DE3930642A1 - Particle size detection device with noise cancellation - Google Patents

Abstract

The detection device determines size from particle effected light scattering in a sensing region illuminated by a laser beam and receiving the particles in a medium, such as air. Background light from molecular scattering is reduced to a level that enables light scattered by particles having a size of about 0,1 micron to be sensed in a high background of molecular scattering such as for example, where molecular scattering can exceed the 0.1 micron particle's scattering by one hundred times. A linear array of detectors are positioned with respect to an imaging system, so that each detector monitors a different portion of the sensing region. An electrical output signal indicative of sensed particle presence within that portion monitored is provided. The output signals from the detectors are parallel processed. The output signals from detectors monitoring non-adjacent portion of the sensing region are combined at noise cancellation units.

Description

The invention relates to the detection of particle sizes, in particular a detection of particle sizes under Ver application of light scattering.

Devices for determining particle sizes are now well known and also that laser beams for it can be used and have been used so far, to take particle size measurements. See for example as the US-PS 34 06 289. Also shows and describes the US-PS 45 71 079 a particle size measurement with a Open cavity laser (active) and U.S. Patent No. 4,594,751 a measurement with a passive cavity; both patents were granted to the current applicant.

So far, a linear array of detectors has been put together with parallel processing of the electrical signals ver uses that by every detector for data acquisition be fathered. See, for example, US Pat. No. 3,941,982 current applicant.

Known particle measuring devices have so far been used for ver uses different purposes, including discovery the presence or size of particles in various which gases, including air. In a special way on airborne particles it made tolerance limits and the effects of contamination of macro particles from the Environment required an effective pollution introduce control to the manufacture of many now to enable devices used. In particular For example, the precision manufacturing of mikro electronic systems largely through development and application of dust-free pressure rooms and clean Facilities made possible.  

For many years, the dust-free overpressure rooms were the Class 100 or class 1000 for practically all used electronic devices more than sufficient. As however, the current generation of microcomputers led to the need for microelectronic Components such as large capacity memory chips that Development of devices during manufacturing are extremely susceptible to contamination.

Macro particle contamination during manufacture development of such devices causes the emission of a usable product is greatly reduced. The verun Cleaning particles can, for example, depending on their nature show the completeness of lithographic images impair or open or depending on their nature shorted circuits of unspoiled districts to lead. Now the manufacture of microchips seems to be the most critical operation in electronic manufacturing industry to be caused by macro contamination particle is disturbed.

The semiconductor industry for VLSI (Very Large Scale Integrated Circuit = very large ICs) is still state of the art propelled for counter for air particles that for Testing of dust-free overpressure rooms can be used. The need for much higher standards reflects that Requirements of the manufacturers of VLSI circuits as well as the Improvements in filtering that make Verun much smaller reach cleaning levels.

Most known aerosol meters have a sample throughput of 0.028 m ³ / min (one cubic foot / min (cfm)). However, to obtain useful statistical values in a class 1 environment, many cubic meters of air must be collected for the sample if the sensitivity is limited to 0.5 microns (0.5 × 10 ^-6 m). As the number of particles increases with decreasing size, most manufacturers of air particle counters chose to determine the size of much smaller particles in order to be able to develop the correct statistical base more easily.

For example, the distribution of the average particle size in a dust-free overpressure space at 0.1 × 10 ^-6 m is almost a hundred times the particle counts of <0.1 × 10 ^-6 m compared to counts of <0.5 × 10 ^-6 m. The more sensitive the particle counter is, the less time is required for the attestation of the room standard.

In addition, the manufacturing devices, whose shape design features that are smaller than 0.5 × 10 ^-6 m. In addition to generating a statistical basis in a very short time, higher sensitivity gives known information about the particle size for more potentially defective generators.

