AN4470 Application note The STPM3x application calibration Introduction The STPM3x is an ASSP family designed for high accuracy measurement of power and energy in power line systems using the Rogowski coil, current transformer or shunt current sensors. The STPM3x devices embed a full set of calibration and compensation parameters which allow the meter to fit tight accuracy standards (EN 50470x, IEC 62053-2x, ANSI12.2x for AC watt meters) using low cost components, after a fast calibration procedure explained in this document. According to energy meter measurements, the customer has to pay for energy consumption. The correct operation of the meter, as well as its accuracy and reliability are very important features both for the customer and the electricity company. That’s why the quality control of meter is so important and strict. Special care has to be given both to the design stage and the calibration procedure. The former allows the right dimensioning of analog front-end components so to fit the current dynamics and the meter constant pulse. The latter impacts on many meter key ratings directly. October 2015 DocID026176 Rev 2 1/19 www.st.com Contents AN4470 Contents 1 2 Calibration principles and underlying theory ................................ 3 1.1 Principles of digital energy measurement system ............................. 3 1.2 Accuracy and stability influence factors ............................................ 5 Measuring system design ............................................................... 6 2.1 3 Design example ................................................................................ 7 System calibration........................................................................... 9 3.1 Amplitude calibration ....................................................................... 10 3.1.1 3.2 3.3 Phase-shift calibration ..................................................................... 13 3.2.1 Step-by-step phase-shift calibration procedure ................................ 15 3.2.2 Example:phase-shift compensation ................................................. 16 Offset calibration ............................................................................. 16 3.3.1 4 2/19 Step-by-step amplitude calibration procedure .................................. 12 Step-by-step offset calibration procedure......................................... 17 Revision history ............................................................................ 18 DocID026176 Rev 2 AN4470 Calibration principles and underlying theory 1 Calibration principles and underlying theory 1.1 Principles of digital energy measurement system Digital energy measuring system, based on the STPM3x, is composed of: Analog section with high-resolution sigma-delta analog/digital converters (ADCs) Digital section with powerful digital signal processor (DSP) to perform power and energy measurement, as well as other secondary parameters The main scheme of this system is indicated in the below figure. Figure 1: Digital power and energy measurement system ANALOG SECTION v(t) V(φ U ) Voltage transducer AFE DIGITAL SECTION V’(φ U ) + V off ΣΔ modulator Decimation CHV voff (t) θ i(t) Current transducer I(φ I ) Compensation AFE VREF CLK Voltage reference Time base ΣΔ modulator I’(φ I’) PHV Output data DSP CHC + I off Decimation PHC Compensation OFAF OFS OFA OFR ioff (t) STPM3x GIPG1706151231LM Voltage and current paths include the following blocks: Sensors for voltage and current Signal conditioning (to optimize signals to match the required ADC input level) ADCs Common section consists of the following elements: System DC reference voltage System time base, provided by a quartz crystal oscillator or by an external (MCU) clock A/D converters collect samples of phase current and phase-to-neutral voltage synchronized to the sample clock. Outputs of the analog section are samples of voltage and current in digital form with an exact time relationship. The digital section consists of DSP providing real time calculation based on the voltage and current sampled values to calculate power, energy, RMS values and other parameters through standard mathematical formulas. A correction algorithm hardwired in DSP corrects amplitude and phase-angle errors of the measured samples, while correction parameters are calculated during the calibration phase. From the same set of corrected samples, power, energy and all other parameters are