DE2712534A1 - Active sound attenuation using secondary wave - has stored programme used to obtain signal representing secondary wave - Google Patents

Abstract

The active sound attenuation is used to attenuate a primary wave by means of a secondary wave. A transducer is used to obtain a signal representing the primary wave to be attenuated and a second transducer is used to produce the secondary wave from a second signal. This second signal is obtained from the first signal by combining the latter with a stored programme. The latter may be modified in accordance with the required attenuation of the primary wave, as indicated by a third transducer situated after the first two transducers in the direction of propagation of the primary wave.

Description

Method and device for active sound attenuation

The invention relates to a method and an apparatus for active damping of longitudinal pressure and compression waves and relates in particular on the attenuation of sound carried in the air, for example in a canal or other confined space.

As a sound wave from a series of condensations and thinnings exists, it has long been assumed that the energy content of the wave can be reduced by having the primary wave with a specially generated Secondary wave combined in such a way that the dilutions of the secondary wave coincide with the compressions of the primary shaft and vice versa.

This principle, known as active damping, is used for reduction of pressure fluctuations that exist in the medium and thus withdraws from the primary wave Energy. This active method of soundproofing is in great Scope described in U.S. Patent No. 2,043,416, and is in that particular area Much work has been done in recent years.

The secondary wave must be opposite to the primary wave to be canceled by it can be generated exactly, and one has already considerable successes in the active one Has achieved sound attenuation of primary waves of simple sinusoidal shape, is the quality of attenuation achieved so far when the primary wave is a natural one occurring noise is in which the primary wave in terms of amplitude and the frequency varies widely in any time-dependent manner less satisfactory.

The invention therefore relates to a method and an apparatus to generate a secondary wave for active damping of any given Primary wave that has a high degree of damping regardless of the complexity of the primary wave create and provide, in preferred embodiments, a self corrective Facility, the response of the system based on its previous embodiment is modified.

In a first aspect of the invention, the apparatus relates to Attenuation of a primary wave in a system by a specially generated secondary wave a first converter for deriving a first electrical signal that is temporal correlates the primary wave to be attenuated represents, a second converter for Generation of the secondary wave from a second electrical signal, a storage device, in which a program of time-correlated operational steps is included, as well as a device for obinierun the first electrical signal with the program for generating the second electrical signal.

In one embodiment of the device according to the invention, the program the temporal response of the system to a specific change at the input of the first converter characterized and specified, and the steps of the program can be derived be by getting the impulse speech of the system is intended for a variety of different delta functions, soft as Input channels are in contact with the first transducer.

In another embodiment of the device according to the invention can be a cancellation converter downstream in the direction of propagation of the primary wave of the first and second transducers, the output of the cancellation transducer is used to modify the second electrical signal. This rlodisizierunq of the second electrical signal can be made in that the program the operating steps occurring in the combination and / or the amplitude of the second electrical signal.

The storage device can only be analog or digital or a combined analog-digital embodiment, and expediently the first electrical signal is stored, with the combination or convolution is carried out with at least one multiplier between the Storage device for the first signal and the storage device for the characterization of the system is working. If a large number of mutip] ies is available, is it is expedient to assign one of these for the respective stored operating step take, but a single multiplier can also be used, its channels be used repeatedly over time.

According to a further aspect of the invention, a method for Schalldämpfunq in a limited space the procedural steps of deriving a first electrical signal that temporally correlates one entering the room and represents the primary sound wave to be attenuated, and the use of the first electrical signal for deriving a second electrical signal, wherein the second electrical signal in a second transducer to generate a secondary wave is used in the space, which is at least partially the primary wave in the space cancels, and is characterized in that the second electrical signal from the first electrical signal by folding and combining the first electrical Signals is derived with a program of time-correlated operating steps.

The program of time-correlated operating steps can be set in advance be and remain unchanged. The determination of the previously set steps can be based on the response of the system, i.e. the other two, the connecting electrical Components and the limited space, based on characterizing impulses.

