Application Note Poly-Phase Energy Metering IC M90E32AS APPLICATION OUTLINE This document describes system application issues when using the M90E32AS (polyphase energy metering ICs) to design poly-phase energy meters. The M90E32AS is applicable in class 0.5S or class 1 poly-phase meter design and also supports Three-Phase Four-Wire (3P4W, Y0) or Three-Phase Three-Wire (3P3W, Y or Δ) connection modes. The M90E32AS uses 3.3V single power supply. In a typical 3P4W design, there are three transformers®ulators to provide power supply. The AC power supply outputs 3.3V to chip digital power supply DVDD after rectifier and voltage regulation. The analog power supply AVDD should be connected directly to digital power supply DVDD. The M90E32AS has on-chip power-on-reset circuit. The RESET pin should be connected to DVDD through a 10kΩ resistor and a 0.1μF filter capacitor to ground. The M90E32AS has highly stable on-chip reference power supply. The Vref pin should be decoupled with a 4.7μF capacitor and a 0.1μF ceramic capacitor. The M90E32AS employs 16.384MHz as the system frequency. The M90E32AS has builtin crystal oscillator circuit and 10pF matching capacitance. Users only need to connect a 16.384MHz crystal between OSCI and OSCO pins in application. The M90E32AS provides ADC sampling on three voltage channels (V1, V2 and V3) and three current channels (I1, I2 and I3). These 6 ADC channels can be flexibly mapped in different PCB design. The M90E32AS provides a 4-wire SPI interface (CS, SCLK, SDI and SDO) for external MCU connection. MCU can perform chip configuration and register reading/writing through SPI. The M90E32AS provides four energy pulse output pins: active energy pulse CF1, reactive energy pulse CF2 (can also be configured as apparent energy pulse), fundamental energy pulse CF3 and harmonic energy pulse CF4. They can be used for energy metering calibration and can also be connected to MCU for energy accumulation. The M90E32AS provides three zero-crossing pins ZX0, ZX1 and ZX2 which can select different phase’s voltage or current as zero-crossing judgement. The M90E32AS provides three output pins IRQ0, IRQ1 and WarnOut to generate interrupt and warn out signals at different levels. The default application in this document is 3P4W, otherwise it will be specially indicated. Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 Ta bl e o f C o n t en ts 1 HARDWARE REFERENCE DESIGN ........................................................................................... 4 1.1 3P4W Application ................................................................................................................... 4 1.1.1 1.1.2 1.1.3 1.1.4 1.1.5 Schematics (Current Transformer (CT)) ................................................................................... 4 BOM (CT) ................................................................................................................................. 5 Schematics (Rogowski) ............................................................................................................ 6 BOM (Rogowski) ....................................................................................................................... 7 Circuit Description ..................................................................................................................... 7 1.2 3P3W Application ................................................................................................................... 9 1.2.1 1.2.2 1.2.3 Schematics ............................................................................................................................... 9 BOM ........................................................................................................................................ 10 Circuit Description ................................................................................................................... 