Optical measuring device and method for gas detection

Description

[0001] The invention relates to an optical measuring system and a method for gas detection with a light emitter and at least one light detector as optoelectronic components, and with at least one optomechanical component arranged between the light emitter and the at least one light detector, which has optically effective interfaces for light beams, wherein the light emitter and the at least one light detector are arranged together in a housing or separately in separate housings, and the housing optionally has a housing window that forms an optomechanical component, and wherein the light emitter emits a modulated principal light beam of a mean wavelength λ 0 with a modulation span Δλ,and at least one scattered light beam of the main light beam, partially reflected / scattered from the main light beam at at least one optically effective interface of at least one optomechanical component in the direction of the light emitter and / or light detector, leads to self-mixing in the light emitter and / or to etalons at the light detector through interference with the main light beam, and wherein the interference of the main light beam with the scattered light beam results in an interfered main light beam, and wherein the measurement signal of the light detector comprises a noise signal component and a main signal component, and the noise signal influences a measurement signal of the light detector demodulated with a demodulation frequency.

[0002] Such optical measurement systems are used for various measurement tasks, such as length measurement or spectroscopy. In these systems, a laser diode is often used as the light emitter and a suitable photodiode as the light detector. Laser absorption spectroscopy, for example, is used in gas detection. In this process, the main light beam emitted by the light emitter is detected by the light detector after passing through a gas or gas mixture, and the received signal is fed to a signal analyzer, particularly a lock-in amplifier, for evaluation. The signal analyzer separates constant interference patterns from the measurement signal of the light detector. However, it cannot completely eliminate time-varying interference patterns from the received signal, so the detection sensitivity for the gas to be detected decreases significantly due to the increased noise.Triggers for time-varying interference patterns include temperature influences that alter the length of the optical path for the main light beam from the light emitter to the light detector. Furthermore, reflections and / or scattering of the main light beam at internal surfaces of the measuring system housing, at interfaces of beam-shaping and / or beam-directing optical or optomechanical components located within the housing (such as lenses or mirrors), or at the inner or outer surface of the housing window can cause such interference patterns. These reflections or scatterings can result in stray light beams directed towards the light emitter and / or the light detector.Such scattered light rays cause self-mixing in the light emitter and / or etalons at the light detector by interfering with the main light beam, forming interference patterns and creating an interfered main light beam that is received by the light detector. These interference patterns are also temperature-dependent and can therefore change over time. Self-mixing and etalons are caused by the different optical path lengths of the scattered light rays from the optical and / or optomechanical component that partially reflects the main light beam to the light emitter or light detector, respectively, compared to the optical path length of the main light beam. For etalons, the distance between the reflecting or...The scattering optical and / or optomechanical component up to the light detector is decisive for self-mixing; the distance of the corresponding component to the aperture of the light emitter, which forms a "receiver" for the scattered light beam entering the resonator through the aperture and interferes with the main beam partially reflected back on the inside of the laser, is decisive for self-mixing.

[0003] In laser absorption spectroscopy for gas detection, the wavelength modulation method is frequently used. In this method, the wavelength and usually also the intensity of the main light beam from the light emitter, for example, a continuously tunable diode laser, is modulated at a frequency f, with the wavelength being varied across a possible absorption spectrum of the sample being analyzed. The laser light is absorbed by the gas sample when the wavelength of the light corresponds to, or varies across, the gas's resonance frequency. When the main light beam, after passing through the gas sample, reaches the light detector, e.g., a photodiode, the output signal of the light detector includes AC voltage components at the modulation frequency f and at the superimposed higher harmonic frequencies mf, where m is a positive natural number.Demodulation of the output signal of the light detector at a harmonic frequency mf shifts the measurement to a higher frequency band mf with lower 1 / f noise, thereby improving the measurement sensitivity of the optical measurement system.

[0004] When using lasers as light emitters, the relatively large coherence length makes interference between the main light beam emitted by the light emitter and scattered light beams, caused by unwanted reflections or scattering of the main beam, particularly problematic. Two different phenomena must be distinguished here. The phenomenon known as self-mixing, where light emission is influenced by radiation fed back into the laser aperture, has a particularly strong effect during measurement because the feedback radiation is amplified within the laser. In practice, the main cause of self-mixing is often the scattering or reflection of the main light beam at the housing window, which is designed to allow the main light beam to exit and protects the laser light emitter from environmental influences such as contamination or ambient humidity.This phenomenon also occurs with an inclined housing window due to scattering, as some of the light can be fed back into the laser aperture directly or indirectly, for example, through scattering at the inner housing wall. Other surfaces located further away from the light emitter aperture can also cause self-mixing (e.g., reflection / scattering at the photodiode). Typically, however, their influence is less pronounced. The second phenomenon that occurs in optical measurement systems is interference on the detector, which arises from the different optical path lengths of the main and scattered beams and is called etalons. Etalons are caused, analogous to self-mixing, by scattered light rays from all optoelectronic and optomechanical components in the measurement system, which then interfere with the main beam on the detector.To minimize the distortion of the light detector's measurement signal and thus the measured values ​​due to self-mixing and / or etalons, either the self-mixing or the etalons themselves must be reduced, or their influence on the measured value must be minimized.

[0005] Various measures are known in the art to mitigate the aforementioned disturbances. These include, in particular, the anti-reflective coating of all housing windows, the wedge-shaped design of the housing windows, and the tilted, i.e., inclined, arrangement of the housing windows in the beam path to prevent scattering as much as possible or to deflect it from the beam path of the main light beam, which could cause interference with the main light beam and thus self-mixing. Furthermore, absorbing coatings in the housing of the optical measuring system and / or apertures arranged in the beam path can reduce reflections and scattering. Despite these effective measures, it is not possible to prevent all reflections and scattering of the main light beam that lead to self-mixing, which generally has a strong negative impact on the measurement signal of the light detector in optical systems with coherent radiation.During operation, the housing windows and surfaces of such optical measuring systems often become contaminated, for example, by dust or condensation. This contamination significantly increases the scattering of the main light beam, and the aforementioned interference effects intensify over the system's lifetime. Therefore, for the lifespan and service cycle time of such optical measuring systems, it is advantageous to design the system in such a way that the unavoidable contamination has as little impact as possible on the system's measurement signal.

[0006] When measuring spectra, all the aforementioned disturbances are typically visible as periodic signals, which are referred to as "fringes" by those skilled in the art. Self-mixing and etalons in tunable diode laser spectroscopy (TDLS) systems lead to a distortion of the measured gas absorption line, which cannot be easily eliminated by calibration or computational methods. Periodic disturbances with a period on the order of the width of the measured absorption line are particularly problematic.

[0007] In principle, the phenomenon of self-mixing can be greatly reduced with an optical isolator. Unfortunately, good isolators for non-telecom wavelengths are very expensive, so they are rarely used in industrial applications. Furthermore, it should be noted that the optical isolator itself also has optical interfaces, and the optical interface of the optical isolator facing the laser can also generate self-mixing through reflection / scattering. Persson et al. propose a method called "intensity referencing," which reduces the effect of self-mixing using balanced detection. However, this method only achieved interference amplitude reductions of approximately a factor of 10 (Appl. Phys. B 87, 523-530 (2007)). Webster describes a very simple method to prevent etalons by inserting a plane-parallel plate, transparent to the laser wavelength, at an angle into the optical path.Periodically tilting the inclined plate during measurement ensures that the interference signal averages out due to the periodically changing optical path length. This solution reduced the interference signal by a factor of 30 (Opt. Soc. Am. B 2, 1464 (1985)). To avoid wear on mechanically moving components, thus increasing the sensor's lifespan, piezo-based actuators can be used. However, these typically only allow for small path difference variations, making them efficient only for averaging interferences with a very small free spectral range compared to the gas absorption line width. For example, Silver and Stanton use a piezoelectric transducer (patent US4934816) that varies the longitudinal deflection of the mirror in a multipath cell, thereby averaging out interferences caused by the mirror arrangement.However, such piezoelectric transducers are usually relatively expensive. Reid et al. have shown that by applying an additional, slower frequency (compared to the actual modulation frequency) to the modulation signal, it is possible to average such interferences without changing the optical path lengths (Appl. Opt. 19, 3349-3354 (1980)). However, this principle is only practically applicable to interferences with a very small free spectral range, since averaging must be performed over at least one period. Nearly the same result can also be achieved by low-pass filtering, for example, by using signal averaging in data processing. In this case, the free spectral range must differ significantly from the width of the gas absorption line; otherwise, the absorption signal will also be affected during averaging. Shang-Qian Wu et al. describe a similar approach.Ehlers et al.: "Efficient Reduction of Fringe Noise in Trace Gas Detection with Diode Laser Multipass Absorption Spectroscopy", JAPANESE JOURNAL OF APPLIED PHYSICS, JAPAN SOCIETY OF APPLIED PHYSICS, JP, Vol. 39, No. 7A, July 1, 2000 (2000-07-01), pages 4034-4040, to choose the length of a multipass cell large enough so that the free spectral range is small and the demodulated measurement signal can thus be "filtered" in post-processing. Ehlers et al.: "Use of etalon-immune distances to reduce the influence of background signals in frequency-modulation spectroscopy and noise-immune cavity-enhanced optical heterodyne molecular spectroscopy", J. Opt. Soc. Am. B 31, 2938-2945, 2014 describes improvements in the field of FM spectroscopy, i.e., in the field of frequency modulation spectroscopy. The previous proposed solutions are mostly only efficiently applicable to a specific problem.For example, an optical isolator in front of the laser will only suppress self-mixing, but not etalons on the detector. Most of the cited solutions aim to reduce interference on the averaged measurement signal from the detector.

