Apparatus and method for conductivity measurement including probe contamination compensation

[0002] 1. Field of Invention

[0003] The present invention relates generally to conductivity measurement systems and the low-cost probes thereof, such as used in commercial dishwashers, and particularly to conductivity measurement systems which compensate for probe contamination.

[0004] 2. Description of Related Art

[0005] Industrial dishwashers use conductivity measurement systems to maintain proper detergent concentrations in the dishwashers' wash water. Conductivity measurement systems are well known and typically include a probe that has first and second electrodes submerged in the wash water. A signal from a source circuit is applied to the electrodes to induce a current between the electrodes. This current, which is mirrored in the source circuit, is determined by dividing the voltage in the source circuit by the impedance of the source circuit. The conductivity of the wash water is then determined by dividing the current between the electrodes by the voltage across the electrodes.

[0006] Current flow in an aqueous solution, e.g., the wash water, is facilitated by the flow of ions between the electrodes. In an industrial dishwasher, the ions are provided by the detergent. Thus, increasing the detergent concentration results in a corresponding increase in the conductivity of the wash water. The relationship between wash water conductivity and detergent concentration for a particular detergent is typically stored in a look-up table, thereby allowing detergent concentration to be easily derived from wash water conductivity.

[0007] As current is induced between electrodes in an aqueous solution, ions begin accumulating on one of the electrodes. The ions accumulating on the electrode surface each occupy a finite space such that after a time period t there is no more available surface area on the electrodes on which ions may accumulate. This phenomena, known in the art as polarization, reduces current flow between the electrodes and may result in erroneous conductivity measurements which, in turn, lead to erroneous detergent concentration measurements. Thus, when the electrodes become polarized, the detergent concentration of the wash water is perceived by the dishwasher to be too low, thereby leading to the addition of detergent to wash water that may, in reality, already be of a desired detergent concentration.

[0008] Further, when used as described above, the electrodes undesirably accumulate non-conductive particles thereon which, in turn, reduce the effective area of the electrodes. As a result, contamination of the electrodes speeds the above-described polarization of the electrodes and, therefore, diminishes the useful life of the electrodes.

[0009] In theory, the effect of polarization upon conductivity measurements can be eliminated by calculating conductivity the instant current is induced between the electrodes, since at time t=0 ions have not yet accumulated on the electrodes. Here, the voltage between the electrodes must be measured just as the source signal that induces current in the wash water is asserted. Unfortunately, such an approach is not feasible. First, there are significant characteristic variations between ion species during the first 1-2 microseconds of aqueous current flow. Since the ion species of the detergent is typically unknown, measurements taken within the first 1-2 microseconds are unreliable. Second, it is very difficult to fabricate a circuit which can produce a source pulse and then immediately capture an analog reading produced by the source pulse.

[0010] U.S. Pat. No. 4,756,321 discloses an industrial dishwashing system in which a continuous AC signal is applied to first and second electrodes submerged in wash water to induce a current between the electrodes. The resulting current is measured over time, and then used to calculate the conductivity of the wash water. Conductivity is then converted into a logarithmically scaled detergent concentration. Here, the continuous current flow between the electrodes results in a continually increasing polarization of the electrodes. As a result, the electrodes must be either cleaned or replaced at regular intervals. The servicing of the electrodes is not only expensive, but also reduces operating efficiency of the dishwasher. Further, this system's inability to measure or predict electrode contamination makes it even more difficult to optimize the useful life of the electrode.

[0011] Another approach involves driving the electrodes with a pulsed DC signal as described, for instance, in U.S. Pat. No. 4,119,909. In that system, the pulsed DC signal induces short pulses of current between the electrodes in the wash water. Use of short current pulses reduces polarization and, thus, increases the useful life of the electrodes, as compared to the averaging technique disclosed in U.S. Pat. No. 4,756,321. However, conductivity measurements provided by this approach are nevertheless influenced by polarization. Further, this system, like that disclosed in U.S. Pat. No. 4,119,909, is unable to measure or predict electrode contamination. It is therefore difficult to accurately determine when or at what rate the measured conductivity deviates from the actual conductivity and, as a result, the accuracy with which this approach maintains the detergent concentration at a target level is compromised. It is thus also difficult to maximize the intervals at which the electrodes are cleaned or replaced and, therefore, difficult to maximize the useful life of the electrodes.

