Method and signal generator for generating a servo signal for an acousto-optical element

Abstract

The invention presents methods for generating a feedback signal for an acousto-optical element (105) such that, despite amplification with a nonlinear transfer function, it exhibits the desired signal shape and that, during feedback control of an acousto-optical element (105), the nonlinearities of the amplifier (104) have no effect. The invention offers two methods based on pre-distorting the raw signal—before it is amplified to form the feedback signal—by comparing an amplified signal with the unamplified signal. One method uses the feedback signal itself, and the other uses a reference signal. (Figure 2)

Description

Title of the invention: Method and signal generator for generating a control signal for an acousto-optical element. Summary of the invention

[0001] The present invention relates to two methods and a signal generator for generating a servo signal for an acousto-optic element, as well as an arrangement of such a signal generator with an acousto-optic element and a microscope with such an arrangement. State of the art

[0002] A particular challenge in certain areas of microscopy is to provide, independently of the method used, excitation light having one or more predetermined wavelengths. Depending on the type of microscopy method and / or the type of sample, one or more beams of excitation light may be required, which generally must have predetermined spectral characteristics.

[0003] Thus, for example, in the field of fluorescence microscopy, it is important to use light with the wavelength that excites fluorescence. Different wavelengths are necessary, in particular, when the sample contains fluorescent materials with different emission wavelengths.

[0004] In the field of confocal scanning microscopy, it is particularly interesting to adapt the intensities for certain wavelengths or to turn on or off certain wavelengths.

[0005] To this end, wavelength-selective elements can be used that operate on the basis of the acousto-optic effect. Such acousto-optic elements generally comprise an acousto-optic crystal that is set into oscillation by means of an acoustic signal generator, also known as a converter or transducer. Typically, such a converter comprises a piezoelectric material and two or more electrodes in contact with this material. By electrically applying high frequencies, typically in the range of 10 MHz to 10 GHz, to the electrodes, the piezoelectric material is excited into oscillation, thereby generating an acoustic wave that propagates through the crystal. Acousto-optic crystals are distinguished by the fact that the generated acoustic wave modifies the optical characteristics of the crystal.

[0006] Examples of such acousto-optical elements are adjustable acousto-optical filters (AOTF), acousto-optical modulators (AOM), deflectors acousto-optic (AOD), acousto-optic ray splitters (AOBS) and acousto-optic ray assemblers (AOBM).

[0007] When using acousto-optic elements, their control presents a particular challenge. Usually, the high-frequency electrical signals for the converter are generated by a frequency generator (for example, a voltage-controlled oscillator (VCO), a phase-locked loop (PLL) or a synthesizer using the DDS (direct digital synthesis) method) and are amplified using a high-frequency amplifier in such a way that the amplitude is large enough to set the crystal into oscillation.When several different control frequencies are applied at the same time, light rays of several wavelengths can be deflected at the same time (e.g. with an AOTF, AOBS, AOBM, AOM) or a wavelength of an incident light ray can be deflected simultaneously into several light rays of different directions (e.g. with an AOD).

[0008] To generate a control signal with multiple frequencies, several individual frequency generators, for example DDS synthesizers, each generating, for example, a sinusoidal signal, can be combined, for example, by analog mixing of the output signals. Alternatively, digital superposition is also possible, as shown in document WO 2011 / 154501 AL

[0009] However, high-frequency amplifiers most often exhibit a non-fully linear amplification characteristic across the entire spectrum. Therefore, even if an input signal to the amplifier is already matched to the requirements of the acousto-optic elements, the amplified signal may exhibit phase shifts, amplitude shifts, and harmonics with mixing frequencies. This can result in unwanted wavelength portions being selected by the acousto-optic elements. The object of the present invention is therefore to take into account the amplifier's nonlinearities for the control of an acousto-optic element. Presentation of the invention

[0010] According to the invention, two methods and a signal generator for generating a feedback signal for an acousto-optic element, an arrangement comprising such a signal generator and an acousto-optic element, and a microscope with such an arrangement are proposed, which have the characteristics of the independent claims. Advantageous embodiments are the subject of the dependent claims and the following description.

[0011] The invention presents methods for generating a feedback signal for an acousto-optical element such that, despite amplification using a nonlinear transfer function, it exhibits the desired signal shape and that amplifier nonlinearities do not affect the feedback of the acousto-optical element. The invention proposes two methods based on predistorting the raw signal—before it is amplified to constitute the feedback signal—by comparing an amplified signal with the unamplified signal. One method uses the feedback signal itself, and the other uses a reference signal.