[0008] Based on the prior art described above, the invention aims to propose a method for significantly reducing the influence of the interference signal of the light detector's measurement signal on the demodulated measurement signal, so that the main signal is clearly recognizable in the measurement signal and thus the measurement sensitivity of the optical measurement system is greatly improved.

[0009] The claimed subject matter is defined in the independent claims. Advantageous further developments are described in the dependent claims.

[0010] The invention is based on the core idea of ​​using the properties of wavelength modulation spectroscopy (WMS) to reduce interference. This is achieved by adjusting the distances in the measurement system to match the modulation range of the light emitter, thus preventing interference in the light detector's measurement signal from being transferred to the demodulated signal (e.g., the 2f signal). By carefully selecting the distances in the measurement system, all dominant interferences in the demodulated measurement signal can, in principle, be suppressed or at least reduced.

[0011] According to the invention, the spatial orientation and / or distance of at least one of the optoelectronic and / or optomechanical components of the optical measuring system relative to another optoelectronic and / or optomechanical component is adapted to the modulation span Δλ of the modulated main light beam of wavelength λ₀ and is designed such that a Fourier component of the selected demodulation frequency for the modulation span Δλ is minimal or zero for all phases, wherein an optical path length difference ΔS for the main light beam relative to the scattered light beam exists at the selected modulation span Δλ of the modulated main light beam of wavelength λ₀, wherein the free spectral range with a period A of the measurement signal results from the path length difference ΔS, which is selected such thatthat, with the aid of the given modulation span Δλ of the modulated main light beam of wavelength λ₀, the effect of the interference on the measurement signal of the demodulated signal is canceled out or at least minimized for all phases. The period A is determined by the spacing of the transmission maxima. It results from the interference of the main signal with the interference signal. This means that the corresponding spacings and the modulation span Δλ are matched to each other. The modulation span Δλ is defined as the peak-to-peak amplitude of the modulation signal, which can have any modulation shape. This affects any optical and / or optomechanical components at which reflections or scattering can occur, forming scattered light beams.which propagate towards and are directed towards the light emitter and / or the light detector. The at least one scattered light ray is considered part of the uninterrupted or the interfered main light ray. Thus, both the main light ray and the scattered light rays originate in the aperture of the light emitter, since the scattered light rays are split off from the main light ray during its partial reflection / scattering.

[0012] The invention preferably relates to an optical measuring system comprising a laser diode as a light emitter and a photodiode as a light detector, which form optoelectronic components of the measuring system, and comprising at least one housing window arranged between the light emitter and the light detector, and comprising means for controlling the light emitter and for processing the measurement signal of the light detector. The light emitter, the light detector, and the at least one housing window for the passage of the main light beam constitute optoelectronic and / or optomechanical components of the measuring system, wherein the light emitter and the light detector are arranged in a common housing with one housing window or in separate housings, each with one light-transmitting housing window. In addition to the light emitter, light detector, and at least one housing window, the measuring system may also include further optoelectronic, optomechanical, and / or mechanical components.The optical measuring system includes elements arranged inside or outside the at least one housing. These elements can be, for example, mirrors for convolution of the beam path or a gas cell with windows into which the gas substance to be examined is introduced. The light emitter emits a modulated main light beam with a mean wavelength λ₀ and a modulation range Δλ. Typically, the light emitter with a mean wavelength λ₀ and a defined modulation range Δλ is modulated at a frequency of f wavelengths, and the signal received by the light detector is demodulated. The modulation can be performed with any waveform, for example, a triangular or cosine waveform. In In practice, the light emitter is preferably modulated with a point-symmetric waveform, where the modulation pattern contains odd Fourier components (1f, 3f, 5f, etc.). Characteristically, the Fourier component of the measurement signal, i.e., the measurement signal at two, four, and / or m times the modulation frequency, is determined such that m = 2, 4, 6, ...., thereby suppressing the amplitude modulation component in the measurement signal. In In practical applications, the Fourier component is usually determined at twice the modulation frequency or at an even integer multiple m of the modulation frequency. If the Fourier component is measured at twice the frequency, the result is the so-called 2f signal; at four times the frequency, the 4f signal, and so on.

[0013] In accordance with the core concept of the invention, the relative position and / or orientation of the optoelectronic and / or optomechanical components in the beam path, where undesired reflections or scattering can occur, is optimized with respect to the position of the light emitter and the light detector of the optical measuring system such that the resulting interferences have a minimal impact on the measurement signal. In particular, the distance between the housing window and the light emitter, and preferably also to the light detector, is selected such that the influence of these effects is minimal. Furthermore, the distances of other optoelectronic and / or optomechanical components to each other and to the light emitter and / or the light detector can also be selected such that the disturbances they cause have a minimal impact on the measurement signal emitted and demodulated by the light detector.

[0014] In principle, there are two independent ways to reduce the aforementioned disturbances by optimizing the position of the optoelectronic and / or optomechanical elements in the beam path. Firstly, by appropriately selecting the largest possible distance between an optoelectronic and / or optomechanical component of the optical measuring system and the aperture of the light emitter, the optical path length of the scattered light beam can be increased to such an extent that the power fed back into the aperture of the light emitter (self-mixing) is attenuated, particularly if the reflection or scattering is diffuse. In particular, increasing the optical path length leads to a reduced noise component of the measurement signal. If the cause of the reflected scattered light rays is point-like scattering bodies (dust, scratches, etc.),When the aperture is located at a certain distance, the scattered light typically exhibits a diffuse characteristic where the power decreases proportionally to (1 / I) 2 < (I = distance). Thus, for example, increasing the distance between the aperture of the light emitter and distant optoelectronic or optomechanical components such as windows, lenses, mirrors, etc., can attenuate the magnitude of the interference signal relative to the proportion of the main signal in the measurement signal. The same applies to mechanical and / or optomechanical components such as apertures, internal housing parts, etc., that are touched or struck by the main light beam.

[0015] Furthermore, in addition to diffuse scattering, non-diffuse scattering of the scattered light can also be achieved by a suitably chosen, specifically defined distance between the optoelectronic and / or optomechanical and / or mechanical components of the optical measuring system according to the invention. This reduces the influence of the interference signal on the measurement signal emitted by the light detector, which may be disturbed, for example, by self-mixing or etalons, upon reaching the light detector. The interference signal of the measurement signal generally consists of a number of interference signal components caused by reflections or scattering at the various optoelectronic and / or optomechanical and / or mechanical components of the optical measuring system, with the respective scattered light rays forming a scattered light beam that extends in the direction of the light emitter or the light detector.By adjusting the distances of the respective components to each other, it is possible to reduce the respective Fourier component of the corresponding interference signal of the scattered light beam, i.e., the Fourier component of the selected demodulation frequency for the selected modulation span Δλ disappears, or at least is greatly minimized.