10 2 INTERFACE ................................................................................................................................ 11 3 POWER MODES ........................................................................................................................ 13 3.1 Normal Mode ........................................................................................................................ 13 3.2 Partial Measurement Mode................................................................................................... 13 3.3 Detection Mode..................................................................................................................... 14 3.4 Idle Mode .............................................................................................................................. 14 3.5 Transition And Application of Power Modes ......................................................................... 14 4 CALIBRATION ........................................................................................................................... 17 4.1 Calibration Method................................................................................................................ 17 4.2 Calibration in Normal Mode .................................................................................................. 17 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.2.6 4.2.7 4.2.8 Metering Enable and CRC Calibration .................................................................................... 18 PL Constant Configuration (PL_Constant) ............................................................................. 19 Metering Method Configuration (MMode0) ............................................................................. 20 PGA Gain Configuration (MMode1) ........................................................................................ 21 Offset Calibration of Voltage/ Current/ Power ......................................................................... 21 Voltage/ Current Measurement Calibration ............................................................................ 22 Energy Metering Calibration ................................................................................................... 24 Fundamental Energy Metering Calibration ............................................................................. 25 4.3 Calibration in Partial Measurement Mode............................................................................. 26 4.4 Calibration in Detection Mode............................................................................................... 26 4.4.1 4.4.2 Current Detection Module Configuration ................................................................................ 26 Current Detection Threshold Calibration ................................................................................ 27 5 FUNCTION REGISTERS CONFIGURATION ............................................................................ 28 5.1 Startup Current Configuration............................................................................................... 28 5.2 Sag Function......................................................................................................................... 29 6 COMPENSATION METHOD ...................................................................................................... 30 6.1 Current Segment Compensation Descriprion ....................................................................... 30 6.2 GainIrms Segment Compensation Example......................................................................... 32 6.3 PhiIrms Segment Compensation Example ........................................................................... 33 M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 2 6.4 Frequency-Phase Compensation (PhiFreqComp)................................................................ 