calculated in real time through standard mathematical formulas. Calculated values are stored in 32-bit registers, from which output pulses are generated with frequency proportional to the measured power. DocID026176 Rev 2 3/19 Calibration principles and underlying theory Basic definitions and formulas are given below: Active power Equation 1: P = V · I · cosφ Apparent power Equation 2: S = V·I Reactive power Equation 3: 2 2 S – P = V·I· sinφ Q = Power factor Equation 4: P PF = cos φ = ---S where: V, I = effective values of voltage and current φ = φv - φI current-to-voltage phase-angles φv φI voltage and current to common reference phase-angles Measured active power Equation 5: P'= V'· I' · cosφ' +Pof f where: V' = V (1 + εv) I' = I (1 + εi) φ' = φ+ θ εv = voltage amplitude error εi = current amplitude error θ = current-to-voltage phase-angle error Poff = power offset (due to Voff, Ioff residual signals) Neglecting term εv*εi, the measured active power is: 4/19 DocID026176 Rev 2 AN4470 AN4470 Calibration principles and underlying theory Equation 6: P'=V· I · (1+ 1.2 v+ i ) cos (φ+ɵ)+ Pof f Accuracy and stability influence factors All components, which have some influence on system accuracy and stability, can be found in the input analog section. Only a limited number of internal components determines system accuracy: Voltage and current sensors Signal conditioning section Oscillator frequency Internal reference voltage source Analog-to-digital converter gain To reach the desired stability and linearity, high quality components have to be used. Moreover, the circuit has to be carefully designed to minimize some issues such as: shorttime repeatability, linearity or immunity degrade. Besides, external influences can affect meter accuracy, such as: Capacitive and inductive coupling to inputs and between phases (crosstalk) High frequency electrical and magnetic fields (EMC) Common-mode voltage between inputs and to earth Low frequency magnetic fields Measuring setup (wiring, earth connection ground loop) Source (stability of V, I, φ, signal quality) Long-time drift Humidity Undesired external influences should be reduced to minimum through the shielding of the analog part or compensated in hardware or software. If system is not immune to external influences, it can only work under very special conditions and results cannot be reproduced in other locations, where there may be a different measuring setup. In this case, also statistical effects, due to noise, have higher impact on short-time repeatability. External influences on total system accuracy can be more important than the basic specified error. The STPM3x does not introduce any crosstalk error neither between voltage and current inputs nor among different phases. However, the voltage front-end handles considerable amplitude voltages, which make it a potential source of noise. Disturbances could be emitted into current measurement circuitry, interfering with the signal to be measured. Typically, this shows a non-linear error at small signal amplitudes and non-unity power factors. At unity power factor, voltage and current signals are in phase, and crosstalk between voltage and current channels appears as a gain error, which can be calibrated. When voltage and current are not in phase, crosstalk has a non-linear effect on measurements, which cannot be calibrated. Crosstalk is minimized by a well-planned PCB and the correct use of filter components. DocID026176 Rev 2 5/19 Measuring system design 2 AN4470 Measuring system design The maximum voltage and current measurement, the number of pulses per kWh (indicated as CP, constant pulses) and the measurement accuracy are the main ratings of the meter. A correct analog front-end component choice allows the line signal to fit the device input dynamics; selectable gain of internal current amplifier scales the input signal according to sensor sensitivity. A typical application example is shown in Figure 2: "Application example". Figure 2: Application example L N R1 Vip LED CP pulses Vin R2 STPM3x Iip CT CP pulses SPI/UART Iin MCU Converted data LOAD 230 VRMS 80 A RMS Input ± 300 mV ± 35 mV Sensors ± 300 mV ± 300 mV 230 VRMS 80 A RMS AFE gain Output GIPG0304141257LM The choice of external components in the transduction section of the application is a crucial point of the application design, affecting the precision and the resolution of the whole system. A compromise has to be found among the following needs: 