[0016] Frequently, the housing window of the at least one housing of the optical measuring system, in which the light emitter and / or the light detector are arranged, is the main cause of stray light rays from the main light beam. These stray rays interfere with the main light beam to form an interfered main light beam, resulting in a measurement signal that comprises a noise component and a main signal component, with the noise component being coupled to the main signal. This applies regardless of whether the light emitter and the light detector are arranged in separate housings or in spaced-apart housing sections of a common housing, and applies to both the light emitter and the light detector. Accordingly, according to the invention, the distance between the light emitter, the at least one housing window, and / or the light detector is preferably determined.between the light detector, the at least one housing window and the light emitter, and thus the optical path length for the main light beam and / or the at least one scattered light beam, is selected such that the period A of the interference signal caused by these components in the measurement signal meets the conditions specified above.

[0017] As already mentioned, the optical measuring system according to the invention for gas detection comprises a light emitter, a light detector, and at least one housing window. The optimal distances between any two of these optoelectronic and / or optomechanical components of the optical measuring system, preferably between the light emitter and the at least one housing window and / or the light detector, can be empirically determined or calculated for any selected waveform with which the main light beam is modulated.

[0018] In the empirical determination of suitable distances, the distance between any two of the optoelectronic or optomechanical components of the optical measuring system is changed in small increments, and the measurement signal output by the light detector is demodulated. For example, a Fourier component of the measurement signal is determined at twice the modulation frequency. Distances can be identified at which the Fourier-1 component of the selected demodulation frequency disappears or is at least significantly minimized. At such points, the demodulated measurement signal essentially contains the main signal, as the effects of the interfering signal on the main signal are largely eliminated.

[0019] When calculating favorable distances L between any two of the optical or optomechanical components of the optical measuring system, such as the light emitter and the housing window or the light detector and the housing window, it can be assumed that these components form a low-finance Fabry-Perot etalon. For calculating the optical path length difference ΔS, a distinction must be made between the phenomena of self-mixing and etalons. In the following, S1 denotes the optical path length from the light emitter to a reflector and the optical path length from the reflector to the light detector. This refers to a symmetrical arrangement, in which the optical path length from the light emitter to the light detector is consequently 2*S1, with reference to which the invention is described. It is obvious that in an asymmetrical arrangement, i.e.,With different distances between the emitter and the detector relative to the reflector, the different optical path lengths add up. For the following description of the invention, a non-interfering main light beam, an interfered main light beam, and a scattered light beam are assumed, wherein the main light beam, depending on the configuration, travels path length S1 and the scattered light beam travels partial path lengths S1', S1", S1‴, etc.

[0020] For the sake of clarity, the phenomena of "self-mixing" and "etalons" will be described separately in the following sections, even though they can occur simultaneously. The formulas that follow represent only one or the other phenomenon.

[0021] In the case of etalons, starting from the aperture of the light emitter, the main light beam and the scattered light beam travel different optical paths until they interfere with the main light beam on the detector. The optical path length difference of the main light beam compared to the at least one interfering scattered light beam is thus given by ΔS = |S scatter - S main |. If part of the main light beam is reflected at the photodiode surface of the light detector and then returns to the detector via the housing window of the light detector located at a distance L, the optical path length difference, ΔS = |S scatter - S main |, is equal to twice the distance between the photodiode and the housing window of the light detector.Thus, ΔS = |S scatter - S main | = |(2*S1 + 2*S1') - 2*S1| = 2*S1' = 2L is obtained as the path length difference, where 2*S1 represents the distance between the aperture of the light emitter and the detector surface and S1' corresponds to the distance L of the housing window of the light detector to the light detector.

[0022] In the case of self-mixing, interference occurs between the scattered light beam, which is fed back into the aperture of the light emitter, and the uninterrupted main light beam within the light emitter, for example, a laser, so that the respective interfered main light beam emerges from the aperture of the light emitter. Since, in the case of self-mixing, the uninterrupted main light beam is entirely within the light emitter, Smain must be set to 0 for self-mixing, resulting in the optical path length difference ΔS = |Sscatter - Smain| = |Sscatter - 0| = Sscatter, which arises solely from the path of the scattered light beam outside the aperture of the light emitter. Im If the light is directly backscattered from the housing window of the light emitter, the optical path traveled by the scattered light ray is Sscatter = 2L, i.e., twice the distance between the aperture of the light emitter and the point where the main light ray hits the housing window, and thus ΔS = Sscatter = 2L. If the light is reflected indirectly back into the aperture of the light emitter, e.g., via a housing wall, the path of the scattered light ray is correspondingly longer.

[0023] The transmission spectrum of such an arrangement exhibits narrow transmission maxima at "fringes" for wavelengths that fulfill the resonance condition, while other spectral ranges are almost completely canceled out in the transmission. This occurs through constructive or destructive interference of the light rays traveling in the resonator or detector. It should be noted that there are no "true" Fabry-Perot etalons in the optical measuring system according to the invention, since the design ensures that parallel surfaces are avoided as much as possible. However, a fringe can also be caused by scattered light rays. The principles known from the Fabry-Perot etalon for parallel resonator surfaces can be applied analogously to inclined resonator surfaces and thus also to the optical measuring system according to the invention.

[0024] The distance between the transmission maxima is called the free spectral range of the resonator (FSR) and is denoted by A. This free spectral range A depends on the optical path length difference ΔS between the principal light beam and the at least one scattered light beam on the light detector. The free spectral range A is known to be calculated as follows: Λ = λ 0 2 ΔS , where λ 0 is the mean wavelength of the modulated main light beam.

[0025] The following general relationship applies to the amplitude A of the interference signal of the light detector of the optical measuring system according to the invention, which depends on the optical path length difference. A detector = A fringe cos 2 π Λ λ + ψ fringe .

[0026] This has a maximum amplitude A fringe and a phase ψ fringe underlying.

[0027] The modulation of the light emitter can be described by the formula λ = λ 0 + Δλ 2 ⋅ waveform 2 πft . can be described, where waveform is the modulation shape and f is the modulation frequency for the main light beam. The amplitude A of the AC component of the considered demodulated measurement signal from the light detector is thus given by A detector t λ 0 Δλ = A fringe cos 2 π Λ Δλ 2 ⋅ waveform 2 πft + ψ fringe ′ , where the following applies to this phase: ψ fringe ′ = ψ fringe + 2 πΔS λ 0 .

[0028] The amplitude A of an exemplary lock-in reference signal for the considered 2f signal is obtained according to the formula A lockin , 2 f t = A lockin cos 4 πft .

[0029] Generalized and therefore valid for all mf signals, i.e. for 2f, 4f, 6f signals, etc., the formula is: A lockin , mf t = A lockin cos 2 πmft .

[0030] The lock-in amplifier used multiplies the light detector signal. A detector ( t, λ 0 , Δ λ ) with the reference signal A lockin, 2 m(t) and then integrates a number of N periods, where the measurement time per period is inversely proportional to the modulation frequency f. For an mf signal, this yields the dependence of the amplitude A A mf λ 0 = f N ∫ 0 t = N f A detector t λ 0 Δλ ⋅ A lockin , mf t dt .

[0031] The optical path length difference ΔS is now chosen such that for a modulation signal with a given waveform and modulation span Δλ, the above formula applies to all possible phases. ψ fringe ′ A mf minimized or to A mt = 0 will be.

[0032] The calculation is analytically very complex depending on the waveform, but numerically very simple. The formulas above apply to all phases and are therefore independent of the phase setting, i.e., the phase relationship between the main light beam and the at least one scattered light beam.

[0033] It follows that the Fourier component of the demodulated measurement signal of order 2f, 4f, 6f, etc., exhibits a behavior dependent on the distance L between the respective optical or optomechanical components, whereby the amplitude of the interference signal in the demodulated measurement signal becomes zero at certain intervals due to precisely defined optical path length differences ΔS of the main light beam and the at least one scattered light beam. If the optical path length difference ΔS of the main light beam and the at least one scattered light beam is not optimal, a phase-dependent demodulated measurement signal results.

[0034] With an optimized distance L, the demodulated measurement signal from the light detector is phase-independent. Therefore, temperature-induced phase changes, for example, have no or only a greatly reduced effect on the demodulated measurement signal.