34 7 TEMPERATURE COMPENSATION .......................................................................................... 35 7.1 On-chip Temperature Sensor Configuration......................................................................... 35 7.2 Temperature Compensation Based on Ugaint Example ...................................................... 36 7.3 Temperature Compensation Based on Reference Voltage .................................................. 37 M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 3 D C B A IB- IB+ IA- IA+ UN UC UB UA 1 1 R53 2.4 R51 2.4 R44 2.4 R38 2.4 1K R55 GND 1K R49 1K R46 GND 1K R35 240K GND R26 240K 240K R25 R17 240K 240K 240K R16 R4 R3 240K R29 240K R20 240K R7 240K R30 240K R21 240K R8 C20 18nF C18 18nF C16 18nF C11 18nF IBN IBP IAN IAP IC- IC+ Voltage Sampling 240K R28 240K R19 240K R6 240K R31 240K R22 240K R9 2 Current Sampling (with CT) 240K R27 240K R18 240K R5 2 R45 2.4 R39 2.4 R32 1K R23 1K R10 1K C8 18nF VCP 1K R47 GND 1K VBP C6 18nF R36 GND GND GND C3 18nF VAP 3 3 C17 18nF C12 18nF ICN ICP IAP IAN IBP IBN ICP ICN R40 1K 4 C13 18nF C10 0.1uF AVDD AGND I1P I1N I2P I2N I3P I3N IC IC Vref AGND C5 0.1uF DVDD33 GND 0.1uF 0.1uF C4 10uF C2 C1 R2 10K 5 C14 18nF R43 1K GND C15 18nF 16.384MHz X1 IC NC PM1 PM0 TEST IRQ1 IRQ0 WarnOut CF4 CF3 CF2 CF1 R11 10K DVDD33 36 35 34 33 32 31 30 29 28 27 26 25 R12 10K R13 10K CF4 CF3 CF2 CF1 5 ZX2 ZX1 ZX0 IRQ1 IRQ0 WarnOut CF4 CF3 CF2 CF1 PM1 PM0 SDI SDO SCLK CS 6 6 Connect to MCU GND R14 10K Poly Phase Metering AFE Chip (ATM90E32AS) R41 1K U5 ATM90E32AS 1 2 3 4 5 6 7 8 9 10 11 12 AVDD33 C7 0.1uF GND GND C9 4.7uF GND 4 48 47 46 45 44 43 42 41 40 39 38 37 DVDD DGND NC NC DGND VDD18 VDD18 RESET SDI SDO SCLK CS Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 V1P V1N V2P V2N V3P V3N DGND OSCI OSCO ZX0 ZX1 ZX2 M90E32AS [Application Note] VAP GND 510 R33 U4 510 R24 U3 510 R15 U2 510 R1 U1 GND GND GND GND PS2501 PS2501 PS2501 PS2501 1 2 3 4 5 CON-5 1 2 3 4 5 JP1 8 D4 CF1 D3 CF2 D2 CF3 D1 CF4 GND 7 Poly Phase Metering AFE ATM90E32AS (3P4W with CT) 1 1.0 of 1 A3 8 ATM90E32AS_3P4W_CT.SchDoc Title: File: * Project: Revision: Page: 11:24:59 * Document Number: Date: Size: * ApprovedBy: 3/19/2014 Felix Yao DrawnBy: * SMART ENERGY Atmel China CheckedBy: 510 R48 510 R42 510 R37 510 R34 Energy Pulse Output Indicate CF4 CF3 CF2 CF1 Energy Pulse Output (Isolated with Optocoupler) CF4 CF3 CF2 CF1 7 D C B A 3P4W Application VBP 1.1 VCP HARDWARE REFERENCE DESIGN 13 14 15 16 17 18 19 20 21 22 23 24 4 1 1.1.1 Schematics (Current Transformer (CT)) 1.1.2 BOM (CT) Table-1 3P4W BOM (CT) Component Type SMT Capacitor Designator Quantity Parameter Tolerance C3 C6 C8 C11 C12 C13 C14 C15 C16 C17 C18 C20 12 18nF ±5% NP0 (anti-aliasing filter capacitor) C1 C4 C5 C7 C10 5 0.1μF ±10% X7R C9 1 4.7μF ±10% X7R C2 1 10μF ±10% X7R R38 R39 R44 R45 R51 R53 6 2.4Ω ±1% 1/8W 25ppm R1 R15 R24 R33 R34 R37 R42 R48 8 510Ω ±5% 1/8W 100ppm R10 R23 R32 R35 R36 R40 R41 R43 R46 R47 R49 R55 12 1kΩ ±1% 1/8W 25ppm (anti-aliasing filter resistor) R2 R11 R12 R13 R14 5 10kΩ ±5% 1/8W 100ppm R3~R9, R16~R22, R25~R31 21 240kΩ ±1% 1/8W 25ppm LED D1 D2 D3 D4 4 - - SMT Optocoupler U1 U2 U3 U4 4 PS2501 - Crystal X1 1 16.384MHz ±20ppm SMT Resistor IC U5 1 M90E32AS - Connector JP1 1 CON-5 - M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 5 D C B A IB- IB+ IA- IA+ UN UC UB 1 C22 18nF 240K 240K R28 240K R19 240K R6 240K R29 240K R20 240K R7 240K R30 240K R21 240K R8 1K 1k IBN IBP IAN IC- IC+ 1K R45 1K R39 240K R31 240K R22 240K R9 2 C8 18nF VCP C6 18nF VBP 1K C25 GND 18nF R47 1K VAP C3 18nF R36 GND GND GND C23 18nF R32 1K R23 1K R10 1K Current Sampling (with Rogowski Coil) C20 18nF C28 GND 18nF R55 R53 C16 18nF C11 18nF IAP Voltage Sampling C18 18nF 1K 240K R27 240K R18 240K R5 C26 18nF R49 1K 1K R51 1K R44 1K 1K C24 GND 18nF R46 R35 R38 GND R26 240K 240K 