DESCRIPTION
Title:
Electronic watt-hour meter.
Technical field:
Measurements of active and reactive electricity consumption, suitable for use in single and three-phase supply systems.
Background art:
Most of the electronic watt-hour meters, available at present, use different methods to multiply two signals proportional to the instantaneous values of the supply voltage and the load current to obtain the instantaneous power. The time integration of this power is proportional to the consumed energy. The main differences between these methods are in the multiplication and integration stages. The different methods of multiplication are:
1- Variable conductance multiplication:
The logarithms of the voltage and current signals are obtained and added ; the antilogarithm of the result is proportional to the instantaneous power.
2- Hall effect multiplication:
A semiconductor, carrying the current signal, is placed in a transverse magnetic field proportional to the supply voltage. A voltage, perpendicular to the current and the magnetic field is induced. This voltage is proportional to the instantaneous power.
3- Time division multiplication:
This is a sampling technique that produces a pulse train. Each pulse has its width proportional to one signal (voltage or current) and its height proportional to the other signal. The pulse area is therefore proportional to the power at the sampling point.
4- Digital multiplication:
The voltage and current signals are sample and converted into digital form. A microprocessor is used to perform the multiplication and calculate the power at the sampling point.
The integration method used depends on the output of the integration stage. Analogue integration is used when the multiplier has an analogue signal as its output, and digital integration is used when the multiplication is performed by a microprocessor. In this case , the microprocessor calculates the energy from knowing the power at each sample and the time between samples.
Disclosure of invention:
The invention follows a different approach and does not use any of the multiplication methods described above. The basic principle of the technique used will now be explained:
The instantaneous value of the supply voltage is given by the formula:
(1)
where:
'V' is the rms value of the supply voltage, and 'w' is the angular frequency.
The instantaneous load current can also be represented by the formula:
(2)
where:
'I' is the rms value of the load current, and
'θ' is the phase angle between the load current and the supply voltage.
If a signal proportional to the load current is integrated for a half-cycle, the integrator output at the end of the integration half-cycle is given by the formula:
(3)
where:
'Ko' is the ratio between the current signal and the load current,
'K1' is the integrator time constant,
' α ' is the integration starting angle which is the angle between the integration starting point and the first zero-crossing point of the supply voltage to the positive values.
Formula (3) shows that when the integration is performed during the positive half-cycle of the supply voltage (α = 0) or during the negative half-cycle of the supply voltage ( α = π), then 'A' will be proportional to 'I cos θ / K1 w'.
Formula (3) further shows that if the integration is performed when the supply voltage slope is positive
( α = - π / 2 ) or when the supply voltage slope is negative ( α = + π / 2 ), then 'A' will be proportional to 'I sin θ / K1 w'.
Since 'A' has been obtained by integrating the load current for a half-cycle, any even harmonics present in the load current will have no effect on the value of 'A'. However, the effect of any odd harmonics component of an order 'm' depends on its phase relation with the main component. This will vary between zero and a maximum value equal to the effect of an equal main component current divided by 'm'. Therefore, these odd harmonics and any D.C. component in the load current should be attenuated to the required level.
The following methods can be used for watt-hour and var-hour measurements:
1- A signal proportional to the load current is integrated as described above. Every integration halfcycle, the integrator output changes by the value 'A' as given by formula (3). After the end of '6' integration half-cycles, the current signal is switched off and a reference D.C. voltage 'Vo' is integrated. The polarity of this D.C. voltage is chosen so that the integrator output moves linearly towards zero voltage. The counting time 'Tc' which is the time needed for the integrator output to reach zero is:
The value of 'Vo' is chosen so that the integrator output should reach zero within a fixed number of halfcycles 'F' during which the load current is not integrated. Therefore a measurement will be obtained every 'G+F' half-cycles. The time equivalent of these
'G+F' half-cycles will be called the measuring time 'Tm' and is given by:
Tm = (G+F) π / w (5)
The pulses produced by a pulse generating means (oscillating at frequency 'fc' ) are counted during the counting time. If 'fc' is made proportional to the supply voltage ( fc = K2 V), then, the number of these pulses is: n = fc Tc
From (5) and (6):
n = K Tm V I cos (θ - α ) (7) where:
Formula (7) shows that 'n' is proportional to the active or reactive energy consumed by the load during the measuring time. If the counts 'n' are accumulated for a period of time, the total number 'N' will be proportional to the watt-hour or var-hour (depending on the integration half-cycle starting point as described earlier) consumed by the load during that period. As can be seen from (7), 'N' is independent of the supply frequency and the integrator capacitor.