[0035] A distance-dependent intensity curve exhibits successive peak-shaped intensity regions, comprising a main maximum with a highest amplitude peak and a number of intermediate maxima, each with reduced amplitude peaks. The position of the main maximum relative to the intermediate maxima depends on the order of the demodulated mf measurement signal used. For a 2f signal, the main maximum begins as the first region at the origin of the coordinate system, with the intermediate maxima located to the right of the main maximum. For a 4f signal, an additional intermediate maximum lies between the origin and the main maximum, and for a 6f signal, two intermediate maxima are located to the left of the main maximum, and so on. The position of the main maximum is determined by the m / 2 order of the demodulated measurement signal used. For a 2f signal, the main maximum forms the first peak of the intensity curve, for a 4f signal the second, for a 6f signal the third, etc.The width of the intermediate maxima ΔX is determined by the mean wavelength λ₀ and the modulation span Δλ. This results in the width of the principal maximum being 2ΔX. The distance L is then chosen such that the optical path length difference ΔS of the principal light beam to the at least one scattered light beam is proportional to the width ΔX, i.e., a multiple thereof. The minima – zeros – of such an intensity curve, with the exception of the two zeros that define the principal maximum, are spaced ΔX apart.

[0036] Preferably, a significant reduction in the effect of the interference signal on the main signal in the measurement signal is already achieved if the optical path length difference ΔS is greater than 0.5*ΔX*(m+1) by suitable selection of the distance L between the optoelectronic and / or optomechanical components, with m=2, 4, 6, ..., corresponding to the order 2f, 4f, 6f,... of the measurement signal used.

[0037] In general, the optimal optical distance L between the optoelectronic and / or optomechanical components, at which the effect of the interference signal is canceled out, is given by the following relationship for the optical path length difference ΔS, for an order m of the demodulated measurement signal used: ΔS = n*ΔX, where n = 1, 2, 3, ..., and where n ≠ m / 2, where n is a natural integer greater than zero, with m = 2, 4, 6, ..., corresponding to the order 2f, 4f, 6f, ... of the measurement signal used. The restriction to n ≠ m / 2 excludes the possibility that, with respect to a 2f signal, n cannot be 1, for a 4f signal it cannot be 2, for a 6f signal it cannot be 3, etc. This excludes the intensity region with the highest amplitude peak, i.e., the main maximum, which has twice the width of the adjacent intermediate maxima. This defines ΔX as the smallest distance between the minima of the intensity curve.This is because if n=m / 2 is allowed, the worst-case scenario occurs, meaning the interference signal reaches its maximum. When evaluating the 2f signal, this occurs for n=1 (m=2) and therefore for ΔS=ΔX. For this reason, ΔS>ΔX should be chosen, as the interference signal decreases significantly between ΔS=ΔX and ΔS=2*ΔX.

[0038] The above explanations concern the optimal adjustment of the distance L for canceling out the interference signal. In practice, however, it may suffice to minimize the interference signal by allowing a bandwidth around the zeros. This means that for ΔS=n*ΔX, n is not an integer. The condition with n=1, 2, 3,... then merely serves to find the points at which the interference signal becomes minimal or zero.

[0039] In preferred embodiments of the invention with any modulation form of the main light beam, the optoelectronic and / or optomechanical components are arranged relative to each other and / or to the light emitter such that ΔS is greater than 0.8 times or less than 1.2 times n*ΔX, with n ≠ m / 2 and with m = 2, 4, 6,..., according to the order 2f, 4f, 6f,... of the measurement signal used. Sufficient reduction of self-mixing and etalons can already be achieved if the spacing of the optoelectronic and / or optomechanical components is chosen such that ΔS is greater than 0.7 times or less than 1.3 times n*ΔX, with n ≠ m / 2 and with m = 2, 4, 6,..., corresponding to the order 2f, 4f, 6f,... of the measurement signal used.

[0040] In one embodiment of the invention, the main light beam is modulated with a triangular waveform. Thus, to minimize or eliminate the effect of the interference signal on the main signal in the measurement signal, the suitable distance L between the optoelectronic and / or optomechanical components is obtained when ΔS = n*ΔX, with ΔX = λ 0 2< / Δλ, where the restriction applies that n ≠ m / 2.

[0041] The Fourier component under consideration vanishes in a triangular modulation, for example, whenever the optical path length difference ΔS = n*ΔX = n* λ The value 0 2< / Δλ is given by , where n = 1, 2, 3,... is a natural integer greater than zero. It should also be noted that n ≠ m / 2 is chosen and m = 2, 4, 6,..., corresponding to the order 2f, 4f, 6f, ... of the measurement signal used.

[0042] In the case of triangular modulation and demodulation with 2f, the distance L is preferably chosen such that the period A of the demodulated measurement signal corresponds to half the modulation span Δλ of the main light beam. Here, n=2 is chosen in the formula ΔS=n*λ 2< / Δλ. In the case of more distant optical or optomechanical components, the period A is chosen such that a multiple nA of the period A corresponds to the modulation span Δλ.

[0043] In an embodiment of the invention with a triangular modulation of the main light beam, the optoelectronic and / or optomechanical components can be arranged relative to each other and / or to the light emitter such that, due to the selected distance, the period A is less than 0.8 times the modulation span Δλ, or a multiple nA of the period A is greater than 0.8 times or less than 1.2 times the modulation span Δλ of the main light beam, with as before n ≠ m / 2 and with m = 2, 4, 6 and so forth. Sufficient reduction of self-mixing and etalons can already be achieved if the distance L of the optoelectronic and / or optomechanical components is chosen such that a multiple nA of the period A is greater than 0.7 times or less than 1.3 times the modulation span Δλ, also with n ≠ m / 2 and m = 2, 4, 6, etc.

[0044] In one embodiment of the invention, ΔS corresponds to twice the distance L between the light emitter and the housing window (ΔS=2L), since the dominant scattered radiation is scattered directly back to the light emitter and scattering via the housing wall can be neglected (blackened inner wall of the housing).

[0045] In a favorable embodiment of the invention, the distance between an inner surface of the housing window of the light emitter and the light emitter, and / or between an inner surface of the housing window of the light detector and the light detector, is typically 8 mm, preferably at least 10 mm, and more preferably at least 15 mm. This minimum distance ensures a significant reduction in the effect of the interference signal on the main signal in the measurement signal, since the optical path length difference ΔS is significantly greater than 0.5*ΔX*(m+1). The specified distance refers to the distance of the housing window to the aperture of the light emitter (laser diode) or to the chip surface of the light detector (photodiode).

[0046] Scattering of the main light beam emitted by the light emitter always occurs at the two interfaces of the housing window, i.e., inside and outside, regardless of the thickness of the housing window. This generates two interference signals that can interfere with the main signal in the light emitter (laser cavity), leading to a modulation of the radiation output power. The same applies to scattered light beams at the housing window of the detector, where the scattered light beam interferes with the main light beam incident on the detector surface. Therefore, it is advantageous to position the light emitter, the light detector, and the inside of the housing window relative to each other in such a way that the Fourier component of the interference signal caused by the inside disappears. This ensures that only the interference signal caused by the outside of the housing window can interfere with the main signal.To ensure that the corresponding interference signal caused by the outer surface of the housing window also disappears from the Fourier component under consideration, the distance between the outer surface of the housing window and the light emitter / detector must also meet the aforementioned requirements regarding the period A of the interference signal in the measurement signal. This can be achieved by appropriately selecting the thickness of the housing window. The distance between the outer and inner surfaces of the housing window, determined by this thickness, also satisfies the requirements regarding the period A of the interference signal in the measurement signal. If contamination of the inner surface of the housing window can be ruled out, the design can alternatively be adapted to the outer surface of the housing window only.

[0047] All the measures described above are effective when the main light beam is modulated by any waveform. In addition to modulating the main light beam with a triangular waveform, modulating it with a cosine waveform has also proven advantageous. However, modulation with a triangular waveform is more effective and therefore more advantageous than modulation with a cosine waveform when stray light rays from more distant optoelectronic or optomechanical components are to be suppressed.

