Apparatus for optically determining a concentration of a substance in a fluid

Abstract

The invention proposes an apparatus and a method for optically determining a concentration of a substance in a fluid, comprising a measurement amplifier for amplifying measured photocurrents, the measurement amplifier having at least one linear transimpedance amplifier for detecting and amplifying the photocurrents.

Description

[0001] Device for the optical determination of a substance concentration of a fluid

[0002] Description

[0003] The invention relates to a device and a method according to the independent claims.

[0004] In the process industry, particularly in biotechnology, optical measurements, such as IR (infrared) absorption or turbidity measurements, are used for quality assurance and process control. Such measurements are often based on the attenuation of a light beam through interaction with a fluid (absorption, scattering), and in some cases also on an increase in the detected intensity (scattering, fluorescence, change in polarization state, change in propagation velocity).

[0005] Document DE 10 2013 1 1 1 696 A1 discloses a measuring device for measuring the flow of a fluid in a pipeline. This measuring device has, among other things, a measuring amplifier which detects and amplifies a measuring signal such as a measuring voltage. However, such prior art measuring devices have numerous disadvantages, as explained below. The monitored variable can, for example, be in particular the cell density in bioreactors or a color change for the identification and / or separation of products or product phases in pipelines or the ultraviolet absorption (one or two wavelengths) for controlling protein separation processes in biotechnology. Ideally, the measuring device is as universally applicable as possible for such measurements.

[0006] The evaluation / visualization of measurement data and process control are increasingly performed using, for example, programmable logic controllers in conjunction with graphic displays to represent the user interface. A local user interface on the transmitter can often be omitted with an appropriate, particularly bidirectional, connection between the measuring device and the controller.

[0007] Sensors, amplifiers, and transmitters are available in various designs, e.g., optek submersible probes ASD 12 or ASD25. In these examples, an NIR LED sends a light beam through a window into a measuring gap filled with a fluid medium (pipeline, biotech fermenter). The light passes through the medium and is attenuated by absorption and / or scattering. The light beam leaves the measuring volume and strikes a silicon photodiode, which generates a photocurrent approximately proportional to the light intensity. The attenuation of the light can be used to determine the properties of the medium. e.g., for cell or bacterial fermenters, there is a correlation between cell density and the attenuation of the photocurrent.

[0008] It is typical for these applications that the intensity to be measured covers many orders of magnitude and that the radiated optical intensity is or should be limited, so that it is advantageous to be able to record the smallest possible photocurrents with high accuracy, linearity and reproducibility. With some sensors it is possible to modulate the radiated optical intensity over time, in particular , for example, to switch it off. If a measuring transducer is to be universally usable, it is advantageous if it can also be used for sensors whose intensity radiated into the interaction zone is approximately constant over time and, in particular, cannot be specifically modulated over time. In order to achieve universal applicability of the measuring amplifier, the response time should nevertheless be as short as possible, with times of around 100ms - 200ms being sufficient for detecting the phase boundary of fluids or for controlling valve units, e.g.in chromatographic substance separation, are considered sufficient and proven in operation.

[0009] The aforementioned measuring devices are intended for process control and monitoring. They should therefore provide a measurement signal that is at least quasi-continuous over time, even if the measured variables, as described above, vary over a wide range. The device should be compact, robust, and cost-effective to manufacture even in medium quantities. Furthermore, a wide temperature range is desired without compromising measurement accuracy.

[0010] For photometric measurement tasks, e.g. in spectrophotometers, where small photocurrents have to be detected, systems are often used in which the measuring light is temporally modulated with high dynamics (electrical or mechanical chopping or shredding) in order to be able to separate noise that changes slowly over time (evaluation e.g. with lock-in amplifiers). In some cases, such arrangements also use two-beam solutions in which the source intensity temporarily interacts with the measurement volume and temporarily is fed to the photodetector for normalization purposes via a path that is not influenced by the measurement volume. In these solutions, the modulation frequency of the measuring radiation (and thus of the photocurrents) must be sufficiently high in order to be able to follow the change in the measurement variable sufficiently quickly, even taking into account the necessary signal processing.This principle requires additional effort for modulating the light source, for transmitting and / or detecting the frequency and phase of the modulation, as well as for filtering the signal (e.g. in the sense of phase-sensitive rectification).

[0011] However, omitting such measures increases the demands regarding the magnitude and constancy of otherwise uncorrectable disturbances. Important aspects of suitable circuits for detecting photocurrents are discussed in US20220214212A 1 (PTL 1 ). However, the task is aimed at applications in which continuous measurements are only required to a limited extent and interrupting the measurement for configuration purposes and determining disturbances is possible.

[0012] In the context of the solutions presented there, for example, slowly changing DC components should be corrected by reference measurements. However, this presupposes that it is possible to interrupt the optical measuring radiation at suitable times in order to detect the influence of disturbances and correct them during the measurement. The teachings of this document are only partially applicable to applications in which the measuring radiation cannot or should not be deliberately prevented, or in which the measuring signal must be continuously available even in the event of significant changes in the measured variable. Measuring arrangements with regularly or temporarily interrupted measuring radiation are efficient for correcting slowly changing disturbances, but do not offer an intrinsic solution with regard to providing a sufficient dynamic measurement range.Photocurrent amplifiers that convert the primary photocurrent along a nonlinear, usually logarithmic characteristic (e.g., . into a voltage) are advantageous in this regard. This voltage can then be measured, for example, using a conventional, linear ADC (analog-to-digital converter). Such solutions can also be found in PTL 1.