2- The current signal is integrated as described before, and a positive or negative reference voltage is integrated, whenever needed, in order to keep the integrator output at zero. The values of these reference voltages are chosen to produce equal slopes 'Vo / K1' with different direction at the integrator output. The pulses produced by a voltage to frequency converter (oscillating at frequency proportional to the supply voltage 'fc = k2 V') are used to increment a counting means when the positive reference voltage is integrated and to decrement the counting means when the negative reference voltage is integrated (or vice versa). After 'G' integration half-cycles, the contents of the counting means will be changed by the value 'n' represented by formula (6) above. The time equivalent to 'G' integration half-cycles is:
T = G π / w (8) From (6) and (8):
n = N = B T V I cos ( θ - α ) ( 9 ) where :
'B' is a constant equal to
Formula (9) shows that 'N' is proportional to the active or reactive energy consumed by the load during any length of time 'T'.
In practice, the integrator output is kept within two limits. Only when the integrator output moves outside the allowed range, the suitable reference voltage is integrated to bring it back inside the limits. Also, the switching of these reference voltages is synchronized with the pulses 'fc'. The total error introduced by this arrangement is less than the energy
needed to move the integrator output from one limit to the other. This error becomes negligible within a couple of seconds after the measurement starts. On the other hand, the synchronization of the reference voltages integration with 'fc ' , allows their integration time to be equivalent to an integer number of pulses. This eliminates any error due to a missing fraction of a pulse during counting and enables the use of lower values of 'fc' without impairing the measurement accuracy.
Another improvement is to invert the current signal and integrate the non-inverted current signal during the integration half-cycles and the inverted current signal during the other half-cycles. This compensates any D.C. component, offset voltage and temperature drifts in the current input and filter circuits.
Brief description of drawings
Fig. (1) is the expected wave-form shown for the case of the first method with an inductive load and 'α = 0', 'G = 1', and 'F = 1'. In this figure:
(a) is the supply voltage wave-form,
(b) is the load current wave-form,
(c) is the input of the integrator, and
(d) is the output of the integrator. The counting time 'Tc' and 'A' are shown.
Fig. (2) is the expected wave-form in the case of the second method with an inductive load and without current inversion. In this figure:
(a) is the supply voltage wave-form,
(b) is the load current wave-form,
(c) is the integrator input,
(d) is the integrator output, and
(e) is the pulses 'fc'.
Fig. (3) is a block diagram of the watt-hour and var-hour meters. In this figure:
(1) is the Voltage Sensing circuit. It provides the supply voltage signal which is a suitable low voltage A.C. signal proportional to the supply voltage,
(2) is the Half-cycle Finder circuit which determines when the load current signal is to be integrated. In the case of the watt-hour meter, it tests whether the supply voltage is positive or negative; and in the case of the var-hour meter, it tests whether the supply voltage slope is negative or positive. This circuit receives its input from the Voltage Sensing circuit and the information from its output are sent to the Control circuit,
(3) is the Voltage-Controlled Oscillator (V.C.O.) circuit. It generates the measuring pulses which are counted during the counting time as determined by the Control circuit. The frequency of these pulses 'fc' is proportional to the supply voltage. In the case of the second method, the Control circuit synchronizes the switching of the reference voltages with these pulses, (4) is the Current Sensing circuit. It provides the load current signal which is a low voltage A.C. signal proportional to the load current. In the case of the second method the inverse of this signal may also be required. The phase angle between this signal (or its inverse) and the load voltage signal should be equal to the phase shift between the main component of the load current and the main component of the supply voltage,
(5) is the Reference voltages circuit. It provides the reference voltages that are needed to force the
integration circuit output to move in the direction selected by the control circuit,
(6) is the switching circuit. It receives the signals from the Current Sensing and the Reference Voltages circuits, and connect the suitable signals (as required by the control circuit) to the Integration circuit input.
