Device and Method for Measuring Electrical Power

The present invention relates to an apparatus, according to the preamble of Claim 1, for measuring electrical power.

The invention also relates to a method for measuring electrical power.

An application of the invention is also a measuring device and method for electrical energy.

100 million kilowatt-hour meters are manufactured each year. Recent years have seen the emphasis move to so-called inductive electronic kilowatt-hour meters. In addition, the significance of remote reading is increasing. Besides actual kilowatt-hour meters, power measurement also takes place in several machines and apparatuses. Though pressures exist to introduce power measurement in nearly all devices, this cannot yet be done due to the price of a power meter and the lack of an economical interface. If a generally used and economical interface and power-measurement component were to be available for machines and devices both in factories and in homes, the market potential for such components would be in the order of billions of items annually. Existing kilowatt-hour meters are still not integrated on the component level. This is due to the limitations set by high voltages and large currents, as well as the requirements for accuracy and a wide dynamic range.

At present, a company producing energy meters will sell its products to a power company, which installs them in enterprises and domestic households. If the business changed so that a component manufacturer were to sell power-measurement units to, for example, a domestic-appliance manufacturer, power measurement would very quickly be integrated in a single circuit.

At present, new kilowatt-hour meters using in domestic households cost the power company about € 30, if remote reading, for example, is not connected to it. The most common technology is to use a current transformer. This solution is expensive, because a current transformer cannot be used as such, as it becomes saturated with direct current

(present regulations require the meter to withstand direct current). One solution used is to connect two current transformers in series, but this increases the price of the meter. Another solution is to install a resistive current divider in the current transformer, in which case both the alternating and the direct current are reduced and the direct current does not saturate the transformer. In a method previously developed by the applicant of the present application, the measurement of current was performed inductively. This method is economical and functional, but requires an expensive IC circuit for multiplication. In addition, the inductive coil must be large enough to ensure that the voltage caused by the magnetic field will be sufficiently powerful. Resistance shunts have also been used in kilowatt-hour meters, but have a problem in producing sufficient voltage over the resistance without overheating the resistance with large currents. This has been a common solution, especially in single-phase kilowatt-hour meters. In all of the aforementioned solutions an IC circuit is used for multiplication.

Hall sensors have been used for a long time in kilowatt-hour meters, but their great thermal dependence and poor sensitivity have made this method difficult, hi a Hall sensor, multiplication takes place directly in the component, because the Hall voltage is the product of the magnetic field and the current travelling through the Hall component.

In the period 1978 - 1980, the applicant of the present application developed an electronic kilowatt-hour meter called a watt guard. The product was intended to provide domestic households with an economical meter for monitoring power consumption in various devices. Current measurement was based on a resistance shunt while multiplication took place using a pulse width-height converter. In the period 1984 - 1987, the applicant of the present application developed a kilowatt-hour meter based on microprocessor technology. This was the world's first microprocessor-based meter and is now used in electricity-quality meters. In the period 1996 - 1998, the applicant of the present application developed a kilowatt-hour meter based on gradiometric induction coils. In connection with it, an integrated circuit was also developed, in which the multiplication of the current and voltage took place using sigma-delta converters and digital multipliers.

The invention is intended to eliminate the defects of the prior art disclosed above and for

this purpose create an entirely new type of apparatus and method for measuring electrical power.

The invention is based on implementing the sensor component as a silicon micro- mechanical structure, in such a way that the mutual multiplication of the current and voltage takes place directly in the measuring component.

In one preferred embodiment of the invention, a time integral of the power is created, in order to determine the energy consumption.

More specifically, the apparatus according to the invention is characterized by what is stated in the characterizing portion of Claim 1.

The method according to the invention is, in turn, characterized by what is stated in the characterizing portion of Claim 12.

Considerable advantages are gained with the aid of the invention.

With the aid of the method according to the invention, sensitivity can be varied using both a spring factor and with the aid of the current travelling in the coil. The current is measured without galvanic contact. The component measures the active power directly using a single component. The method does not require an expensive IC circuit, instead a cheap CMOS circuit is sufficient. The dynamic range of the meter is wide. Power meters for different orders of accuracy can be developed from the component. The sensor is not sensitive to direct current and, if we use gradiometric reading, it is also not sensitive to an external alternating field. The invention permits not only active power, but also idle power to be measured. If the current travelling in the MEMS component is made from direct current, the component will become a magnetometer and can be used to measure the current. In addition, the device according to the invention is advantageously mass-produced.

hi the following, the invention is examined with the aid of examples and with reference to the accompanying drawings.

Figure Ia shows a schematic top view of one power meter according to the invention.