With the introduction of lasers, the possibility of measuring the size of particles up to 0.1 × 10 ^-6 m via light scattering became an everyday practice, since the entire energy of the laser beams can be concentrated to a small dimension of high intensity. Various devices now on the market provide a sensitivity of 0.1 × 10 ^-6 m, but none of them can collect samples at a throughput of 0.028 m ³ / min (1 cfm) and, in fact, generally cannot exceed collect a throughput of 0.0028 m ³ / min (0.1 cfm) with a sensitivity of 0.1 × 10 ^-6 m.

For various processes, the manufacturers of semiconductors also need gases of high purity with low micropollutants of macro particles. This is necessary in order to be able to carry out the measurements with line pressures of mostly (up to 1 × 10 ⁶ Pa (150 PSI)). Some of these gases are also high molecular weight gases which scatter more light than air (ie a mixture of mainly oxygen and nitrogen), therefore the molecular scatter must be reduced even if flow rates are less than 0.028 m ³ / min (1 cfm) are sufficient.

The possible statistical advantage with a smaller particle size in environments with high molecular scattering is thus partially lost. Obviously, a combination is required which achieves high sensitivity in such an environment (e.g. including a high throughput up to or greater than 0.028 m ³ / min (1 cfm)) with high sensitivity (to detect particles whose size is at most 0 , 1 × 10 ^-6 m).

The invention provides one part noise reduction chdetektor, in particular a high noise reduction Sensitivity in an environment with high molecular scatter ung.

Background noise due to molecular scattering is ver reduces, with the noise reduction by several linear detectors, each of which is a portion of the entire measuring range, which at the intersection of the Flow path of a mixture of gas and particles and one Laser beam is arranged and an electrical output signal generated for the sampled or measured particles, the output signals of the detectors with the output signals of the detectors are processed in parallel, which monitor non-adjacent sections of the measuring range, which are united by noise suppression devices.

The drawings show a complete embodiment of the Invention of the best kind so far for the practical Application of the basics of the invention was developed:

FIG. 1 is a concentration distribution curve for particle sizes according to the classification system and the areas of sizes, which are published today in the US Federal Standard 209 (FS209);

Fig. 2 is an illustration of a distribution curve for particle sizes after the planned revision of FS209;

Fig. 3 is a representation of a distribution curve for particle sizes, which show the data of the actual count in currently built dust-free overpressure rooms for the semiconductor production of very large IC's (VLSI);

4A is a partial schematic view showing the cross section of a circular flow, which serves to build up a measurement range.

Figure 4B is a partial schematic view showing a rectangular-shaped flow cross-section which serves to build up a measurement range.

Fig. 5 is a schematic representation of a simplified side section of a particle size measuring device with an open cavity laser, the invention being used with this;

Fig. 6 is a simplified side view showing the arrangement of the detector unit for monitoring the Meßbe range via an imaging device or an imager according to an embodiment of the inven tion;

Fig. 7 is a simplified electronic block diagram showing the electronic signal processing according to an embodiment of the invention;

Fig. 8 is a simplified block diagram showing the use according to an embodiment of the inven tion of the noise reduction circuit in connection with the parallel processing device of Fig. 7;

Fig. 9 is a schematic electronic circuit diagram with Dar position of the suppression circuit of Fig. 8.

Current particle control in rooms with electronic manufacturing is mainly based on the application of US Federal Standard 209 (FS209), the VDI standard 2083 or similar standards. Figure 1 shows the current classification system and the size concentration ranges used in FS209. The left vertical column of Figure 1 (as well as Figures 2 and 3) shows the total amount of particles / cubic feet which, as shown, are equal to or greater than the specified particle size. Counts below 10 particles / cubic foot (0.35 particles / l) are unreliable (and therefore shown in dashed lines), except in the case of a large number of samples. The smallest particle that must be measured has a diameter of 0.5 µ, with the maximum purity level being 100 particles / cubic foot.

These standards for dust-free pressure rooms exist but for several years, and the current requirement Advances in electronic products showed that this covered Level for a convenient production of microelectronics facilities is unsatisfactory because the Particle control requirements have now become stricter  are. The need for revised definitions of the Levels of contamination from particles are thus open obviously and therefore plans are being developed, FS209 to change to measure small particles and clean higher levels.

