In a static meter of electric energy, the metering of current and voltage must meet strict requirements of precision and stability. Moreover, the dynamics of the metering of current must be considerably high because of the wide current range of the meters. In static meters of electric energy, in which the analog signals derived from the voltage and from the current of the site of consumption to be metered are converted into digital form in an A/D converter, from which the converted digital signal is passed further into a micro-controller for calculation of momentary power, the dynamics of the A/D converter must be wide. A/D converters that possess wide dynamics are, however, expensive components, for which reason, until now, their use has been restricted mainly to special meters.

In static meters of electric energy, traditionally a wiring system has been used in which a metering signal proportional to the voltage to be metered of each phase and a metering signal proportional to the current to be metered of each phase have been passed into an input channel of their own in a multiplexer. On the other hand, the output interface of the multiplexer has been connected to the input interface of an amplifier which can be set or programmed. The output gate of the amplifier has again been connected to an A/D converter, and the output interface of the A/D converter has been connected to a micro-controller. By means of an amplifier which has been fitted between the output of the multiplexer and the A/D converter and which can be set or programmed, a suitable level of amplification is sought until the result of the A/D conversion is acceptable so as to widen the range of measurement of the A/D converter of a narrower range of dynamics.

Such a procedure of metering based on an amplifier that can be programmed is, however, poorly suitable for static meters of electric energy, because, after each result of conversion, it is necessary to perform a comparison and, based on the comparison, a new amplification must be set in the amplifier that can be set or programmed. In order that the amplifier that can be set or programmed could be controlled by means of a micro-controller, the micro-controller must be provided with separate control pins for this operation.

The range of measurement of an A/D converter can also be widened so that the metering signal proportional to the current of each phase of the object to be metered is divided into a number of branches, each of which branches is passed through an amplifier with fixed amplification into the input channels in a multiplexer. In metering of the voltage, a result of transformation produced by one channel with the amplification one is sufficient, for the phase voltage remains highly stable as compared with the phase current.

By means of an arrangement in which the current channels at the input side of the multiplexer are divided into a number of branches provided with amplifications of different magnitudes, a method of metering is obtained which is quicker and better suited for a meter of electric energy than if an amplifier is used which can be programmed or set and which is placed between a multiplexer and an A/D con- verter.

In the FI Patent 86,224, a static meter of electric energy and a method for calibra- tion of said meter are described. The meter includes an electronic metering mechan- ism, in which there is an A/D converter with amplifiers connected ahead of said converter for different ranges of measurement as well as a controller part, in which there is a micro-computer which carries out the processing of the signals of the metering mechanism while taking into account the correction values of the different elements in the wiring system. In each particular case, the correction values repre- sent an overall value for the range of measurement concerned and are included in the characteristic quantities of the meter stored in the meter, each of which characteristic

quantities represents a numerical value corresponding to one specified amount of energy. Thus, here a metering signal proportional to the voltage of the object to be metered is connected directly to one input channel in the multiplexer. On the other hand, a metering signal proportional to the current of the object to be metered is divided into four branches, each of which branches is passed through an amplifier to the input channels of the multiplexer. Thus, in the input channels of the multiplexer, four current metering signals amplified by means of four amplifications of different magnitudes are available.

In said FI Patent 86,224, each amplifier is used within one current range meant for it only. The selecting of the current range at each particular time concerned is carried out by a central processor unit, which, based on the momentary output signal of the amplifier, controls another amplifier into operation when the value of the current has become higher or lower. In said patent publication, it is not described in more detail in what way this is carried out. Based on the specification of the patent, it is, however, obvious that the current ranges of the amplifiers have been deter- mined as effective values of the current or as quantities proportional to said values.

Thus, here the effective value of the current is determined in the central processor unit from a digital current signal after an A/D converter at specified time intervals, which time interval must probably be at least equal to the duration of one mains cycle. After the effective value of current has been determined, the current channel corresponding to said effective value of current is used in calculation of the momen- tary powers until a new effective value of current is determined at the end of the time interval, which new effective value of current again determines which current channel will be used during the next time interval, etc. Thus, the current channel to be used in calculation of the momentary powers is always defined in advance for the following time period of a duration of at least one mains cycle.