1. Maximizing signal-to-noise ratio in the voltage and current channel 2. Choosing kS current-to-voltage conversion ratio and the voltage divider ratio, to achieve calibration for a given CP 3. Choosing kS to take advantage of the whole current dynamic range according to the desired maximum current and resolution Rules for a good application design are described in this section. After the design phase, any tolerance of the real components respect to these values or the device internal parameter drift can be compensated by calibration. This stage is necessary to get the desired CP after calibration. To reach CP target output constant pulse, the analog front-end component dimensioning can be carried out in two ways: 6/19 Choosing the value of R1 voltage divider resistor, given R2 and kS current sensor sensitivity DocID026176 Rev 2 AN4470 Measuring system design Choosing kS given R1 and R2 voltage divider resistors Calculations for these two methods are developed below: First method: constant kS Given R2 (smaller voltage divider resistor), kS (current sensor sensitivity) and CP, target meter constant pulse (pulses/kWh), as calculation inputs, R1 voltage divider resistor value derives from the following formula: Equation 7: Second method: constant R1 Given R1, R2 (voltage divider resistors) and CP target meter constant pulse (pulses/kWh) as calculation inputs, kS current sensor value derives from the following formula: Equation 8: CP value can be scaled by a division factor through LPWx[3:0] bits in DSP_CR1, DSP_CR2 for the two channels according to the device p/n. The resistor (in the first method) or the current channel sensor (in the second method) has to be chosen as closer as possible to the target value; small tolerance is compensated by calibration. 2.1 Design example This example shows the correct dimensioning of a meter using a current transformer with the following specifications: Table 1: Example1 design data Parameter Value VN nominal voltage 230 VRMS IN nominal current 5 ARMS IMAX maximum current 40 ARMS CP constant pulses 1000 imp/kWh The values of voltage divider resistors are 770 kΩ and 470 Ω. Setting C P = 64000 pulses/kWh (at LED_PWM = 1, the device default value) and according to the previous calculation, the following values are obtained: DocID026176 Rev 2 7/19 Measuring system design AN4470 Table 2: Example1 calculated data Parameter Value Current sensor sensitivity VMAX IMAX 1 Vre f R1 + R2 VM A X = --- ------------------ -------------------- = 347.8 V 2 A R2 2 V · · · V ref 1 1 I MAX = --- · ------------------- · ------------------- = 60.5 A 2 A · 2 kS I To set the desired LED pulse output, a division factor can be set through LPWx[3:0] bits in DSP_CR1 and DSP_CR2 configuration registers. Any tolerance, producing C P small variation respect to 1000 imp/kWh, is compensated by calibration. 8/19 DocID026176 Rev 2 AN4470 3 System calibration System calibration The calibration procedure is a key feature among main meter requirements. In fact, it impacts directly on accuracy, cost, manufacturing and reliability of the meter. After the final assembly phase, an energy meter requires a calibration procedure due to unknown tolerances respect to nominal values of the following analog blocks: Voltage and current sensors Oscillator frequency Internal or external reference voltage source Analog-to-digital converter gain The STPM3x device is composed of independent channels for line voltage and current respectively. Each channel includes its own 12-bit digital calibrator to adjust the signal amplitude, digital filter to remove any signal DC component; moreover the device embeds phase calibration registers for each line and power offset compensation registers. Calibration is carried out in three steps: Amplitude calibration is mandatory for class accuracy higher than Class 2 Phase-shift calibration is mandatory for CT-based meters Power offset calibration (optional for class accuracy higher than 0.2) To calibrate, the following equipment has to be interfaced: Precision current and voltage source (Gen) Higher class precision energy meter (HPM) (optional) Meter under calibration (MUC) Calibration process controller (CPC) UART/SPI interface to the STPM3x device Please see Figure 3: "Meter calibration setup". Figure 3: Meter calibration setup SPI / UART Interface pulses error MUC IN HPM* Calibration process controller VN M1* VRMS VRMS M2* I RMS (*) I RMS optional equipment optional signal GIPG0304141302LM