Figure Ib shows a side cross-section of the power meter according to Figure Ia.

Figure 2 shows a schematic view of a second energy meter according to the invention.

Figure 3 shows a schematic view of a third measuring arrangement according to the invention.

Figure 4a shows a schematic top view of the measuring arrangement according to the invention.

Figure 4b shows a cross-section on the plane A-A of the solution of Figure 4a.

The present invention discloses a new way to make a kilowatt-hour meter, which is not sensitive to direct current and in which the current and voltage are multiplied directly in the measuring component.

Figures Ia and Ib show a schematic view of the micro-mechanical power meter. The solution according to Figure Ia is intended to measure the power travelling in a phase conductor 3. The voltage being measured, i.e. the potential between the phase conductor 3 and the neutral conductor 7, is taken from the voltage divider 50 between the terminals 8 and 9 of the coil, and thus converted into current in the coil 2. The switch 51 and phase reverser 52 can be used to mechanically stabilize the sensor component, allowing the direction of the current travelling in the coil 2 to be changed periodically. In the figures, a rocker-type moving plate 1 is drawn, on both sides of which the coils 2 required to create a magnetic field are integrated. The rocker is thus supported on its base on a beam 5. The current conductor 3 that is the object of the measurement runs close to the coils, creating a gradient in the magnetic field at the sensor component 4. The position of the plate 1 relative to the conductors 3 and 7 is measured capacitively with the aid of the electrodes 6 shown in Figure Ib, and once the spring factor of the rocker 1, 5 is known, the force acting, which is in turn directly proportional to the power travelling in the

current conductor 3, can be determined directly from the location of the change. In a manner to be described hereinafter, the electrodes 6 can alternatively be used for force feedback, in which case the force acting is obtained through the feedback magnitude (current or voltage).

According to Figure 2, a second alternative is to coat the plate 21 with a metal layer 27 and induce in it an eddy current using an immobile coil 22 in the component, which is located on the base 26 at a distance from the plate 21. The coil 22 is connected in the same way as the coil 2 of Figure 1. The selection of the type of component used is based on the accuracy requirement, the dynamic range, and the manner of manufacturing the component. If we place a current conductor 23 close to the component, the current will induce a magnetic field in the plate 21. The current travelling in the plate 21 induces a magnetic dipole in the current being measured. The force acting on the plate is the vector product F = I ₁₁ XB ₁ of the current i _u travelling through the coil 22 and the magnetic field

B . . If the end result is regarded as a scalar quantity, we note that in the case of sine signals the force F = aUI cos φ, where φ is the phase difference between the current and the voltage and α is a constant. In other words, by measuring the integral of the force in time we can measure the electrical energy. The essential feature in this invention is that the multiplication operation of the two quantities required for power measurement takes place directly in the measuring component. On the other hand, the high sensitivity eliminates the need to bring a large current galvanically to the measuring component, as it is instead sufficient for the current conductor to run near the component. There are other attractive properties too associated with this method, but they will appear hereinafter when we describe the electronics of the meter.

The following describes one possible realization of the sensor. Force is measured, for example, by making the structure a rocker, in which there is a coil 2 creating a magnetic dipole on both sides of the rocker, for example according to Figure 1. On both sides there are a number of electrodes 6, with the aid of which the location of the rocker 4 is measured capacitively, while on the other hand the location of the rocker is kept constant by means of electrical feedback. The power being measured induces a force in the rocker, but the feedback voltage is adjusted in such a way that the rocker 4 remains on

average at the point of equilibrium. When the rocker is in equilibrium, the mean effective value of the compensating voltage will be the same as the active power being measured. The position of the rocker is measured capacitively. If there are electrodes of different size in the feedback, the feedback voltage can be scaled. This means that, if the power being measured is low, the feedback is routed to the rocker through an electrode with a low capacitance. The small electrode means that a high voltage will be required to achieve equilibrium. In the case of high power, the feedback is routed to a large electrode. Feedback electrodes of different sizes are shown in, for example Figure 2, with the reference number 25 while the measurement electrodes are shown with the reference number 24. This permits an extension to the dynamic range. In other words, by using, for example, pulses with a constant voltage and a duration made constant for compensation, we obtain the power directly from the frequency of the pulse queue. In addition, the idle power forces us to run the pulses to the opposite side of the rocker. The measurement and feedback are made against the earth plane 28. The difference in the end result depicts the active power while the number of 'negative' pulses depicts the share of the idle power. This means that the same component can be used to measure both active and idle power. In addition, by using pulse of differing during, or different electrodes, we can extend the dynamic range of the meter.