[0048] It is self-evident that the optical measuring system according to the invention for measuring a spectrum may also incorporate measures already known from the prior art for avoiding stray light rays or reducing their intensity, such as a housing window inclined relative to the beam path, anti-reflection coatings on the housing window, wedge-shaped windows, blackening of the inside of the housing, an optical isolator in the beam path, an active temperature sink for the light emitter, and optionally a passive temperature sink for the housing, in order to make the optical path length for the main light beam and, if applicable, the occurring stray light rays as independent of temperature as possible.

[0049] The invention is explained in more detail below with reference to an embodiment illustrated in the drawing. Further features of the invention will become apparent from the following description of the embodiment in conjunction with the accompanying drawings. The drawings show: Figure 1 shows an optical measuring system according to the invention, comprising a first housing with an inclined housing window and a laser diode arranged therein as a light emitter, and a second housing with an inclined housing window and a photodiode arranged therein as a light detector; Figure 2 shows the superposition of the measuring signal by interference (self-mixing) of main and scattered light beams at three different distances ( Fig. 2a - 2c ) between the light emitter and the housing window when the main light beam is modulated by a triangular waveform; Figure 3 the effects of varying the distance between the light emitter and the housing window according to Figur 2 on the 2f signal derived from the detector's measurement signal ( Fig. 3a - 3c Figure 4 shows the amplitude of the interference signal in the 2f signal with respect to the distance determining the optical path length difference when the main light beam is modulated by a triangular waveform; Figure 5 shows a comparison of the zero crossings of the measurement signal for a triangular (A) and cosine (B) waveform as a function of the distance between the light emitter and the housing window and a constant modulation span; Figure 6 shows a comparison of the amplitude dependence of the 2f signal on the distance between an optoelectronic or optomechanical component and the light emitter using a triangular (A) and a cosine (B) waveform; Figure 7 shows a schematic representation of the path of the scattered light beam for direct backscattering / reflection ( Fig. 7a ) from the housing window of the light emitter into the laser aperture (self-mixing effect) and indirect backscattering / reflection (self-mixing) from the housing window via a housing wall into the laser cavity ( Fig. 7b ) and the path of the main light beam and the scattered light beam during scattering at the gas cell entrance window and interference of the two light beams on the light detector ( Fig. 7c ) as well as backscattering from the aperture of the photodiode to the housing window of the light detector and direct backscattering to the photodiode ( Fig. 7d ) and resulting etalon on the detector, and Figure 8 the distance-dependent amplitude of the interference signal for a 2f signal ( Fig. 8a ), a 4f signal ( Fig. 8b ) and with a 6f signal ( Fig. 8c ), each involving a modulation of the main light beam by means of a triangular waveform.

[0050] The Figur 1 Figure 1 shows an optical measuring system 1 according to the invention for measuring a spectrum during gas detection. The system comprises a first housing 2 with an inclined, translucent housing window 3, in which a laser diode is arranged as a light emitter 4 on an active temperature sink 5. The measuring system 1 also has a second housing 6, separate from the housing 2, with an inclined housing window 7, in which a light detector 8 is arranged. The two housings 2 and 6 are spaced apart from each other, opposite each other, and aligned with each other. The light emitter 4 and the light detector 8 form the optoelectronic components 20 of the illustrated measuring system 1. A main light beam 9 is emitted from the light emitter 4 and is directed centrally onto the housing window 3. The main light beam 9 exits the housing 2 and penetrates a gas or material arranged between the housing 2 with the light emitter 4 and the housing 6 with the light detector 8.Gas mixture 10, before it enters the housing 6 containing the light detector 8 through the housing window 7. The housing 2 and the housing 6 are arranged relative to each other such that the emitted main light beam 9 falls on the light detector 8 after passing through the gas or gas mixture 10. In the case of the... Figur 1 In the illustrated embodiment, the dominant disturbance is caused by reflection or scattering of the main light beam 9 at the housing window 3. The light emitter 4 is preferably a laser diode with a laser aperture arranged in the direction of the housing window 3. The main light beam 9 is partially reflected, for example, at the inner surface 12 of the housing window 3, whereby the resulting scattered light beam 11 couples into the light emitter 4 and thus causes self-mixing there. In principle, however, this applies to all reflective or scattering surfaces and also to beam paths with multiple reflections or scatterings, as long as the reflected scattered light beams 11 return to the active zone of the light emitter 4.

[0051] The Figuren 7a - 7d The figures illustrate, in a schematic diagram, the optical path length Smain of the principal light beam 9 and Sscatter of the scattered light beam 11, in the case that at least one scattered light beam 11, partially reflected back from the principal light beam 9 towards the light emitter 4 and / or the light detector 8, leads to self-mixing in the light emitter 4 and / or to etalons at the light detector 8, i.e., to an interfered principal light beam 9. The scattered light beam 11 is thereby directed directly or indirectly, i.e., deflected at least once, towards the light emitter 4 or the light detector. For clarity, the light beams 9 and 11 are shown side by side in the figures, although they actually lie on top of each other. Figuren 7a bis 7d Figure 1 shows an embodiment in which the light detector 8 is arranged in a housing 6 next to the light emitter 4, which is housed in a housing 2, and a reflector 16 is arranged opposite them. Figuren 7a , 7b The different optical path lengths of the scattered light beam Sscatter 11 in the housing 2, between the light emitter 4 and the housing window 3 of the housing 2, which interferes with the main light beam (not shown) located inside the light emitter (laser diode). Alternatively, the reflector 16 can be omitted if the housing 6 is positioned in its place. A gas cell 17, with two opposing cell windows 18, 18', is arranged between the housing 2 with the housing window 3 and the housing 6 with the housing window 7. A gas cell 17 containing a gas or gas mixture 10 to be detected is located in the gas cell. Figuren 7c and 7dillustrate the different optical path lengths S main of the main light beam 9 and S scatter of the scattered light beam 11 up to the light detector 8 housed in the casing 6.

[0052] Figur 7a Figure 1 shows the light emitter 4, positioned at a distance L from the housing window 3 of the housing 2. The interfered main light beam 9 emanating from the light emitter 4 reaches the light detector 8 via the optical path length 2*S1. The main light beam 9 emitted by the light emitter 4 is first partially reflected back towards the light emitter 4 at the inside of the housing window 3 of the housing 2, forming a scattered light beam 11. The scattered light beam 11 travels an optical path length S1' from the light emitter 4 to the housing window 3 and an optical path length S1" from the housing window 3 back to the light emitter 4. The optical path lengths S1' and S1" correspond to the distance L between the housing window 3 and the light emitter 4. Therefore, the optical path Sscatter from the light emitter 4 to the housing window 3 and back to the light emitter 4 is Sscatter = S1' + S1" = 2L.Since, by definition, the main light ray 9 is completely located in the light emitter (laser) during self-mixing, Smain = 0 and the path length difference is ΔS = |Sscatter - Smain | = |Sscatter - 0 | = 2L.

[0053] In the light emitter 4, self-mixing occurs as an interference phenomenon between the main light beam 9 and the scattered beam 11, so that the light detector 8 detects a measurement signal of the interfered main light beam 9, which comprises a main signal component and a noise signal component. The scattered light beam 11 originates from the portion of the main light beam 9 emitted by the light emitter 4 that is reflected back at the housing window 3. At the time of switch-on, the light emitter 4 emits an uninterrupted main light beam 9, and later the main light beam 9 that has already interfered with the scattered light beam 11 inside the light emitter 4. The interfered main light beam 9 passes through the gas cell 17 and is then reflected back to the light detector 8 at the reflector 16, passing twice through the two inclined cell windows 18, 18'.The self-mixing takes place in the light emitter 4 itself, so that the optical path length difference ΔS in the housing 2 is given as ΔS = |S scatter - S main |, with S main = 0, as |S scatter |, since the main light ray 9 extends only within the light emitter 4.

[0054] Figur 7b Figure 1 also shows the light emitter 4 arranged at a distance L from the housing window 3 of the housing 2. The main light beam 9 emitted by the light emitter 4 is partially reflected back inside the housing window 3 of the housing 2 towards an inner wall 19 of the housing 2, forming a scattered light beam 11. The scattered light beam 11 is then deflected by the inner wall 19 towards the light emitter 4 by further scattering. The scattered light beam 11 thus travels the optical path length S1' from the light emitter 4 to the housing window 3, then the optical path length S1‴ from the housing window 3 to the inner wall 19, and finally the optical path length S1" from the inner wall 19 to the light emitter 4. The optical path length S1' corresponds to the distance L between the light emitter 4 and the housing window 3.This results in an optical path length difference ΔS from the light emitter 4 to the housing window 3 and back to the light emitter 4 of ΔS = S1'+ S1"+S1‴ > 2L. After interference, an interfered main light beam 9 emerges from the light emitter 4, which is composed by self-mixing of the non-interfered main light beam 9 and the scattered light beam 11 and reaches the light detector 8 via the optical path length 2*S1.