Alternatively, an adaptation strategy can be used to Example that of the successive approximation using a further converter in the limited space to ensure the success of the reversal operation certain. The changes in the program of the operating steps can be carried out by Jtand depending on the output signal of a converter in the limited Space, but a fully or semi-automatic correction can also be used.

Is the inventive method for damping sound waves used, which are transnorted by a gas and propagate in a channel, so the first electrical signal can be derived from the output of a microphone, which is located in the channel upstream of a loudspeaker from which the Secondary wave goes out. With such an arrangement, a part of the speaker generated energy cause a feedback through the channel to the microphone, which gives an incorrect representation of the primary wave to be damped. the Elimination of this feedback signal has caused problems and its elimination a number of methods have been given whereby the channel is in the range between the speaker and microphone lined up and a highly oriented microphone is used, which is essentially dusty to the feedback signal.

According to a further aspect of the invention, this can be done by the microphone The feedback signal picked up can be compensated for by using a suitable electrical Signal is subtracted from the microphone, that by a second fold which is the path between the input terminals of the loudspeaker, the line between the loudspeaker and the microphone and the output terminals of the Microphones simulated.

When the first convolution system through the above-mentioned adjustment method is optimized, the first convolution becomes undesirable to some extent Remove effects from the feedback path just described. In this way a reasonable degree of erasure can be achieved by the first convolution, especially in Achieve connection with the above-mentioned passive attenuation and / or the directional microphone.

The invention is intended in the following on the basis of exemplary embodiments and will be explained in more detail with reference to the accompanying drawings.

The drawings show schematic representations in FIGS. 1 to 6 to explain the arrangement used for active sound absorption; Fig. 7 to 8c representations to explain a way to achieve a characterization a system in the method according to the invention and the device for implementation the procedure; and FIGS. 9 to 14 show representations for explaining systems for use in the method according to the invention.

The basic principles of active sound attenuation are discussed below are explained in more detail with reference to FIGS. 1 to 6.

In the illustration according to FIG. 1, a line or a channel is 1 with any cross-section and with a blower or fan 2 recognizable, which is at any point along the length of line 1. Sound, i.e. the primary wave from fan 2 propagates in the direction of the arrow along the Leituna 1 off and tries to energize microphone 3. A loudspeaker 4 reacts in the Way that it tries to prevent the excitation of the rlikrofone and emits thereby a scarf] waveform with antiphase.

The maintenance of this waveform are controlled by the directional properties the loudspeaker system, the type of line 1 and the sound frequencies depend. Of main interest is the sound, which is in the form of the secondary wave to the right in the same direction as the sound propagates from the loudspeaker, because it is in antiphase and spreading in the same direction. If its Size is correct and a plane wavefront is built up quickly, then one becomes Deletion or cancellation of the primary wave from fan 2 take place.

Fig. 1 shows an arrangement in which the microphone and loudspeaker very much close and essentially equidistant from the sound source formed by the fan 2 are arranged. However, the principle applies to arrangements in which the Transducers are arranged at a distance in the direction of sound propagation, as shown in Fig. 2 to 5 is indicated.

The amount of the speaker 4 after ig. 1 emitted sound can be changed by changing the acoustic attenuation on the sound path between the loudspeaker 4 and the microphone 3 changes, with an increase in attenuation increases the level of the sound emitted by loudspeaker 4. Alternatively, you can The amount and size of the sound can be controlled electronically, as indicated in FIG is.

In the arrangement according to FIG. 2, two loudspeakers 4a and 4 can be seen 4h, which are arranged so that the output of the speaker 4b amplified by the Signal of an amplifier 5 can be controlled. Adjusting the reinforcement thus enables the degree of cancellation to be adjusted downstream. It can be one different numbers of speakers and a wide variety use different geometrical arrangements for active sound absorption, including speakers that are on other walls of the pipe, in branch pipes or within the cross section of the line. The degree of cancellation at a point downstream of the loudspeaker or loudspeakers can also can be influenced by the acoustic properties of the line, however take this effect into account by using an electronic circuit 6 in the system is built in to compensate for the line properties, as shown in Fig. 4 and 5 is indicated.