In micro-mechanical components, charging of the surfaces and mechanical instabilities often appears due to drift. If we reverse the direction of the voltage, the direction of the magnetic dipole will also reverse and through it the force acting on the rocker. If we reverse the direction of the voltage at intervals of, for example, 20 periods, we will be able to eliminate drift almost entirely. The direction of the voltage is changed using, for example, a micro-mechanical or semiconductor switch. A circuit for changing the direction of the voltage is also shown in Figure Ia.

If we make allowance for both the elimination of drift and the force feedback, we obtain a power meter, the accuracy of which depends only on the stability of the reference. In addition, because the force feedback is made directly digitally, there will be no increase in the inaccuracy of the meter due to signal digitalization. If the mutual locations of the current conductor and the component do not change, we can easily build an economical meter with the order of even 0.1.

Figure 3 shows an arrangement, in which the current conductor is formed in such a way that at the sensor the gradient of the magnetic field is small. Thanks to the symmetry of the current conductor, the field of the second magnetometer is the same but with the opposite sign. This arrangement means that the sum of these two power meters is independent of the external homogeneous 50-Hz magnetic field. The shaping of the current conductor also means that the power reading in the first order will not change, even if the component moves relative to the current conductor, for example, due to thermal expansion.

Kilowatt-hour meters should withstand an extremely powerful 50-Hz (or 60-Hz) external magnetic field, without the meter showing a wrong reading. One way to eliminate the external field is to use a rocker-type MEMS component 4 according to Figure 1, but place the current coils 2 (as in the figure) on both sides of the rocker, in such a way that only the gradient of the field induced by the current conductor 3 leads to a force turning the rocker but the external homogeneous field is cancelled. In order that the sensitivity will be sufficient and the component will not be sensitive to the mutual positions of the current conductor and the component, the component should be reasonably large.

It is also possible to make an arrangement according to Figure 3. hi it, two identical MEMS kilowatt-hour sensors 31 and an IC circuit 34 connected to them is placed are placed in a case 30 located in the vicinity of the current conductor 33. A sufficient sensitivity will be obtained even though the MEMS component 31 was to be less than 1 mm x lmm in size. Because the elements of the gradiometric measurement are in different components, we can extend the cover by at least 5 mm - 8 mm, without the component costs being substantially increased. It should also be noted that the MEMS component 31 need not be packed in a vacuum, because in this application we can accept gas attenuation.

Figures 4a and 4b show a fourth solution according to the invention, in which the micro- mechanical component 41 is located inside the current conductor 43. The current conductor is preferably thinned in the vicinity of the line A-A, in order to increase the strength of the magnetic field. The coil 45 at precisely as possible at the centre of the

current conductor 43, when the sum flux of the magnetic field running through the coil 45 will be zero and the coil 45 can be used as a reference coil to eliminate external interference. The coil 46, for its part, is intended to be located at the maximum point of the magnetic field. The element 41 is typically like the rocker component 4 shown in Figures Ia and Ib and is possibly also electrically connected in the same way. The measuring and controlling IC circuit 44 is preferably manufactured on the same substrate with the micro-mechanical circuit 41 using, for example, the SOI (Silicon On Insulator) technique. The location of the encased element 40 in a slot made in the conductor 43 is shown in greater detail in Figure 4b.

If the component is used in a 1 -phase kilowatt-hour meter, it is preferable to place the entire measurement of power in an IC circuit inside the component, but in a 3-phase meter it is preferable to place only the electronics essential in terms of power measurement in a single IC circuit and connect a processor, which collects data from three components and controls the operations of the components, to the kilowatt-hour meter. The production costs are a single gradiometric power meter could be € 0.3 - 0.5 and the sales price correspondingly in the order of € 1.5 - 2. The cost in a 1 -phase kilowatt-hour meter is reasonable, because the measurement of power does not demand many external elements, but the total cost in a 3-phase meter is already significant. However, if we make a remotely readable kilowatt-hour meter, which contains a process and memory, the situation is again economical, in terms of the totality.

The present invention discloses a method for using micro-mechanical components in power and kilowatt-hour meters. In the method, the input of current and voltage is converted into a force, which is measured capacitively. The force is compensated preferably by a pulse queue using feedback electrodes of different sizes. The method compensates for possible non-linearity while the dynamic range is made extremely wide. The drift that may relate to the MEMS component is compensated by changing the phase of the alternating current proportional to the voltage travelling in the rocker. The effect of an external magnetic field on the operation of the meter can be eliminated, for example, by placing two components in the same case, in such a way that the homogeneous field will effectively induce a single large force in both sensors.