HU231579B1 - Second-order gradient loudspeaker, second-order gradient line array and surface array formed therefrom, and second-order gradient loudspeaker system - Google Patents

Description

180°

128669-17035B SZT

Second-order gradient loudspeaker, and SECOND-ORDER GRADIENT LINEAR RADIATOR AND SURFACE RADIATOR formed therefrom, ALSO SECOND-ORDER GRADIENT SOUND RADIATOR SYSTEM

The invention relates to a second-order gradient loudspeaker, which comprises at least two first-order passive gradient - e.g. cardioid - loudspeakers, the two first-order gradient loudspeakers being arranged at a given distance from each other, essentially behind each other, and furthermore, there is a gap of a given width between the two first-order gradient loudspeakers, the volume of which constitutes an acoustic capacitance Ck. The invention also relates to a second-order gradient line radiator and surface radiator formed from such second-order gradient loudspeakers, and to a second-order gradient loudspeaker system obtained as a combination of such second-order gradient loudspeakers and a power amplifier.

Here and hereinafter, the term “loudspeaker” refers to a single electroacoustic transducer, i.e. one without additional elements, which converts electrical energy into acoustic energy, i.e. generates sound waves in the air. The term “loudspeaker” is used for a device obtained by assembling one or more loudspeakers with one or more mechanical and/or acoustic and/or electrical elements (e.g. sound box, horn, sound guide, electronic elements, etc.) that improve or modify sound radiation.

Ensuring adequate speech intelligibility is a serious challenge when sounding spaces with high reverberation, typically train stations, large-volume halls, sports facilities (swimming pools, sports halls), road and railway tunnels, airports, etc. Informing those present is vital, but this can only be done with the help of a public address system that is capable of producing adequate speech intelligibility. Speech intelligibility is of particular importance in emergency situations, where announcements are often the only way to effectively convey information. In open spaces, it is similarly important to ensure sufficient directionality in order to minimize environmental noise pollution. In order for information to be communicated intelligibly in these situations and locations, it is necessary to increase the so-called directionality of loudspeaker systems. Echo is also a major issue when it comes to music sound reinforcement, especially if the music contains vocals and a lot of bass. Bass reflections can cause particularly unpleasant subjective effects.

GO

- 2 Speech intelligibility is now measured numerically, with a value between 0 and 1 provided by the so-called STI (Speech Transmission Index) [see IEC 60268-16:2011 Sound system equipment - Part 16: Objective rating of speech intelligibility by speech transmission index]. The value of the STI is related to the so-called critical distance, also known as the Hall radius. Generally, speech intelligibility is adequate within the critical distance. In a reverberant room, the square of the critical distance is directly proportional to the loudspeaker control factor Q and inversely proportional to the reverberation time of the room [see K.B.Ginn: Application of B&K Equipment to Room Acoustics; 1978, Bruel, Naerum Offset, p.29 and H.Kuttruff: Room Acoustics 1979, Applied Science Publishers LTD, London, X.2]. Consequently, for good speech intelligibility, a loudspeaker with a higher Q control factor should be used, the longer the reverberation time of the given room.

The directivity curve of a standard loudspeaker is nearly circular at low frequencies in both the horizontal and vertical directions, so its directivity factor Q ~ 1 exists [see L.L.Beranek: Acoustics. 1986 Edition, Cambridge, USA, pages 111,114]. In reverberant rooms, i.e. with a reverberation time of 4-10 seconds, they are practically useless for sound reinforcement tasks. The reason for this is that the direct sound of the loudspeaker is reduced so much at a distance of a few meters compared to the so-called diffuse sound representing the reverberation that the direct sound carrying the information is suppressed by the diffuse sound, therefore the speech can only be heard but not understood. As is known, speech intelligibility can be increased by increasing the directivity factor Q of the loudspeakers operating in the room, or by installing sound absorbers in the room, thereby reducing the reverberation time. However, the latter is a costly procedure and cannot be implemented in historic buildings and churches.

One way to increase the Q control factor is to use a first-order gradient loudspeaker.