(7) is the Integration circuit which integrates the outputs of the switching circuit,
(8) is the Level Sensing circuit. It informs the
Control circuit with the condition of the integrator output. It senses when the integrator output reaches zero in the case of the first method, and when the integrator output is outside a predetermined range in the case of the second method,
(9) is the Counter and Display which counts the measuring pulses received from the Control Means, divides the count by a suitable scale, and displays the meter readings,
(10) is the Control circuit. It receives the outputs of the Level Sensing, the V.C.O., and the Halfcycles Finder circuits; and controls the following:
a) the function of the switching circuit,
b) the switching of the counting pulses to the Counter and Display,
c) the selection of the counting direction (up/down)
Best mode for carrying out the invention
The best mode for carrying out the invention is to use the second method described above. The supply voltage signal is obtained using a suitable voltage divider and a band-pass filter tuned to the main component frequency. This filter increases the accuracy of detecting the zero-crossing points of the supply voltage.
The supply voltage signal is rectified and filtered, the resulting D.C. voltage is used to drive a voltage to frequency converter whose output will oscillate at a frequency proportional to the supply voltage. The converter pulses from this circuit are used
by the control circuit to initiate and stop the integration of the reference voltages, this permit the use of low oscillation frequencies, and one of the readily available voltage to frequency converter can be used. The Control circuit passes these pulses together with the counting direction signal to the counter during the integration of the reference voltages.
The supply voltage signal is also used to determine the integration half-cycles. In the case of the watthour meter, a zero-crossing detector is used to define the positive and negative half-cycles. In the case of var-hour meter, a 90 degree phase shift is introduced by differentiating the supply voltage signal then the zerocrossing points are found to determine when the supply voltage slope is negative or positive. The differentiator output will be clean enough since its input signal is filtered by the band pass filter.
A shunt is used as a current sensor, the load current signal is obtained by amplifying and filtering the voltage drop across this shunt. The filter used is a band pass filter similar to the one used with the supply voltage signal, this arrangement will compensate any phase shift introduced by the first filter. The inverse of the load current signal is also obtained. The load current signal is integrated during the integration half-cycles, and its inverse is integrated during the other half-cycles. The integrator output is monitored and when it drifts outside a first range, a constant reference voltage is applied to the integrator input at the first rising (or falling) edge of a pulse from the voltage to frequency converter. The polarity of this reference voltage is selected to bring the integrator output back to within a second range smaller than the first. After the integrator output enter the second range, the reference voltage is switched off at the first rising (or falling) edge of a pulse at the output of the voltage to frequency converter. During the reference voltage integration, the counter is incremented or decremented (according to the sign of the
reference voltage) at every falling (or rising) edge of the voltage to frequency converter output.
The counter content is displayed after a suitable scaling
Industrial applicability
Traditionally, induction type watt-hour meters are used for electricity consumption measurements. However, economics and more flexible tariff requirements increases the demand for electronic meters.
the simplicity of the method allow the production of low cost electronic single phase electricity meters suitable for use for residential electricity consumption measurements.
Three-phase measurement is possible by using tow or three units that share the same reference voltages, counter and display. In this case, proper arrangements are made to prevent integrating a reference voltage in more than one circuit at the same time.
Electronic meters are suitable for use as a basis of flexible load management system and for remote meters reading.
When a high number of meters are needed in one area, space and cost can be saved by sharing the voltage sensing, the half-cycle finder, and the voltage to frequency converter circuits between all the meters connected to the same phase, in addition to sharing the reference voltages circuit between all meters.