[0013] In absorption measurements, the photocurrent is often approximately linked to the concentration of the absorbing substance via a Lambert-Beer law. With regard to the substance concentration, this results in a coherent linear measurement range that fully utilizes the ADC resolution. However, precise logarithmic amplifiers are technically demanding, sometimes require thermostatting, deliver difficult-to-predict rise times, and / or have input stages that disrupt the linearity of the connected photocurrent sources (photodiodes) through input (leakage) currents and input offset voltages. Linear transimpedance amplifiers (TIAs) are state-of-the-art for detecting photocurrents. The characteristic curve is defined by a negative feedback resistor (transimpedance) that is constant over a measuring range, and the photocurrent is converted proportionally into a voltage.

[0014] Achieving large measuring ranges, however, practically requires measuring range switching by influencing the transimpedance, as also described in PTL 1. Discontinuous switching of measuring ranges, however, disrupts signal filtering and makes it difficult to follow dynamic signals sufficiently quickly and accurately, which is why PTL 1 uses a solution with a nonlinear (e.g., logarithmic) characteristic. In addition, the switching elements must be supplied with energy. Semiconductor switches, as presented as possible solutions in PTL 1, are low-wear and fast, but they inject parasitic charges that are difficult to compensate for. These charges are indistinguishable from photocurrents and disadvantageously limit the achievable measuring range towards small currents. The photocurrent converted into a voltage can be converted into a digital value using known A / D converters.This seems to be advantageous even when analog interfaces (e.g. 4-20mA current interfaces) are used to transmit the measured value to higher-level controls.

[0015] State-of-the-art 24-bit sigma-delta converters are available as integrated components with comparatively low power consumption and a good price-performance ratio. They also offer highly effective internal filters to protect against interference at common line frequencies (50 or 60 Hz). However, inexpensive converters in this class with good filtering and low power consumption reduce data rates and increase latency to such an extent that signal rise times of 100 ms - 200 ms are difficult to reproduce. This also delays range switching, which leads to problems when tracking time-varying signals for visualization, control, or regulation purposes.

[0016] For example, a delayed switching can lead to the measurement signal exceeding control limits in the measurement chain, and thus to temporarily only strongly and non-linearly distorted measurement values ​​being available, which can disrupt a filtered value in the long term.

[0017] It is therefore the object of the present invention to eliminate the disadvantages of the prior art and, in particular, to provide an improved device for optically determining a substance concentration in a fluid and an improved method for optically determining a substance concentration in a fluid. This object is achieved by the subject matter of the independent claims. Advantageous developments of the invention are specified in the subclaims. The scope of the invention also includes all combinations of at least two of the features specified in the description, the claims and / or the figures. In the specified value ranges, values ​​lying within the stated limits are also intended to be disclosed as limit values ​​and to be claimable in any combination.

[0018] The invention relates to a device for optically determining a substance concentration of a fluid, comprising a measuring amplifier for amplifying measured signals, in particular photocurrents, wherein the measuring amplifier has at least one linear transimpedance amplifier for detecting and amplifying the signals, in particular the photocurrents.

[0019] The feature substance concentration in this disclosure includes not only the concentration of a substance in the narrow sense, but also the concentration of a substance in a specific electronic and / or spatial configuration as well as a relative ratio of such concentrations or an excess of such a configuration.

[0020] The invention further relates to a method for optically determining a substance concentration of a fluid, wherein a measuring amplifier amplifies measured signals, in particular photocurrents, wherein the measuring amplifier has at least one linear transimpedance amplifier for detecting and amplifying the signals, in particular the photocurrents.

[0021] According to the invention, it is particularly advantageous to provide continuous measurement data without jumps and with low temporal latency. Furthermore, it is advantageous to dispense with logarithmic amplifiers. All features described below for the device according to the invention apply analogously to the method according to the invention, and vice versa.

[0022] Furthermore, a measuring range switch and effective filtering are provided.

[0023] In a preferred embodiment, the device further comprises: a source emitting a source spectrum, a wavelength-selective means arranged in front of a measuring volume, a measuring space delimiting the measuring volume at least in one beam path, and a detector for measuring a wavelength-related absorption of a measuring spectrum passing through the measuring volume.

[0024] In another preferred embodiment, the device further comprises: a source emitting a source spectrum, a wavelength-selective means arranged in front of a measuring volume, a measuring space delimiting the measuring volume at least in one beam path, and a detector that measures light scattered in the measuring volume.

[0025] The term light or optical radiation should not be restricted to the visible range, but rather refer to one or more sub-ranges from the spectrum from ultraviolet to infrared radiation. In addition to conventional photodiodes (semiconductor diodes, e.g. based on silicon or germanium), all detectors that convert light into a photocurrent in a predictable manner can be used to convert the optical intensity into a photocurrent. The invention allows both positive and negative currents to be processed at an active measuring input; in the case of photodiodes, both a cathode and an anode can therefore be connected to the active measuring input. In a preferred embodiment, the measuring amplifier is configured to continuously record the photocurrents and make them available to a higher-level controller.This particularly advantageously enables continuous, uninterrupted and jump-free monitoring, which enables the control and / or regulation of process-relevant variables, the timely control of separation processes and the complete process documentation for the purpose of quality assurance.

[0026] In another preferred embodiment, the latency of the measuring amplifier is < 500 ms. This low latency advantageously enables, in particular, better and more stable control of the process parameters and is advantageous with regard to the control of separation processes, e.g., by switching a fluid valve upon detection of a phase boundary.