The term “gradient” is used in physics to refer to the pressure gradient. The general form of ideal, theoretical gradient polar patterns can be written as (1+ a cos Θ) / (1+ a), where a defines the type of gradient polar pattern, ranging from cardioid (a = 1) to “8” (bidirectional) (a—>») [see L.L. Beranek: Acoustics. 1986 Edition, Cambridge, USA, p. 181], while Θ is the angle between the gradient speaker axis and the radiation direction. The gradient can be generated by active or passive electronic elements or by passive

GO

- 3 acoustic elements. Accordingly, here and hereinafter we will speak of active gradient formation and active gradient loudspeaker in the case of electronic elements, or passive gradient formation and passive gradient loudspeaker in the case of passive acoustic elements, in accordance with the terminology accepted in the art.

As is known from the literature, the characteristics of gradient formation are often examined as a function of frequency. Accordingly, here and in the following, the frequency variable ‘f’ is represented by the variable (kd), which is a variable in the

It is proportional to the frequency according to the relationship 2nfd (kd) = k · d =---c; here k is the wavenumber, k = ^j- f is the frequency, c is the speed of sound in air, and d is a reference distance, in our case e.g. the distance between two sound gates.

The gradient method - although limited in its application due to frequency response and directional characteristics - increases both horizontal and vertical directivity and provides cardioid, hypercardioid or “8” shaped directional characteristics in both planes (with a maximum value of Q = 3 - 4). In practice, in order to simplify the problems and difficulties, electronic elements, i.e. active phase shifters, are used for gradient formation, which means that the first-order gradient formation is implemented with two loudspeakers, two amplifiers and active signal shaping circuits [see e.g. the Geo D System of Nexo SA and the Axys and subwoofer product brochures of Duran Audio BV]. Here and further on, such an arrangement will be referred to as an active first-order gradient loudspeaker.

Another way to increase the directivity factor Q is to use so-called line radiators. This does not change the horizontal directivity, but the vertical directivity can be increased depending on the length of the line radiator.

A first-order passive hypercardioid line and surface loudspeaker can be found in the Hungarian patent application P1200426 a.sz., published, primarily for the sound reinforcement of large spaces with good speech intelligibility. According to the solution, the first-order passive cardioid loudspeakers are designed in a new, optimal way, taking into account the effect of the dimensions of the loudspeaker housing, and are then arranged in a line above each other. The purpose of the presented solution is to increase the intelligibility by

GO

-4Improving the vertical beamwidth of sound radiation, which reduces diffuse, secondary sound propagation, which is primarily responsible for the deterioration of speech intelligibility in reverberant, high reverberation spaces. In the considered solution, it is recommended to design a surface radiator with the presented line radiators in order to further enhance the directivity.

US 2010/0220871 A1 discloses a loudspeaker in which two loudspeakers separated by a space of “L” are arranged side by side. The loudspeakers are conventional closed-box loudspeakers, i.e. they are not of the gradient type. The loudspeakers are driven from a common source in such a way that the phase position of the drive signals arriving at the loudspeaker terminals is different. The phase difference is ±90° in the band around 1000 Hz, which is ensured by filters placed in the path of the drive signal. The phase difference is essentially constant in a relatively wide band, and is produced by a filter placed in the path of the drive signal (so-called all-pass filter). The time delay created by the acoustic and electrical paths, and the difference between them, affects the resulting acoustic radiation characteristic. By combining these appropriately, horizontal and/or vertical directivity can be improved.

The patent document HU-220,406 B, as a document designating the closest prior art, discusses a highly directional loudspeaker, one embodiment of which comprises two first-order passive cardioid loudspeakers arranged one behind the other at a given distance. The two loudspeakers are connected directly to the source in antiphase with respect to each other. In another embodiment, the front first-order passive cardioid loudspeaker is connected directly to the signal source in a passive manner, while the rear first-order passive cardioid loudspeaker is connected in antiphase to the signal source via a delay unit. As a result of this choice of acoustic arrangement and supply, a strong beam (second-order hypercardioid directional characteristic) occurs in a relatively wide frequency band. The directional effect can be further improved by organizing it into a sound column.