A second microphone 7 is downstream in Ig. 4 and 5 can be seen.

A signal from the microphone 7 indicates that at this point a complete Cancellation has only been done inadequately, and can be used either manually or automatically the gain of the amplifier 5 and the compensation circuit 6 to set.

The amplifiers in turn can have circuits to prevent Contain oscillations caused by the delay in the feedback line be generated. The speaker 4b can be arranged to reflect Waves deflects from the microphone 3 and / or the line can with sound absorbing Material to be lined.

Additionally or alternatively, a similar active system be arranged on the opposite wall1 or there can be a plurality of systems in different positions on the different walls of the pipe or at locations within the cross-section of the pipe. Farther can have a variety of cancellation microphones 7 at the most diverse downstream located places be provided.

When the delay of the sound path in the feedback loop between the speaker or speakers and the microphone is chosen so that they are compared with the period of the highest frequency waveform of interest, is not short, then the system can, for example, in that shown in FIG Shape can be modified.

In the embodiment shown in Fig. 5 is the acoustic feedback path between the loudspeaker 4a and the microphone 3, wherein the latter generates an electrical signal S.

Every signal occurring at point X has two feedback loops.

The first goes through the loudspeaker 4a, the microphone 3 and the line 8 to a summing point 9. The second way is via a circuit 10 and a Delay element 11 to the summing point 9. The delay element 11 and the Circuit 10 are combined to provide between the loudspeaker 4a and to compensate the microphone 3 occurring Zeitverzögerunq, and the delay element 11 and circuit 10 together simulate the characteristics of the loudspeaker 4a, the microphone 3 and the airways from the loudspeaker 4a to the microphone 3. The two Feedback paths are summed in antiphase at 9 and if both are the same have temporal characteristics, they cancel each other out.

The total numbers, 5, 6, 10 and 11 can either be from Jiand or automatically monitored, for example via the override microphone 7.

As already mentioned, the microphone 3 does not have to be directly opposite the Speakers 4a be arranged. It can be left of anywhere this point, on any wall or in any position in the cross-section on line 1, although this is an additional delay in the line between points 9 and X required.

The frequency and phase response of the loudspeaker (s) can be improved by using a separate turn in the loudspeaker and this into the feedback loop of the driver amplifier, not shown for the loudspeaker.

A general problem that affects all of the previously described Arrangements for the active elimination of sound or vibrations occurs is the Desire to know how to modify a signal waveform when they move from a generator to a scanning device, including the transducer goes through itself; it is the solution to this problem in particular; on which the Invention essentially aims.

Fig. 6 shows the basic situation in the case of an air line. When a If the signal SL is present at the loudspeaker 4, a signal 5M is generated by the microphone 3. However, different frequency components will have different delays and attenuations due to the response of the transducers themselves and also due to the different ones possible paths within the line 1, two of which are indicated by arrows in Fig. 6 are indicated. Thus, it is unlikely that the signal 5 will only be is a delayed and attenuated version of the signal S, and if we were previously one Only want to make a statement about the signal SM from the knowledge of the signal 5L, so it is necessary to characterize the path.

This can be done by using suitable test signals, e.g.

applies and determines how they are changed when used as a signal 5 sts occur. For example, an applied delta function SL can be modified in this way as indicated in FIG. 7.

One can expect a delta function of different amplitude is modified in a corresponding manner and a wave with the same shape, however different amplitude generated.

The response of line 1 to any other signal, in this case the noise from the fan, can then be predicted by superposing the response pulses will. An example of this is shown in FIGS. 8a to 8c.

One can think of the noise or the sound from the fan as one composite sum of a series of delta function pulses with different Imagine amplitudes as indicated in Fig. Toa. Each of these impulses will a timed response signal cause the test response signal in its form is similar, but with its amplitude to the instantaneous amplitude of the fan noise accordingly, as shown in Fig. 8b. The response signals can be added be to the response of the line to the overall waveform of the fan predict (see. Figure 8c).