[0027] In another preferred embodiment, the measuring range of the measuring amplifier is < 4 decades. By limiting the dynamic range of the measuring range between the switching points, sufficient reserve remains, particularly advantageously, to ensure sufficient resolution at the lower end of the measuring range and sufficient measurable signal overflow at the upper end of the measuring range, so that even dynamically changing signals can be tracked despite range switching without significant loss of information at the switching point.

[0028] In another preferred embodiment, the measuring amplifier has a measuring range switch. This advantageously allows for a particularly wide overall measurement dynamic range while eliminating the need for logarithmic amplifiers. It is further preferred that the measuring range switch comprises a reed relay. This advantageously makes it possible, in particular, to implement a measuring range switch suitable for detecting unmodulated photocurrents, since it permanently ensures high switching dynamics (high isolation against interference currents and low on-resistance).

[0029] In another preferred embodiment, the measuring amplifier has input protection by means of diodes connected to an input potential of the measuring amplifier. Low-leakage amplifier inputs (usually based on isolated field-effect transistors) are sensitive to electrostatic discharges. A technically reliable application is advantageously enabled, whereby the potential difference across the protection diodes should be as small as possible to minimize leakage currents.

[0030] Low-noise input stages for transimpedance amplifiers contain sensitive components that can be damaged even by minor loads (e.g., electrostatic discharge / ESD). Reliable operation therefore requires suitable protective circuitry. Protecting the amplifier inputs with semiconductor diodes, which divert voltage spikes directly to the amplifier's positive or negative supply voltage, leads to unacceptable, temperature-dependent interference currents through these diodes due to the high voltage across them.

[0031] One solution involves connecting two parallel, oppositely polarized diodes to a voltage very close to the voltage of the respective input. Solutions that combine protective diodes, a voltage follower, and bypass diodes in a single housing or integrated circuit can also be used. Another preferred embodiment provides a metallic housing with a low-impedance connection to a reference input of the measuring amplifier. This advantageously achieves, in particular, low sensitivity to electrical interference fields expected in a technical environment.

[0032] In another preferred embodiment, a galvanic isolation of a photocurrent input of the measuring amplifier from a power supply unit and / or a communication unit is provided. This galvanic isolation advantageously minimizes interference currents in sensitive parts of the device or the measuring amplifier, and thus also interference voltages that these currents cause at inductors and conductor resistances. This, in particular, reduces interference influences on the reference potentials (e.g., ADC relative to the non-inverting input of the transimpedance amplifier).

[0033] In another preferred embodiment, the measuring amplifier has an attenuation for interference at 50 / 60 Hz > 100 dB. Such a high attenuation of interference at 50 / 60 Hz is particularly advantageous for enabling unaffected measurement of even small photocurrents.

[0034] In another preferred embodiment, the device or method is configured to detect unmodulated radiation. The advantage is, in particular, that signals from sensors that do not provide for source modulation or use sources that are not easily modulated can be evaluated.

[0035] Further advantages, features and details of the invention will become apparent from the following description of preferred embodiments and from the drawings, each of which shows a schematic view:

[0036] Fig. 1 shows a schematic diagram of a measuring amplifier of a device according to the invention,

[0037] Fig. 2 shows a measuring amplifier circuit board for a measuring amplifier of a device according to the invention, with a plan view from above on the left, a side view in the middle and a plan view from below on the right,

[0038] Fig. 3 shows a diagram in which a relative coefficient size b is plotted as a function of a coefficient index a,

[0039] Fig. 4 shows a transmission curve,

[0040] Fig. 5 shows a housing of a measuring amplifier of a device according to the invention and

[0041] Fig. 6 shows an exemplary embodiment of a device according to the invention. In the figures, advantages and features of the invention are identified by reference numerals that identify them according to embodiments of the invention, with components or features with the same or equivalent function being identified by identical reference numerals.

[0042] According to the invention, photocurrents are detected by a linear transimpedance amplifier (TIA) as shown in Fig. 1. The photocurrents are fed to an inverting input 101 of an operational amplifier 102, and there is a negative feedback of at least one resistor 103 as a transimpedance from an output to the inverting input 101. The operational amplifier 102 and the resistor 103 form the linear transimpedance amplifier.

[0043] To implement multiple measuring ranges, one or more resistors 104, 105 are connected in parallel to resistor 103. A characteristic curve (output voltage of a measuring amplifier as a function of an input current) is designed to be approximately linear, thus eliminating any nonlinear, particularly exponential or logarithmic, negative feedback.

[0044] In order to measure the smallest possible photocurrents, the minimization and / or compensation of leakage currents is necessary.

[0045] Reed relays 106, 107 can advantageously be used as switching elements. These enable a dynamic range unattainable with semiconductor switches (high insulation resistance in the non-actuated state, very low residual resistance in the switched state). Furthermore, the isolation between the control circuit and the switched circuit is better than could be achieved with semiconductor switches. The theoretically limited, reliably achievable number of switching cycles does not represent a limitation on service life in most real-world applications, taking into account the temporal behavior of the measured variable, especially if a measuring range switching is designed with sufficient hysteresis (e.g., at least 10%, better 30%, preferably more than 40% of the more sensitive measuring range).