In light of the above, our aim with the invention is to increase the directivity of loudspeakers in both the horizontal and vertical planes. For this, we found the second-order gradient principle to be the most suitable, which can be achieved by forming a double gradient. Second-order gradient loudspeakers increase directivity in both the horizontal and vertical planes, and also co

- 5 provide second-order cardioid, hypercardioid or “8” polar patterns in both planes (with a maximum Q factor of 5 - 8).

In light of this, our aim with the invention is to eliminate or at least alleviate the application problems of the gradient principle so that the loudspeaker created using it remains as simple in structure as possible. Our further aim with the invention is to create a line or surface loudspeaker with second-order gradient loudspeakers, thereby further increasing the Q control factor of the loudspeaker at medium and low frequencies, which in this case generally means frequencies below 2000 Hz. This requires the satisfaction of specific geometric and acoustic conditions.

The second-order gradient principle increases the difficulties of using first-order gradient loudspeakers due to the double gradient formation. The most unpleasant phenomenon of these is that when moving towards higher frequencies, for example, the cardioid directivity curve is significantly distorted, and the frequency response drops sharply towards low frequencies, with a slope of essentially 12 dB/octave, which results in a narrow transmission band [see, for example, H.F. Olson: Gradient Loudspeakers, JAES Vol. 21., 1973. Marz. 83-93]. In practice, in order to simplify the problems and difficulties, electronic, i.e. active phase shifters are used for both gradient formations, which means that the known second-order gradient loudspeakers contain active first-order gradient loudspeakers. That is, four loudspeakers, four amplifiers and active circuits are used. The four or more loudspeakers, which are necessarily placed close to each other, interact with each other, and their operation is significantly disturbed by the diffraction sound reflection caused by each other and/or the effect of at least three narrow, nearly closed spaces/space parts (cavities) formed between them. Furthermore, such a solution is expensive due to the above.

The invention is based on several findings. If the first-order gradient formation is implemented with passive acoustic elements and a passive acoustic phase shifter, the implementation of the second-order gradient loudspeaker becomes simpler and thus less expensive, since in this way only two loudspeakers and at most two amplifiers need to be used. In the case of such a design, therefore, fewer loudspeakers are present, so there is less adverse interaction between them, as a result of which only one narrow, nearly closed space (i.e. a cavity forming an acoustic capacitance) and/or one diffraction-induced interaction problem needs to be solved/handled. This cost

GO

- 6 The effect further narrows the transmission band and is thus detrimental. During our investigations, we have recognized how it is possible to turn this detrimental interaction into a beneficial one and thereby expand the transmission band. This is a particularly important realization, since gradient formation, especially second-order gradient formation, can only be used in a narrow transmission band. We have also recognized that if the delay time required in the second gradient formation is reduced at increasing frequencies, the width of the transmission band can be increased.

The objects of the invention are achieved by a second-order gradient loudspeaker according to claim 1. Possible further preferred exemplary embodiments of the second-order gradient loudspeaker according to the invention are defined in claims 2-16. A preferred option for utilizing the invention is a second-order gradient loudspeaker system according to claim 17 or 18, a second-order gradient line radiator according to claim 19 and a second-order gradient surface radiator according to claim 20.

The most important advantage of the second-order gradient loudspeaker according to the invention is that its directivity Q factor at any frequency in the 30-1000 Hz frequency range significantly exceeds the directivity Q factor of conventional loudspeakers, as well as conventional line and surface loudspeakers. This improves speech intelligibility - especially in reverberant rooms - and significantly increases the area from which the voiced speech can be heard. The directivity Q factor of the second-order gradient loudspeaker according to the invention is high precisely in the 30-1000 Hz frequency band, where the room generally has the longest reverberation time and thus the increased directivity factor is most needed. In outdoor spaces, the increased directivity of the loudspeakers according to the invention reduces the sound, i.e. noise pollution of the environment. Another advantage of the second-order gradient loudspeaker according to the invention is that it can be produced at a relatively low cost.