According to a preferred aspect of the invention, the in Fig. 8b, store the fundamental waveform reproduced as a program of time-correlated steps. If this program is combined or folded with a given signal SJ, so it shows what kind of output signal 5 M from microphone 3 at one to the loudspeaker 4 to be applied signal SL will be expected.

In the case of a simple arrangement, such as that shown in FIG 2 is reproduced with only one loudspeaker, this is due to the superposition program derived from response pulses, as described with reference to the figures 8a to 8c, the time response for upstream Characterize sound propagation, including the speaker, for example of the loudspeaker 4a, and the 21ikrofones 3, and can thus be used to compensate for the Feedback from speaker 4a to microphone 3 is used in the manner discussed above will.

Another "characterization" is for that with the microphone output 3 program to be folded required to be downstream for cancellation or deletion of the loudspeaker, however, this can be achieved through a similar adjustment process as used for upstream "characterization".

Alternatively, the downstream "characterization" can be empirical derived from the upstream "characterization" or even from the upstream "Characterization" and the properties of the transducers by a method of convolution division be calculated.

A complete elimination of the feedback signal can be achieve in the manner indicated in Fig. 3 by one of the Suspension loudspeaker 4 derived signal S1 from signal S2 from microphone 3 is electrically subtracted. the The signal S1 can be derived from the cancellation loudspeaker signal by a second Convolution can be made that the response of the speaker, the line and compensated for the microphone.

9 to 11 show various methods for setting the Characterization and for the execution of the folding reproduced.

In Fig. 9 the input signal, e.g. from microphone 3 in Fig. 3, which represents the primary wave to be damped, a delay element, e.g. a Shift register 15 supplied. The delay element should be one length of the input signal store, which is essentially equal to the duration of the stored characterization is.

In the example of FIG. 9, only three stages are shown, however in practice there are more than these three e.g.

32 steps.

A memory for weight coefficients that is linked to the input signal The program of steps or stages to be folded is represented in FIG. 9 with 111, W2, W3 indicated, and when the input signal through the shift register 15 passes through, the multipliers M1, M2 and M3 give their output to an addition circuit 16, which provides the basis of the signal to the loudspeaker 4b.

The folding can be done digitally or on an analog basis, and instead of a multiplier for each weight coefficient and an addition circuit a single multiplier M (see FIG. 10) can be used at which the channels are used repeatedly over time. The scanning speed must of course be great compared to the speed at which the input signal runs through register 15.

The weight coefficients in memory can be easily based on the response behavior of the system to the execution of a single delta function can be adjusted, the compensation being improved by the empirical adjustment is going to a characterization to give that any Exactly cancels the input signal.

Fig. 11 shows how the weight coefficients in a circulating Register 16 should be saved to keep the characterization up to date bring to. As indicated in FIG. 11, this amounts to regular or irregular Test response signal supplied to intervals via an addition circuit 17, for example 10t of the input signal, so that only gradual modifications of the weight coefficients occur in order to take into account extraneous or background noises.

Since extraneous or background noises do not interfere with the test signal or the signal movement are correlated in the delay element, they are with the same probability can be added or subtracted at each point in the delay register so that the mean value is zero.

The register 16 can act as a shift register 18 and the weighted addition circuit 17 be designed as a time multiplexer 19, as shown in Fig. 12, the multiplexer 19 in communication with the test response signal only about 10% of the time stands.

Another embodiment of a system of the type described above consists of a single delay unit with discrete stages, each Level contains the digital equivalent of the relevant value of the input signal. The information in memory is also in digital form, and so is the multiplier is a digital multiplier.

The delay unit or the sneaker or both may alternatively also contain analog data instead of digital information. For example The delay unit can consist of a charge-controlled shift register or alternatively consist of a series of "sample and flold" circuits.

Fin example of an analog embodiment of the memory for the Weight coefficient is simply a series of voltage supplies connected Potentiometers, each potentiometer having the required weight coefficient is set accordingly.

A particular embodiment of the invention has a number of "Sample-and-1Told" circuits used to generate an analog shift register are cascaded so that the stored in any element Similarly, information is passed on to the next element if a "sample" or Ahfragesignal is received. Scanning signals are applied to each element in turn, starting at the end of the register which contains the oldest signal. In this way, information is not overwritten until it has been scanned is.