DocID026176 Rev 2 9/19 System calibration AN4470 Gen equipment generates voltage and current line signals at the same frequency and a phase-shift between them. HPM and MUC equipment measures the same signals, and HPM computes the error by comparing LED frequency output. If HPM is not available, amplitude calibration can be performed having either a precise voltage/current generator or a voltage/current RMS meter. Calibration process controller is an automated system which runs calibration process routines to configure the STPM3x device on MUC before calibration, controls Gen, monitors HPM equipment, reads from the device, calculates the correction parameters and writes them into the device. Since the STPM3x hasn’t any non-volatile memory, CPC should take into account the permanent storage of calculated calibrators. CPC can be interfaced to the STPM3x through its SPI/UART peripherals. If an STPM3x evaluation board is used, the following interfaces are available: The STPM3x parallel programmer The STEVAL-IPE023V1 USB isolated interface RS232 interface (as the one embedded in the STPM3x evaluation board) The STPM3x evaluation software, running automatic calibration procedure, can be found on www.st.com; it can be used with all above listed interfaces. Further information is available in the UM1719. For all available tools and software please visit www.st.com. 3.1 Amplitude calibration Any energy measure performed by the device (active wideband and active fundamental, reactive or apparent power and energy) is calculated digitally (without error) from current and voltage signals. This means that every measure is automatically calibrated if current and voltage channels are calibrated. CP (power sensitivity constant pulse) target value is achieved by amplitude calibration of these signals. Independent and precise line signal generators could be used for this calibration, because line frequency and phase between line signals have not a significant impact, observing RMS values. If the line generator is precise and stable enough, theoretically, the additional precision energy meter (HPM) is not necessary to perform the calibration; in fact signal amplitudes (voltage and current RMS value) are calibrated and DC offset is rejected, thanks to the almost ideal linearity of the STPM3x. This may simplify the generation of reference line signals of accurate output values. If accuracy is not guaranteed, reference values of line signals can be obtained by RMS meters. Meter calibration is achieved by calibrating the device, just one measuring point, at nominal values, such as: 230 VRMS, 5 ARMS, 50 Hz. Calibrating voltage and current in a single operating point leads to a very short (one second in an automated environment) calibration time. Each voltage and current channel of the device (according to the p/n) have to be compensated following the same procedure. Given the device internal parameters in Table 3, and having one between R 1 or kS calculated as stated in Equation 7 and Equation 8, voltage and current RMS register target values, XV and XI respectively, are calculated by DSP as follows: Voltage register value at VN nominal voltage 10/19 DocID026176 Rev 2 AN4470 System calibration Equation 9: Current register value at IN nominal voltage Equation 10: 17 IN · A I · cal I · kS· 2 X I = --------------------------------------------------------------V ref Table 3: STPM3x internal parameters Parameter Value Voltage reference Vref = 1.20 [V] Decimation clock DCLK = 7812.5 [Hz] kint = 1 (if ROC bit = 0 in DSP_CR1,2) Integrator gain (for Rogowski coil only) kint = 0.8155773 (if ROC bit = 1 in DSP_CR1,2) Voltage channel gain AV = 2 Current channel gain AI= 2/16 AV voltage ADC gain is constant, while AI current ADC gain is chosen according to the sensor used and to the desired current input dynamics. The calibration procedure has as final result k V and kI correction parameters which, applied to the STPM3x voltage and current, introduce signal path attenuation or amplification compensating small tolerances of analog components. kV and kI calibration parameters are the decimal representation of the corresponding voltage and current 12-bit calibrators: CHVx[11:0], CHCx[11:0] (where x = 1 or 2 respectively for primary and secondary channel according to the device p/n) from DSP_CR5 to DSP_CR8 registers. Through hardwired formulas, kV and kI fine-tune measured values from 0,75 to 1 in 4096 steps, according to CHV and