Specifically, the two first-order passive gradient loudspeakers used to implement the second-order gradient loudspeakers of the invention each have a cardioid polar pattern, and the delay time τ

- essentially d/c in magnitude, i.e. the polar pattern of a second-order gradient loudspeaker is second-order cardioid, or

- the polar pattern of the second-order gradient loudspeaker is the most strongly directed second-order hypercardioid, with a maximum directivity factor of Q=8,

- 7 where d is the distance between two first-order gradient loudspeakers arranged one behind the other, and c is the speed of sound measured in air at a given temperature.

In the following, we present some embodiments and applications of the second-order gradient (e.g. second-order cardioid) loudspeaker that is the subject of the invention, as examples, and we describe its operation and properties in more detail based on the attached drawing, where the

- Figure 1 shows a known first-order cardioid loudspeaker; the

- Figure 2 shows the electrical equivalent of the first-order cardioid loudspeaker of Figure 1;

- Figure 3 shows the transfer function, i.e. the frequency response, of a first-order (i.e. single) gradient formation providing a cardioid characteristic curve Ai(kd,O) at different angles Θ; - Figure 4 shows a series of directional characteristic curves of a first-order (i.e. single) gradient formation providing a cardioid characteristic curve;

- Figure 5 shows the Q control factor of a gradient formation providing a cardioid characteristic curve as a function of (kd), i.e. it depicts the frequency dependence of the Q control factor;

- Figure 6 shows a cross-section and side view of a second-order gradient loudspeaker according to the invention;

- Figure 7 shows the directional characteristic curve of a second-order cardioid characteristic curve distorted due to diffraction;

- Figure 8 shows a block diagram of an embodiment of a second-order gradient loudspeaker system according to the invention, implemented with a second-order gradient loudspeaker, a power amplifier and passive filters;

- Figure 9 shows a block diagram of an embodiment of a second-order gradient loudspeaker system according to the invention, implemented with a second-order gradient loudspeaker, power amplifiers and active filters;

- Figure 10 shows a second-order gradient line array formed by stacking second-order gradient speakers according to the invention; and

- Figure 11 shows a second-order gradient surface radiator formed by placing second-order gradient line radiators according to the invention next to each other.

Figure 1 shows a known first-order passive cardioid loudspeaker.

The single gradient formation is done with acoustic elements. In the cavity with acoustic capacity Co behind the LS loudspeaker, the box resonances are damped by the sound-absorbing material E,

GO

- 8 which is a fibrous material (rock wool or glass wool) and/or foamed plastic. The acoustic impedance Zi(Ri,Mi) consists of, or may consist of, the acoustic resistance Rí and the acoustic mass Mi components, as shown in the figure, where the acoustic resistance R1 is usually dominant. The acoustic resistance is protected from mechanical effects by a protective cover P, which is usually a perforated plate.

Figure 2 shows the electrical equivalent of the first-order cardioid loudspeaker of Figure 1.

The LS loudspeaker is represented by the acoustic impedance Zm(Rm,Mm,Cm), which is composed of the acoustic mass Mm and the acoustic capacitance Cm, and the acoustic resistance Rm. The gradient formation is provided by a network of passive acoustic elements (phase shifter, phase rotator), namely the acoustic mass Mi, the acoustic resistance Rí, and the acoustic capacitance Co.

Figure 3 shows the transfer function, i.e. the frequency response, of a (simple) gradient formation providing a cardioid characteristic curve Ai(kd,O) as a function of the parameter (kd) (“frequency”) at different angles O. If O denotes the angle between the loudspeaker axis and the direction under investigation, according to the literature, the general shape of ideal, theoretical cardioid directional characteristics can be written as (1+cos O)/2.

The upper limit of the cardioid gradient formation is clearly visible at (kd)t = π for O=0° (0 radians). This (kd)t breakpoint, or upper limit, depends on the type of the directional characteristic curve, i.e. the value of α (varying between π and 2π). There is a discontinuity in the transmission here, in the figure Ai(n,0) = 0.