Although now a revision of FS209 under the pen leadership of the Institute of Environmental Science this revision is not yet complete. One of the main goals of this revision is to create a space or an area, the degree of purity or the purity to define air class at the level of class 10 and measuring particles of sizes below 0.5 µ diameter to enable adequate stati technical definitions to arrive at the lower levels. It is expected that the standards that apply in all industries areas will be used, probably soon revi and that the revision of FS209 is only the first seems to be from a number of new standards. Did Class 1 levels are currently available will also have to be produced in the near future be included.

Since the particle concentration increases rapidly with the decrease in particle size, the recommendation of measurements of smaller particles seems to be expedient. Therefore, the plans for the revised FS209 include measurements of 0.3 × 10 ^-6 m and for cleaner areas 0.1 × 10 ^-6 m. Fig. 2 shows the currently planned classification format, which allows measurements of sizes below 0.5 × 10 ^-6 m. However, it should be noted that the slope of the classification curves of FIG. 2 is based on the same slope of the current FS209.

A change in the slope of the classification curves is being considered based on recent data obtained in less dust-free VSLI rooms in the United States. A laser counter was used for measurements from 0.1 × 10 ^-6 m upwards, the data of which are shown in FIG. 3. It appears that the particle size distribution in highly dust-free positive pressure spaces has a much steeper increase below the size of 0.1 x 10 ^-6 m than would be expected from the current classification curves according to FS209, with one possibility of making the increase even steeper to be interpreted as in FIG. 3.

If the classification ^{curves are} extended in the size range of 0.1 × 10 ^-6 m, the time and effort required to obtain valid data for rooms of class 100 and better are reduced, regardless of whether FS209 is simply extra-polished or according to the newer data is revised according to FIG. 3.

Developing a device that is sensitive of 0.1 µ in a high molecular environment has scattering, as achieved by the invention was not easy to carry out. With regard an environment with molecular scattering due to high throughput require rates up to or more than one cubic foot / min these throughputs have a fairly large flow area due to the sensitive volume of an optical measuring device tung.

Even at speeds of sound, it is impossible to achieve the cross-section of 1.8 mm ² and full speed of sound with practical vacuum sources. Thus, the more normal flow velocities are 1/3 to 1/2 sound speed, and the flow cross-sections increase to 4 to 5 mm ² at 0.028 m ³ / min (1 cfm) with normal instruments. This large flow cross section must be illuminated evenly if the particles are to be grasped exactly in size.

Thus, the illuminated area cannot be near a narrowly focused area of a laser beam, but must usually be arranged in an elongated laser beam if the flow cross-section is circular as in Fig. 4A or if it is stretched along the beam axis if the flow cross-section is downstream Fig. 4B is rectangular. Therefore, an almost optimal construction, as used in the invention, has a flow cross section of 1 × 10 mm ² , and the converging lenses are like those of US Pat. Nos. 45 71 079 and 45 94 751.

The laser beam itself must generate approximately 1 W of power within the measuring range in order to produce a sensitivity of 0.1 × 10 ^-6 m. A HeNe laser resonance cavity operating at 633 nm is the most convenient way to generate the required power, although any 1 W light laser can be used (eg argon ions). Since light sources can be developed which provide a sufficient energy density to measure small particles such as 0.1 × 10 ^-6 m and allow cross sections for 1 cfm, the problem appears to be easily solvable.

This would be the case if the particles were the only source of light scattering in the measuring range. Through careful construction, stray light can be effectively excluded. However, there are approximately 10 ¹⁶ air molecules / mm ³ that accompany the airborne particles as they pass through the laser beam, and these air molecules cannot be excluded. Calculations show that such an air volume scatters over 100 times more energy than a normal particle of 0.1 × 10 ^-6 m (see RG Knollenberg: "The Measurement of Particle Sizes Below 0.1 Micrometers", in The Journal of Environ mental sciences, January-February 1985). This large accumulation of molecular scattering media is always present, whereby this signal source mainly represents direct current.