Each analog value can in turn be sent to a multiplying digital-to-analog converter are fed in via an analog multiplexer or multiple coupler, whose binary address inputs are connected to a counter. This counter is also intended for the waveform generator and in this way ensures that each scanned element is always multiplied by the corresponding element stored in the direct access memory. The waveform for the Convolution that is stored in the direct access memory can be made in simpler or in a more complicated way by a central processing unit or some other logic circuit be modified, depending on a comparison of the levels of the rest or not deleted signals before and after the waveform has been modified is. The residual signal can with a downstream microphone and a Sound level detector can be monitored.

Fig. 13 shows suitable circuitry for performing the required Convolution and has a tapped analog delay line 40 consisting of a Series of 32 sample-and-llold circuits is constructed, with each output one then supplied identical stage. The outlets, the taps along the delay line 40, the circuit 41 with Resistances and potentiometers are summed so that the center position of each potentiometer shifers 42 represents the position for an output zero. A movement of a given Grinder 42 upward in the arrangement shown in FIG. 13 causes a Subtract the stored element from the output signal, and move upwards causes a corresponding addition of a part of the element.

In each case represents the amount of movement of the potentiometer wiper 42 from its middle position a multiplication factor with a positive or negative sign, which corresponds to the respective direction of displacement. the "Sample-and-Hold" circuits are made up of 32 lines from two to 4 lines. 16 line decoders 43 and 44 are monitored, only one line of these is in the "high" state, corresponding to that addressed by a binary counter 45 Pulse which is supplied with a 32-KH2 clock pulse on a line 46. A Signal with the high state switches on the transmission gate, which acts as a box Tr is reproduced and connects the sample and "hold" capacitor to the output the previous stage. The 32nd "sample-and-i3old" circuit is addressed first and the transmission gate TG 32 is switched on and connects the sample and 1? old capacitor of stage 32 (C32) with the output of the previous stage. TG 32 is then opened, when day 31 closes and C 31 is loaded from its previous stage until finite the input is sampled when TG 1 turns on and an input pulse in the capacitor C1 stores. In this way, input samples or pulses travel down the line 11o from right to left, step by step, until the sequence of 32 Sample-and-Jold "switches.

In a further development of this system are the potentiometers replaced in the circuit 41 by an analog multiplexer, which the respective sample for the subsequent multiplication or multiplication by a digitally controlled Analog multiplier, where the multiplication factor for each element in a memory corresponding to the positions of the 32 potentiometer sliders 42 is stored and changed in any order will. In addition, can the described analog delay line 40 through an integrated circuit that performs the same function, or by an equivalent complete digital circuit.

FIG. 14 shows a preferred arrangement of the system according to FIG. 10, that works like this: An analog input signal that represents the primary wave to be canceled is fed to an analog shift register 20, which in turn 32 Has "sample and hold" circuits. The outputs 20a, 20b, 20c etc. are from an analog multiplexer 21 synchronous with the sampling of the outputs 22a, 22b, 22c etc. of a 32-level direct access memory (RA? I) 22, which stores the convolution waveform in digital form. The saved waveform in RAM memory 22 may have been derived so that one of the system's delta functions in the manner described above.

The multiplexer 21 connects the outputs 20a, 22a, 20b, 22b etc.

in pairs sequentially with a multiplication digital-to-analog converter 23, a first ana] ogeinaang 24, a second dipital entrance 25 and a Has analog output 26. In the particular case discussed, the traversal is All 32 contacts of the shift register 20 and the RAH memory 22 in one rtillisecond completed, and before the next pass of the contacts 20a, 22a begins, the shift register 20 is updated to reflect changes in the ring signal, bringing the shift register up to date 20 takes place in the direction of arrow U.

The convolution requires that the integral of the wave form over the entire Pass is taken, and a low pass filter 27 acts as an integrator, the Cutoff frequency of the low-pass filter 27 is a function of the reciprocal of the transit time of the multiplexer is.