[0046] The switching elements require space, are expensive, and are potential sources of interference currents. Their number, particularly in the area of ​​influence of the inverting input of the TIA, should therefore be minimized. To further reduce interference currents, actuating windings are preferably connected to a terminal 108 at a potential that is very close to the potential of the non-inverting terminal 109 of the transimpedance amplifier, in particular by being connected to it with a low impedance. In the non-driven state, the entire actuating winding is at this potential, which in the steady-state state of the transimpedance amplifier is very close to the potential of the inverting input. This eliminates any significant voltages that could drive unwanted leakage currents between the actuating winding and the photocurrent input.

[0047] Less sensitive measuring ranges can be achieved by connecting resistors in parallel with the negative feedback resistor for the most sensitive measuring range. Switching elements should be omitted in the path of the highest negative feedback resistor. This means that all switching elements are inactive in the most sensitive measuring range and the influence of the control of the switching elements can be minimized. When relays are used as switches, all actuating windings are completely at non-critical voltage in the most sensitive measuring range. In the preferred embodiment, the transimpedance amplifier is equipped with three measuring ranges (MB4, MB3, MB2). An integrated component with low input offset voltage and low input current and FET input transistors is used as the operational amplifier 102. The desired level of performance can thus be achieved without the need for a discrete design.The effective transimpedances in the respective measuring ranges are approximately 100 megohms (103) (most sensitive measuring range, MB4), to which a 560 kOhm (104) (MB3) and an additional 3000 Ohm (105) (MB2) can be connected in parallel. The TIA's dynamic range is approximately + / - 5V. If available, a switch to another measuring range occurs above 4.75V or below 0.0125V.

[0048] This ensures a sufficient distance between the switching level and the upper control limit (signal limitation) or lower control limit (lack of resolution) to enable tracking of even changing signals. Furthermore, a significant switching hysteresis results, which prevents frequent switching of the measuring ranges under real process conditions. To limit the bandwidth of the transimpedance amplifier and reduce the tendency to undesirable oscillation, it is advisable to connect capacitors 1 10, 1 1 1, 1 12 in parallel with the negative feedback resistors. These must be matched to the characteristics of the amplifier used and, above all, to the input capacitance at the inverting input 101.

[0049] If high input capacitances are to be expected, larger negative feedback capacitances must be selected. If the cable length for connecting the photodiode to the transimpedance amplifier is variable, a large negative feedback capacitance must be selected for stability reasons. Using short cables with low capacitances can lead to a significant change in the signal bandwidth. The strong variation in the signal bandwidth can be counteracted by deliberately increasing the effective input capacitance by connecting a capacitor in parallel. Although this reduces the maximum achievable bandwidth, this is offset by a significantly reduced change in the bandwidth as a function of the cable length (or other sources of input capacitance), which simplifies further signal processing, e.g. by frequency filtering.

[0050] In a preferred embodiment: 33pF for MB4, an additional 220nF for MB3, and an additional 1nF for MB2. These allow operation with an input capacitance of 50nF, which mitigates the further influence of external cable capacitances up to 30m with 200pF / m.

[0051] Low-noise input stages for transimpedance amplifiers contain sensitive components that can be damaged even by minor loads (e.g., electrostatic discharge / ESD). Reliable operation therefore requires suitable protective circuitry. Protecting the amplifier inputs with semiconductor diodes, which divert voltage spikes directly to the amplifier's positive or negative supply voltage, leads to unacceptable, temperature-dependent interference currents through these diodes due to the high voltage across them.

[0052] An alternative solution is to connect two parallel, oppositely polarized diodes to a voltage very close to the voltage of the respective input. Solutions that combine protection diodes 114, a voltage follower 115, and bypass diodes 116, 117 in a single package or integrated circuit can also be used. However, the voltage follower induces an additional voltage offset across the protection diodes due to its input offset voltage (and thus generates additional interference currents). This can be avoided if the protection diodes are directly connected to the non-inverting input 109 via a low-resistance resistor 118 or a bridge. In this case, the voltage follower is both decoupled and protected against overload by a resistor (possibly already integrated). The voltage follower 115 and the resistor 119 can also be omitted if the integrated solution allows it.

[0053] In the area of ​​the transimpedance amplifier and its input lines 109, 113, 101, the measuring amplifier is sensitive to the influence of electric fields and interference currents. This applies particularly to the circuit section connected to the inverting input of the transimpedance amplifier with low impedance (the active measuring input—in the preferred embodiment connected to a cathode of photodiode 120, which in the preferred embodiment is not part of the measuring amplifier but of the external measuring sensor—the connection to the reed relays, feedback components, especially in the most sensitive measuring range).

[0054] Electric field components can be minimized by electrical shielding measures 121, which are connected to the non-inverting input 109 with low impedance. In the preferred embodiment, a shielding housing made of sheet metal is used to protect the sensitive circuit components (in particular the transimpedance amplifier including the feedback components, the switching elements and the ESD protection circuit). Some mechanical properties of the preferred embodiment can be seen from Fig. 2. The shielding housing with TIA 201 is soldered onto the circuit board during the circuit board assembly process, with emphasis being placed on a tight connection to the circuit board. In addition, the circuit board itself is designed to be tight in the area of ​​the shielding housing (e.g., no open vias). Access to the interior of the shielding housing is possible via a plug-on cover 202 made of sheet metal.This allows the shield housing to be used as a lost mold for an insulating potting compound. This potting compound protects the sensitive areas of the transimpedance amplifier from contamination and the ingress of humidity. The circuit board is slightly larger than the shield housing and has tooth-like contact surfaces along its edge for mechanical and electrical connection. Together with the shield housing and potting compound, it forms a module that is soldered onto the main circuit board 203.

