This algorithm, typified in the ADE7753, is accurate enough but suffers from the disadvantage that a residual error makes the calibration of the meter more difficult, increasing one of the hidden costs of manufacturing. The operation of the algorithm is illustrated in figure 1.
The input signals from the sensors are the instantaneous voltage
where V is the peak voltage and is related to the line frequency (see) and the instantaneous current
where I is the peak current, is related to the line frequency and is the phase difference between voltage and current (see).
After a pre-conditioning (gain adjustment and DC component removal) stage, these signals are fed into the inputs of two dedicated A/D converters and converted in digital form. Both signals are then sent to a multiplier block, which multiplies both digital voltage and current signals:
yielding p(t), which is the instantaneous power signal (see).
The resulting instantaneous power is calculated by the sum of two components. The first,
is the active power (DC part), while the second component
is the AC part of the instantaneous power (see).
In order to extract the active power, a low pass filter (LPF) removes the AC part, leaving only the active power available for the energy measurement. However, since the low pass filter is not an ideal ‘brick wall’ filter, a part of the AC component (sinusoidal ripple) of about -22dB is still present after the signal processing. Therefore, to perform the meter calibration accurately, the power read must be averaged over the time or done synchronously with the line period.
ST’s new STPM01 power metering chip eliminates this problem by implementing a new and innovative algorithm developed and patented by Iskraemeco, a major energy meter manufacturer, using a proprietary ST technology called BCD6 (Bipolar-CMOS-DMOS). This technology, originally developed for demanding applications in the automotive and computer peripheral fields, allows sophisticated digital signal processing blocks, precision analog circuits and multiple power transistors to be integrated on the same robust, low-cost chip (see figure 2).
The new algorithmic approach implemented inside the STPM01, which allows the device to reach new levels of performances in terms of accuracy, is described below. In the STPM01, after the pre-conditioning and the A/D conversion, the digital voltage signal (which is dynamically more stable compared to the current signal) is processed by a differentiator stage which produces the following transformation:
The resulting signal, together with the pre-processed and digitized current signal
is then available for the calculation process. These digital signals are also fed to two additional stages that perform the integration of the signals, giving:
Now four signals are available. Combining (pairing) them by means of two multiplying stages, two results are obtained:
After these two operations, another stage performs the subtraction between the results p2 and p1 and a division by 2, yielding the active power:
and removing the AC part
from the instantaneous power.
An important feature of the algorithm and its implementation in the STPM01 is that it is equally suitable for use with current sensors such as Rogowski coils whose output is not the instantaneous current i(t) but its rate of change, . In this case, the initial voltage signal differentiator stage is switched off and the signals coming from the A/D conversions and their consequent integrations will be:
The signals process flow will be the same as shown in the previous case, but with the formulas above, the result will be the same.
The absence of any AC component allows a very fast calibration procedure: all that is required is to set (using the internal device programming registers) the voltage and current sensor conversion constants, using the effective voltage and current (Vrms, Irms) readings that are provided by the STPM01’s built-in communication port. This avoids the need to perform time-averaged readings of the active power or for line synchronization.
The STPM01, which ST is fully supporting with PCB reference designs, evaluation kits and free software, illustrates how advanced technologies such as BCD6 are transforming the metering market by allowing advanced Digital Signal Processing algorithms to be transparently embedded in robust devices aimed at the cost-sensitive metering market.