It is the variations (alternating current) in the molecular scatter signal (alternating current noise) that mask the desired particle signal of 0.1 × 10 ^-6 m. The noise itself is proportional to the molecular stray DC signal. However, the noise is higher for laser sources than calculated from Shott's noise assumption, and the laser cavities are easily randomly modulated by the flow itself (creating noise). In such cases, the AC noise is directly proportional to the DC component rather than the square root. Measurements show that the molecular scattering noise in most systems of 1 cfm is 5 to 10 times stronger than that of a particle of 0.1 × 10 ^-6 m, making detection and measurement with the known devices impossible.

A high molecular environment or background Scattering exists, for example, if the molecular scattering is a hundred times is larger than the scattered particles of 0.1 µ. These high molecular scattering can not only be generated if the gas volume under consideration is large, as is the case for high Throughputs is required but also, for example, due to high molecular density (high pressures above the Ambient pressure) or due to the presence of large Gas molecules. In any case, if a background with High molecular scatter is encountered, the Vor direction according to an embodiment of the invention used to allow high sensitivity in the presence of the background with strong moles Specular scatter is achieved.  

The invention uses an illumination device to illuminate a measuring area or a measuring volume through which a gas (usually air), the particles of which are to be measured, has to flow. The illuminating device is preferably like that shown in FIG. 5, wherein a illuminating device 10 with a laser beam according to US Pat. Nos. 45 71 079 and 45 94 751 is used.

As shown in FIG. 5, access to the laser cavity 12 can be achieved by closing a plasma tube 14 at one end with a Brewster window 16 . A curved mirror 18 (with a radius of, for example, r = 100 cm) is arranged at the other end of the tube 14 . A photodetector 20 and a preamplifier 22 are conventionally arranged on the laser beam axis outside the mirror 18 to provide a measure of the relative intensity of the illumination.

For particle size measurement, the particles whose size is to be measured are injected by beams 24 so that they pass the laser beam 25 between the Brewster window 16 and an external laser mirror 26 (where r = 70 cm, for example).

As shown in FIG. 5, the block 28 for taking optical samples is arranged next to the injection area for the particles, wherein in the block 28 light-collecting optical elements or an imager 29 is positioned, the two lenses 30 and 31 above and close to that May have particle injection field.

A photodetector 33 with a plurality of detectors 34 (FIGS . 6 and 7) is arranged next to the block 28 so that it lies in the focal plane of the lenses 30 and 31 . The electrical rule's output signals of the photodetector 33 are applied to the parallel processor 35 . The output of the parallel processor 35 may be passed through an analyzer 36 to produce a low resolution display output which is passed through a converter 37 which provides a high resolution digital display output for the particle size. In addition, the converter 37 outputs a display output signal for exceeding a minimum size.

Photodetector 39 and preamplifier 40 can optionally be arranged on the other side of the mirror 26 in the laser beam axis for a reference measurement such as the photodetector 20 and preamplifier 22 .

In this invention, the noise is reduced by having an imaging device or an imager with detector reduce the backlight from molecular scattering to easy-to-use levels. As FIG. 6 shows, the detector 33 has a linear arrangement of rectangular elements or detectors 34 (eleven of which are shown in FIG. 6 for the purpose of explanation). The linear arrangement or battery of detectors connected to the parallel processor for two-dimensional data acquisition is described in US Pat. No. 3,941,982.

This arrangement of elements (which can be photo diodes according to FIG. 7) "see" corresponding volumes (ie a section of the entire measuring range) within the laser beam. The light scattered by a particle is again formed on a single element as a bright image together with a background of diffuse molecular scattering, which is generated by all air molecules in the monitored section of the entire measuring range of a certain element.

This sets the amount of molecular background scatter reduced by the number of elements required for the arrangement or battery have been selected. For light noise sources, that are not detected by the Shott noise also occurs a direct reduction in noise that is proportional to the number of elements of the arrangement. For shott Noise sources is the reduction in noise proportions tional to the square root of the number of elements of the An order.