Based on the above, the breakpoint frequency ft can be calculated from the cardioid gradient (kd)t = π breakpoint, i.e.:

2nf _t d (kd) _t = π =----cc

ft = "2d"

It also follows from this that to avoid distortions, the operating frequency limit must be chosen to be lower than the upper limit value formed by the breakpoint frequency ft. It is advisable to choose the operating limit frequency to be (kd)t/1.5, i.e. here (kd) ~ 2 is fulfilled.

Substituting this and then arranging the relations, we get that

GO

- 9 1 ^T< 2f _t

Here we have taken into account that the condition for the ideal theoretical cardioid polar pattern is that the d/c time constant is equal to the delay time τ between the two sound gates [see e.g. L.L.Beranek: Acoustics. 1986 Edition, Cambridge, USA, page 150].

Figure 4 shows a series of directivity curves of a first-order (i.e., single) gradient formation providing a cardioid characteristic curve. The transfer function Ai(kd,O) shows a series of relative directivity curves Airel(kd,O) for different (kd) parameters. It can be seen that in the case of (kd)=2 the cardioid characteristic curve is distorted, but still acceptable. In the case of (kd)=1.5 the cardioid characteristic curve is almost flawless.

Figure 5 shows the Q control factor of a gradient formation providing a cardioid characteristic curve as a function of (kd), i.e. it depicts the frequency response of the Q control factor. The figure shows that for (kd)=2, Q ~ 1.5, for (kd)=1.5, Q ~ 2. The values show the distortion of the directivity curve, since for the ideal cardioid, Q=3.

Figure 6 shows a cross-section and side view of a second-order gradient loudspeaker according to the invention. The figure shows a second-order gradient loudspeaker according to the invention. The second-order gradient loudspeaker is formed by two first-order passive gradient loudspeakers 1, 2. The front first-order passive cardioid loudspeaker 1 has an LSi loudspeaker and an acoustic impedance Zi(Ri,Mi), and the rear first-order passive cardioid loudspeaker 2 has an LS2 loudspeaker and an acoustic impedance Z2(R2,M2). In the cavity with an acoustic capacity Co behind the loudspeakers LS1, LS2, the damping of the box resonances is ensured by a sound-absorbing material E. The distance d is between the first-order passive gradient loudspeakers 1, 2, while a gap (cavity) of width h is formed between the two loudspeakers LS1, LS2, which forms an acoustic capacity Ck. According to the invention, this cavity is preferably filled in part or in whole with a sound-absorbing material E. The sound-absorbing material E may be a fibrous material (rock wool or glass wool) or foamed plastic. The protective cover P is covered from the inside with an acoustic resistance material T, preferably a non-woven textile material (e.g. vetex), where the sound-absorbing material E and the acoustic resistance material T together form an acoustic impedance ZK(RK, MK), where ZK is the acoustic co

- 10 impedances are formed by acoustic resistance Rk and acoustic mass Mk components. The first-order passive gradient loudspeakers 1, 2 are driven by an audio frequency electrical signal modified through at least one element selected from the group consisting of an APF all-pass filter, an LPF low-pass filter and a DL delayer in such a way that the rear passive gradient loudspeaker 2 is in antiphase with respect to the front passive gradient loudspeaker 1, and its driving electrical signal is delayed by a delay time τ in total with respect to the driving electrical signal of the front passive gradient loudspeaker 1, where τ is d

< τ <--a c

Figure 7 shows the directional characteristic curve of a second-order cardioid characteristic curve distorted by the cutoff frequency of gradient formation and by diffraction.

The figure shows measurement results. The second-order cardioid directional characteristics are distorted not only by the breakpoint frequency f _t of the gradient formation, but also by the interaction of the two loudspeakers 1, 2 placed close to each other and by the diffraction effect; see Figure 6, where the two first-order passive gradient loudspeakers 1, 2 are arranged close to each other, behind each other. At the lower frequency shown in Figure 7, (kd)=0.6, an almost ideal second-order cardioid directional characteristic is visible. At a higher frequency, (kd)=2, the directional characteristic is already distorted, partly due to the proximity of the breakpoint frequencies f _t and (kd)t, partly due to the diffraction effect of the first passive gradient loudspeaker, and partly due to the effect of the gap (cavity) of width h (Figure 6).