[0012] An apparatus and method for measuring conductivity of an aqueous solution are disclosed which compensate for polarization and provide warning of electrode contamination. In accordance with the present invention, one or more DC pulses are applied to first and second electrodes submerged in an aqueous solution such as, for instance, the wash water of an industrial dishwasher. The voltage at the first electrode is measured at first and second predetermined times after initiation of the one or more DC pulses. The difference between the respective first and second measured voltages is calculated and then compared to a predetermined threshold value. If the difference voltage exceeds the predetermined threshold value, thereby indicating that the electrodes are sufficiently contaminated so as to soon require cleaning or replacement, an alarm signal is asserted. In this manner, present embodiments maximize the useful life of the electrodes and, thus, minimize servicing costs.

[0013] Further, present embodiments provide conductivity measurements compensated for polarization. Linear regression of the first and second measured voltages is used to calculate the voltage at the first electrode at the beginning of the one or more DC pulses, i.e., at time t=0. The resistivity of the solution is calculated using Ohm's Law, and then converted into conductivity according to the known K factor of the solution. In some embodiments, conductivity is provided in logarithmically scaled measurement units, known in the art as Beta units. Since the conductivity of the solution is calculated according to the electrode voltage at the beginning of the DC pulse, the measured conductivity of the solution is not influenced by polarization. In this manner, present embodiments effectively compensate for polarization, and thereby produce a more accurate conductivity measurement, as compared to the prior art. As a result, present embodiments greatly reduce the likelihood of incorrect detergent concentrations resulting from erroneous conductivity measurements.

[0014] FIG. 1 is a block diagram of an industrial dishwasher in which a conductivity measurement system in accordance with the present invention is used;

[0015] FIG. 2 is a block diagram of a micro-controller suitable for use in the conductivity measurement system of FIG. 1, including a schematic diagram of a four-layer capacitor modeling the polarization of wash water;

[0016] FIG. 3 is a timing diagram showing various signals associated with the operation of one embodiment of the present invention; and

[0017] FIG. 4 is a flow chart illustrating operation of a conductivity measurement system in a preferred embodiment of the present invention.

[0018] Like components in the Figures are similarly labeled.

[0019] Principles of the present invention are described below with reference to the industrial dishwasher 20 disclosed in U.S. Pat. No. 4,756,321, incorporated herein by reference, for simplicity only. It is to be understood that embodiments of the present invention may be used in other industrial dishwashers, or for any application in which it is desired to measure the conductivity of an aqueous solution. Accordingly, the present invention is not to be construed as limited to specific examples herein.

[0020] Referring to FIG. 1, an industrial dishwasher 20 of the type described in U.S. Pat. No. 4,756,321 is shown to include a conductivity measurement system 100 in accordance with the present invention. System 100 is connected to a probe 52 having first 54aand second 54belectrodes submerged in a tank 35 of wash water used to wash dishes 38. In response to signals received from the probe 52, the system 100 provides control signals to a rinse pump 28, a detergent pump 32, and a sanitation pump 31 so as to ensure proper concentrations of a rinse agent, detergent, and a sanitation agent, respectively, within the wash water. For a discussion of the general operation of the dishwasher 20, as well as the advantages realized thereby, see U.S. Pat. No. 4,756,321.

[0021] Referring to FIG. 2, the measurement circuit 100 includes a micro-controller 102 having an output terminal OUT coupled to the first electrode 54avia a resistor R₁, where resistor R_ONmodels the on-resistance of the micro-controller 102. The resistor R₁should be of a value suitable for the conductivity range of the wash water. In one embodiment, where R_ONis 60Ω, a value of 200Ω is chosen for resistor R₁, as explained in detail below. The micro-controller 102 also has an input terminal IN coupled to the first electrode 54avia a resistor R₂which serves as a series protection resistor for the ADC input terminal. Although the resistor R₂should thus be as large as possible in order to provide maximum protection for the ADC input terminal, the resistor R₂must also be small with respect to the input impedance of the ADC 104 in order to preserve signal strength. In one embodiment, where the micro-controller 102 is able to operate accurately with a source impedance as high as 10 kΩ, a value of 4.7 kΩ is selected for the resistor R₂. The second electrode 54bis tied to ground potential. Block 200 is an electrical representation on the wash water in the tank 35, where capacitors C1 and C2 form a four-layer capacitor which models polarization of the wash water, and the impedance element Z models the impedance of the wash water. Increases in electrode contamination are modeled by reducing the size of the capacitors.

[0022] The micro-controller 102 includes an analog-to-digital converter (ADC) 104, a look-up table 106, a memory 108, a central processing unit 110, and a timing circuit 112. The micro-controller 102 is connected to a voltage supply V_DCand ground potential.