[0055] The sensitive areas, both inside and outside the encapsulation, are carefully protected, especially from circuit components that may assume a different potential. Conductive structures 121 are used as shields (e.g., traces), which are connected either directly to the non-inverting input 109 of the TIA or to its reference voltage. They enclose the traces of the sensitive circuit components almost completely, or at least largely. To ensure their effectiveness, these metal structures have open surfaces (no coating).

[0056] To minimize leakage currents within the printed circuit board material, vias are inserted into the shielding conductors. Their primary function is to absorb leakage currents, which would otherwise not be necessary for carrying operating currents. This allows leakage currents to be controlled at a level that makes it possible, even at elevated temperatures and elevated humidity, to dispense with the use of printed circuit board base materials other than glass-fiber-reinforced epoxy resins in the preferred embodiment. In particular, the use of fluoroplastics as a base material is not necessary. In the preferred embodiment, it is thus possible to reduce the uncompensated interference currents of the measuring transducer to less than 500 fA over a wide range of temperatures and humidity.Since small photocurrents are to be measured, effects of dielectric absorption can have a disruptive effect, particularly in the photocurrent circuits (capacitances between the anode line and cathode line of the photodiode, or between the inverting input and non-inverting input of the TIA, as well as capacitances in the TIA's feedback path), especially in sensitive measurement areas. In the preferred embodiment, capacitors with a plastic dielectric are used at these points, in particular with a dielectric made of PP, PET, or PPS, whereby SMD capacitors can preferably be used. To optimize the noise behavior, the signal amplification already takes place in the transimpedance amplifier; further voltage amplification before or in the analog-to-digital converter (ADC) 123 is omitted.

[0057] The input offset voltage of the transimpedance amplifier appears as an offset voltage at its output. If the dynamic range of the transimpedance amplifier allows it, it is sensible to select a higher value than the permissible voltage range of the subsequent circuit (e.g., ADC input dynamic range) and to reduce it before its input using a voltage divider 122, because the interfering input offset voltage is reduced in the division ratio. In the preferred embodiment, the output voltage of the TIA is nominally halved. This voltage divider can then simply be expanded to form an RC element by using a capacitor (not shown in Fig. 1) in parallel with the differential measurement input of the ADC, which limits the signal to a reasonable bandwidth before the ADC conversion. This can reduce the risk of the ADC being overloaded due to interference. Likewise, this reduces interference components caused by undersampling.In the preferred embodiment, the time constant is approximately 3.3 ms. Depending on the polarity of the photocurrent (e.g., the anode or cathode of the photodiode is connected to the inverting input), the TIA produces negative or positive output voltages. If, for shielding reasons (e.g., the photodiode housing is connected to the cathode), the anode is connected to the inverting input of the transimpedance amplifier, an unfavorable situation arises with regard to the ADC because the voltages to be measured are negative relative to the reference potential. The mirroring of the measurement signal at the reference voltage or the shifting by a constant amount induced additional noise (e.g., the input offset voltage of an operational amplifier).

[0058] In the preferred embodiment, an ADC is therefore used that can detect both positive and negative input voltages (nominally between -5V and +5V relative to the reference point 109), although the usable measuring range actually uses essentially only positive voltages. This reduces the component complexity. The resulting reduction in the usable AD converter resolution from 24 bits to 23 bits is at least partially compensated for by the reduced influence of additional disturbances. This means that disturbances (e.g. leakage currents and input offsets of the transimpedance amplifier, but also electromagnetic interference with a vanishing mean value over time) do not result in small signals lying below the detection range of the AD converter and thus no longer being reliably detected. The application of compensation currents, which ensure a fixed polarity of the input signal and which are proposed in PTL 1 under this aspect, can therefore be omitted.

[0059] Furthermore, this results in a largely symmetrical design of the supply voltages of both the TIA and the ADC relative to the reference signal. This simplifies the filtering of the supply voltage 124, which can be used jointly for both the ADC and TIA, and reduces the residual influences of the supply voltage on the measurement result. As shown schematically in Fig. 1, the AD converter 123 can be connected to a microcontroller 126, for example, via a serial SPI bus. In a preferred embodiment of the invention, the signal sampling occurs with a modulator frequency of the sigma-delta ADC of 256 kHz.From a multitude of filter functions implemented in the converter, a combination is selected that delivers the data with low latency (the filter chain oscillates within one filter / decimation cycle, the signal delay corresponds in continuous operation to the time period between two data points on the output) and a relatively high output data rate of about 400 data points per second with a data width of 24 bits each.

[0060] The omission of further decimation and / or filtering in the AD converter, in particular the omission of the use of sinc 3 Filters ensure a continuous data stream with relatively low latency. By using state-of-the-art microcontrollers (e.g., ARM Cortex-M4 derivatives), it is possible to inexpensively provide sufficient computing power, program memory, and data storage to calculate even relatively high-order digital filters.

[0061] In a preferred embodiment, a 60th-order FIR filter (61 symmetric real positive coefficients for the normal form 1, 16-bit coefficient resolution) is used. The relative distribution of the coefficients can be seen in Fig. 3; deviations from the normalization from 1 can be taken into account elsewhere in the calculation.