240K R25 R17 240K R16 R4 240K 3 C17 18nF C12 18nF ICN ICP IAP IAN IBP IBN ICP ICN R40 1K 4 C13 18nF C10 0.1uF AVDD AGND I1P I1N I2P I2N I3P I3N IC IC Vref AGND C5 0.1uF 0.1uF 0.1uF C4 10uF C2 C1 R2 10K 5 C14 18nF R43 1K GND C15 18nF 16.384MHz X1 IC NC PM1 PM0 TEST IRQ1 IRQ0 WarnOut CF4 CF3 CF2 CF1 R11 10K DVDD33 36 35 34 33 32 31 30 29 28 27 26 25 R12 10K R13 10K CF4 CF3 CF2 CF1 5 ZX2 ZX1 ZX0 IRQ1 IRQ0 WarnOut CF4 CF3 CF2 CF1 PM1 PM0 SDI SDO SCLK CS 6 6 Connect to MCU GND R14 10K Poly Phase Metering AFE Chip (ATM90E32AS) R41 1K U5 ATM90E32AS 1 2 3 4 5 6 7 8 9 10 11 12 AVDD33 C7 0.1uF GND GND C9 4.7uF GND DVDD33 GND VAP 4 VBP R3 3 VCP 2 48 47 46 45 44 43 42 41 40 39 38 37 Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 DVDD DGND NC NC DGND VDD18 VDD18 RESET SDI SDO SCLK CS M90E32AS [Application Note] V1P V1N V2P V2N V3P V3N DGND OSCI OSCO ZX0 ZX1 ZX2 6 13 14 15 16 17 18 19 20 21 22 23 24 UA 1 GND 510 R33 U4 510 R24 U3 510 R15 U2 510 R1 U1 GND GND GND GND PS2501 PS2501 PS2501 PS2501 1 2 3 4 5 CON-5 1 2 3 4 5 JP1 8 D4 CF1 D3 CF2 D2 CF3 D1 CF4 GND 7 8 Poly Phase Metering AFE ATM90E32AS (3P4W with Rogowski Coil) 1.0 ATM90E32AS_3P4W_Coil.SchDoc Title: File: of 1 * 1 A3 Project: Revision: Page: 11:25:09 * Document Number: Date: Size: * ApprovedBy: 3/19/2014 DrawnBy: * SMART ENERGY Felix Yao Atmel China CheckedBy: 510 R48 510 R42 510 R37 510 R34 Energy Pulse Output Indicate CF4 CF3 CF2 CF1 Energy Pulse Output (Isolated with Optocoupler) CF4 CF3 CF2 CF1 7 D C B A 1.1.3 Schematics (Rogowski) 1.1.4 BOM (Rogowski) Table-2 3P4W BOM (Rogowski) Component Type Designator Quantity Parameter Tolerance 21 18nF ±5% NP0 (anti-aliasing filter capacitor) 5 1 1 8 0.1μF 4.7μF 10μF 510Ω 18 1kΩ ±1% 1/8W 25ppm (anti-aliasing filter resistor) 5 21 10kΩ 240kΩ ±5% 1/8W 100ppm ±1% 1/8W 25ppm LED C3 C6 C8 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25 C26 C27 C28 C1 C4 C5 C7 C10 C9 C2 R1 R15 R24 R33 R34 R37 R42 R48 R10 R23 R32 R35 R36 R38 R39 R40 R41 R43 R44 R45 R46 R47 R49 R51 R53 R55 R2 R11 R12 R13 R14 R3~R9, R16~R22, R25~R31 D1 D2 D3 D4 4 - - SMT Optocoupler U1 U2 U3 U4 4 PS2501 - Crystal X1 1 16.384MHz ±20ppm IC U5 1 M90E32AS - Connector JP1 1 CON-5 - SMT Capacitor SMT Resistor ±10% ±10% ±10% ±5% 1/8W X7R X7R X7R 100ppm 1.1.5 Circuit Description The recommended circuit for the M90E32AS three-phase four-wire (3P4W) application is as shown in 1.1.1 Schematics (Current Transformer (CT)). The M90E32AS can use CT and Rogowski coil in current sampling. The recommended circuit for 3P4W application with Rogowski coil is as shown in 1.1.3 Schematics (Rogowski). It is recommended to use two-order filtering when sampling with Rogowski coil. The other parts are same with the CT application circuit. It is recommended to refer to the circuit of 1.2.1 Schematics when using Rogowski coil in 3P3W design, only need to change CT to Rogowski coil. The recommended type of Rogowski coil is: PA3202NL (Pulse Electronics). Poly-phase voltage is sampled over resistor divider network with recommended ratio of 240KΩ x 7:1KΩ. The anti-aliasing filter capacitor is recommended to be 18nF. Poly-phase current and N line current are sampled over current transformer (CT). The CT ratio and load resistance should be selected based on the actual metering range. The anti-aliasing filter resistance/capacitor is suggested to be 1KΩ/18nF for the current sampling circuit. The CF1~CF4 pins are provided with driving capacity of 8mA which can drive LED and optocoupler parallelly. The other digital pins are provided with driving capacity of 5mA which can drive optocoupler directly. M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 7 Application note: how to select CT and CT load resistance Condition: M90E32AS ADC input voltage range is 120μVrms ~ 720mVrms M90E32AS ADC input gain PGA_GAIN = 1, 2, 4 Assume: Metering range of the energy meter is Imin ~ Imax CT current output ratio is N:1 CT load resistance is