We have developed a method that reduces this disadvantageous effect. According to this, the gap (cavity) of width h, which forms the acoustic capacitance Ck, is at least partially filled with sound-absorbing material E and surrounded by a protective cover P, where the sound-absorbing material E and the protective cover P together form an acoustic impedance Zk(Rk, Mk), and furthermore the acoustic resistance Rk and the time constant of the acoustic capacitance Ck satisfy the

R^C _K <~— ² · ^f condition.

Figure 8 is a block diagram of an exemplary embodiment of a second-order gradient loudspeaker system according to the invention, implemented using a single amplifier and passive filters,

GO

-11 whose Amp power amplifier receives the audio frequency signal from a Sig. signal source. Both first-order passive gradient loudspeakers 1, 2, i.e. both the front and the rear, receive the electrical signal from the common Amp power amplifier. The second-order APF1 all-pass filter, LPF1 low-pass filter and DL1 delay unit connected in a chain are connected to the LS1 loudspeaker of the first-order passive gradient loudspeaker 1. The second-order APF2 all-pass filter, LPF2 low-pass filter and DL2 delay unit connected in a chain are connected to the LS2 loudspeaker of the first-order passive gradient loudspeaker 2 in anti-phase. The APF1, APF2 all-pass filters, the LPF1, LPF2 low-pass filters and the DL1, DL2 delay units are all implemented with passive electrical elements. Here and further on, as is otherwise obvious to the skilled person, an all-pass filter is understood to mean a filter that passes the signal without any modification in its amplitude, i.e. its frequency response is straight, but changes the phase of the signal passing through it as a function of frequency. The first-order passive gradient loudspeakers 1, 2 are driven by a modified audio frequency electrical signal through the listed elements, where the electrical signal is modified by the individual elements in different ways: the all-pass filters APF1, APF2 in phase and propagation time, the low-pass filters LPF1, LPF2 in phase, propagation time and amplitude, while the delay units DL1, DL2 in propagation time and phase, e.g. to different extents as a function of frequency.

Figure 9 is a block diagram of an exemplary embodiment of a second-order gradient loudspeaker system according to the invention, implemented using two different amplifiers and active filters, in which a signal source Sig. provides the audio frequency signal. Both first-order passive gradient loudspeakers 1, 2, i.e., both front and rear, receive the electrical signal from two different power amplifiers Amp1 and Amp2. The second-order all-pass filter APF1, low-pass filter LPF1 and delay unit DL1 connected in a chain are connected to the input of the power amplifier Amp1, and the output of the power amplifier Amp1 is connected to the speaker LS1 of the first-order passive gradient loudspeaker 1. The second-order all-pass filter APF2, the low-pass filter LPF2 and the delay unit DL2 connected in a chain are connected to the input of the power amplifier Amp2, and the output of the power amplifier Amp2 is connected in anti-phase to the speaker LS2 of the first-order passive gradient 2 loudspeaker. Each of the all-pass filters APF1, APF2 and the low-pass filters LPF1, LPF2, and each of the delay units DL1, DL2 are implemented with digital and/or analog electrical elements, preferably implemented in a DSP digital signal processor, as is obvious to the person skilled in the art.

GO

-12 Figure 10 shows a second-order gradient line array formed by stacking second-order gradient speakers according to the invention.

Figure 10 shows a second-order gradient line array formed from second-order gradient loudspeakers. It is clear that the second-order gradient loudspeaker shown in Figure 6 is multiplied by any number and the units are arranged one above the other. In this case, five second-order gradient loudspeakers are placed one above the other. Thus, the second-order cardioid characteristic curve shown horizontally is enhanced vertically with additional directivity gain, i.e. the directivity factor Q is further increased. We obtain a vertically enhanced, so-called beam-shaped directivity curve.

