Device and method for calibrating a microphone arrangement

Abstract

The invention relates to a device (100) for calibrating a microphone arrangement (20, 30), comprising: - a flow channel (10), wherein the flow channel (10) has at least one cylindrical wall segment (11), - at least one microphone arrangement (20) for detecting a reference sound field, wherein the at least one microphone arrangement (20) is arranged around a central axis (12) of the at least one cylindrical wall segment (11), wherein the at least one microphone arrangement (20) has at least two microphones (21, 22), - at least one loudspeaker (40) for outputting at least one excitation signal (41), wherein the at least one loudspeaker (40) is rotatable along a circular path around the central axis (12) of the cylindrical wall segment (11), - at least one drive unit (50) for rotating the at least one loudspeaker (40), - at least one control unit (60) for controlling the calibration, and a corresponding procedure.

Description

[0001] The invention relates to a device for calibrating a microphone arrangement and a corresponding method.

[0002] A microphone array is used, for example, in the experimental investigation of sound fields. For example, a sound field generated by an engine can be investigated using multiple microphones in a flow channel. This type of investigation is relevant, for example, in the development of noise reduction approaches for engines. The microphone array typically consists of several microphones arranged, for example, in the flow channel, for better recording of the sound field.

[0003] The microphones in a microphone array may, for example, exhibit differences in runtime and level before initial operation. These differences may arise from manufacturing tolerances or additional peripherals on the microphones, for example, and may falsify the test results. To compensate for these differences, the microphones must be calibrated. Common calibration methods, for example, use a loudspeaker to generate an acoustic excitation signal that excites a sound field with various overlapping eigenmodes – the so-called vibration modes. If the vibration modes overlap coherently, for example, the microphones in the microphone array record sound pressures that vary spatially and temporally, meaning that the recorded sound field is not suitable for direct calibration.

[0004] In order to still be able to use the sound field recorded by the microphone array for calibration, an analytical or numerical model sound field is usually used, which models, for example, the sound pressures expected at the microphones as a result of an excitation signal and thus enables a comparison with the recorded sound field. The usual calibration procedures are therefore based on a model of the sound field. However, this modeling is only an approximation - e.g., due to physical or geometric assumptions made - and can therefore lead to inaccuracies in the calibration. Furthermore, the modeling requires additional time, which prolongs the calibration process.

[0005] DE 10 2008 041 916 B3 discloses a test device in a flow channel for aerodynamic and acoustic measurements on engines. The test device is characterized by a modular design of the disturbance channel and the other components. For acoustic measurements, measurement modules equipped with microphones can be positioned at various locations in the flow channel.

[0006] Another such test device is known from CN 1 12 268 708 A. It discloses an acoustic measurement system with two microphone arrays, one of which records the sound field inside the flow channel. During frequency-domain analysis, the sound field is broken down into components that are assigned to different modes.

[0007] The technical problem therefore arises of creating a device and a method for calibrating a microphone arrangement which reduce inaccuracies in the calibration of the microphone arrangement and which shorten the time required for calibration.

[0008] The solution to the technical problem is achieved by the subject matter having the features of the independent claims. Further advantageous embodiments of the invention are set forth in the subclaims.

[0009] A device for calibrating a microphone arrangement is proposed, comprising: - a flow channel, wherein the flow channel has at least one cylindrical wall segment, - at least one microphone arrangement for detecting a reference sound field, wherein the at least one microphone arrangement is arranged around a central axis of the at least one cylindrical wall segment, wherein the at least one microphone arrangement has at least two microphones, - at least one loudspeaker for outputting at least one excitation signal, wherein the at least one loudspeaker is rotatable along a circular path around the central axis of the cylindrical wall segment, - at least one drive unit for rotating the at least one loudspeaker, - at least one control unit for controlling the calibration, wherein the device is designed to carry out the following steps: - Outputting the at least one excitation signal by means of the at least one loudspeaker, wherein the outputting of the at least one excitation signal generates a reference sound field in the flow channel, wherein the reference sound field has a fundamental mode and a plurality of further vibration modes, - Rotating the at least one loudspeaker along the circular path around the central axis of the at least one cylindrical wall segment by means of the at least one drive unit at a rotation frequency, wherein the rotation of the at least one loudspeaker shifts oscillation frequencies of the plurality of further oscillation modes depending on the rotation frequency, so that only one oscillation frequency of the fundamental mode corresponds to an excitation frequency of the at least one excitation signal, - detecting the generated reference sound field by means of the at least two microphones of the at least one microphone arrangement, - detecting a waveform of the fundamental mode in the detected reference sound field as a function of the excitation frequency by means of the control unit, wherein the detected waveform of the fundamental mode deflects the at least two microphones of the at least one microphone arrangement in phase and amplitude, - Leveling of runtime and / or level differences depending on the detected waveform using the control unit.