Such a method described in the FI Patent 86224, in which the current channel used in calculation of the momentary power is always determined in advance for a time period of a length of at least one mains cycle, is not suited for use in a situation in which the change in the amplitude of the current and the speed of change of the

momentary value of the current are great or high. A great change in the amplitude of the current and a high speed of change of the momentary value of the current during one mains cycle may cause saturation of the A/D converter before the micro- controller controls a current channel of lower amplification into use at the beginning of the next mains cycle or of the next mains cycle after that. At a time when the A/D converter is in a saturated state, calculation of momentary power cannot be carried out, which has again the consequence that the result of energy metering is distorted.

By means of the method of metering in accordance with the present invention, a static meter of electric energy is obtained which meets strict requirements of preci- sion and stability and in which the dynamics of current metering are high.

Further, by means of the method of metering in accordance with the present inven- tion, a static meter of electric energy is obtained in which the multiplexer can be controlled all the time in the same way and in which the current channel used in calculation of each momentary power can be selected on the basis of a simple algorithm. Thus, the metering and calculation of momentary power become highly precise and reliable. In a meter of electric energy that carries out the method, there cannot occur any overloading of the A/D converter arising from a great change in the amplitude of the current or from a high speed of change in the momentary value of the current. In a meter of electric energy that carries out the method, it is, however, possible to utilize the A/D converter constantly optimally. The very rapid adaptability of the present method to great changes in the amplitude of the current and to high speeds of change in the momentary value of the current is a property whose significance will be increased further, because the currents to be measured are often and to an increasing extent non-sinusoidal.

The principal characteristics of the method of metering in accordance with the present invention are disclosed in the characterizing part of claim 1.

In the method in accordance with the present invention, the principle mentioned above is applied, in which a metering signal proportional to the voltage of each phase of the object to be metered and a metering signal proportional to the current of each phase of the object to be measured are passed to the input channels of a multiplexer. Said metering signal proportional to the current of each phase has been divided into at least two parallel branches, which have been connected, by the intermediate of an amplifier with a fixed amplification of different magnitude, each of them to an input channel of its own in the multiplexer. Thus, in the input channels in the multiplexer, of the metering signal proportional to the current of each phase, at least two metering signals amplified by means of amplifications of different magnitudes are available. The output channel of the multiplexer has been connected directly to the input channel of an A/D converter, and the output channels of the A/D converter have been connected to a micro-controller.

In the method in accordance with the invention, the micro-controller controls the multiplexer all the time in the same way, i. e. each input channel in the multiplexer is alternatingly connected to the output channel of the multiplexer. Thus, in the micro-controller, all the time, corresponding to each digital sample of the voltage of each phase of the object to be metered, at least two digital samples are available, which have been amplified from the current of the corresponding phase with different amplifications. Based on an algorithm programmed in the micro-controller, it is judged which of the current channels is chosen as the channel whose digital sample is used at each particular time in calculation of the momentary power.

This algorithm can be based on the idea that the digital sample of the current channel of highest amplification is always examined first, and, if this sample does not cause saturation of the A/D converter, this sample is used in calculation of the momentary power. On the other hand, if the digital sample of the current channel of highest amplification causes saturation of the A/D converter, a transfer takes place to the current channel of next highest amplification, and so forth until the result of conver- sion is acceptable, i. e. within the range of operation of the A/D converter.

In a situation in which the current of each phase is divided into three branches of different amplifications, the current channel used for calculation of momentary power can also be selected so that the comparison is started from the channel of middle amplification. Based on the digital sample of the current channel of middle amplification, it is possible to conclude directly which of the channels is acceptable.

If the digital sample of the channel of middle amplification is very close to zero, for calculation of momentary power the channel of highest amplification is used. On the other hand, if the digital sample of the channel of middle amplification is very close to the highest (or lowest) possible value, for calculation of momentary power the digital sample of the channel of lowest amplification is used. In this case, it is required that the ranges produced by the channels overlap each other to some extent, as is usually the case.