[0062] This makes it possible to generate very high attenuation levels in the range of mains-related interference (50 Hz and / or 60 Hz) and the resulting harmonics (e.g., integer multiples of the fundamental frequency generated by rectifying the mains voltage, with the interference amplitudes decreasing sharply with increasing frequency). Latency remains within acceptable limits for the applications described above: it can be seen that, due to the design principle, the effect of the input signal on the filter output can be delayed by a maximum of 61 / (400 / s), or approximately 150 ms. A step is mapped to at least 50% at the output after 75 ms. Fig. 4 shows the calculated effect of the filter.

[0063] With the selected filter, in combination with the transfer function of the analog signal chain and the transfer function of the ADC in the selected operating mode, the fundamental frequencies of the interference at 50 Hz can be attenuated by more than 160 dB (20 log (In / Out)). Interference with the full input amplitude is attenuated to such an extent that the amplitude at the filter output is nominally less than 1 bit (at 24-bit resolution). This high attenuation is available up to frequencies above 350 Hz. Thus, very high attenuations of up to 7 times the fundamental frequency (50 Hz) or up to 6 times the fundamental frequency (60 Hz) are achievable.

[0064] To achieve a wide dynamic measurement range, the transimpedance amplifier is switched. At the switching limits, the scaling of the ADC output signal changes. In the preferred embodiment, the necessary normalization is taken into account at the output of the digital filter. At the moment of switching, all ADC values ​​already in the filter are renormalized to the newly applied scaling to prevent unnecessary filter settling. After switching between two measuring ranges, the transimpedance amplifier exhibits a transient response. In the preferred embodiment, this is slowest when switching to the most sensitive measuring range. Here, it is helpful to select a relatively high bandwidth, but still within the stable operating range of the TIA, in order to achieve fast settling. In the preferred embodiment, fewer than 10 output values ​​are affected by the switching.

[0065] The interference can be minimized by replacing the disturbed ADC values ​​with estimated values ​​after switching. In the preferred embodiment, the last output value of the digital filter (taking correct scaling into account) is used as the substitute value. Depending on the expected temporal progression of the input variable, this value represents a better estimate for the substitute values ​​than, for example, the last available ADC value would do, since short-term interference has already been filtered out. Under real conditions, currents of 50 pA with a noise component of 0.5 pA can be measured with the selected arrangement. The uncorrected interference variables are less than 500 fA relative to the input in the most sensitive measuring range and at an ambient temperature of 65 °C. Currents of up to and including 1.5 mA can be measured.

[0066] The measuring ranges are individually characterized in the factory with regard to slope (effective transimpedance) and intercept (input offset of the TIA and leakage currents), and the measured values ​​are corrected and standardized accordingly before filtering. There is no need for re-characterization over the lifetime of the device. At the switching points between the measuring ranges, residual jumps occur which can be neglected compared to the noise component after filtering. The measuring transducer usually also provides the energy for operating the light sources in electrical form. In the preferred embodiment, an almost constant voltage is provided, which can be switched, in particular by the microcontroller. The current consumed by the light source is monitored, in particular measured with a time resolution. In the event of a malfunction (e.g.In the event of a fault (e.g., excessive current consumption, source not connected), the supply voltage can be switched off by the microcontroller. Furthermore, it is possible to record a current consumption modulated over time by the light source control and thus transmit information (e.g., on the operating status, operating time, or ambient conditions of the light source) to the measuring amplifier. In the event of a fault, the supply voltage can be switched on periodically to check for the persistence of the fault pattern and to resume normal operation if the fault disappears.

[0067] For communication with a higher-level control unit, the transmitter has an interface. In the preferred embodiment, this is an electrical, differential, particularly 2-wire RS485 interface with an additional line for the reference potential. Data is transmitted bidirectionally in a manner compatible with standard MODBUS installations (MODBUS RTU with data rates of 9600, 19200, and 38400 bps). The transmitter can be identified (device type, individual identification / serial number) and configured (bus address, transmission parameters) via the bidirectional digital communication. Furthermore, the user can set a start value to which measured values ​​are output relative to the calculated start value. The measured values ​​are available both with and without calculation of the start value, and a logarithmic ratio value related to the start value is also provided.The calculated measured values ​​are updated with a cycle time of approximately 100 ms. Response times are typically below 150 ms, and below 200 ms at the switchover points. In the preferred embodiment, the local user interface of the measuring amplifier is limited to a few LEDs for signaling the supply voltage and operational readiness, the reception of digital data, and the signaling of detected error states. These LEDs are visible through translucent openings in the front panel 502. A display for displaying variable symbols, text, or numbers is omitted; the entire configuration and functional scope is made accessible via the digital interface. A reset button is also provided, which enables both a restart and a reset to factory settings.

[0068] The measuring amplifier in the preferred embodiment is used to operate sensors such as the optek ASD 12, in which LED light sources and photocurrent sensors are arranged spatially adjacent in a metallic housing and are subject to similar interference (EMC). In addition, the LED supply and the photocurrent derivation are carried in a common cable. There is therefore no galvanic isolation between the voltage supply for the light sources and the measuring circuit for the photocurrents. However, galvanic isolation is provided between the photocurrent measuring inputs and the supply voltage, between the photocurrent inputs and the digital communication interface (RS485), and between the digital external communication interface and the power supply, with at least one functional isolation being provided with a nominal voltage of 250V AC.For improved immunity to interference, especially against surge voltages, the coupling between the potential islands is designed with low capacitance, specifically with a parasitic capacitance of < 5 nF. The measuring amplifier has several connection options for PE (protected earth) protective conductors to enable connection to the local equipotential bonding. This connection can be used to locally apply a shield of the RS485 cable to PE, as well as to enable a high-impedance connection of the photoelectric circuit to ground potential. The latter maintains the positive effect of potential isolation while allowing the dissipation of tribologically generated electrostatic charges, which can generate disruptive displacement currents in the measuring circuit or sporadic discharges with a broad interference spectrum.