CHC values. For example: CHV = 0 generates a correction factor -12.5% (kV = 0.75) and CHV = 4065 determines a correction factor +12.5% (k V = 1) following below equations: Voltage correction factor Equation 11: CHV k V = 0.125 · ------------ + 0.75 2048 Current correction factor Equation 12: CHC k I = 0.125 · ------------ + 0.75 2048 When system is connected and powered on, having the applied VN and IN nominal values, a certain number of readings has to be performed to average voltage and current RMS DocID026176 Rev 2 11/19 System calibration AN4470 values. After RMS register samples have been read and averaged, obtaining V AV and IAV values, voltage and current channel calibrators are calculated as follows: Voltage calibrator Equation 13: XV CHV =14336 ------------ 12288 VAV · Current calibrator Equation 14: XI CHC =14336 ------------ 12288 IAV · where XV and XI are those calculated in Equation 9 and Equation 10. kI and kV correction parameters can fine-tune measured values only within the calibration range of ±12.5% of voltage or current channel. If after the calibration, CHV or CHC calculated values are out of range (less than 0 or more than 4095), the application cannot reach the target value of CP power sensitivity. In this case, design and calibration phase should be repeated choosing a smaller C P value. If one or more calibrator values are out of range, energy meter board could be not able to perform these measurements, maybe because component tolerance is too big, or due to some issues during the layout phase, so the application has to be redesigned. Otherwise, calibrator values can be written into the STPM3x, the average RMS readings are very close to XI and XV target values and LED output frequency is very close to HPM frequency output. Since the RMS resolution is a bit lower than the energy resolution, it is still possible, after CHC and CHV calculation, to fine tune these calibrators in order to reduce the error on active power shown from the HPM. 3.1.1 Step-by-step amplitude calibration procedure The following steps summarize the calibration procedure explained above 1. Design the application as stated in Section 3: "System calibration" so that the relationship among R1, R2, kS and CP is coherent with equation 7 and equation 8 2. Reset the STPM3x to have registers in the default state 3. Configure the device through CPC according to the chosen application. The following registers have to be configured (one or both primary and secondary channels, according to the application and to the device p/n): ROCx (in DSP_CR1 or DSP_CR2) 0: for CT or shunt 1: for Rogowski coil GAINx (in DFE_CR1 or DFE_CR2) to set the correct current gain channel CHVx and CHCx (in DSP_CR5 and DSP_CR8) have to be set to default (0x800) obtaining a calibration range of ±12.5% of voltage or current channel 4. Apply stable and accurate nominal values of VN and IN voltage and current signals, with PF =1 to one or both primary and secondary channels. For the stability of the source please refer to the equipment documentation; add 0.5 seconds to the maximum so that the STPM3x RMS values are stable 12/19 DocID026176 Rev 2 AN4470 System calibration 5. Perform RMS register sample acquisition (DSP_REG14 and/or DSP_REG15) through CPC; average the values to obtain VAV and IAV; minimum suggested values are 20 samples in 5 line cycles 6. Calculate CHVx and CHCx calibrators using equation 13 and equation 14 7. Write calibration values to the device and store them in a non-volatile memory The whole procedure requires one second in an automated environment. 3.2 Phase-shift calibration The STPM3x does not introduce any phase-shift between voltage and current channels. However, voltage and current signals come from transducers, which could have inherent phase errors. For example, a phase error from 0.1 ° to 0.3 ° is common for a current transformer (CT). These phase errors can vary from part-to-part, and have to be corrected in order to perform accurate power calculations. Errors associated with phase mismatch are particularly evident at low power factors. The phase compensation block provides a digital correction of the phase-shift for primary and secondary channel independently. This block introduces a delay between current and voltage samples which is fine-tuned by PHCx[9:0] and PHVx[1:0] phase calibration bits in DSP_CR4. The delay (in degree) introduced by these registers on the waveforms is given below: Table 4: Phase-delay