After selection of channel, the momentary power is calculated by multiplying the conversion results corresponding to the phase voltage and to the phase current by each other. After this, scaling of the momentary power back to the level correspon- ding to the normal level can be carried out so that the value of momentary power is divided by the amplification of the current channel that was used. The result is a normal power reading, which has, however, been metered by means of a metering method which corresponds to an A/D converter of wide dynamics.

During one mains cycle, i. e. 20 milliseconds, as a rule, about 4... 200 power readings are calculated, depending on the desired band width. Thus, the momentary power can be calculated even at intervals of 100 microseconds. Thus, the momentary power is calculated at a number of different points by means of a sinusoidal current signal, in which connection the amplitude of the sinusoidal current signal varies in each calculation. Thus, near the zero point of the sinusoidal current signal, with a current channel with highest amplification, results suitable for calculation of momen- tary power are obtained, and near the peak point of the sinusoidal current signal, with a current channel with a lower amplification, results suitable for calculation of momentary power are obtained. Thus, the metering algorithm mixes the results obtained from different channels with each other during one cycle after calculation

of power. In such a case, in fact, all channels are used as mixed with each other, so that the metering errors produced by errors of amplification of the channels are mixed over the entire metering range.

The metering method in accordance with the invention can be applied to a single- phase or three-phase static meter of electric energy, by whose means an effective power, an idle power, or an apparent power is metered at a site of production or consumption of electric energy. The metering method can be applied with the mains frequency of 50 Hz commonly used in Europe and with the mains frequency of 60 Hz commonly used in the USA, or with any other mains frequency whatsoever. The metering method can be used in a situation in which the voltage and the current of the object to be metered vary sinusoidally, but it is also very well suited for use in a situation in which the voltage and/or the current of the object to be metered vary non-sinusoidally.

Thus, the metering method can be applied, for example, in a static single-phase meter intended for households, which is accomplished by using a 4-channel multiplexer and a 10-bit A/D converter so that one channel transforms the phase voltage and the remaining three channels transform the phase current while using, for example, the amplifications 1,8 and 32.

The metering method can also be applied in a static three-phase meter, which is accomplished by using a 12-channel multiplexer and a 10-bit A/D converter, in which case three channels are used for transformation of the phase voltages and the remaining nine channels are divided evenly for each phase current. In such a case, each phase current is amplified, for example, with the amplifications 1,8 and 32.

The current channels can also be divided into two branches only, in which case each phase current is amplified, for example, with the amplifications 1 and 8. In a three- phase meter, in such a case, a 9-channel multiplexer is needed, and in a single-phase meter a 3-channel multiplexer is needed. When just two current branches are employed, in the calculation of momentary power the selection of the current

channel to be used at each particular time is made quicker, but at the same time the dynamics of the current metering in the meter suffer slightly.

The current channels can, of course, also be divided into four or more branches if necessary. The higher the number of the branches into which the current channels are divided, the higher are the dynamics obtained for current metering, but at the same time the selecting of the current channel to be used in calculation of the momentary power at each particular time becomes slower.

In the method in accordance with the invention, of course, the metering signal proportional to each phase voltage of the object to be metered can be divided into a number of branches if this is necessary. Normally, the voltage of the object to be metered, however, remains to such an extent stable that this is unnecessary.

The maximal frequency of conversion of the A/D converter sets the upper limit for the frequency of sampling, and the higher the number of the current channels and/or the voltage channels that are used in the input of the multiplexer, the lower is the maximal frequency of sampling.

The metering method in accordance with the invention will be described in the following in detail with reference to the accompanying figures, the application of the metering method in accordance with the invention being, however, not supposed to be confined to said illustrations alone.

Figure 1 illustrates a static single-phase meter of electric energy as a block diagram, in which meter the metering method in accordance with the invention can be applied.

Figure 2 illustrates a static three-phase meter of electric energy as a block diagram, in which meter the metering method in accordance with the invention can be applied.

Figure 3 illustrates a sinusoidal voltage and a current amplified with three different amplifications, all of them present at the same phase.