RCT So the parameters meet the formula as below: 120 μVrms < PGA_GAIN × RCT × I min N PGA_GAIN × RCT × I max < 720mVrms N 8 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 D C B A IA- IA+ UC UB 1 240K 240K R44 2.4 1K R46 GND 1K R35 240K 240K R38 2.4 R26 R25 GND R4 R3 240K R29 240K R7 240K R30 240K R8 C16 18nF C11 18nF IAN IAP IC- IC+ Voltage Sampling 240K R28 240K R6 240K R31 240K R9 2 Current Sampling (with CT) 240K R27 240K R5 R45 2.4 R39 2.4 R32 1K R10 1K 1K R47 GND 1K VCP C8 18nF R36 GND GND C3 18nF VAP 3 3 C17 18nF C12 18nF ICN ICP GND ICP ICN IAP IAN R40 1K 4 C13 18nF AVDD AGND I1P I1N I2P I2N I3P I3N IC IC Vref AGND C5 0.1uF DVDD33 GND 0.1uF 0.1uF C4 10uF C2 C1 R2 10K 5 R43 1K GND C15 18nF 16.384MHz X1 IC NC PM1 PM0 TEST IRQ1 IRQ0 WarnOut CF4 CF3 CF2 CF1 R11 10K DVDD33 36 35 34 33 32 31 30 29 28 27 26 25 R12 10K R13 10K CF4 CF3 CF2 CF1 5 ZX2 ZX1 ZX0 IRQ1 IRQ0 WarnOut CF4 CF3 CF2 CF1 PM1 PM0 SDI SDO SCLK CS 6 6 Connect to MCU GND R14 10K Poly Phase Metering AFE Chip (ATM90E32AS) U5 ATM90E32AS 1 2 3 4 5 6 7 8 9 10 11 12 AVDD33 C7 0.1uF C10 0.1uF GND C9 4.7uF GND 4 VAP 2 VCP UA 1 48 47 46 45 44 43 42 41 40 39 38 37 DVDD DGND NC NC DGND VDD18 VDD18 RESET SDI SDO SCLK CS V1P V1N V2P V2N V3P V3N DGND OSCI OSCO ZX0 ZX1 ZX2 13 14 15 16 17 18 19 20 21 22 23 24 GND 510 R33 U4 510 R24 U3 510 R15 U2 510 R1 U1 GND GND GND GND PS2501 PS2501 PS2501 PS2501 1 2 3 4 5 CON-5 1 2 3 4 5 JP1 8 D4 CF1 D3 CF2 D2 CF3 D1 CF4 GND 7 Poly Phase Metering AFE ATM90E32AS (3P3W with CT) 1 1.0 of 1 A3 8 ATM90E32AS_3P3W_CT.SchDoc Title: File: * Project: Revision: Page: 11:24:48 * Document Number: Date: Size: * ApprovedBy: 3/19/2014 Felix Yao DrawnBy: * SMART ENERGY Atmel China CheckedBy: 510 R48 510 R42 510 R37 510 R34 Energy Pulse Output Indicate CF4 CF3 CF2 CF1 Energy Pulse Output (Isolated with Optocoupler) CF4 CF3 CF2 CF1 7 D C B A 1.2 3P3W Application 1.2.1 Schematics M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 9 1.2.2 BOM Table-3 3P3W BOM Component Type Designator Quantity Parameter Tolerance 8 18nF 5 1 1 4 8 0.1μF 4.7μF 10μF 2.4Ω 510Ω ±5% NP0 (anti-aliasing filter capacitor) ±10% X7R ±10% X7R ±10% X7R ±1% 1/8W 25ppm ±5% 1/8W 100ppm 8 1kΩ LED C3 C8 C11 C12 C13 C15 C16 C17 C1 C4 C5 C7 C10 C9 C2 R38 R39 R44 R45 R1 R15 R24 R33 R34 R37 R42 R48 R10 R32 R35 R36 R40 R43 R46 R47 R2 R11 R12 R13 R14 R3~R9, R25~R31 D1 D2 D3 D4 5 14 4 10kΩ 240kΩ - ±1% 1/8W 25ppm (anti-aliasing filter resistor) ±5% 1/8W 100ppm ±1% 1/8W 25ppm - SMT Optocoupler U1 U2 U3 U4 4 PS2501 - Crystal X1 1 16.384MHz ±20ppm IC U5 1 M90E32AS - Connector JP1 1 CON-5 - SMT Capacitor SMT Resistor 1.2.3 Circuit Description This circuit is the recommended circuit for the M90E32AS three-phase three-wire (3P3W) application. Phase B is the reference ground in 3P3W application. In 3P3W system, Uab stands for Ua, Ucb stands for Uc and there is no Ub. Phase B voltage, phase B current and N line sampling current are not needed in 3P3W application. Pin 5, 6, 15 and 16 should be connected to GND. All NC pins should be left open. The other parts of 3P3W application circuit are similar to 3P4W and can be treated in the same way. 10 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 2 INTERFACE The M90E32AS provides a four-wire SPI interface (CS, SCLK, SDI and SDO). The SPI interface in Slave mode is mainly used for register read/write operation. A complete SPI read/write operation is of 32 bits, which contains 16-bit address and 16-bit data. In the 16-bit address, bit0 ~ bit9 correspond to valid register address A0 ~ A9, and bit10 ~ bit14 are reserved (these bits are don’t-care). Bit15 indicates the SPI operation is read or write. SPI Operation Description Highest Bit (Bit15) Read Read register data 1 Write Write data to register 0 The transmission of address and data bits is from high to low, which means MSB first and LSB last. Note that the M90E32AS read/write only supports single address operation, rather