[0010] The inventors have recognized that by rotating the loudspeaker, the Doppler effect is created in the reference sound field and that this effect affects the multitude of other vibration modes, but not the waveform of the fundamental mode.

[0011] The fundamental mode exhibits a waveform whose amplitude is constant in a circumferential direction of the cylindrical wall segment. The fundamental mode runs particularly along the central axis, meaning that the wave crests and troughs of the waveform are oriented orthogonally to the central axis. Due to these properties, the oscillation frequency of the fundamental mode is not shifted by the rotation of the loudspeaker around the central axis.

[0012] The further vibration modes can, for example, be higher-order vibration modes and, in particular, have waveforms whose respective amplitude is not constant in the circumferential direction of the cylindrical wall segment. Thus, the waveforms of the further vibration modes do not deflect the microphones of the microphone arrangement in particular with the same phase and amplitude. The waveforms of the further vibration modes, in particular, run at least partially in a radial direction and / or circumferential direction of the cylindrical wall segment, i.e., the wave crests and troughs of the respective waveform are, in particular, non-orthogonal to the central axis. For example, the waveforms of the further vibration modes propagate spirally in the cylindrical wall segment or flow channel. Due to these properties, the vibration frequencies of the further vibration modes are shifted by the rotation of the loudspeaker around the central axis.

[0013] The flow channel can be designed, for example, as a wind tunnel. The flow channel is, in particular, rotationally symmetrical. The flow channel can, for example, have a cylindrical wall. The cylindrical wall segment is, in particular, a section of the wall of the flow channel. The central axis is, in particular, oriented parallel to an axial flow direction of the flow channel. The central axis of the wall segment is, for example, an axis of symmetry of the cylindrical wall segment. The cylindrical wall segment has, in particular, at least two bores in the circumferential direction for arranging the at least two microphones of the microphone arrangement.

[0014] The microphone arrangement is oriented in particular in a plane orthogonal to the central axis of the flow channel. The microphone arrangement is arranged in particular in the circumferential direction of the cylindrical wall segment. For example, the at least two microphones can be arranged flush in the wall of the at least one cylindrical wall segment, e.g., in the previously explained bores. In particular, the microphone arrangement is arranged in a circle around the central axis. The microphones can be designed, for example, as condenser microphones. A detection direction of the at least two microphones can be the radial direction of the cylindrical wall segment.

[0015] The loudspeaker can be designed, for example, as a compression loudspeaker. The loudspeaker can, for example, output the excitation signal at a level in a range of 80 dB to 120 dB. The loudspeaker outputs the excitation signal, in particular, parallel to the central axis. A plane of the circular path of the loudspeaker is oriented, in particular, orthogonal to the central axis of the flow channel. The plane of the circular path and the plane of the microphone arrangement are oriented, in particular, parallel to each other.

[0016] The drive unit can be designed, for example, as an electric motor. The rotation frequency of the loudspeaker can, for example, have a value in a range from 1 Hz to 200 Hz or from 60 rpm to 12,000 rpm. The drive unit is connected, in particular, to the at least one loudspeaker via a connecting element. The connecting element can be designed, for example, as a cantilever or as a circular disk, with the at least one loudspeaker being arranged, for example, at a free end of the connecting element.

[0017] The control unit can be designed as a microcontroller, for example, or include one. Controlling the calibration can include the control unit performing, in particular, the steps of detecting and leveling.

[0018] The excitation signal can, for example, be a sine tone with an excitation frequency of 2 kHz and an excitation level of 100 dB. The excitation signal can, of course, also have a different excitation frequency and / or excitation level than the one specified above. The at least one excitation signal can, for example, be generated by the control unit and sent to the at least one loudspeaker for output.

[0019] The generated reference sound field can, for example, be an airborne sound field in the flow channel, meaning that the sound propagation medium is, for example, air. Due to the superposition of the vibration modes, the reference sound field can, for example, exhibit spatially and temporally varying sound pressures. The sound pressure of the reference sound field can vary, in particular, along the circumferential direction of the cylindrical wall segment.