Fig. 1 illustrates a static single-phase meter of electric energy, in which the metering method in accordance with the invention can be applied. The phase voltage U1 of the object to be metered is transformed by means of a voltage divider 201, and the phase current I1 is transformed by means of a current transformer 211 in order that a voltage metering signal Uul proportional to the voltage and suitable for the range of measurement of the meter of electric energy and a current metering signal Iil proportional to the current and suitable for the range of measurement of the meter of electric energy should be obtained. If the voltage and/or the current of the object to be metered is/are inside the range of measurement of the meter of electric energy, voltage 201 and current 211 transformers are, of course, not needed, but the meter of electric energy can be connected directly to the voltage U1 and to the current I of the object to be metered.

A current metering signal Iil proportional to the current 11 of a single-phase object of consumption to be metered has been divided into three branches a11,a12,a13, each of which branches allXal2al3 has been connected to the inputs 1... 4 of a multiplexer 40 by means of amplifiers of fixed amplifications, preferably amplifiers 3011,3012, 3013 based on an operation amplifier. To the inputs of the amplifiers 3011, 3012, 3013, current metering signals li, have been connected, and each of the amplifiers 3011, 3012,3013 has a fixed amplification of different magnitude. The amplification of the first amplifier 301, can be, for example, 1, the amplification of the second amplifier 3012, for example, 8, and the amplification of the third amplifier 3013 can be, for example, 32. Thus, the amplitude of the current signal Iill obtained from the output of the first amplifier 301l is equal to the amplitude of the current metering signal Iil the amplitude of the current signal Iil2 obtained from the output of the second amplifier 3012 is 8 times the amplitude of the current metering signal lil, and the amplitude of the current signal Iil3 obtained from the output of the third amplifier 3013 is 32 times the amplitude of the current metering signal Iil.

The output 15 of the multiplexer 40 has been connected to the input of the A/D converter 50. The digital parallel interface at the output side of the A/D converter 50 has again been connected to the input interface of the micro-controller 60.

By means of the micro-controller 60, the connecting of the input channels 1... 4 of the multiplexer 40 to the output 15 of the multiplexer 40 is controlled, and, based on the voltage metering signal Uul that has been converted to digital form and on the current signals Iin... Iii3 that have been amplified from the current metering signal Iil, the momentary power is calculated by means of multiplication. The input channels 1... 4 of the multiplexer 40 are connected to the output 15 of the multiplexer 40 alternatingly. This can be accomplished so that, during each cycle, the channels 1... 4 are connected to the output 15 of the multiplexer 40 at invariable time intervals, after which cycle there follows a short pause before the channels 1... 4 are connected during the next cycle, again at the same invariable time inter- vals, to the output of the multiplexer 40. By means of the duration of the pause between the coupling cycles, it is possible to adjust the desired sampling interval.

Thus, in the micro-controller 60, corresponding to each digital sample of the voltage metering signal Uul, there are always three digital samples of the current signal Iill, Iil2nil3 amplified with different amplifications available. The momentary powers are summed in a programmed register, and, when the sum register exceeds a reading of a certain magnitude, a reading of a magnitude equal to said invariable is reduced from the sum register, and a pulse of a certain width is formed in the output of the micro-controller 60. When the sum register has an overflow at a certain power to be metered, said pulse is repeated at certain intervals. In this way, a frequency output proportional to the power is obtained. The pulse frequency f, which is obtained from the micro-controller 60 and which is proportional to the power, is passed to a metering apparatus 70 which records the energy.

Calculation of momentary power from momentary values of voltage and current, summing of momentary powers, formation of a pulse proportional to the power, and calculation of energy are all carried out in a way in itself known to a person skilled in the art, and they are not described herein in more detail.

In the micro-controller 60, the current channel 2... 4 for calculation of momentary power at each particular time is also selected. Corresponding to each digital sample of a voltage metering signal Uul, there are three digital samples of current signals

Iil amplified with different amplifications, of which samples one is selected for calculation of the momentary power concerned. This selecting is carried out from among the digital samples of the amplified current signals IinJinJiis so that the digital sample of the current signal is selected which produces the highest non- saturated result of conversion in the A/D converter.