[0023] Referring also to the timing diagram of FIG. 3, the micro-controller 102 generates at its output terminal OUT a DC pulse having a duration of T and an amplitude equal to V_DC. The voltage at the electrode 54ais measured at times t₁and t₂, where t₁<t₂<T, thereby giving measured voltages V₁and V₂, respectively, which are stored in the memory 108. If there is no polarization within the wash water, the measured voltages V₁and V₂will be equal, as illustrated by case A of FIG. 3. If, on the other hand, there is polarization, the current flow between the electrodes 54aand 54bwill decrease between times t₁and t₂and, therefore, the voltage V₂will be greater than the voltage V₁, as illustrated by case B in FIG. 3. The rate at which the voltage at the electrode 54achanges is substantially linear and, therefore, compensation techniques discussed below utilize linear algorithms. However, where greater accuracy is desired, more complex, non-linear compensation techniques are used.

[0024] In calculating the conductivity of the wash water, linear regression is used to determine the voltage at electrode 54aat time t=0, i.e., at the beginning of the DC pulse. The voltage on the electrode 54aat time t=0 is given as: $V_0=\frac{\left(V_1*t_2\right)-\left(V_2*t_2\right)}{\left(t_2-t_1\right)}$

[0025] Since the voltage V₀corresponds to time t=0, the voltage V₀is not influenced by the effects of polarization. Thus, the conductivity of the wash water, which is calculated using Ohm's law and the known K factor of the electrodes 54aand 54b, is not influenced by polarization within the wash water. In this manner, present embodiments compensate for polarization.

[0026] The difference between the first and second measured voltages, V_diff=V₂−V₁, is indicative of the extent to which the electrodes are contaminated. Accordingly, if the difference value exceeds a predetermined threshold corresponding to the maximum degree of acceptable electrode contamination, the micro-controller 102 generates an alarm signal alerting an operator of the system 20 that the electrodes 54aand 54bneed to be cleaned or replaced. Further, in some embodiments, if the measured voltage V₀at time t=0 is greater than a second predetermined threshold, the micro-controller 102 generates an alarm signal alerting the operator that the electrodes 54aand 54bare sufficiently contaminated so as to require servicing. In this manner, present embodiments facilitate servicing of the electrodes 54aand 54bbefore contamination becomes sufficient to degrade conductivity measurement accuracy.

[0027] In preferred embodiments, an inexpensive micro-controller such as, for instance, the 16C72, 16C73, or the 16C74, all available from Microchip Semiconductor Corp. is used for the micro-controller 102 in order to minimize cost. Since these micro-controllers are typically unable to take quick successive analog measurements, the first and second voltage measurements, V₁and V₂, are sampled during two separate pulses, as explained below, where V_DC=5 volts, R_ON=60Ω, R₁=200Ω, R₂=4.7 kΩ, t₁=10 μs, t₂=15 μs, T=20 μs, and K=0.4.

[0028] Referring also to the flow chart of FIG. 4, the micro-controller 102 generates at time t=0 a first DC pulse having an amplitude of 5 volts (step 1). The DC pulse induces an electric field between the first and second electrodes 54aand 54bwhich, in turn, results in current flow between the electrodes 54aand 54bin the wash water. At time t=10 μs, the micro-controller 102 samples the voltage at the first electrode 54avia resistor R₂(step 2). The resultant analog voltage V₁is provided to the micro-controller 102 via its input terminal IN and is thereafter converted to a digital voltage D₁via the ADC 104. The digital voltage D₁is stored in the memory 108 (step 3). At time t=20 μs, the micro-controller 102 terminates the first pulse, and the electrode 54adischarges to ground potential (step 4). After a predetermined period of time such as, for instance, 125 μs, the micro-controller 102 generates at its output terminal OUT a second DC pulse having an amplitude of 5 volts (step 5). The micro-controller 102 samples the voltage at the first electrode 54aat a time 15 μs after the second pulse is initiated (step 6). The resultant analog voltage V₂′ is converted to a digital voltage D₂′ via the ADC 104, and stored in the memory 108 (step 7).

[0029] As mentioned above, the conductivity measuring system 100 alerts an operator of the dishwasher 20 when electrode contamination exceeds acceptable levels. Here, the CPU 110 of the micro-controller 102 calculates the difference between the first and second stored digital (binary) voltages, D_diff=D₁−D₂′ (step 8), and then compares the difference voltage D_diffto a predetermined threshold voltage D_th(step 9). If the difference voltage D_diffexceeds the predetermined threshold voltage D_th, the micro-controller 102 activates an alarm signal to alert an operator of the dishwasher 20 that the electrodes need to be serviced (step 10).