[0020] The detection of the reference sound field can comprise detecting the local sound pressure of the reference sound field acting on a microphone using the microphone at a time of flight and at a level. For example, a microphone diaphragm of the microphone is deflected due to the sound pressure and a measurement signal proportional to the deflection can be generated by electronic conversion. The measurement signal can, for example, have a sound pressure level for a respective time of flight, i.e., be designed as a temporal profile of the detected reference sound field. Each microphone of the microphone arrangement generates in particular at least one measurement signal. The control unit can be configured to receive the measurement signals of the at least one microphone arrangement and to further process them.

[0021] In particular, the terms phase and amplitude refer to physical properties of one or more waveforms of the generated reference sound field. As explained above, the waveforms of the other vibration modes do not deflect the at least two microphones in the same phase and amplitude. However, the waveform of the fundamental mode deflects the at least two microphones in the same phase and amplitude. This circumstance is used, as explained below, to calibrate the microphone arrangement. The terms delay and level, on the other hand, refer in particular to the phase(s) and amplitude(s) of one or more waveforms of the recorded reference sound field, for example, as recorded by electronic conversions using the at least two microphones.

[0022] Detecting the waveform of the fundamental mode can comprise transforming the recorded reference sound field from the time domain to the frequency domain. For example, the measurement signals from the at least two microphones can be transformed from the time domain to the frequency domain using a fast Fourier transformation. The waveform of an oscillation mode, e.g. the fundamental mode, can be detected in particular as a function of an amplitude of the waveform. In the frequency domain, the measurement signals have, for example, frequency-dependent Fourier coefficients, where a Fourier coefficient can correspond to the amplitude of a respective waveform. Since the oscillation frequencies of the other oscillation modes differ from the oscillation frequency of the fundamental mode due to the Doppler effect, the waveform of the fundamental mode can, for example, be particularly easily distinguished from the waveforms of the other oscillation modes in the frequency domain.In particular, the amplitude assigned to the excitation frequency in the frequency domain corresponds to the amplitude of the waveform of the fundamental mode. The amplitudes of the other recorded waveforms can be filtered out of the recorded reference sound field, for example, by frequency filtering, and the waveform of the fundamental mode can be detected in the recorded reference sound field in this way. Detection particularly comprises determining, for each microphone in the microphone array, a time point and a corresponding level corresponding to the detected waveform of the fundamental mode, for example, by determining the time point and level at which the waveform of the fundamental mode was detected in the measurement signal of the respective microphone.Detecting the waveform of the fundamental mode has the advantage that no modeling of the sound field is necessary for calibration and therefore no assumptions about the sound field have to be made that could lead to inaccuracies.

[0023] Leveling is particularly necessary because, for example, the microphones, for structural reasons, record the sound pressure of the waveform of the fundamental mode at different levels, i.e., with level differences, and at different propagation times, i.e., with propagation time differences. Leveling of propagation time and / or level differences particularly includes determining at least one propagation time and / or level difference between the at least two microphones in the microphone arrangement. If, however, no propagation time and / or level difference is determined, then leveling is not necessary. A propagation time difference can, for example, be a time offset between the measurement signals of the at least two microphones. The propagation time difference can, in particular, be structurally determined and, for example,caused by the electronic periphery of a microphone, which causes the time offset between the measurement signals, even though the waveform of the fundamental mode deflects the at least two microphones in phase, i.e. in particular simultaneously. A level difference can, for example, be a difference between the levels for the amplitude of the waveform of the fundamental mode recorded by the at least two microphones, even though the waveform of the fundamental mode deflects the at least two microphones with the same amplitude. For example, a first microphone in the microphone arrangement can record the amplitude of the detected waveform with a first level and a further microphone in the microphone arrangement can record the amplitude of the detected waveform with a further level, whereby a difference between the recorded levels corresponds to the level difference.However, since the detected waveform of the fundamental mode deflects the microphones of the microphone array simultaneously and with the same amplitude, such a time lag and / or level difference can be leveled out to calibrate the microphone array. The at least one time lag and / or level difference can be leveled out, for example, by shifting one of the microphones of the measurement array by the determined time lag and / or level difference, i.e., by providing it with an offset. In this way, the at least one microphone array can be calibrated so that no time lag and / or level difference is present in future investigations.