[0069] To facilitate easy installation, the transmitter is provided in the preferred embodiment with a narrow plastic housing (Fig. 5), which allows mounting on standard DIN rails using a spring-loaded catch 501. It has a main circuit board 203 (Fig. 2) that is perpendicular to the mounting direction of the DIN rail and parallel to the main extension directions of the housing. A further circuit board 204 is mounted parallel to the main circuit board. This board carries the transimpedance amplifier with its shielding housing and its switching elements and feedback elements. For contacting the external circuit components, the housing is provided with four terminal blocks 205, each with three or four screw terminals.

[0070] The longitudinal direction of the terminal blocks (the direction in which the terminal points are arranged) is perpendicular to the main circuit board, with the terminal blocks containing contact elements that are soldered directly to the circuit board. The terminal of the active measurement input of the photoelectric circuits (in the preferred design, the anode branch) is placed between the inactive measurement input (in the preferred design, the cathode branch) and the shield connection, which is arranged at approximately the same potential. This arrangement of the terminal block and the signals allows for large air or creepage distances around the active measurement input. Where these distances are small, it at least ensures that the potential differences between conductive parts and the active measurement input are small. The active measurement input is thus well shielded against interference currents. The power supply is provided with a DC voltage with a nominal 24 V + / - 20% and a power of approximately 5 watts.

[0071] In the first preferred embodiment described above, the transmitter serves to supply a single-channel absorption system, e.g., optek absorption probes ASD 12 or ASD25.

[0072] In a second exemplary embodiment, the measuring transducer is used to operate 2-channel absorption sensors. The second embodiment is very similar to the first embodiment. However, in this case, the connected light source provides radiation at two different wavelengths, particularly in the ultraviolet range, whereby the radiation is largely constant for both wavelengths when switched on. The measuring sensor provides a photocurrent signal for each wavelength via photodiodes. The measuring transducer is accordingly equipped with two photocurrent inputs, each with a TIA and an ADC, which transmit the digital signals to a common microcontroller.Since the photodiodes in this arrangement are spatially adjacent and thus exposed to the same interference, it is not necessary to galvanically isolate the photoelectric circuits from each other; both TIAs can be operated at a common reference potential and common supply voltages can be used for both TIAs and both ADCs.

[0073] The two TIAs are each arranged in their own shielded housings. The shield electrodes and shield connections can be used jointly for both photocurrent circuits. The data transmitted to the higher-level controller via the digital bus is expanded compared to the first preferred embodiment. Start values ​​can be acquired for both photocurrent inputs jointly or separately on command via the bus. For both inputs, both raw data and data calculated with the start value are available, both in current-proportional and in absorption units (logarithmic scaling). In addition, a measured value proportional to the quotient of the two currents as well as the difference between the absorption values ​​for both photocurrent inputs are also provided. The housing used is twice as wide as for the first preferred embodiment; otherwise, the comments on the latter also apply mutatis mutandis to the second embodiment.

[0074] Further configurations and / or further embodiments can be derived from the solution approaches presented above. In some cases, the integration of the measuring transducer together with the measuring sensor (or parts thereof) in a common housing (transmitter) is advantageous. In these cases, particularly for use in biotechnology, it is usually sensible to design the housing to be hermetically sealed, easy to clean and steam sterilizable. In addition to one or two photocurrent inputs, further photocurrent inputs can also be implemented. Depending on the measuring sensor, these can be arranged either galvanically isolated from one another or in one or more non-separated groups. Such arrangements also allow, for example, the recording of the incident intensity for standardization purposes if the radiation source emits an intensity that varies too greatly over time.

[0075] For range switching of a TIA, other types of electromagnetic relays, including micromechanical ones, can be used alongside the preferred reed relays for reduced requirements. If the requirements for switching dynamics or interference currents are lower, e.g., because the light source can be modulated over time, in particular, largely switched off, the use of semiconductor switches may be the preferred solution. Slowly changing disturbances can then be assessed and corrected with the source switched off. The findings regarding range switching and transient response then also apply accordingly to this operating state.

[0076] Arranging the TIAs in separate housings represents a flexible, modular solution. If measuring inputs are not galvanically isolated from one another, it can also be advisable to accommodate several TIAs in a shielded housing. The number of measuring ranges and the measuring range limits can be varied. If several measuring inputs are configured, these can also differ in this respect. With regard to the power supply for the light sources, it is also possible to do without this or to configure them multiple times to supply and monitor multiple sources. The use of the digital filter described above represents a good solution, particularly under unfavorable EMC conditions or under rapidly changing process conditions. If the requirements are less stringent, however, filters with a lower order and lower attenuation at the critical interference frequencies can also be used.

[0077] For low requirements, an ADC operating mode can also be selected which generates a higher latency or a lower data rate. In some applications, it can be useful to provide discrete digital inputs and / or outputs in addition to a digital interface for fast or local applications, e.g. to signal the acquisition of start values ​​or the reaching of limit values. These outputs can then be fully or partially galvanically isolated from other connections. The measuring transducer can also generate further measuring functions from the acquired photocurrents, e.g. the calculation to measured variables such as transmittance, turbidity in turbidity units (FTU, NTU, PPM), color units such as Hazen, or ICUMSA is possible.