[0012] According to a first aspect of the invention, a raw signal is generated using at least one correction term with an IQ modulation having a setpoint portion I and a setpoint portion Q, and is then amplified to constitute the control signal, the setpoint portion I and / or the setpoint portion Q being corrected using said at least one correction term, said at least one correction term being obtained by analyzing the control signal. It is also conceivable to generate the raw signal using several IQ modulations, each with a setpoint portion I and a setpoint portion Q and a carrier frequency, and to assemble the results—in particular additively—to obtain the raw signal. This aspect of the invention essentially corresponds to regulating the signal shape of the control signal during the control process, such that very good results can be obtained.The non-linearity of the amplification transfer function is regulated by the correction of the raw signal. More specifically, changes in the transfer function during operation, for example due to temperature variations, etc., are taken into account and regulated.

[0013] Generating the raw signal using IQ modulation also has the advantage that the control signal for the acousto-optic element can be generated with the desired frequency spectrum by first transforming the frequency spectrum, using an inverse Fourier transform, into a signal in the time domain, which is then modulated, using IQ modulation, onto a carrier signal with a carrier frequency. The real part of this modulated signal is then the raw signal with the desired frequency spectrum around the carrier frequency. Alternatively, several frequency spectra can be transformed into several signals in the time domain, from which several modulated signals can then be obtained, the real parts of which are assembled to constitute the raw signal.

[0014] Particularly in cases where the control signal must have several frequencies, that is to say at least two, a single frequency spectrum which is composed of at least two frequencies, or several frequency spectra which together have said at least two frequencies, and one or more single sideband modulators are needed to easily obtain the desired servo signal for the acousto-optic element.

[0015] During IQ modulation, in a known manner, two baseband signals (I and Q) are modulated, by mixing, onto the carrier signal. One of the two baseband signals (Q) is then obtained from the transmitted signal using a Hilbert transform. For further details, reference is made to the relevant specialized literature.

[0016] The invention uses this simple generation of a servo signal for the acousto-optical element, in order to perform compensation of the non-linear amplification in the high-frequency amplifier.

[0017] Preferably, the analysis of the control signal includes an IQ demodulation to obtain a current part I and a current part Q, said at least one correction term being determined from a comparison of the current part I with the setpoint part I and / or from a comparison of the current part Q with the setpoint part Q. This form of signal analysis is particularly suitable for the proposed method, since it provides direct comparison quantities for regulation.

[0018] According to one embodiment, IQ demodulation is performed by transforming the amplifier's output signal into a digital signal, performing a complex Fourier transform, and then a coordinate transform. The subsequent steps of the process can be implemented computationally through the digital transformation. The various frequency components can be determined using the Fourier transform, whereby, after a transformation into Cartesian coordinates, the baseband signals (actual values) are obtained. From these baseband signals, by calculating differences from the initial baseband signals (setpoint values), correction values ​​are obtained. These correction values ​​can then be added to the initial baseband signals to obtain corrected baseband signals and thus a corrected raw signal.

[0019] According to another aspect of the invention, a raw signal is pre-distorted using a correction function and amplified according to a transfer function to constitute the control signal, the correction function being obtained by amplifying, according to the transfer function, a reference signal into an amplified reference signal and comparing the amplified reference signal with the reference signal. The raw signal can be generated in any way, in particular using the IQ modulation described above.

[0020] This variant has the advantage that the characteristics of the nonlinear transfer function can be determined beforehand and can be used to determine the correction function, which is notably the inverse of the function of transfer, before the actual control signal is generated. During operation, it is not necessary to perform complex calculation operations, so this method is less demanding in terms of the signal generator's capacity, but still gives very good results.

[0021] According to an advantageous implementation, the transfer function is approximated as a polynomial, in particular of the third degree. For suitable high-frequency amplifier transfer functions, such an approximation is usually sufficient and is still simple to analyze.

[0022] This analysis may preferably include a Fourier transform, in particular an FFT or DFT, to determine the coefficients of the polynomial from the amplified reference signal, after its digitization. These are well-known methods for which prefabricated signal processors are available. Determining the coefficients is particularly simple when the reference signal comprises only two frequencies.

[0023] According to an advantageous embodiment, the acousto-optical element is chosen from an acousto-optically adjustable filter, an acousto-optic modulator, an acousto-optic deflector, an acousto-optic ray splitter and an acousto-optic ray assembler.

[0024] Other advantages and embodiments of the invention will become apparent from the description and accompanying drawings.

[0025] It goes without saying that the features stated above and the features which will be explained below, can be used not only in the combinations respectively stated but also in other combinations or separately, without going out of the scope of the present invention.