Figure 11 shows a second-order gradient surface radiator formed by placing second-order gradient line radiators according to the invention side by side.

The second-order gradient speakers shown in Figure 6 are placed not only above each other, but also next to each other, thus forming a second-order gradient surface speaker.

The practical exemplary embodiment of the second-order gradient surface radiator is implemented with second-order gradient line radiators placed side by side at a distance s, as shown in Figure 10. Figure 11 shows an arbitrary number of i (i is a positive integer) line radiators. Thus, in addition to the gradient formation, the column effect also enhances the directivity factor from both directions and narrows the directivity curve, significantly increasing the value of the directivity factor Q.

co

Claims (20)

Patent claims 1. A second-order gradient loudspeaker comprising two first-order gradient loudspeakers (1, 2) arranged one behind the other at a given distance (d), where there is a gap (h) between the two first-order gradient loudspeakers (1, 2) providing an acoustic capacity (Ck), the first-order gradient loudspeakers (1, 2) having a passive design, i.e. they are implemented as an acoustic network consisting of passive acoustic elements having an acoustic mass (M), an acoustic resistance (R) and an acoustic capacity (C), where the acoustic network comprises at least one acoustic phase shifter member providing a passive gradient formation, and further where the first-order passive gradient loudspeakers (1, 2) are driven by a modified audio frequency electrical signal, characterised in that each of the first-order passive gradient loudspeakers (1, 2) is an all-pass filter (APF), a low-pass filter (LPF) and is driven through at least one element selected from the group consisting of delay units (DL) in such a way that the electrical signal of the rear first-order passive gradient loudspeaker (2) is in phase with the front first-order passive gradient loudspeaker (1) and delayed by a delay time (τ) compared to the electrical signal of the front first-order passive loudspeaker (1), where the value of the delay time (τ) is selected from the interval d 0 < τ <--a · c, where c is the speed of sound measured in air at a given temperature; and α is a factor determining the type of the directivity curve of the second-order gradient loudspeaker from cardioid (α =1) to “8”-shaped (bidirectional; α——»). 2. A second-order gradient loudspeaker according to claim 1, characterized in that the first-order passive gradient loudspeakers (1, 2) have a cardioid polar pattern and the delay time (τ) is substantially d/c. 3. A second-order gradient loudspeaker according to claim 1 or 2, characterized in that the delay time (τ) is adjusted to the cutoff frequency ft of the gradient formation. -14 1 ^Γ< 2 ⁷ 7 with the relation. 4. A second-order gradient loudspeaker according to claim 1, characterized in that the first-order passive gradient loudspeakers (1, 2) have a cardioid polar pattern and the delay time (τ) is zero. 5. A second-order gradient loudspeaker according to any one of claims 1-4, characterized in that the acoustic mass (M2) in the acoustic network of the rear first-order passive gradient loudspeaker (2) is greater than the acoustic mass (Mi) in the acoustic network of the front first-order passive gradient loudspeaker (1). 6. A second-order gradient loudspeaker according to any one of claims 1-5, characterized in that the gap between the first-order passive gradient loudspeakers (1, 2) having an acoustic capacitance (Ck) is at least partially filled with a sound-damping material (E). 7. A second-order gradient loudspeaker according to claim 6, characterized in that the sound-absorbing material (E) is at least one of rock wool, glass wool, foamed plastic or a combination thereof. 8. A second-order gradient loudspeaker according to claim 6 or 7, characterized in that the gap between the first-order passive gradient loudspeakers (1, 2) having an acoustic capacitance (Ck) is surrounded by a protective cover (P). 9. A second-order gradient loudspeaker according to claim 8, characterized in that the protective cover (P) is covered from the inside with an acoustic resistance material (T), preferably a non-woven textile material, particularly preferably a vetex material. 