Because, according to one embodiment of the invention, a number N of independent detectors is used, the electrical output signal of each must be sampled individually to determine when a particle image was "observed" by an element. Fig. 7 shows several detectors 34 , which form the detector device 33 , where each detector is connected to a different signal processing circuit 41 of the parallel processor 35 . In particular, each detector 34 in each signal processing circuit 41 is connected to a preamplifier 42 (each of which has a feedback resistor 43 which is connected between the output and the negative input). Each processing circuit 41 also has a comparison circuit 44 and a storage element for peak voltages (holding circuit) 46 .

Each preamplifier 42 develops an amplified electrical signal of the occurrence of noise and particles, which is present at the associated comparison circuit 44 . The threshold level of each comparison circuit is set to the peak amplitude of the minimum detectable particle size signal. The output signals of the comparison circuit are combined together in the logic control unit 48 in an OR circuit, the logic control circuit 48 emitting a pulse for the passage time of a particle for clock purposes and an indication of when a particle of the minimum recognizable size is scanned becomes.

For dimensions larger than the minimum recognizable size, the resolution of the size depends on the size of the desired resolution. For example, if a simple count of all particles in a single size channel, <0.1 x 10 ^-6 m, is desired, the signal processing circuits 41 can simply be used with the logic controller 48 .

For multichannel high-resolution applications (eg, where a resolution of 0.01 × 10 ^-6 m at 10 channels between 0.1 × 10 ^-6 m and 0.2 x 10 ^-6 m is desired), the peak amplitudes of each must Detector can be saved individually before processing. For this high resolution (where, for example, the second threshold level is 0.11 × 10 ^-6 m), the output signal of each pre-amplifier 42 is fed to its associated peak voltage memory (holding circuit) 46 , which stores the peak amplitude. The peak amplitudes stored in the holding circuits 46 are then read individually from the converter 37 , which has the logic control unit 48 and a scanning analog-digital converter 49 ( FIG. 7).

For multi-channel applications with low resolution, it is sufficient to sum all the output signals of the amplifiers 42 in that each output signal via an associated resistor 50 to a summing amplifier 52 (whose feedback resistor 43 is connected between the output and the negative input) of the analyzer 36 also includes a conventional pulse height analyzer 54 to which the output signals of amplifier 52 are applied. Such a low resolution application assumes that the second threshold level used is sufficiently higher than the first so that the added noise is still less than the peak amplitudes that exceed the second threshold level. For example, if the second threshold level is 0.2 × 10 ^-6 m, it would be approximately 40 times larger than the threshold level of 0.1 × 10 ^-6 m, the added noise being much lower than such a high threshold level.

Each signal processing circuit thus determines together with your detector when a particle appeared and also the size of the scatter event that occurs between two elements can be split when the picture is on the border or is slightly out of focus.

A further improvement in the detection can be achieved by installing a noise suppression circuit in the parallel processor circuit of FIG. 7.

As shown in FIG. 8, the noise suppression device 58 consists of a plurality of noise suppression circuits 60 (designated as 60 a and 60 b in FIG. 8), to which the output signals of the detectors are applied, which monitor mutually dependent or non-adjacent sections of the entire measuring range which is swept by the detector.

As shown in Fig. 6, the detector 33 has a plurality of detectors which are connected so that each area monitored by each detector covers a portion of the entire measurement area, each area having some neighboring areas and the rest of the other areas being removed or not is adjacent.

According to Fig. 8 the detector 34 monitors a an area which is adjacent to the detector 34 b, however, the c is not adjacent away from the detector 34 or. In the same way, the detector 34 b monitors an area which is adjacent to that of the detector 34 c , but which is or is not adjacent to the detector 34 d , and finally the detector 34 c monitors an area adjacent to the detector 34 d .

Although only four detectors are shown in Fig. 8, it is apparent that this circuit pattern is performed by all detectors used and that pairs or groups of detectors which monitor distant or non-adjacent areas can be easily selected for noise suppression .

