing an open loop gain of the amplifying means in the predetermined low frequency range to a substantially uniform high level as compared to gains outside the predetermined low frequency range. The compensating means compensates the phase response in a region outside the predetermined frequency range while maintaining a phase margin sufficient to maintain the stability of the region, and adjusts the frequency of a portion close to the corner frequency. The slope of the response characteristic is set to 12 dB or more per octave, and the open loop gain is calculated from the corner frequency. Establishes a frequency characteristic that reduces the response to substantially one critical frequency, thereby providing high gain in the predetermined frequency range, avoiding oscillations, and controlling the controlled output signal and the input signal. A system characterized by minimizing an error with a desired controlled output signal dictated by: 2. The system of claim 1, wherein said phase margin is at least π / 6. 3. The amplifying means, wherein the slope of the open loop gain from the corner frequency to a first frequency between the corner frequency and the critical frequency is greater than 6 decibels per octave; The slope up to the second frequency above the critical frequency is 12 per octave.
3. The method of claim 1 or 2, further comprising means for setting a change in frequency response such that the slope of frequencies below decibels and higher than the second frequency is substantially zero decibels per octave. System. 4. 4. The system of claim 3 wherein the slope between the corner frequency and the first frequency is at least 12 dB per octave. 5. 4. The system of claim 3, wherein a slope between the first frequency and the second frequency is not greater than 6 decibels per octave. 6. 6. The system of claim 5, wherein the slope between the first and second frequencies is substantially 6 decibels per octave.