[0078] If the microcontroller is heavily loaded (e.g., due to the implementation of multiple photocurrent inputs using complex digital filters), it may be useful to set up multiple subsystems of the invention described above, each equipped with its own microcontroller. The data is preferably combined via a digital connection using known technology (serial transmission via UART, SPI, IIC, CAN bus, or similar, parallel digital bus connection) and made available to the higher-level controller. Screw terminals can also be replaced with spring terminals or plug-in sockets for spring terminals or screw terminals, or with connectors. Communication with the higher-level controller can also be achieved via alternative digital communication systems.

[0079] Suitable physical transmission paths include wired transmissions (RS485, Ethernet-like physics, APL, T l-L), but also optical transmission via fiber optic or wireless connections (WiFi, mobile communications). Protocols that can be used include TCP / IP, Profinet, Ethernet IP, OPC UA, and MQTT. If the data is available to the microcontroller at a comparatively high data rate and low latency, as described above, this offers expanded options for predictive measured value evaluation and range switching. Additional filters and algorithms can be connected in parallel to the actual filter for generating measured values. These filters and algorithms, for example, evaluate the signal curve with regard to the first temporal derivative or higher temporal derivatives with lower latency. This then makes it possible to make better decisions regarding measuring range switching and to estimate precisely fitting substitute values ​​for measured values ​​​​affected during the transient phase.

[0080] The use of a high-order FIR filter represents a manageable solution given the computing power available in the state of the art. Replacement with similarly effective IIR filters is possible. Within the scope of the invention, microcontrollers can also be replaced by microprocessors with external memory or peripheral components, as well as by controller implementations in FPGAs.

[0081] Fig. 6 schematically shows a device 600 according to the invention. The device 600 has a sensor 601 with a measuring gap 602 and a media seal (O-ring) 603. The sensor 601 is connected via a cable 604 via connectors to terminals 605 for the power supply and to terminals 606 for the photocurrent input of a measuring amplifier 607 (view obliquely from below), 608 (view obliquely from above). The measuring amplifier 607, 608 has terminals 609 for communication with a controller and terminals 610 for the power supply and the PE.

[0082] The basic functioning of such a device is known to those skilled in the art. For the specific functioning of the measuring amplifier 607, 608 in particular, reference is made to the above explanations. The embodiments described above serve merely to illustrate the invention and do not limit the basic inventive concept in any way.

[0083] List of reference symbols

[0084] 101 inverting input

[0085] 103 Resistor 102 Operational amplifier 104,105 Resistors

[0086] 106,107 Reed relay 108 Connection 109 Non-inverting connection 110,111,1 12 Capacitances 113 Capacitor

[0087] 114 Protection diodes 115 Voltage follower 116,117 Bypass diodes 118,119 Resistor 120 Photodiode 121 Screen structure

[0088] 122 Voltage divider 123 ADC 123

[0089] 124 Filter Supply

[0090] 125 digital bus

[0091] 126 microcontrollers 201 shielded housing with TIA

[0092] 202 Cover 203 Main circuit board

[0093] 204 Module circuit board 205 Terminal blocks a Coefficient index b Relative coefficient value 501 Bracket, detent 502 Front panel 600 Device 601 Transducer

[0094] 602 Measuring gap media seal (O-ring)

[0095] Cable with connector

[0096] Connections for supply

[0097] Photocurrent input connections

[0098] Measuring amplifier (view from below)

[0099] Measuring amplifier (view diagonally from above) Connections for communication with control system Connections for supply and PE

Claims

Patent claims 1. Device (600) for optically determining a substance concentration of a fluid, comprising a measuring amplifier (607,608) for amplifying measured photocurrents, characterized in that the measuring amplifier (607,608) has at least one linear transimpedance amplifier (102, 103) for detecting and amplifying the photocurrents.

2. Device according to claim 1, wherein the measuring amplifier (607,608) is configured to continuously detect and provide the photocurrents to a higher-level controller.

3. Device according to at least one of the preceding claims, wherein a latency of the measuring amplifier (607,608) is < 500 ms.

4. Device according to at least one of the preceding claims, wherein a measuring range of the measuring amplifier (607,608) is < 4 decades.

5. Device according to at least one of the preceding claims, wherein the measuring amplifier (607,608) has a measuring range switch.

6. Device according to claim 5, wherein the measuring range switching comprises a reed relay (106, 107).

7. Device according to at least one of the preceding claims, wherein the measuring amplifier (607,608) has input protection by diodes which are connected to an input potential of the measuring amplifier (607,608).

8. Device according to at least one of the preceding claims, comprising a metallic housing with a low-impedance connection to a reference input of the measuring amplifier (607,608).

9. Device according to at least one of the preceding claims, comprising a galvanic isolation of a photocurrent input of the measuring amplifier (607,608) from a supply unit and / or a communication unit.

10. Device according to at least one of the preceding claims, wherein the measuring amplifier (607,608) has an attenuation for disturbances of 50 / 60 Hz > 100 dB. 1 1 . Device according to at least one of the preceding claims, configured to detect unmodulated radiation.

12. Method for the optical determination of a substance concentration of a fluid, in particular with a device according to at least one of the preceding claims, wherein a measuring amplifier (607,608) amplifies measured photocurrents, characterized in that the measuring amplifier (607,608) has at least one linear transimpedance amplifier for detecting and amplifying the photocurrents.

Images