[0024] In one embodiment, at least one microphone of the at least one microphone arrangement is designed as a reference microphone. In this way, for example, a detected propagation time and / or a level detected by the reference microphone can be used as a reference value for leveling the microphone arrangement. This improves the accuracy of the calibration. The reference microphone can in particular be designed as a calibrated microphone. In this way, it can be ensured that, for example, there is no propagation time and / or level difference in the reference microphone. The reference microphone is in particular one of the at least two microphones of the microphone arrangement or a further microphone of the at least one microphone arrangement.

[0025] In one embodiment, the rotation of the at least one loudspeaker around the central axis of the wall segment is a balanced rotation. In this way, for example, the drive unit can be protected from damage due to imbalance and it can be ensured that the rotation of the loudspeaker actually takes place along a circular path. In particular, the device can comprise a balancing element for balancing the rotation. For example, the at least one loudspeaker can be arranged at a free end of the previously explained connecting element and a balancing element designed as a counterweight can be arranged at an opposite free end of the connecting element. The balancing element can, for example, rotate along the circular path offset by 180° to the loudspeaker. This makes it possible to compensate for the weight of the loudspeaker and thus balance the rotation of the loudspeaker. The balancing element can, for example,be designed as a further, particularly structurally identical, loudspeaker. The further loudspeaker can also output the at least one excitation signal. This makes it possible, for example, to easily increase the sound level when outputting the at least one excitation signal.

[0026] In one embodiment, the excitation frequency of the at least one excitation signal is adjustable as a function of the rotation frequency of the loudspeaker, wherein the adjustment makes the excitation frequency proportional to the rotation frequency. By adjusting the excitation frequency, a proportionality relationship between the excitation frequency and the rotation frequency is specified. The adjusted proportionality relationship also determines the ratio of the frequency shift of the other oscillation modes to the excitation frequency. In this way, the oscillation frequency of the fundamental mode can be more easily distinguished from oscillation frequencies of the other oscillation modes, e.g. in the frequency spectrum. This simplifies the detection of the waveform of the fundamental mode. A proportionality relationship between the excitation frequency and the rotation frequency is also referred to as the rotation order.For example, the rotation frequency has a value of 10 Hz and the excitation frequency has a value of 2 kHz, so that the proportionality ratio of excitation frequency to rotation frequency is 200, resulting in a rotation order of 200. In particular, a current rotation frequency is recorded. By recording the current rotation frequency, fluctuations in the rotation frequency, e.g., due to current fluctuations in the drive unit, can be taken into account when setting the excitation frequency. The current rotation frequency can be recorded, e.g., using a rotary encoder in the drive unit, whereby the excitation frequency can be adjustable depending on the recorded rotation frequency. The excitation frequency can be set using the control unit.

[0027] In one embodiment, the at least one loudspeaker is configured to output an excitation signal with a frequency in a range of 20 Hz to 20 kHz, and / or the at least two microphones of the microphone array are configured to detect a frequency in a range of 20 Hz to 20 kHz. In this way, a frequency range that is perceptible to the human ear can be output and / or detected. This is advantageous for investigating engine sound fields and for reducing engine noise perceptible to the human ear.

[0028] In one embodiment, the diameter of the flow channel has a value in a range from 0.5 m to 5 m. In this way, at least parts of an engine, e.g., one or more fan stages, with a corresponding diameter can be examined in the flow channel. In particular, the cylindrical wall segment has a diameter in a range from 0.5 m to 5 m. In particular, all sections of the flow channel have a diameter in a range from 0.5 m to 5 m.

[0029] In a possible alternative embodiment, the diameter of the flow channel has a value in a range of 2.5 cm to 50 cm. This allows for the investigation of miniaturized setups with ring-shaped sound pressure sensors.

[0030] In one possible embodiment, the excitation signal has an excitation frequency in the range of 20 kHz to 1 GHz. This allows the excitation signal to be output in the ultrasonic range. This is particularly advantageous for the aforementioned miniaturized setups.

[0031] In one embodiment, the device comprises at least one further microphone arrangement, wherein the at least one further microphone arrangement has at least two microphones, wherein the at least one further microphone arrangement is arranged around the central axis of the at least one cylindrical wall segment. In this way, multiple microphone arrangements can be calibrated simultaneously. The technical effects explained for the at least one microphone arrangement also apply to the further microphone arrangement. In particular, the detected waveform of the fundamental mode also deflects the at least two microphones of the at least one further microphone arrangement in phase and amplitude.