[0055] Figur 7c Figure 1 shows the optical path length Smain of the principal light beam 9 and the optical path length Sscatter for the scattered light beam 11 between the light emitter 4 and the light detector 8, assuming that the modulated principal light beam 9 emanating from the light emitter 4 is partially backscattered at the cell window 18 of the gas cell 17 facing the housings 2, 6, in the direction of the light detector 8. The principal light beam 9 travels the optical path length Smain = 2*S1 from the light emitter 4 to the light detector 8. The scattered light beam 11 travels the optical path length S1' from the light emitter 4 to the cell window 18, and the optical path length S1" from the cell window 18 to the light detector 8. S1' and S1" can be the same or different.The optical path length difference ΔS between the optical path length Smain of the principal light beam 9 and the different optical path length Sscatter of the scattered light beam 11 between the light emitter 4 and the light detector 8 is thus given by ΔS = Smain - Sscatter = 2*S1 - (S1'+S1"). Therefore, an etalon appears at the light detector 8 as an interference phenomenon between the principal light beam 9 and the scattered light beam 11 in the form of an interfered principal light beam 9, which consists of the uninterrupted principal light beam 9 and the scattered light beam 11, so that the light detector 8 comprises a measurement signal with a principal signal component and an interference signal component.

[0056] Figur 7d Figure 1 shows the optical path length Smain of the principal light beam 9 and the optical path length Sscatter for the scattered light beam 11 between the light emitter 4 and the light detector 8, assuming that the modulated principal light beam 9 emanating from the light emitter 4 is not partially backscattered by the cell window 18 facing the housings 2, 6 towards the light detector 8, forming a scattered light beam 11, but is only partially reflected back by the aperture of the light detector 8 (photodiode) towards the housing window 7 of the light detector 8. The housing window 7 reflects or scatters the scattered light beam 11 back towards the photodiode 8. This results in interference between the principal light beam 9 and the scattered light beam 11, forming an interfered principal light beam 9 consisting of these two beams.The main light beam 9 and the scattered light beam 11 travel the optical path length Smain = Sscatter = 2*S1 from the light emitter 4 to the light detector 8. The main light beam 9 is partially reflected back from the light detector 8 towards the housing window 7. The scattered light beam 11 additionally travels the optical path length S1' from the light detector to the housing window 7 and the optical path length S1" from the housing window 7 to the light detector 8. The optical path lengths S1' and S1" correspond to the distance L between the light detector 8 and the housing window 7. This results in an optical path length difference ΔS between the main light beam 9 and the scattered light beam 11 of ΔS = |(2*S1) - (2*S1 + S1' + S1")| = S1' + S1" = 2L, so that the light detector 8 comprises a measurement signal with a main signal and a noise signal.

[0057] Typically, in optical measuring systems of this type, scattered light rays 11 occur, and thus interferences with the respective main light ray 9 occur according to the Figuren 7a bis 7d The reference phenomena cause the main signal and the interference signal to overlap, resulting in a distorted measurement signal from the light detector. By appropriately selecting the distance between the optically effective interfaces of the respective optical and / or optomechanical components of the optical measurement system, the influence of the interference signal on the measurement signal can be significantly reduced. Figur 1 shows one in the Figuren 7a - 7d Variant not shown, similar to the Figur 7a , in which the housing 6 is positioned in place of the reflector 16 and replaces the reflector 16. The gas cell 17 is also not shown. The principal light beam 9 is partially reflected at the housing window 3 of the housing 2 with the light emitter 4, forming a scattered light beam 11 that extends in the direction of the light emitter 4. Corresponding scattered light beams can also be generated at the housing window 7 of the housing 6 with the light detector 8, but these are not shown. The principal light beam 9 is reflected or scattered at an inner surface 12 and an outer surface 13 of the housing window 3. The housing window 3 has a thickness 14, which is determined by the distance between the inner surface 12 and the outer surface 13 of the housing window 3. The light emitter 4, the light detector 8, and the two housing windows 3 and 7 form optoelectronic and optomechanical components 15 of the optical measuring system 1.The scattered light beam 11 reflected from the main light beam 9 towards the light emitter 4 causes a measurement signal (in the laser) through self-mixing with the main light beam. This measurement signal comprises a noise signal and a main signal. The light emitter 4 is positioned with its aperture at a defined distance L from the housing window 3. This distance L changes the optical path length difference ΔS = 2*L and influences the amplitude and period A of the noise signal caused by the housing window 3 within the measurement signal.

[0058] The effect of a disturbance on the measurement signal is shown below using three different distances L between the aperture of the light emitter 4 and the scattering inner surface of the window 12 as examples. Figuren 2a - 2c and 3a - 3c discussed here. Fig. 7a The case shown is for direct backscattering. The first distance L = 0.3 cm ( Fig. 2a , 3a) corresponds approximately to the distance between the light emitter 4 and the inside of the window 12. The distance L = 0.584 cm ( Fig. 2b , 3b ) was chosen to demonstrate the worst-case scenario, i.e., to generate a noise signal with a period A on the order of the usual width of a gas absorption line. With L = 1.17 cm ( Fig. 2c , 3c The case is shown in which no interference, i.e., no noise signal, is found in the demodulated measurement signal. The representations are based on a triangular waveform with a fixed modulation width Δλ = 0.195 nm at λ₀ = 1.512 µm.

[0059] The Figuren 2a bis 2c The figures show the simulated measurement signal of the light detector 8 for the three distances L, or optical path length differences ΔS = 2*L. The simulated measurement signal is the normalized power of the light emitter in a range from 0.9985 to 1.000. The figures are intended to illustrate the different waveforms. The peak at 0 GHz represents the actual absorption signal to be measured, in this case the signal of an NH3 gas peak with a Lorentzian profile, which is now superimposed by the periodic disturbance. As can be seen, the period A decreases with increasing distance L, i.e., with increasing optical path length Sscatter for the scattered light beam 11. However, the amplitude of the disturbance is the same for all three variants, since the reduction of the feedback with increasing distance L of the housing window 3 relative to the light emitter 4 is not taken into account in the corresponding figures.

[0060] The Figuren 3a bis 3c These figures show the effect on the 2f signal derived from the measurement signal, i.e., at m=2. They depict the amplitudes of the respective disturbances in the 2f measurement signal versus the three different optical path lengths Sscatter for the scattered light beam 11, i.e., at different distances L between the aperture of the light emitter 4 and the housing window 3. As can be seen, the disturbance signal decreases from L = 0.3 cm ( Fig. 3a ) to L = 0.584 cm ( Fig. 3b ) rises sharply and then drops sharply again. The representation with L = 0.3 cm ( Fig. 3a ) shows in particular that the disturbance is also visible in the 2f signal. This disturbance limits the measurement accuracy of optical measuring system 1. If the distance L is now increased to L = 0.584 cm ( Fig. 3b ), the effect intensifies, and the NH3 gas signal becomes almost indistinguishable from the interference. However, a significant reduction in the measurement error can be achieved if the distance L between the housing window 3 and the light emitter 4 is chosen such that the condition ΔS = 2 * λ₀ / Δλ = 2 * ΔX is satisfied for the optical path length differences ΔS. For the distance L with L = 1.17 cm ( Fig. 3c ) the measurement signal only shows the signal of the NH 3 peak, i.e. the 2f signal is completely free of the interference disturbance.

[0061] The Figur 4 The 2f measurement signal shows the amplitude A of the interference signal as a function of the distance L for the in Fig. 7a The illustrated case represents direct backscattering. The Fourier component of the demodulated measurement signal is shown. The intensity curve exhibits successive peak-shaped intensity regions, comprising a main maximum with a highest amplitude peak and a number of intermediate maxima, each with reduced amplitude peaks. The large initial main maximum and the subsequent zeros and smaller intermediate maxima are clearly visible. The main maximum begins as the first region at the zero point (L = 0 cm), with the intermediate maxima extending to the right of the main maximum. As can be seen, the interference signal increases sharply from L = 0 cm to L = 0.584 cm and then decreases sharply again, reaching the first local minimum at L = ΔS / 2 = λ 0 2< / Δλ=ΔX.