This comparison can be started from the digital sample of the current signal Iil3 of the current channel 4 of highest amplification. If the digital sample of the current signal Iil3 of the current channel 4 of highest amplification is within the range of operation of the A/D converter 50, the digital sample of the current signal I. n of this current channel 4 is used in calculation of the momentary power concerned. If the digital sample of the current signal Iil3 of the current channel 4 of highest amplification causes saturation of the A/D converter 50, a transfer takes place to the digital sample of the current signal Iil2 of the current channel 3 of next highest amplification. If the digital sample of the current signal I12 of this current channel 3 also causes saturation of the A/D converter 50, the only remaining alternative is the digital sample of the current signal Iill of the current channel 2 of lowest amplifica- tion.

In a situation in which the current of each phase has been divided into three current branches 2... 4, this comparison can also be started from the current channel 3 of middle amplification. From the digital sample of the current signal Iil2 of the current channel 3 of middle amplification it is possible to conclude directly which of the current channels 2... 4 is acceptable. If the digital sample of the current signal Iil2 of the current channel 3 of middle amplification is very close to zero, in calculation of the momentary power concerned the digital sample of the current signal Iil3 of the current channel 4 of highest amplification is used. On the other hand, it the digital sample of the current signal Iil2 of the current channel 3 of middle amplification is very close to the highest (or lowest) possible value, in calculation of the momentary power concerned the digital sample of the current signal Iill of the channel 2 of lowest amplification is used. In this case it is required that the ranges produced by the channels 2... 4 overlap each other to some extent, as is usually the case.

In a situation in which the voltage and current signals are of sinusoidal form, in the vicinity of the zero point of the sinusoidal current signal, the digital sample of the current signal Iil3 of the current channel 4 of highest amplification is used in calculation of momentary power, and in the vicinity of the peak point of the sinusoidal current signal, the digital sample of the current signal Iill of the current channel 2 of lowest amplification is used in calculation of momentary power. If the amplitude of the sinusoidal current signal is low, from the current channel 2 of lowest amplification, no signal present in the range of operation of the A/D con- verter 50 is obtained, so that, in such a case, exclusively the channels 3 and 4 are in use for calculation of momentary power, and finally, when the amplitude of the current signal is very low, exclusively the channel 4 of highest amplification is in use for calculation of momentary power. The selecting of the current channel 2... 4 for use in calculation of momentary power is, however, always carried out in the same way.

The range of measurement of such a static meter of electric energy ranges from about 50 milliamperes up to more than 100 amperes when the current transformer produces, with the highest current, a signal corresponding to the range of input voltage of the A/D converter 50 and when the A/D converter 50 is of the 10-bit type.

The metering method in accordance with the invention is suitable for use in all static single-phase and three-phase meters of electric energy. All the circuits used in the meter consist of commercially available components known to a person skilled in the art. The circuits 40,50 and 60 can be accomplished by means of one component, e. g. the micro-controller H8/3334 manufactured by Hitachi, which comprises an 8- channel multiplexer, a 10-bit A/D converter, an internal program and data memory, and a sufficient number of bit I/O pins.

The circuits 40,50 and 60 can also be accomplished as separate circuits, in which case, as the multiplexer, it is possible to use, for example, the 8-channel multiplexer ADG608 manufactured by Analog Devices. Also the A/D converter can be an

external one, for example the TLC2543 manufactured by Texas Instruments, which apparatus is, in fact, a 12-bit A/D converter which includes an 11-channel multi- plexer. By means of such an A/D converter, it would be possible to accomplish a three-phase static meter of electric energy in which the current channel of each phase is divided into two branches of different amplifications. The amplifications might be carried out, for example, so that the amplification of one branch is 1 and the amplification of the other branch is 8.

Thus, the meter can be constructed out of single circuits or out of circuits in which a number of single circuits have been integrated as a single unit.

Fig. 2 shows a three-phase static meter of electric energy accomplished with three current branches, in which case a multiplexer 40 is needed in which there are three voltage channels 1,5,9 and nine current channels 2... 4; 6... 8; 10... 12, i. e. a total of twelve channels. The situation is fully analogous to the situation illustrated in Fig.