[0032] In one embodiment, the flow channel has at least one anechoic end segment. This way, waveforms in the generated reference sound field are reflected less strongly, for example, compared to an open end segment, so that the calibration of the microphone array is not hampered by reflected sound waves. The anechoic end segment can be arranged, for example, at a free end of the flow channel. The anechoic end segment can be designed, for example, as a funnel that is at least partially filled with damping material.

[0033] Furthermore, it can be provided that the device is designed such that the rotation frequency and / or the excitation frequency are frequency-modulatable.

[0034] Further proposed is a method for calibrating a microphone arrangement, comprising the steps: - Outputting at least one excitation signal by means of at least one loudspeaker, wherein the output of the at least one excitation signal generates a reference sound field in the flow channel, wherein the reference sound field has a fundamental mode and a plurality of further vibration modes, - Rotating the at least one loudspeaker along a circular path around a central axis of at least one cylindrical wall segment at a rotation frequency, wherein the rotation of the at least one loudspeaker shifts oscillation frequencies of the plurality of further oscillation modes depending on the rotation frequency, so that only one oscillation frequency of the fundamental mode corresponds to the excitation frequency, - Recording the generated reference sound field using at least two microphones of at least one microphone arrangement, - detecting a waveform of the fundamental mode in the detected reference sound field as a function of the excitation frequency, wherein the detected waveform of the fundamental mode deflects the at least two microphones of the at least one microphone arrangement in phase and amplitude, - Leveling of runtime and / or level differences depending on the detected waveform.

[0035] The advantages and technical effects mentioned in this disclosure for the device also apply in particular to the proposed method. The device for calibrating a microphone arrangement can, in particular, be designed to carry out one, several, or all steps of the method described in this disclosure.

[0036] In one embodiment, the excitation frequency of the at least one excitation signal is frequency-modulated. In this way, the microphone arrangement can be calibrated for all frequencies of the modulation range. For example, the excitation signal can be frequency-modulated using a sine sweep. The modulation range can, for example, have values ​​in a range from 20 Hz to 20 kHz. The frequency modulation can, for example, be carried out using the control unit explained above. In particular, a reference sound field can be generated in the flow channel for each frequency of the modulation range, wherein the previously explained steps for calibrating the microphone arrangement can be carried out for the respectively generated reference sound field.

[0037] The invention is explained in more detail using exemplary embodiments. The figures show: Fig. 1 an embodiment of a device for calibrating a microphone arrangement, Fig. 2-A is a schematic representation of a fundamental mode of a reference sound field, Fig. 2-B is a schematic representation of another vibration mode of a reference sound field and Fig. 3 an embodiment of a method for calibrating a microphone array.

[0038] In the following, the same reference symbols refer to elements with the same technical features.

[0039] Fig. 1 shows a schematic representation of an embodiment of a device 100 for calibrating a microphone arrangement 20, 30. Fig. 1 shows in particular a longitudinal section through the embodiment of the device 100 along a central axis 12.

[0040] The device 100 comprises a rotationally symmetrical flow channel 10. The flow channel 10 has a cylindrical wall segment 11. An axis of symmetry of the cylindrical wall segment 11 is referred to as the central axis 12. The central axis 12 is oriented, for example, in a flow direction of the flow channel 10. Furthermore, the device 100 comprises another cylindrical wall segment 13.

[0041] The device 100 further comprises the microphone arrangement 20 and the further microphone arrangement 30. The microphone arrangements 20, 30 each have, for example, two microphones 21, 22, 31, 32 designed as condenser microphones. In the wall segment 11, the microphone arrangement 20 is arranged in a circle around the central axis 12, for example in bores provided for the arrangement of the microphones 21, 22. In the further wall segment 13, the further microphone arrangement 30 is also arranged in a circle around the central axis 12. Between the microphones 21, 22, 31, 32 of the respective microphone arrangement 20, 30, an angle of, for example, 180° is spanned in a circumferential direction of the wall segments 11, 13. In this way, the microphones 21, 22, 31, 32 are each arranged uniformly along the circumferential direction of the cylindrical wall segments 11, 13.