Parameter Value Current shift fline ° φC = --------------- · PHCx [ 9:0] ·360 SCLK Voltage shift φ V fline = ---------------- · PHVx [ 1:0 ] ·29 ·360 ° SCLK fline ° 9 φ = --------------- ·( PHCx [ 9:0 ] - PHVx [1:0 ] · 2 )·360 SCLK Global phase shift Where SCLK = 4 MHz and fline is voltage and current signal frequency. As shown in figure 4, a capacitive behavior is determined by the current leading the voltage waveform to a certain angle. In this case, there is the compensation by delaying the current waveform by the same angle through PHCx register. For a 50 Hz line the current channel waveform maximum delayed is: φC ≤ 4.6035° with step ΔφC =0.0045° An inductive behavior has the opposite effect, so that current lags the voltage waveform. In this case, PHV register delays the voltage waveform by the minimum angle to invert the behavior to capacitive and then acting on PHCx register to fine tune the current waveform. PHV impacts on the calculation of power and energies related to both current channels. For a 50 Hz line, the voltage channel waveform maximum delayed is: φV≤ 6.912° with step ΔφV = 2.304° DocID026176 Rev 2 13/19 System calibration AN4470 Figure 4: Phase-shift error Voltage Current capacitive Current indu ctive GIPG160620150948LM From the equation indicated in global phase shift parameter in table "phase-delay", the following correction range is calculated for 50 and 60 Hz line signals: Table 5: Phase error correction range Line frequency Minimum value Maximum value Step 50 Hz -6.120 ° 4.6035 ° 0.0045 ° 60 Hz -8.2944 ° 5.5242 ° 0.0054 ° To compensate phase-shift, stable nominal values of voltage and current signals (VN and IN) shifted by φ = 60 ° angle, have to be applied to MUC. Given e, the error on active power (averaged over a certain number of samples through HPM), the phase-shift angle (θ) between voltage and current can be measured as shown below. Without any phase-shift error, the ideal active power at φ = 60° is: Equation 15: PI = V · I · cos (60) Since voltage and current are shifted by angle θ, the measured power is: Equation 16: PM = V · I · cos (60+ θ) Leading to an error, at PF = 0.5, equal to: Equation 17: PM – PI VIcos 60+θ)-VIcos (60) e = ------------------ = ----------------------------------------------------------------- = 2 cos(60+θ)-1 PI VIcos (60) By measuring the error, the phase-shift derives from the above formula as follows: Equation 18: 14/19 DocID026176 Rev 2 AN4470 System calibration To compensate this error, PHCx and PHVx bits have to be set as below, to introduce a correction factor φ = -θ, according to the following table: Table 6: Phase compensation Parameter Value PHVx = 0x0 PHVx = 0x1 PHCx[9] = 0x0 PHVx = 0x2 PHCx[9] = 0 PHVx = 0x3 PHCx[9] = 0 3.2.1 Step-by-step phase-shift calibration procedure The following steps summarize the calibration procedure explained above: 1. Perform MUC amplitude calibration following steps listed in Section 3: "System calibration" 2. Configure the device through CPC. All registers should be in default state; the following registers have to be configured (according to the channel under calibration and to the device p/n): ROCx (in DSP_CR1 or DSP_CR2) 0: for CT, shunt 1: for Rogowski coil GAINx (in DFE_CR1 or DFE_CR2) to set the correct current gain channel CHVx and CHCx (in DSP_CR5 and DSP_CR8) have to be set as calculated in equation 13 and 14 PHVx and PHCx (in DSP_CR4) are preset to default (0x0) LCSx (in DSP_CR1 and DSP_CR2) is set to 0 or 1 to output on LEDx the desired channel LPSx is set to zero (to output on LEDx the active power signal) 3. Apply stable and accurate nominal values of VN and IN voltage and current signals, shifted by φ = 60 ° angle 4. Read from HPM the e error on the active power from LED frequency 5. Calculate phase-shift error from equation 18 and correction factor from Table 6: "Phase compensation" DocID026176 Rev 2 15/19 System calibration 6. Write PHVx, PHCx to the device and store them in a non-volatile memory 3.2.2 AN4470 Example:phase-shift compensation In a 50 Hz line, after amplitude calibration, the error on active power at PF = 0,5 is measured as: e = 0.038 = 3.8%. From equation 18 the current waveform leads the voltage to θ =1.26 °, so the value to introduce is φ = -θ through PHCx[9:0] = 0x119 (φC=1.2654 °). If the voltage leads current to the same angle, values to introduce are PHV[1:0]=0x1 (φ V = 2.304 °) and PHCx[9:0]=0xE7 (φC =1.0395 °) the current shift, respect to the voltage, is: φV - φC = 1.2654 °. 