[0026] The invention is schematically represented in the drawings with reference to an embodiment and is described below with reference to the drawings. Description of the drawings

[0027] [fig. 1] represents an arrangement consisting of a signal generator and an acousto-optical element according to an embodiment of the invention,

[0028] [fig.2] represents such an arrangement according to figure 1 in the form of a diagram of circuit according to a first embodiment,

[0029] [fig.3] represents such an arrangement according to figure 1 in the form of a diagram of circuit according to a second embodiment,

[0030] [fig.4] represents a schematic overview of the typical elements of a mi confocal chromosome. Detailed description of the drawings

[0031] Figure 1 schematically represents an arrangement 100 comprising a signal generator 101 and an acousto-optical element 200 according to a preferred embodiment of the invention.

[0032] The signal generator 101 is used to control the acousto-optical element 200, which comprises an acousto-optical crystal 202 and a piezoelectric converter 201 for mechanically oscillating the crystal 202. The acousto-optical element is used to exert a desired influence on light, for example, as a frequency selector.

[0033] Figure 2 shows, in the form of a function-highlighting representation, a signal generator 101 according to a preferred embodiment of the invention. In the example shown, the signal generator 101 comprises a functional unit 102 for generating signals, which unit can be implemented as an FPGA. The functional unit 102 is adapted to implement a preferred embodiment of the invention and to generate a raw (digital) signal. Furthermore, the signal generator 101 includes a digital-to-analog converter 103 for converting the raw digital signal into a raw analog signal, and an amplifier 104 for amplifying the raw analog signal into the feedback signal. The feedback signal is applied as a load to the acousto-optic element 105.

[0034] As mentioned in the introduction, amplifiers, such as amplifier 104, do not exhibit a linear transfer function over the entire frequency range, so that amplification causes distortions and the shape of the feedback signal deviates from the shape of the raw signal. The signal generator 101 is therefore adapted, according to the embodiment shown of the invention, to pre-distort (i.e., correct) the raw signal in order to compensate for these non-linearities. In the example shown, the raw digital signal is composed of three transmit signals Si(t), s2(t) and s3(t) generated by IQ modulation. The transmit signals sb s2 and s3 are each composed of two baseband signals I and Q. The two components I, Q are mixed in an I / Q 1025 modulator with a sine or respectively cosine component of a carrier signal having a frequency cOj (j=l, 2, 3), i.e. frequency shifted, and are joined to obtain a modulated signal whose real part is finally transmitted as the output signal Sj(t) (j=l, 2, 3).

[0035] The raw signal is finally obtained as a union (i.e., sum) 1026 of the real parts of said several modulated signals Si(t), s2(t) and s3(t). This digital signal obtained in the functional unit 102 and having an amplitude spectrum 1027 and a phase spectrum 1028 can be further multiplied in an element 1031 with an amplification coefficient or weighting coefficient "gain" and is then emitted as a raw digital signal.

[0036] The amplitude spectrum 1027 and the phase spectrum 1028 of the raw signal produced in this way are also shown in Figure 2.

[0037] Next, it is converted in the digital-to-analog converter 103 into a raw analog signal and is amplified in the amplifier 104 before being emitted, as a servo signal with an amplitude spectrum 1051 and a phase spectrum 1052, onto the load 105, preferably an acousto-optical element.

[0038] The amplitude spectrum 1051 and the phase spectrum 1052 of the control signal are also shown in Figure 2. It can be seen that the amplitudes of the frequencies co1, co2, and co3 have been amplified differently. Furthermore, the phase spectrum 1052 shows a phase shift as a function of frequency.

[0039] To compensate for the non-linearity of the amplifier 104, the output signal of the amplifier 104 is converted in an analog-to-digital converter 106 into a digital signal and is returned to the functional unit 102 for analysis. There, in a demodulator 1032, the current parts I_1_rec and Q_1_rec for the carrier frequency co1, the current parts I_2_rec and Q_2_rec for the carrier frequency co2, and the current parts I_3_rec and Q_3_rec for the carrier frequency co3 are first determined from this signal.

[0040] This is achieved, for example, by a fast Fourier transform and a coordinate transformation. The conversion equations for the coordinate transformation are shown in block 1033.

[0041] Next, in a comparator or difference calculator 1034, the differences AI_1, AQ_1, AI_2, AQ_2, AI_3, and AQ_3 are calculated between the setpoint parts I_1, I_2, I_1, I_3, Q_1, Q_2, and Q_3 and the actual parts I_1 rec, O_1 rec, I_2 rec, O_2 rec, I_3_rec, and Q_3_rec. Block 1035 shows the fundamental difference equation in the complex plane. The differences calculated for AI_1, AQ_1, AI_2, AQ_2, AI_3, and AQ_3 constitute the correction terms for the setpoint parts in the I / Q modulators 1025.