than continuous read or write. The M90E32AS has a special register LastSPIData [78H] for recording the last SPI read/write data. This register can be used for data check for SPI read/write operation. When the system is in strong interference situation, the disturbance signal may cause SPI communication disorder and result in SPI read/write error. In this case, LastSPIData can be used to check the correctness of SPI read/write and strengthen system robustness. For read-clear registers, if the read data is different from the LastSPIData data, the actual data can be obtained by reading the LastSPIData register repeatedly. LastSPIData application is as shown in Figure-1 and Figure-2: SPI Read Buffer Read LastSPIData LastSPIData == Buffer ? Y N Buffer=LastSPIData Read LastSPIData LastSPIData == Buffer ? N Y End Figure-1 LastSPIData Application (Read) M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 11 SPI Write Data Buffer Read LastSPIData LastSPIData == Buffer ? N Y End Figure-2 LastSPIData Application (Write) 12 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 3 POWER MODES Four power modes are supported which correspond to four kinds of power consumption. The power mode is configured by PM1/PM0 pins. 3.1 PM1 PM0 Power Modes 1 1 Normal mode 1 0 Partial Measurement mode 0 1 Detection mode 0 0 Idle mode Power Consumption High ↓ Low Normal Mode In Normal mode, all function blocks are active except for the current detector block. All registers can be accessed. 3.2 Partial Measurement Mode In Partial Measurement mode, the active measurement modules in Partial Measurement mode are the same as Normal mode. In this mode, all the measurements are through the same hardware that does the measurement in the normal mode. To save power, the energy accumulation block and a portion of the DSP computation code will not be running in this mode. In this mode, There are configuration bits in the PMPwrCtrl (0EH) register to get lower power if the application allows: Address Name Bit15 0EH PMPwrCtrl Bit14 Bit13 Bit12 Bit15 ~ Bit0 Bit11 Bit10 Bit9 Bit8 - - - - - - - PMPwrDownVch Bit7 Bit6 Bit5 Bit4 Bit3 Bit2 Bit1 Bit0 - - - - ACTRL_CLK_ GATE DSP_CLK_ GATE MTMS_CLK_ GATE PMClkLow 1. PMPwrDownVch: In Partial Measurement Mode the V0/V1/V2 analog channel can be powered off to save power 0: Power on 1: Power off (default) 2. ACTRL_CLK_GATE: Power off the clock of analog control block to save power. 0: Power on 1: Power off (default) 3. DSP_CLK_GATE: Power off the clock of DSP register to save power. 0: Power on 1: Power off (default) 4. MTMS_CLK_GATE: Power off the metering and measuring block to save power. 0: Power on 1: Power off (default) 5. PMClkLow: In Partial Measurement Mode the main clock can be reduced to 8.192MHz to save power. 0: 16.384MHz 1: 8.192MHz (default) M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 13 3.3 Detection Mode In Detection mode, only the current detector is active and all the registers can not be accessed by external MCU. In this mode, each I/O is in specific state (for details refer to datasheet) and SPI is disabled. So the control and threshold registers for Detection mode need to be programmed in Normal mode before entering Detection mode. Once these related registers are written, there is no need to re-configure them when switching between different power modes. Detection mode related registers are listed as below: Address 10H Name DetectCtrl 11H DetectThA 12H DetectThB 13H DetectThC Current detection is achieved with low power comparators. Two comparators are supplied for each phase on detecting positive and negative current. When any single-phase current or multiple-phase current exceeds the configured threshold, the IRQ0 pin is asserted high. When all three phase currents exceed the configured threshold, the IRQ1 pin is asserted high. The IRQ0/IRQ1 state is cleared when entering or exiting Detection mode. The all three phase currents are considered as the currents of three current channels I1~I3. As there is no phase B current in 3P3W application, IRQ1 will not be asserted high even if both phase A and phase C current exceed the configured threshold. Current detection module can be enabled in Normal mode through configuring the DEtectCtrl(10H) register to facilitate the current detection threshold calibration. 