[0062] The Figur 5 Figure 1 shows a comparison of the dependence of the zero crossings of the amplitude of the demodulated measurement signal at the same modulation span Δλ for a triangular waveform (curve A) and a cosine waveform (curve B). Triangular modulation attenuates the noise component of the demodulated measurement signal more strongly than cosine modulation, but requires slightly larger distances between the light emitter 4 and those optoelectronic and / or optomechanical components 15, such as the housing window 3, which can cause stray light rays 11.

[0063] The Figur 6 Figure 1 shows a comparison of the amplitude dependence of the 2f signal on the distance L between the light emitter 4 and an optomechanical component 15 when the main light beam is modulated by a triangular waveform (curve A) and a cosine waveform (curve B). It is clearly visible that triangular modulation is more effective than cosine modulation. With triangular modulation, the influence of the interference signal decreases significantly more rapidly in intensity with increasing distance between the light emitter 4 and those optoelectronic and / or optomechanical components 15 that can cause scattered light rays 11, such as the housing window 3. The figure shows that with triangular modulation, the optical interference decreases much more sharply with the distance of the optomechanical element than with cosine modulation. Thus, the influence of the interference is significantly lower near the light emitter (laser diode), e.g.,At the housing window 3 of the light emitter 4, which is, for example, 1.8 cm away from the laser diode, the scattering is still almost identical for both curves A and B, but is already significantly reduced when scattering occurs at cell window 18 of the gas cell 17, which is, for example, 5 cm away. The difference is considerable when disturbance occurs at the light detector 8, which is, for example, 60 cm away from the light emitter 4.

[0064] In general, several such feedback loops between the light emitter 4 and one of the other optoelectronic or optomechanical components 15 are possible and usually present in the optical measuring system 1 with different optical path lengths Sscatter. Ideally, the conditions listed above should be met for each individual optical path length Sscatter. In practice, this is often not possible or only possible to a limited extent because certain dependencies generally exist between the different optical path lengths Sscatter, Smain, i.e., those caused by the different distances L of the corresponding components 15. For example, a practically applicable condition for the positioning of an optoelectronic or mechanical component 15 that causes scattered radiation relative to the light emitter 4 is: S Streu − S Haupt = ΔS > λ 0 2 / Δλ , where λ₀ is the mean wavelength and Δλ is the modulation span of the modulated principal light beam 9 in nanometers. Fulfilling this condition is advantageous because the interference signal decreases significantly between an optical path length difference ΔS or ΔS' = λ₀ 2< / Δλ and 2*λ₀ 2< / Δλ. In the case of multiple sources of scattered light beams 11, i.e., multiple reflecting or backscattering optoelectronic or optomechanical components 15 and / or multiple reflecting or backscattering surfaces of such a component 15, arranged at different distances L from the light emitter 4 and thus resulting in multiple differing optical path lengths Sscatter of the scattered light beams 11, the symbol ΔS' denotes the shortest optical path length difference ΔS' that makes a relevant contribution to the interference of the principal light beam 9 with the scattered light beam 11.It can also be seen that the local maxima of the disturbance become smaller with increasing optical path length difference ΔS. This effect is further enhanced by the fact that the amount of feedback light decreases with increasing distance (not in ). Figur 4 taken into account).

[0065] In practical design, the tolerances of the relative positions of the optoelectronic and / or optomechanical components 15, 4, 8 must be taken into account. The fundamental problem with the housing window 3 is that the distance L to the inner surface 12 of the housing window 3 and to its outer surface 13 must be optimized equally to achieve an optimal result. There are two main approaches to this, in order to fulfill the condition that both the inner surface 12 and the outer surface 13 lie at a minimum of the curve A(L), despite the different optical path lengths Sscatter and the resulting optical path length differences ΔS, which are caused by the unequal distance L of the inner surface 12 and the outer surface 13 to the light emitter 4. This can be achieved either by adjusting the thickness 14 of the housing window 3 and / or by tilting the housing window 3 with respect to the beam path.

[0066] One helpful aspect is the insertion of an optical isolator (not shown in the figures) into the beam path, which prevents backscattered light from subsequent optomechanical elements 15 from causing self-mixing. However, the first interface of the optical isolator, located in the beam direction of the light emitter 4 (for example, a quarter-wave plate), still causes scattered radiation that can lead to self-mixing. Therefore, it is also advantageous to optimize the position of this first optical interface according to the spacing rules mentioned above. In In a specific embodiment, a quarter-wave plate can be attached to the inner surface 12 of the housing window 3. This plate rotates back-reflected light from the outer surface, for example, from the outer surface 13 or from the housing window 7 of the housing 6, by an angle of 90° with the light detector 8, so that it cannot interfere with the main light beam 9 of the light emitter 4 or with any stray light beams 11 in the housing 2 in which the light emitter 4 is located. It is advantageous to bond the optical isolator to the inner surface 12 without a gap using an adhesive whose refractive index matches that of the optical isolator or the housing window 3 as closely as possible, so that no additional optical boundary layer is formed between the optical isolator and the housing window. Alternatively, the quarter-wave plate itself can also be used as the housing window 3.

[0067] In the Figur 8 is the distance-dependent amplitude of the interference signal in a 2f signal ( Fig. 8a ), a 4f signal ( Fig. 8b ) and with a 6f signal ( Fig. 8c ), each represented by a triangular waveform during modulation of the main light beam 9. The Figuren 8a bis c The diagram shows the interference signal A(L) as a function of the distance L or optical path length difference ΔS. The large main maximum and the smaller intermediate maxima to its right and / or left, with their intervening zeros, are clearly visible. The position of the main maximum relative to the intermediate maxima depends on the order of the demodulated mf measurement signal used. For a 2f signal ( Fig. 8a ) the main maximum begins as the first region at the zero point, with the intermediate maxima connecting to the right of the main maximum, in the case of a 4f signal ( Fig.8b ) there is also an intermediate maximum between the zero point and the main maximum, in the case of a 6f signal ( Fig. 8c) Two intermediate maxima are located to the left of the main maximum. The position of the main maximum is determined by the order m / 2 of the demodulated measurement signal used. A number of intermediate maxima are also arranged to the right of the main maximum in both the 4f and 6f signals.

[0068] In practice, the laser package is often hermetically sealed, so the inside of the window should not become dirty during its entire lifespan. In this case, it is sufficient to ensure that the inside of the window is optically clean, so that stray light rays from the inside can be disregarded and the design can be focused on the outside of the window, which, unlike the inside, is very susceptible to dirt.

[0069] A typical approach to optimizing a wavelength-modulated measurement system involves first defining the wavelength range used, adapted to the spectrum to be measured; then defining the signal processing method and optimizing the modulation span Δλ with respect to noise, cross-sensitivity, and the like; and finally calculating the optimal positions of all optoelectronic and optomechanical components 15 of the optical measurement system 1 for measuring a spectrum. Once the position of the optoelectronic and optomechanical components 15 has been determined, fine-tuning of the modulation span Δλ can optionally be performed to compensate for manufacturing tolerances.