1, with the difference that the input channels in the multiplexer comprise three phase voltages Uul, Uu2, Uu3 and three phase currents Ii1,Ii2,Ii3. Each phase current has been divided into three current branches a11...a13;a21...a23;a31...a33. The input channels 1... 12 of the multiplexer 40 are connected to the output 15 of the multiplexer 40 alternatingly. This can be accomplished so that, during each cycle, the channels 1... 12 are connected to the output 15 of the multiplexer 40 at invariable time intervals, after which cycle there follows a short pause before the channels 1... 12 are connected during the next cycle, again at the same invariable time intervals, to the output of the multiplexer 40. By means of the duration of the pause between the coupling cycles, it is possible to adjust the desired sampling interval.

Thus, in the micro-controller 60, corresponding to each digital sample of the voltage metering signal Uul, Uu2, Uu3, there are always three digital samples of the current signal In i..-Iii3 ; Ii2i-"Ii23'Ii31---Ii33 amplified with different amplifications from the current metering signal Iil, Ii2, Ii3 of the phase concerned available.

In the following, a brief summary of the operation of a static meter of electric energy as shown in Fig. 1 will be presented.

As the first initial setting, the sampling frequency is set in the micro-controller 60, which frequency can be, at the maximum, about 10 kHz (10,000 cycles per second), which means that a sample is taken at intervals of about 100 Its from the signal to be metered. In a situation in which the mains cycle is 20 ms (frequency 50 Hz), it is possible to take 200 samples during one mains cycle. The sampling frequency is understood as the frequency with which a new sample is taken from each signal present in an input channel 1... 4 in the multiplexer.

In the metering proper, first, the voltage U1 and the current I1 of the site of con- sumption to be metered are transformed to a range Uul and Iil suitable for the meter. After this, the digital numbers of voltage and current after the A/D conver- sion 50 are processed in the micro-controller 60 so that the offsets produced by the amplifiers 3011, 3012, 3013 and by the transformers 201,21l are eliminated from them. Next, the current signal Ii11,Ji12,Ii13 to be used for calculation of momentary power at each particular time is selected, as described above. After the suitable current signal IillsIil2stil3 has been selected, the filtered current sample of said current signal IillIil2Iil3 after the A/D converter is multiplied by a corresponding filtered voltage sample after the A/D converter. After this, the reading that was obtained as a result of the multiplication is divided by the amplification of the corresponding current signal IillIil2Iil3 in order that the scale of the power reading concerned should be the same as the power readings calculated in a corresponding way on the basis of the other current signals IiiiJii2'il3-After this the power readings are summed in a programmed way in a register, in connection with whose overflow a voltage pulse is produced by means of the I/O functions of the micro- controller 60. When the register overflows, a certain number is reduced from it, and the summing is continued again until the next overflow and reduction take place.

Thus, the result is a frequency f proportional to the power in the I/O pin of the micro-controller 60, which frequency can be passed to the apparatus 70 that meters the energy. By means of a frequency output, it is also possible, for example, to pulsate a LED, which indicates the energy measured by the meter.

Fig. 3 illustrates a sinusoidal voltage u and a sinusoidal current i, which are at the same phase and which current has been amplified with three different amplifications il, i2, i3. It is seen from the figure that the results of conversion of the A/D converter 50 are not exactly from the same point of time, but slightly aside from one another. In Fig. 3, a situation is shown in which the voltage u of a single-phase object is transformed first, after which the currents il, i2, i3 amplified with different amplifica- tions are transformed. The momentary value ut0 of the voltage u converted at the moment to corresponds to the momentary values iltl, i2t2, i3t3 of the currents il, i2, i3 converted at the moments tl, t2 and t3. The phase difference of the momentary value i3t3 of the current i3 of the current channel of highest amplification as compared with the momentary value ut0 of the voltage of the voltage channel is, thus, higher than the corresponding phase difference of the momentary value i2t2 of the current i2 of the current channel of lowest amplification. This error can be compensated, for example, by interpolation of a voltage sample which corresponds to the current channel used in the calculation. In such a procedure, the voltage sample correspon- ding to the momentary value i3t3 of the current i3 of the current channel of highest amplification is obtained by interpolation of the voltage sample ut0 forwards by the time t3-to between the voltage and the momentary value i3t3 of the current i3 con- cerned.

Above, just a solution of principle of the method in accordance with the present invention has been described, and it is obvious to a person skilled in the art that numerous modifications can be made to said solution within the scope of the inventive idea disclosed in the accompanying patent claims.