Fig. 8 also shows that the output signal of the detector 34 a via a preamplifier 42 a to the noise suppression circuit 60 a is present, while the output signal of the detector 37 c (ie the detector which is not compared to the area monitored by the detector 34 a) adjacent area monitored) reaches a noise suppressor circuit 60 a via a preamplifier 42 c . Likewise, the output signal of the detector 34 b is via a preamplifier 42 b at the noise reduction circuit 60 b and the output signal of the detector 34 d (ie the detector which monitors an area not adjacent to the area monitored by the detector 34 b ) via a preamplifier 42 d at the noise suppression circuit 60 b (it should be noted that the number of suppression circuits is normally half the total number of detectors used).

After the noise reduction at the noise reduction circuits, the rest of the signal arrives at particle detection circuits. According to FIG. 8, the output of the noise suppression circuit 60 is a so at a comparison circuit 44 a, to a holding amplifier 46 a and a resistor 50 a on a program inspection circuit 36, while the output of the noise suppression circuit b 60 b to a comparison circuit 44, a Holding amplifier 46 b and via a resistor 50 b to the program search circuit 36 (all of these elements have been described with reference to FIG. 7).

Fig. 9 shows the specific circuit diagram of the noise drückungsschaltung 60 (it being noted that the walls ren noise suppression circuits 60 are identical to it) a. As shown, the output signal of the detector 34 a passes through the preamplifier 42 a to the register 62 of the noise suppression circuit 60 a , while the output signal of the detector 34 c is applied via the preamplifier 42 c to the inverter 64 of the noise suppression circuit 60 a , the output signal of the inverter 64 arrives at resistor 66 . The resistors 62 and 66 are connected to one another and activate the output signals of the detectors, the sum output signal of the summing connection 68 to the comparison circuit 44 a , the holding amplifier 46 a and via the resistor 50 a to the program analysis circuit 36 .

Fig. 9 also shows, for example, that when the output signal nalA (of the detector34 a) both noise and an Has useful signal and if the output signalB (from De tector34 c) contains only noise, the noise of the inver output signalB (  toFig. 9) the noise in the out output signalA suppressed when both signals through the Noise reduction circuit can be summed (i.e.,A +  = 0).

Common mode disturbances cannot be coordinated with each other from points on a beam that are very far apart (ie more than a few mm). Although there is a difficulty with all beams, the difficulty is greatest with higher order lasers for operation in multiple types where the coherence length is short. Since the degree of coordination or correlation is related to the spatial phase coherence on the radiation practice when samples are drawn at short intervals, the coherence is greatly improved and the noise is matched to it. Therefore, if an array of detectors 34 is used to probe short distances, and then the output signal from a pair of detectors is inverted, noise reduction according to an embodiment of the invention is achieved (it being noted that the adjacent areas monitoring detectors are not ordered in pairs can be because the images from two adjacent detectors are to be shared).

In operation, a laser beam is passed through the measuring area and a gas with particles is allowed to flow through the measuring area with throughputs which have at least a throughput of approximately 1 cfm. Particles within the measuring range cause light scattering, whereby light scattering due to particles whose diameter is at least 0.1 × 10 ^-6 m can be detected by the detector arrangement. Each detector of the arrangement monitors a certain volume of the measuring range, measures the visible light scatter within the monitored section and generates an electrical output signal for the scanned light scatter (together with the signals generated by the background noise). In order to achieve high-resolution readings, the stored peak amplitudes are then read out individually and converted into a digital output signal.

In an operational embodiment of the invention, a sensitivity of particle sizes of 0.1 microns in an environment with high molecular scatter could be achieved, which is represented by a throughput of 0.02832 m ³ / min (1 cfm).

The invention thus offers an improved detector for Particle sizes that have a high sensitivity in a reverse Exercise with high molecular scatter by using an An order of detectors achieved, which the required Reduce the noise due to the background with Cause molecular scattering.