[0042] In this context, it is shown in Figure 2, by way of example, that the emission signal Si(t) with the frequency coi is generated from the baseband signals I_1 + AI_1 and Q_1 + AQ_1, the emission signal s2(t) with the frequency co2 is generated from the baseband signals I_2 + AI_2 and Q_2 + AQ_2, the emission signal s3(t) with the frequency co3 is generated from the baseband signals I_3 + AI_3 and Q_3 + AQ_3.

[0043] Figure 3 shows, also in a functional representation, a signal generator 101' according to another preferred embodiment of the invention, which is assembled from a functional unit 102', the digital-to-analog converter 103, the amplifier 104 and the analog-to-digital converter 106.

[0044] The functional unit 102' is adapted to implement another preferred form of the invention and to generate a raw (digital) signal.

[0045] Here too, as in Figure 2, the raw signal designated by x is assembled from three emission signals Si(t), s2(t) and s3(t) generated using an IQ 1025 modulation.

[0046] Here, unlike in Figure 2, a correction or pre-distortion is not performed by modifying the baseband signals I_1, I_2, I_3, Q_1, Q_2, and Q_3, but rather in a corrector 1040 by modifying the raw signal x in y. This method is based on determining the transfer function of the amplifier 104 and on pre-distorting the raw signal in the corrector with the inverse of the transfer function.

[0047] In the present example, the transfer function is approximated by a third-degree polynomial y = ax + [3x2 + yx3, so that, by way of example, to correct this non-linear behavior, the raw signal x is pre-distorted in the 1040 corrector with the inverse y = (ax + [3x2 + yx3)1 of the transfer function.

[0048] The coefficients a, [3 and y of the polynomial are determined before the actual control by the use of a reference signal having a known frequency spectrum 1029. This reference signal consists, for example, of a sum of two cosine functions x = A cos cot * B cos (φ²t) with different frequencies coi and φ². Amplifying this raw signal in amplifier 104 results in a frequency spectrum 1051, which is also shown in the figure. Due to the non-linearity of amplifier 104, in addition to the two frequencies coi and φ², harmonics 2πi, 2φ², 3πi, and 3π² are generated, as well as mixing frequencies 2πorπ² and 2π²πi. Other harmonics and mixing terms exist, but these are neglected for the subsequent calculations.

[0049] The output signal of amplifier 104 is converted in analog-to-digital converter 106 into a digital signal and is returned, for analysis, to functional unit 102. Here, a complex Fourier transform 1037 is performed to determine the frequency spectrum. In a device 1038 for determining amplitudes, the amplitudes are determined at the frequencies of interest a>i, 2a>i and 3a>i. The amplitude at a>i results from part I and part Q at frequency a>i as (I_(0i)2 + (Q_tt>i)2- The amplitude at 2a>i results from part I and part Q at frequency 2a>i as (I_2(üi)2 + (Q_2(üi)2 and the amplitude at 3(üi results from part I and part Q at frequency 3(üi) as (I_3(üi)2 + (Q_3(üi)2.

[0050] In a device 1039 for determining coefficients, the amplitudes of the polynomial coefficients are then determined. The amplitude at a > i corresponds to aA. The amplitude at 2a > i corresponds to 3A². The amplitude at 3i corresponds to yA³. Since A is known from the test signal, the polynomial coefficients can thus be determined. In this way, the raw digital signal x can be pre-distorted, as desired, in the corrector 1040 using a correction function.

[0051] Figure 4 schematically shows a confocal microscope with typical components. 500 designates the complete system. The confocal scanning and detection unit is referenced as 505. The associated illumination device is referenced as 506. In the lighting system, an arrangement according to figure 1 is provided.

[0052] With regard to 508, it is a laser light source which is connected to the lighting device 506 by a lighting fiber 507. If desired, the laser light is influenced in the lighting device 506 by means of an acousto-optical element.

[0053] 504 designates an optical adapter for the confocal scanning and detection unit 505 is located on the microscope stand 501. Inside the stand 501 is the specimen table 502 with a sample 503 to be examined. A control unit 509 is connected by corresponding connecting lines to the various components 508, 506, 505, and 501. A computer with control and display programs is referenced at 510; it is also connected to the control unit 509.

[0054] According to a first variant, inside the confocal scanning and detection unit 505, a conventional confocal beam path is arranged which is made up, in a known way, of a single pinhole and a beam scanner, for example a mirror scanner.