[0042] The device 100 further comprises a loudspeaker 40 for outputting an excitation signal 41. An amplifier (not shown) can additionally be connected upstream of the loudspeaker 40 to amplify the excitation signal 41 before output. The excitation signal 41 can be frequency-modulated, for example, using a sine sweep in a range of 20 Hz to 20 kHz. The loudspeaker 40 rotates along a circular path around the central axis 12 during the output of the excitation signal 41.

[0043] To drive the rotation of the loudspeaker 40, the device 100 comprises a drive unit 50 designed as an electric motor. The drive unit 50 is connected to the loudspeaker 40, for example, via a connecting element 51 designed as a cantilever. To balance the rotation of the loudspeaker 40, a balancing element 52 designed as a counterweight is arranged at a free end of the connecting element 51 opposite the loudspeaker 40. An angle of 180°, for example, is spanned between the loudspeaker 40 and the balancing element 52 in the circumferential direction of the wall segment 11.

[0044] To control the calibration, the device 100 further comprises a control unit 60 embodied as a microcontroller. The control unit 60 generates, for example, the excitation signal 41 with a known excitation frequency and a known level. The excitation signal 41 can be sent to the loudspeaker 40 for output. Furthermore, the control unit 60 generates, for example, a signal 53 for driving the drive unit 50 with a known rotation frequency of the loudspeaker 40. The drive signal 53 can be sent to the drive unit 50.

[0045] The device 100 further comprises, for example, a terminal segment 14 configured as a funnel. The terminal segment 14 can be filled with insulating wool, for example, and thus provides low reflection of sound waves propagating in the flow channel 10. For example, the sound waves emitted by the loudspeaker 40 are almost completely absorbed by the terminal segment 14, so that the calibration of the microphone arrangements 20, 30 is not hampered by reflected sound waves.

[0046] By outputting the excitation signal 41, a reference sound field is generated in the flow channel 10. The reference sound field has a fundamental mode G and a plurality of additional vibration modes S. Due to the superposition of the vibration modes G, S, the reference sound field has, for example, spatially and temporally varying sound pressures. The microphone arrangements 20, 30 can detect the sound pressures of the reference sound field present at the microphones 21, 22, 31, 32 and convert them, for example, into measurement signals 23, 24, 33, 34. The measurement signals 23, 24, 33, 34 can be sent to the control unit 60 for calibration purposes.

[0047] Fig. 2-A shows a schematic representation of a fundamental mode G of a reference sound field in an embodiment of a device 100 for calibrating a microphone arrangement 20, 30 (cf. Fig. 1). The fundamental mode G has a waveform W whose amplitude A is constant in a circumferential direction of the cylindrical wall segment 11. This is indicated by black lines which are oriented orthogonal to the central axis 12, i.e. the wave crests and wave troughs of the fundamental mode G are in particular oriented orthogonal to the central axis 12. A distance between the wave crests indicates, for example, a period or oscillation frequency of the fundamental mode G. In particular, the distance between the wave crests indicates the axial spatial frequency of the fundamental mode G. At higher frequencies, the fundamental mode can, for example, also be a superposition of several oscillation forms with different axial spatial frequencies, without exhibiting a circumferential variation.

[0048] Fig. 2-B shows a schematic representation of a further vibration mode S of a reference sound field in an embodiment of a device 100 for calibrating a microphone arrangement 20, 30 (cf. Fig. 1). The further oscillation mode S is, for example, a higher-order oscillation mode and, in particular, has a waveform WS whose amplitude AS is not constant in the circumferential direction of the cylindrical wall segment 11. This is indicated by black lines which are oriented obliquely, i.e., non-orthogonally, to the central axis 12, i.e., the wave crests and troughs of the further oscillation mode S are, in particular, non-orthogonally oriented to the central axis 12. Thus, the waveforms WS of the further oscillation mode S do not deflect the microphones 21, 22, 31, 33 of the microphone arrangements 20, 30 in phase and amplitude. A distance between the wave crests indicates, for example, a period or oscillation frequency of the further oscillation mode S. It is evident that this distance is smaller than that in Fig. 2-A. This indicates a shift in the oscillation frequency of the additional oscillation mode S relative to the oscillation frequency of the fundamental mode G due to the Doppler effect.

[0049] The inventors have thus recognized that rotating the loudspeaker 40 only affects the oscillation frequency of those waveforms WS whose amplitude AS is not constant in the circumferential direction of the cylindrical wall segment 11. However, the oscillation frequency of the fundamental mode G is not shifted by the Doppler effect due to the properties explained above.