To further improve the embodiment of the system, one is Central unit 28 is provided, which receives every residual signal via a line 29, for example the microphone 7 arranged downstream. The central unit 28 is programmed to the memory contents of the RAM memory 22 on the basis of the remainder of the sicrnales modified. For the one used to modify the RAM memory Algorithm by the central unit is available a wide variety of algorithms available according to the circumstances.

For example, the appearance of a residual signal on the line 29 an adjustment setting of the information at each address in the RAM memory one after the other, with groups of addresses one after the other or with all addresses cause together. Any change made to the one stored in PsAM memory Waveform can be assessed to see if the situation improves has, for example, by determining whether the signal on line 29 changes, maintaining corrective changes in the stored information while non-improving changes are deleted. Dic for this The logic used in the adaptation strategy can be refined and developed to the point be where the algorithm is changed when the system learns which is the most sensitive Areas of the waveform stored in R memory are, and are subject to modification these areas are concentrated while a signal remains on line 29.

The method described above and the associated device can be used in a wide range of different industrial applications, for example, the sound attenuation of ventilation systems in cannulas, of AbsaucTsvstemen and in the inlet and outlet chambers of (asturbinen.

For example, in a suction system, the cancellation converter (s) be arranged outside the suction line, for example as a tight fit around the newspaper winding ring, while another rest-signal converter in a suitable position at a distance from the Suction line is arranged. 11 Against the mismatch between the suction outlet and its surroundings becomes the undesirable upstream Signal strongly attenuated, and in this application the cancellation converter (s) be of relatively low performance.

The method described above can also be used to generate "quiet Areas "in an otherwise noisy environment. For this purpose a plurality of "downstream" arranged microphones 7 are used, the be distributed over the desired "quiet area", and the outputs or output signals The respective T2 microphones can be fed to their own power unit will. The outputs of the various power measurement units can then be used in a suitable manner Combined manner and, for example via the newspaper 29, fed to a central unit to adapt the algorithm used for the convolution. A multitude of speakers can be used if the size and / or shape of the desired "Quiet area" requires this.

When the primary sound waves are in the desired "quiet area" when entering widely separated directions, it may be desirable to have one Use scan transducers for each such direction, with each scan transducer has its own system driver and the outputs from the various system drivers fed to a single loudspeaker or loudspeaker field in the "quiet area" will. A residual signal microphone in the "quiet area" can be used for adjustment purposes can be used and similar settings can be made for the algorithms that are responsible for the convolutions in the various system drivers from the output one of the residual signal microphones can be used.

L e r s e i t e

Claims (4)

P a t e n t a n s p r ü c h e Process for sound absorption in one limited space in which a first electrical signal is derived that is temporal correlates represents a primary sound wave entering the room and to be damped, and in which the first electrical signal for deriving a second electrical signal Signal is used in a second converter to generate a secondary wave is used in the space, which is at least partially the primary wave in the space cancels out that the second electrical Signal from the first electrical signal by convolution of the first electrical signal is derived with a program of time-correlated operational steps, which by determining the response behavior of the transducer in the room to characterizing Impulse is derived. 2. The method according to claim 1, d a d u r c h g e k e n n z e i c h n e t that the program is based on the output of a third converter in the limited space is automatically adjusted to the degree of damping of the primary wave improve during cushioning. 3. Device for damping a primary wave in a system a specially generated secondary wave, with a first transducer for deriving a first electrical signal that correlates in time with the primary wave to be damped represents, and with a second converter for generating the secondary wave a second electrical signal, d u r c h q e k e n n -z e i c h n e t, that a memory (22) is provided which contains a program of time-correlated operational steps contains, and that a device (20,21,23) for folding the first electrical Signal provided with the program for generating the second electrical signal is. 4. Apparatus according to claim 3, including one upstream of the first and a third transducer arranged in the direction of propagation of the primary wave converter, d a d u r c h S e k e n n -z e i c h n e t that a device (28) for setting the program of the operating steps used in the raltuna to improve the degree of damping of the primary wave that can be achieved with the device is provided.