[0055] According to a second embodiment, a beam path is located inside the confocal scanning and detection unit 505 with which the sample is simultaneously illuminated by one or more illumination points or by illumination points extended in one direction. Correspondingly, the photons to be detected are selected, for example, with a geometric arrangement of pinholes.

[0056] The sample 503 to be examined is illuminated by means of a microscope optic and is reproduced by means of the same microscope optic, in particular on a sensor arrangement 511 which, according to the embodiment of the confocal scanning and detection unit 505, consists of a photomultiplier or a photomultiplier array. The operating principle of a system 500 shown in Figure 4 is sufficiently known in itself and therefore does not need to be explained here. List of reference signs 100 layouts 101, 101' signal generator 102, 102' functional unit 103 digital-to-analog converter 104 amplifier 105 charge 106 analog-to-digital converter 1025 I / Q modulator 1026 adder 1027, 1028, 1029 spectrum 1051, 1052 spectrum 1031 amplifier (digital) 1032 I / Q demodulator 1033 coordinate transformation 1034 difference calculator 1035 difference in complex plane 1037 Fourier transformer 1038 device for determining amplitudes 1039 device for determining coefficients 1040 corrector 200 acousto-optical elements 201 piezoelectric converter 202 acousto-optical crystal 500 microscope system 501-foot microscope with microscope optics 502 object table 503 sample 504 optical adapter 505 confocal scanning and detection unit 506 lighting device 507 fiber optic lighting 508 laser light source 509 control unit 510 computer with command and display programs 511 sensor arrangement

Claims

Demands

1. A method for generating a feedback signal for an acousto-optical element (105; 200), a raw signal being generated using at least one correction term (AI_1, AI_2, AI_3, AQ_1, AQ_2, AQ_3) with an IQ modulation (1025) having a setpoint portion I (I_1, I_2, I_1_3) and a setpoint portion Q (Q_1, Q_2, Q_3), and being subsequently amplified (104) to constitute the feedback signal, the setpoint portion I (I_1, I_2, I_1_3) and / or the setpoint portion Q (Q_1, Q_2, Q_3) being corrected using said at least one correction term (AI_1, AI_2, AI_3, AQ_1, AQ_2, AQ_3), said at least one correction term (AI_1, AI_2, AI_3, AQ_1, AQ_2, AQ_3) being obtained by an analysis (1032, 1034) of the control signal.

2. Method according to claim 1, the analysis of the control signal comprising a demodulation IQ (1032) to obtain a current part I (I_1_rec, I_2_rec, I_3_rec) and a current part Q (Q_1_rec, Q_2_rec, Q_3_rec), said at least one correction term (AI_1, AI_2, AI_3, AQ_1, AQ_2, AQ_3) being determined from a comparison (1034) of the current part I (I_1_rec, I_2_rec, I_3_rec) with the setpoint part I (I_1, I_2, I_3) and / or from a comparison of the current part Q (Q_1_rec, Q_2_rec, Q_3_rec) with the setpoint part Q (Q_1, Q_2, Q_3).

3. Method according to claim 2, the IQ (1032) demodulation comprising a complex Fourier transform and a coordinate transform.

4. Method for generating a servo signal for an acousto-optical element (105; 200), a raw signal (x) being pre-distorted using a correction function (1040) and being amplified (104) according to a transfer function to constitute the servo signal, the correction function being obtained by amplifying, according to the transfer function, a reference signal into an amplified reference signal and comparing the amplified reference signal with the reference signal.

5. Method according to claim 4, the transfer function being determined from the comparison of the amplified reference signal with the reference signal.

6. Method according to claim 4 or 5, the transfer function being approximated as a polynomial with coefficients, in particular of third degree.

7. Method according to claim 6, the coefficients of the polynomial being determined from the amplified reference signal, using a Fourier transform (1037).

8. A method according to any one of claims 4 to 7, the correction function being obtained from the inverse transfer function.

9. A method according to any one of the preceding claims, wherein an acousto-optical element (200) is controlled by the control signal.

10. Signal generator (101, 101') for controlling an acousto-optical element (105; 200), which is adapted to implement a method according to one of the preceding claims.

11. Arrangement (100, 100') consisting of at least one signal generator (101, 101') according to claim 10 and an acousto-optical element (105; 200).

12. Arrangement (100, 100') according to claim 11, the acousto-optic element (105; 200) being selected from an acousto-optically adjustable filter, an acousto-optic modulator, an acousto-optic deflector, an acousto-optic ray splitter and an acousto-optic ray assembler.

13. Microscope (500) with an arrangement (100, 100') according to claim 11 or 12.