3.4 Idle Mode In Idle mode, all the modules are disabled and all the registers can not be accessed. In this mode, each I/O is in a specific state (for details refer to datasheet) and SPI is disabled. All register values are lost except for current detection related registers. 3.5 Transition And Application of Power Modes The four power modes are controlled by the PM0 and PM1 pins. In application, any power mode transition goes through Idle mode to avoid register value confusion or system status uncertainty in mode transition. All function modules are disabled in Idle mode while the related modules will be enabled after switching from Idle mode to other mode, which is equivalent to reset to the function modules, thus ensuring normal operation of the function modules. It needs to reload registers to ensure normal operation when switching from Idle mode to Normal mode or Partial Measurement mode, while no need to reload registers when switching from Idle mode to Detection mode. Power mode transition is shown as Figure-3: 14 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 Normal Mode PM1:PM0 = 11 Idle Mode PM1:PM0 = 00 Detection Mode PM1:PM0 = 01 Detection Mode related register value will be kept Need to reload all register values All the register values will be lost except for the Detection Mode related registers Partial Measurement Mode PM1:PM0 = 10 Need to reload Partial Measurement related registers Figure-3 Power Mode Transition Note: For description convenience, the intermediary Idle mode will be omitted when refering to power mode transition. The most typical application of power mode transition is no-voltage detection for power meter. The so-called no-voltage state is when all phase voltages are less than the voltage threshold but the load current is greater than the configured current value (such as 5% of rated current). In no-voltage state, the power meter usually uses backup battery for power supply. The system needs to enter low power mode and perform measurement and recording for novoltage state periodically. The recommended flow for power meter with the M90E32AS is as below: 1.Set the current detection threshold to be the minimum load current (such as 5% of rated current) required in no-voltage state. 2. When no-voltage happens, the system enters Idle mode; 3. The system enters Detection mode every once in a while (such as 5s); 4. Once the load current is greater than the configured value, the system enters Partial Measurement mode to measure and record the load current; 5. The system returns to Idle mode after measurement and recording are completed; 6. The system enters Partial Measurement mode every once in a while (such as 60s) to measure and record the load current. M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 15 System voltage sag Enter Partial Measurement mode Enter Idle mode N Measure current>Minimum load current? Y Delay 5s Record current value (No-voltage event) Enter Detection mode Enter Idle mode N IRQ0/IRQ1 output high level? Delay 60s Y Figure-4 Application of Detection Mode and Partial Measurement Mode Application note: Design principle for current detection threshold It is recommended to do system design based on current detection threshold of 1.5mVrms. Example: Assume: The requirement is that the minimum load current detected is 2.5% of rated current. Current specification is 5(60)A; The minimum load current is Id, which corresponds to a 1.5mVrms ADC input signal. The parameters meet the following relations: 16 Minimum Detection