3.3 Offset calibration The STPM3x has power offset compensation register for all measured powers to compensate, for each channel, the amount of power measured due to noise capture in the application. Power offset compensation registers: OFAx[9:0], OFRx[9:0], OFAFx[9:0], OFSx[9:0], compensating active, reactive, active fundamental and apparent power for each channel (according to the p/n) are located in registers from DSP_CR9 to DSP_CR12. The purpose of power offset compensation is to eliminate the error at low power due to noise and external influences. Applying a minimum current IMIN (for example 2% of IN) and measuring power error e, the offset to apply is: Equation 19: pOFF =-(VN· IMIN · e) Power registers are signed values, (MSB is the sign, and negative values are two's complemented); power offset registers are signed registers as well, and LSB value is equal to 4 times power LSB. Power register LSB Equation 20: 2 W Vref · (1 + R1 R2 ) LSB P = ------------------------------------------------------------------------------------- -----------28 k int· A V· A I · k S· cal V· cal I · 2 LSB LSB power offset register Equation 21: 2 2 2 W Vref · (1 + R1 R2 ) LSB PO= LSB P· 2 = ------------------------------------------------------------------------------------- · 2 28 LSB k int· A V · A I· k S· cal V· cal I · 2 LSB value of power and power offset registers is equal in all power types (reactive, apparent, fundamental). 16/19 DocID026176 Rev 2 AN4470 3.3.1 System calibration Step-by-step offset calibration procedure The following steps summarize the calibration procedure explained above: 1. Perform MUC amplitude calibration following steps listed in the above paragraphs 2. Configure the device through CPC. All registers should be in default state; the following registers have to be configured (according to the channel under calibration and to the device p/n): ROCx (in DSP_CR1 or DSP_CR2) 0: for CT, shunt 1: for Rogowski coil GAINx (in DFE_CR1 or DFE_CR2) to set the correct current gain channel CHVx and CHCx (in DSP_CR5 and DSP_CR8) have to be set as calculated in equation 13 and 14 PHVx and PHCx (in DSP_CR4) have to be set as calculated in equation 18 and table 6 OFAx[9:0], OFRx[9:0], OFAFx[9:0], OFSx[9:0] power offset compensation registers from DSP_CR9 to DSP_CR12 are set to zero 3. Apply stable and accurate nominal values of VN voltage signal and IMIN = 0.02 IN 4. Read from HPM the e error on the active power from LED frequency 5. Calculate offset compensation from equation 18 and related register value from equation 19 and related register value from equation 21 6. Write OFAx, OFRx, OFAFx, OFSx to the device and store them in a non-volatile memory DocID026176 Rev 2 17/19 Revision history 4 AN4470 Revision history Table 7: Document revision history Date Revision 07-Apr-2014 1 Initial release. 2 Updated the equations from 8 to 9, from 12 to 13 and from 20 to 21. Updated Table 3: "STPM3x internal parameters", Table 4: "Phasedelay" and Table 5: "Phase error correction range". Updated Section 3: "System calibration", Section 3.2.2: "Example:phase-shift compensation", Section 3.3: "Offset calibration", Section 3.2.1: "Step-by-step phase-shift calibration procedure". 15-Oct-2015 18/19 Changes DocID026176 Rev 2 AN4470 IMPORTANT NOTICE – PLEASE READ CAREFULLY STMicroelectronics NV and its subsidiaries (“ST”) reserve the right to make changes, corrections, enhancements, modifications, and improvements to ST products and/or to this document at any time without notice. Purchasers should obtain the latest relevant information on ST products before placing orders. ST products are sold pursuant to ST’s terms and conditions of sale in place at the time of order acknowledgement. Purchasers are solely responsible for the choice, selection, and use of ST products and ST assumes no liability for application assistance or the design of Purchasers’ products. No license, express or implied, to any intellectual property right is granted by ST herein. Resale of ST products with provisions different from the information set forth herein shall void any warranty granted by ST for such product. ST and the ST logo are trademarks of ST. All other product or service names are the property of their respective owners. Information in this document supersedes and replaces information previously supplied in any prior versions of this document. © 2015 STMicroelectronics – All rights reserved DocID026176 Rev 2 19/19