[0030] The first and second voltages, D₁and D₂′, are then processed by the CPU 110 according to the above-mentioned linear regression algorithm to determine the digital voltage D₀on the first electrode 54aat the beginning of the first DC pulse, i.e., at time t=0 (step 11). In one embodiment, the digital voltages D₁and D₂′ are stored as eight-bit numbers, where the binary number 255 corresponds to the analog value 5 (volts). This voltage at time t=0 and Ohm's Law are then used to calculate the conductance of the wash water (step 12). The conductance is converted to a detergent concentration using the look-up table 106 (step 13).

[0031] For example, where the digital voltages D₁and D₂′ are equal to 65 and 70, respectively, the value D₀(at time t=0) is equal to ((65)(15)−(70)(10))/(10+15)=55. The analog voltage A₀, which corresponds to the digital voltage D₀, is therefore equal to (5)(55)/(255)=1.08 volts. The current flowing in the circuit, i.e., through resistor R₁, is determined using Ohm's (V=IR). Here, the current is equal to (5−1.08)/(60+200)=0.0151 amps. The uncorrected resistance of the element Z, and thus the resistance of the wash water, is equal to (1.08)/(0.0148)=71.5Ω. The corrected resistance of the wash water is determined by dividing the uncorrected resistance by K, i.e., (71.5)/(.4)=178.75Ω, which gives a conductance of 1/(178.75)=0.00559 mhos (or siemens)=5590 μsiemans. This measured conductance corresponds to the beginning of the DC pulse, i.e., time t=0, and is thus not influenced by polarization. In this manner, the conductivity measurement system 100 avoids the polarization-induced, erroneous detergent concentration measurements characteristic of conventional conductivity measurement systems.

[0032] The K factor is indicative of the electrodes' sensitivity and is typically between 0.1 and 10, where electrodes having a small K factor are more suitable for measuring low conductances and, conversely, electrodes having a large K factor are more suitable for measuring large conductances. A cell formed of an electrode pair “sees” a conductance divided by the K factor, e.g., a cell having a K factor equal to 0.1 sees a conductance that is ten times larger than the actual conductance of the wash water.

[0033] In some embodiments, the micro-controller 102 converts the analog voltage A₀into Beta units, a unit of measure especially suited for use in determining the conductance of a detergent solution in an industrial dishwasher. Beta units are well known in the art and are thus not discussed herein. For a detailed discussion of Beta units, see U.S. Pat. No. 4,756,321.

[0034] In the above example, the cell formed by electrodes 54aand 54bhas a K factor equal to 0.4, the Beta Unit range is 60, and there are 3 counts per Beta Unit. The typical conductivity of wash water without detergent is about 600 μmhos, and the typical conductivity of wash water with a maximum detergent concentration is about 12,000 μmhos. Thus, the above cell formed by electrodes 54aand 54bactually “sees” wash water conductivities ranging from 600/0.4=1500 μmhos=0.0015 mhos to 12,000/0.4=30,000 μmhos=0.03 mhos. This conductivity range corresponds to a resistivity range of 1/(0.03 mhos)=33.3Ω to 1/(0.0015 mhos)=666.7Ω. Thus, the maximum Beta count, i.e., 180, corresponds to the minimum wash water conductance, i.e., R_WATER=666.7Ω. Using the voltage divider rule, $\frac{180}{255}=\frac{666.7\Omega }{\left(R_{\mathrm{WATER}}+R_{\mathrm{ON}}+\mathrm{R1}\right)}$

[0035] where R_ON+R₁=277.8Ω. Since as mentioned earlier R_ON=60Ω, a value of about 200Ω is chosen for the resistor R₁. Applicants found that a value of 200Ω for resistor R₁limited the current to a safe level.

[0036] In a preferred embodiment, the above-described pulse sequence is repeated every 250 ms so as to provide four conductivity measurements per second, although the interval between pulse sequences may be adjusted as desired for particular applications.

[0037] In the preferred embodiments, the conductivity measurement system 100 is implemented as software running on a micro-controller. Appropriate program modules may be stored on a CDROM, magnetic disk storage product, or any other computer readable data or program storage product. The software modules in the computer program product may also be distributed electronically, via the Internet or otherwise, by transmission of a computer data signal (in which the software modules are embedded) on a carrier wave.

[0038] While particular embodiments of the present invention have been shown and described, it will be obvious to those skilled in the art that changes and modifications may be made without departing from this invention in its broader aspects and, therefore, the appended claims are to encompass within their scope all such changes and modifications as fall within the true spirit and scope of this invention.