Claims

1. An optical measuring system (1) for gas detection according to wavelength modulation spectroscopy, comprising: a light emitter (4) and at least one light detector (8) forming optical components (20), and at least one opto-mechanical component (15) arranged between the light emitter (4) and the at least one light detector (8), wherein the optical components and the at least one opto-mechanical component have optically effective boundary surfaces for light beams, wherein the light emitter (4) and the at least one light detector (8) are arranged together in a housing (2, 6) or separately in a respective housing (2, 6), and the housing (2, 6) optionally comprises a housing window (3, 7) forming an opto-mechanical component (15), wherein the light emitter (4) emits a modulated main light beam (9) with a mean wavelength λ0 with a modulation span Δλ, and wherein at least one scatter light beam (11) that is partially reflected / scattered from the main light beam (9) at least at one optically effective boundary surface of at least one opto-mechanical component (15) in a direction of the light emitter (4) and / or of the light detector (8) causes self-mixing in the light emitter (4) and / or etalon effects at the light detector (8) through interference with the main light beam (9), wherein the interference of the main light beam (9) with the scatter light beam (11) causes an interfered main light beam (9) which is detected by the light detector (8), and wherein a measuring signal resulting from the detection of the interfered main light beam (9) at the light detector (8) accordingly comprises an interfering signal portion and a main signal portion, the measuring signal is demodulated, and the interfering signal portion in the measuring signal influences the demodulated measuring signal, wherein the spatial orientation and / or the distance (L) of at least one of the optically effective boundary surfaces of an optical and / or opto-mechanical component (20, 15) relative to a second optically effective boundary surface of the same or of a further, second optical and / or opto-mechanical component (20, 15) is adapted to the modulation span Δλ of the modulated main light beam (9) of the wavelength λ0 such that the effect of the interfering signal upon the main signal on a Fourier-component of the demodulated measuring signal for the modulation span Δλ is minimal or zero for all phases, wherein the distances are selected such that amplitudes of all dominant interferences of the Fourier-component of the demodulated measuring signal are suppressed or at least reduced, wherein an optical path length difference ΔS exists for the main light beam (9) relative to the scatter light beam (11) at the selected modulation span Δλ of the modulated main light beam (9) of the wavelength λ0, which is defined by a difference between an optical path length SMain of the main light beam (9) and an optical path length SScatter of the scatter light beam (11), the demodulated measuring signal, due to the optical path length difference ΔS of the main light beam (9) relative to the scatter light beam (11) and due to the interference of the interfering signal with the main signal, has a period Λ as a free spectral range, which is determined by the distance of the transmission maxima of a transmission spectrum of respective two optically effective boundary surfaces of an opto-mechanical and / or optical component (15), of one respective optically effective boundary surface of two opto-mechanical components (15), of one rep-spective optically effective boundary surface of two optical components (20) or of one respective optically effective boundary surface of an optical component (20) and an opto-mechanical component (15), wherein the modulation of the main light beam (9) is made with any periodic waveform, and the modulation span Δλ is the peak to peak amplitude of the modulation signal used for modulating the main light beam (9) of the wavelength λ0.

2. Optical measuring system according to claim 1, characterized in that the optical path length difference ΔS is greater than 0.5*ΔX*(m+1), with m = 2, 4, 6, ..., according to the order 2f, 4f, 6f,... of the employed measuring signal, wherein ΔX is the distance where a Fourier-component of the demodulated measuring signal adjacent to a highest amplitude-peak is repeatedly minimal.

3. Optical measuring system according to claim 1, characterized in that the optical path length difference ΔS is greater than 0.7 times or less than 1.3 times n*ΔX, with n = 1, 2, 3, ...., and with n ≠ m / 2, with m = 2, 4, 6, ..., according to the order 2f, 4f, 6f,... of the employed measuring signal, wherein ΔX is the distance where a Fourier-component of the demodulated measuring signal adjacent to a highest amplitude-peak is repeatedly minimal.

4. Optical measuring system according to claim 3, characterized in that the optical path length difference ΔS is greater than 0.8 times or less than 1.2 times n*ΔX.

5. Optical measuring system according to claim 3, characterized in that the optical path length difference ΔS is equal to n*ΔX.

6. Optical measuring system according to one of the preceding claims, characterized in that the main light beam (9) is modulated by a triangular waveform.

7. Optical measuring system according to claim 6, characterized in that a ΔX is λ02 / Δλ.

8. Optical measuring system according to one of the preceding claims, characterized in that a distance (L) of an inside (12) of a provided housing window (3) of the light emitter (4) to the light emitter (4) and / or a distance (L) of an inside (12) of a provided housing window (7) of the light detector (8) to the light detector (8) is at least 8 mm, preferably at least 10 mm, further preferably at least 15 mm.

9. Method for gas detection according to wavelength modulation spectroscopy, using an optical measuring system (1) comprising a light emitter (4) and at least one light detector (8) as optical components (20), wherein at least one opto-mechanical component (15) is arranged between the light emitter (4) and the at least one light detector (8) and the optical components and the at least one opto-mechanical component have optically effective boundary surfaces for light beams, wherein the light emitter (4) and the at least one light detector (8) are arranged together in a housing (2, 6) or separately in a respective housing (2, 6), wherein the housing (2, 6) optionally comprises a housing window (3, 7) forming an opto-mechanical component (15), and wherein a modulated main light beam (9) with a mean wavelength λ0 and with a modulation span Δλ is emitted from the light emitter (4) and reaches the light detector (8) after passing through a gas or gas mix (10); by at least one scatter light beam (11) that is partially reflected / scattered from the main light beam (9) at least at one optically effective boundary surface of at least one of the opto-mechanical components (15) in a direction of the light emitter (4) and / or of the light detector (8), through interference with the main light beam (9) self-mixing in the light emitter (4) and / or etalon effects at the light detector (8) are caused; wherein by the interference of the at least one scatter light beam (11) with the main light beam (9) an interfered main light beam (9) is formed, which is detected by the light detector (8), and a measuring signal resulting from the detection of the interfered main light beam (9) at the light detector (8) accordingly comprises an interfering signal portion and a main signal portion; the measuring signal is demodulated, and the demodulated measuring signal is influenced by the interfering signal portion in the measuring signal, wherein the spatial orientation and / or distance (L) of at least one of the optically effective boundary surfaces of an optical and / or opto-mechanical component (20, 15) relative to a second optically effective boundary surface of the same or of a further, second optical and / or opto-mechanical component (20, 15) is adapted to the modulation span Δλ of the modulated main light beam (9) of the wavelength λ0 such that the effect of the interfering signal upon the main signal on a Fourier-component of the demodulated measuring signal for the modulation span Δλ is minimal or zero for all phases, wherein the distances are selected such that amplitudes of all dominant interferences of the Fourier-component in the demodulated measuring signal are suppressed or at least reduced, wherein an optical path length difference ΔS for the main light beam (9) relative to the scatter light beam (11) at the selected modulation span Δλ of the modulated main light beam (9) of the wavelength λ0 is selected, which is defined by the different optical path length SMain of the main light beam (9) relative to the optical path length SScatter of the scatter light beam (11), the demodulated measuring signal, due to the optical path length difference ΔS of the main light beam (9) relative to the scatter light beam (11) and the interference of the interfering signal with the main signal, has a period Λ as a free spectral range, which is determined by the distance of the transmission maxima of a transmission spectrum of respective two optically effective boundary surfaces of an opto-mechanical and / or optical component (15), of one respective optically effective boundary surface of two opto-mechanical components (15), of one respeceive optically effective boundary surface of two optical components (20) or of one respective optically effective boundary surface of an optical component (20) and an opto-mechanical component (15), wherein the modulation of the main light beam (9) is made with any periodic waveform, and the peak to peak amplitude of the modulation signal used for modulating the main light beam (9) of the wavelength λ0 is used as modulation span Δλ.

10. Method according to claim 9, characterized in that the optical path length difference ΔS is set greater than 0.5*ΔX*(m+1), with m = 2, 4, 6, ..., according to the order 2f, 4f, 6f,... of the employed measuring signal, and wherein ΔX is the distance where a Fourier-component of the demodulated measuring signal adjacent to a highest amplitude-peak is repeatedly minimal.

11. Method according to claim 9, characterized in that the optical path length difference ΔS is set greater than 0.7 times or less than 1.3 times n*ΔX, with n = 1, 2, 3, ...., and with n ≠ m / 2, with m = 2, 4, 6, ..., according to the order 2f, 4f, 6f,... of the employed measuring signal, and wherein ΔX is the distance where a Fourier-component of the demodulated measuring signal adjacent to a highest amplitude-peak is repeatedly minimal.

12. Method according to claim 11, characterized in that the optical path length difference ΔS is set greater than 0.8 times or less than 1.2 times n*ΔX.

13. Method according to claim 11, characterized in that the optical path length difference ΔS is set as n*ΔX.

14. Method according to one of the preceding claims 9 - 13, characterized in that the main light beam (9) is modulated by a triangular waveform.

15. Method according to claim 14, characterized in that a ΔX is selected, that results in λ02 / Δλ.