Claims (11)

1. Particle measuring device for a particle size detector with a first device which enables a medium containing particles to be guided through a certain measuring range and a second device which illuminates the measuring area so that the particles in the measuring range cause light scattering, characterized in that , in that the measuring particles (10) the following blocks:
a detector arrangement ( 33 ) which receives and detects the light scattered in the measuring region ( 25 ) and has a plurality of detectors ( 34 ) which are arranged in such a way that certain detectors ( 34 a , b , c ) of certain sections ( 24 ) of the measuring area ( 25 ) receive scattered light, the certain sections ( 24 ) of the measuring area ( 25 ) being spaced apart or at a distance from each other and each detector ( 34 a , b , c , d ) emits an output signal which is both noise can also contain useful signals which contain the detected light, which has been scattered by the particles in the specific area ( 24 ) of the measuring area ( 25 ), which is detected by the specific detector ( 34 a , b , c , d ) , such as
a processor ( 35 ) with a noise suppression arrangement ( 58 ) which is connected to the individual detectors ( 34 a , b , c , d ) and causes the noise-containing output signals of the individual detectors ( 34 a , b , c , d ) are essentially deleted so that the output signals of the processor ( 35 ) practically only show the particles which cause the light scattering in the measuring area ( 25 ). 2. Device according to claim 1, characterized in that it ( 10 ) enables the effective operation of the particle size detector with a strong molecular background noise. 3. Device according to claim 1 or 2, characterized in that the detector arrangement ( 33 ) and the processor ( 35 ) enable effective detection and processing of the light scattered by particles in the measuring area ( 25 ), the particles being one Have a diameter of at least 0.1 µ and larger. 4. Device according to one of the preceding claims, characterized in that the detector arrangement ( 33 ) and the processor ( 35 ) an effective detection and processing of the light scattered by the particles in the measuring region ( 25 ) he possible, the flow of the medium with the particles through the measuring area ( 25 ) with a throughput of at least one cubic foot (about 0.028 m ³ / min.). 5. Device according to one of the preceding claims, characterized in that the processor ( 35 ) is a parallel processor which has a plurality of processing circuits ( 41 ), each with a different one of the detectors ( 34 a , b , c , d ) is connected to process their output signals separately and that the output signals of the processing circuits ( 41 ) are connected to the noise suppression arrangement ( 58 ). 6. Device according to one of the preceding claims, characterized in that the noise reduction arrangement ( 58 ) has a plurality of noise reduction circuits ( 60 a , b ), each of which with two specific detectors ( 34 a + 34 c ; 34 b + 34 d) connected is. 7. Device according to claim 6, characterized in that each noise suppression circuit ( 60 a , b ) has a summing element ( 62 , 66 ) and a signal inverter ( 64 ) which is connected to one ( 34 c) of these detectors ( 34 ) invert their output signal before feeding them into the summing elements ( 62 , 66 ). 8. Device according to one of the preceding claims, characterized in that the detector arrangement ( 33 ) have second predetermined detectors ( 34 a , 34 d ) which receive light from second sections ( 24 ) of the measuring range ( 25 ), where the second sections ( 24 ) arranged at a distance from one another and adjacent to the respective first section ( 24 ) of the measuring range ( 25 ), and in that the processor ( 35 ) has a second noise suppression circuit ( 60 b ) which is connected to the second group the detectors ( 34 a , 34 d ) is connected to cause noise suppression in their output signals. 9. Device according to claim 8, characterized in that the detector system ( 30 ) receives and detects the light scattered over the substantially entire measuring range ( 25 ), with certain first and second sections of the measuring range ( 25 ) alternating. 10. The device according to claim 8 or 9, characterized in that the noise reduction arrangement ( 58 ) has a plurality of noise suppression circuits ( 60 a , b ), the number of which is equal to the number of detectors ( 34 a , b , c , d ). 11. Device according to one of the preceding claims, characterized in that the processor ( 35 ) has amplifiers ( 42 , 44 ) which are connected between the detectors ( 34 a , b , c , d ) and the noise suppression system ( 58 ).