10. A second-order gradient loudspeaker according to claim 3 or 4, characterized in that the gap (h) between the first-order passive gradient loudspeakers (1, 2) having an acoustic capacitance (Ck) is at least partially filled with a sound-damping material (E) and surrounded by a protective cover (P), wherein the sound-damping material (E) and the protective cover (P) together with the acoustic resistance material (T) provide the resulting acoustic impedance (Zk(Rk, co - 15 Mk)) and the time constant of the acoustic resistance (Rk) and the acoustic capacitance (Ck) satisfy the following condition: 1 Η _κ ·0 _κ <-— 2· ^f t 11. A second-order gradient loudspeaker according to any one of claims 1-10, characterized in that the delay of the anti-phase audio frequency electrical signal of the rear first-order passive gradient loudspeaker (2) is provided by a time (t) delay unit (DL2). 12. A second-order gradient loudspeaker according to any one of claims 1-10, characterized in that the rear first-order passive gradient loudspeaker (2) is driven by the anti-phase audio frequency electrical signal through at most two second-order low-pass filters (LPF2), and furthermore the front first-order passive gradient loudspeaker (1) is driven directly by the electrical signal, so that the delay time (t) is provided by the phase shift (delay time) of the second-order low-pass filter (LPF2). 13. A second-order gradient loudspeaker according to any one of claims 1-10, characterized in that the rear first-order passive gradient loudspeaker (2) is driven by the anti-phase audio frequency electrical signal through two second-order low-pass filters (LPF2), and the front first-order passive gradient loudspeaker (1) is driven by the electrical signal through a single second-order all-pass filter (APFi), where the filters are sized such that the delay time (t) is provided by the difference between the delay time of the second-order low-pass filter (LPF2) and the delay time of the all-pass filter (APFi). 14. The second-order gradient loudspeaker according to any one of claims 1-10, characterized in that the rear first-order passive gradient loudspeaker (2) is driven by the anti-phase audio frequency electrical signal through at least one element selected from the group consisting of an n-th order low-pass filter (LPF2), an n-th order all-pass filter (APF2) and a delay unit (DL2), and the front first-order passive gradient loudspeaker (1) is driven by at least one element selected from the group consisting of an n-th order low-pass filter (LPFi), an n-th order all-pass filter (APFi) and a delay unit (DLi). GO - 16 are driven by the audio frequency electrical signal, where the filters are sized and the filter parameters are set so that the delay time (τ) is provided by the difference between the delay times of the filters and the delay units 15. A second-order gradient loudspeaker according to any one of claims 1-14, characterized in that the delay time (τ) is frequency dependent. 16. A second-order gradient loudspeaker according to claim 15, characterized in that the delay time (τ) varies with frequency depending on the setting of the filters, preferably the delay time (τ) is smaller at higher frequencies. 17. A second-order gradient loudspeaker system comprising a loudspeaker and a power amplifier (Amp), characterized in that the loudspeaker is a second-order gradient loudspeaker according to any one of claims 1-16, wherein each of the first-order passive gradient loudspeakers (1, 2) is driven by an electrical signal from the output of the power amplifier (Amp), and each of the filters (APF1, LPF1, APF2, LPF2) and the delay units (DL1, DL2) are implemented with passive electrical elements. 18. A second-order gradient loudspeaker system comprising a loudspeaker and two different power amplifiers (Amp1, Amp2), characterized in that the loudspeaker is a second-order gradient loudspeaker according to any one of claims 1-16, wherein the first-order passive gradient loudspeakers (1, 2) are driven by an electrical signal from the output of the two different power amplifiers (Amp1, Amp2), and further, the filters (APF1, LPFi, APF2, LPF2) and the delay units (DL1, DL2) are implemented with passive and/or active electrical elements, and the active electrical elements are formed by digital signal processors (DSPs) in the case of both the filters (APF1, LPFi, APF2, LPF2) and the delay units (DL1, DL2). 19. A second-order gradient line array, characterized in that it comprises at least two second-order gradient loudspeakers according to any one of claims 1-16 arranged one above the other. CO 20. A second-order gradient surface array, characterized in that it comprises at least two second-order gradient line arrays according to claim 19, conveniently arranged next to each other at a given distance (s).