[0050] Fig. 3 shows an embodiment of a method for calibrating a microphone arrangement 20, 30. The steps of the method explained below can be carried out by means of the device 100 (cf. Fig. 1).

[0051] In a step S1, an excitation signal 41 is output by means of a loudspeaker 40. By outputting the excitation signal 41, a reference sound field is generated in a flow channel 10. The reference sound field has a fundamental mode G and a plurality of further vibration modes S (cf. Fig. 2-A and Fig. 2-B).

[0052] In a step S2, the loudspeaker 40 is rotated along a circular path by means of a drive unit 50. The circular path is arranged around a central axis 12 of at least one cylindrical wall segment 11 of the flow channel 10. A plane of the circular path can be oriented orthogonal to the central axis 12. By rotating the loudspeaker 40, oscillation frequencies of the plurality of further oscillation modes S are shifted depending on a rotation frequency of the loudspeaker 40. This is due to the Doppler effect. In particular, the oscillation frequencies of the further oscillation modes S are shifted by an amount proportional to the rotation frequency. Only the fundamental mode G is not affected by the Doppler effect, so that an oscillation frequency of the fundamental mode G continues to correspond to the excitation frequency.

[0053] In a step S3, the generated reference sound field is captured using two microphones 21, 22 of the microphone arrangement 20. Additionally, the generated reference sound field can be captured using two microphones 31, 32 of the further microphone arrangement 30.

[0054] In a step S4, a waveform W of the fundamental mode G is detected in the acquired reference sound field. For example, the measurement signals 23, 24, 33, 34 can be transformed from the time domain to the frequency domain using a fast Fourier transform. The oscillation frequencies of the further oscillation modes S can be easily identified in the frequency domain compared to the oscillation frequency of the fundamental mode G because they are shifted by an amount that is proportional to the rotation frequency. The oscillation frequency of the fundamental mode G, on the other hand, corresponds to the excitation frequency of the excitation signal 41 and can thus also be easily identified in the frequency domain. By filtering the measurement signals 23, 24, 33, 34, the waveforms WS of the further oscillation modes S can be filtered out, for example, from the measurement signals 23, 24, 33, 34, and the waveform W of the fundamental mode G can thus be detected.

[0055] In a step S5, at least one propagation time and / or level difference is leveled off as a function of the detected waveform W. For example, a sound time difference can be determined as a temporal offset between the measurement signals 23, 24. In this case, for example, a first microphone 21 can record the detected waveform W at a first point in time and a further microphone 22 of the microphone arrangement 20 can record the detected waveform W at a further point in time. The time interval between the first point in time and the further point in time then corresponds to the sound time difference of the microphones 21, 22. However, since the detected waveform W of the fundamental mode G deflects the microphones 21, 22 of the microphone arrangement 20 at the same time and with the same amplitude A, such a propagation time difference can be leveled off to calibrate the microphone arrangement 20.For example, the propagation time difference can be leveled by offsetting one of the microphones 21, 22 of the measuring arrangement 20 by the determined propagation time difference. In this way, the microphone arrangements 20, 30 can be calibrated simultaneously, i.e., in a single step of the process. List of reference symbols 10 flow channel 11 cylindrical wall segment 12 Central axis 13 further cylindrical wall segment 14 Final segment 20 Microphone arrangement 21 Microphone 22 Microphone 23 Measurement signal 24 measurement signal 30 additional microphone arrangements 31 Microphone 32 microphone 33 measurement signal 34 measurement signal 40 speakers 41 Excitation signal 50 drive unit 51 connecting element 52 Balancing element 53 Drive signal 60 control unit 100 device A Amplitude of the waveform of the fundamental mode AS Amplitude of the waveform of another oscillation mode G Basic mode S further vibration mode S1 ... S5 step W Waveform of the fundamental mode WS Waveform of another oscillation mode

Claims

[1] Device (100) for calibrating a microphone arrangement (20, 30), comprising: - a flow channel (10), wherein the flow channel (10) has at least one cylindrical wall segment (11), - at least one microphone arrangement (20) for detecting a reference sound field, wherein the at least one microphone arrangement (20) is arranged around a central axis (12) of the at least one cylindrical wall segment (11), wherein the at least one microphone arrangement (20) has at least two microphones (21, 22), - at least one loudspeaker (40) for outputting at least one excitation signal (41), wherein the at least one loudspeaker (40) is rotatable along a circular path around the central axis (12) of the cylindrical wall segment (11), - at least one drive unit (50) for rotating the at least one loudspeaker (40), - at least one control unit (60) for controlling the calibration, wherein the device (100) is designed to carry out the following steps: - outputting (S1) the at least one excitation signal (41) by means of the at least one loudspeaker (40), wherein the output of the at least one excitation signal (41) generates a reference sound field in the flow channel (10), wherein the reference sound field has a fundamental mode (G) and a plurality of further vibration modes (S), - Rotating (S2) the at least one loudspeaker (40) along the circular path around the central axis (12) of the at least one cylindrical wall segment (11) by means of the at least one drive unit (50) at a rotation frequency, wherein the rotation of the at least one loudspeaker (40) shifts oscillation frequencies of the plurality of further oscillation modes (S) as a function of the rotation frequency, so that only one oscillation frequency of the fundamental mode (G) corresponds to an excitation frequency of the at least one excitation signal (41), - detecting (S3) the generated reference sound field by means of the at least two microphones (21, 22) of the at least one microphone arrangement (20), - detecting (S4) a waveform (W) of the fundamental mode (G) in the detected reference sound field as a function of the excitation frequency by means of the control unit (60), wherein the detected waveform (W) of the fundamental mode (G) deflects the at least two microphones (21, 22) of the at least one microphone arrangement (20) in phase and amplitude, - Leveling (S5) of runtime and / or level differences depending on the detected waveform (W) by means of the control unit (60). [2] Device (100) according to claim 1, characterized by that at least one microphone (21, 22) of the at least one microphone arrangement (20) is designed as a reference microphone. [3] Device (100) according to one of the preceding claims, characterized by that the rotation of the at least one loudspeaker (40) about the central axis of the wall segment (11) is a balanced rotation. [4] Device (100) according to one of the preceding claims, characterized bythat the device (100) is designed such that the excitation frequency of the at least one excitation signal (41) can be adjusted as a function of the rotation frequency of the loudspeaker (40), wherein the adjustment makes the excitation frequency proportional to the rotation frequency. [5] Device (100) according to one of the preceding claims, characterized by that the at least one loudspeaker (40) is designed to output an excitation signal (41) with a frequency in a range from 20 Hz to 20 kHz and / or that the at least two microphones (21, 22) of the microphone arrangement (20) are designed to detect a frequency in a range from 20 Hz to 20 kHz. [6] Device (100) according to one of the preceding claims, characterized by that a diameter of the flow channel (10) has a value in a range of 0.5 m to 5 m. [7] Device (100) according to one of the preceding claims, characterized bythat the device (100) comprises at least one further microphone arrangement (30), wherein the at least one further microphone arrangement (30) has at least two microphones (31, 32), wherein the at least one further microphone arrangement (30) is arranged around the central axis (12) of the at least one cylindrical wall segment (11). [8] Device (100) according to one of the preceding claims, characterized by that the flow channel (10) has at least one low-reflection end segment (14). [9] A method for calibrating a microphone arrangement (20, 30), comprising the steps of: - outputting (S1) at least one excitation signal (41) by means of at least one loudspeaker (40), wherein the output of the at least one excitation signal (41) generates a reference sound field in a flow channel (10), wherein the reference sound field has a fundamental mode (G) and a plurality of further vibration modes (S), - Rotating (S2) the at least one loudspeaker (40) along a circular path around a central axis (12) of at least one cylindrical wall segment (11) of the flow channel (10) by means of a drive unit (50), wherein the rotation of the at least one loudspeaker (40) shifts oscillation frequencies of the plurality of further oscillation modes (S) as a function of the rotation frequency, so that only one oscillation frequency of the fundamental mode (G) corresponds to an excitation frequency of the at least one excitation signal (41), - detecting (S3) the generated reference sound field by means of at least two microphones (21, 22) of at least one microphone arrangement (20), - detecting (S4) a waveform (W) of the fundamental mode (G) in the detected reference sound field as a function of the excitation frequency, wherein the detected waveform (W) of the fundamental mode (G) deflects the at least two microphones (21, 22) of the at least one microphone arrangement (20) in phase and amplitude, - Leveling (S5) of runtime and / or level differences depending on the detected waveform (W). [10] Method according to claim 9, characterized by that the rotation frequency of the at least one loudspeaker (40) and / or the excitation frequency of the at least one excitation signal (41) is frequency modulated.

Images