Load Current Id Related Current In Maximum Current Imax ratio to rated current 2.5% In In 12In corresponding A/D input signal 3mVrms 60mVrms 720mVrms actual current 250mA 5A 60A M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 4 CALIBRATION 4.1 Calibration Method Normally voltage, current, mean power, phase angle, frequency and so on are regarded as measurement values, while active energy, reactive energy and so on are regarded as metering values. Measurement and metering function both need calibration before normal use as shown in below table. Power Mode Normal mode Parameter Need Calibration Calibration Method voltage/current √ offset/gain calibration power/frequency/phase angle/ power factor X - full-wave energy metering √ offset/gain/phase angle calibration fundamental energy metering √ offset/gain calibration harmonic energy metering X - Partial Measurement mode voltage/ current/ power X - Detection mode current detection √ threshold calibration In typical application of three-phase power meter, voltage, current and full-wave energy must be calibrated. The others can be calibrated according to actual application, no need to calibrate if no use. The calibration flow follows the sequence of measurement first then metering. Metering calibration is realized by first calibrating gain and then calibrating phase angle compensation, only single-point calibration is needed over the entire dynamic range. Reactive does not need to be calibrated since it is guaranteed by chip design. Frequency, phase angle and power factor do not need calibration, since their accuracy is guaranteed by chip design. 4.2 Calibration in Normal Mode The basic functions, such as measurement, metering, harmonic analysis and so on are only active in Normal mode. So calibration in Normal mode is basic and a must. The related registers need to be configured before calibration. Calibration flow is as shown in Figure-5. M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 17 Calibration Initialize Voltage/Current Offset Calibration Voltage/Current Gain Calibration Power Offset Calibration Energy Gain Calibration Phase Angle Calibration End Figure-5 Active Energy Metering Calibration Flow in Normal Mode 4.2.1 Metering Enable and CRC Calibration The device can automatically monitor the CRC changes versus a golden CRC which is latched after the first time the CRC computation is done. The latching event is triggered by none "0x55AA" value written to the CfgRegAccEn register (which means configuration done), followed by a new C C result available event. Once golden CRC is latched, the CRC_CMP signal is enabled. Subsequent CRC result will be compared with the latched CRC to generate the CRC error status. CRC error status can be read, and if configured, can goes to WARN or IRQ0 pins to alert the MCU in the case of CRC error. 18 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 4.2.2 PL Constant Configuration (PL_Constant) Energy accumulation and metering are usually referenced by energy unit, such as kWh. However, within the M90E32AS, energy calculation or accumulation are based on energy pulse (CF). kWh and CF are connected by Meter Constant (MC, such as 3200 imp/kWh, which means each kWh corresponds to 3200 energy pulses). The chip’s PL_Constant is a parameter related to MC. One PL_Constant corresponds to 0.01CF. PL_Constant should be configured according to different MC in application. The M90E32AS provides four energy pulse outputs: active energy pulse CF1, reactive energy pulse or apparent energy pulse CF2, fundamental energy pulse CF3 and harmonic energy pulse CF4. Their Meter Constants are all set by PL_Constant in union rather than separately. The PL_Constant registers consist of the PLconstH[31H] and PLconstL[32H] registers, corresponding to high word and low word of PL_Constant respectively. PL_Constant is calculated as below: PL_constant = 450,000,000,000 / MC 450,000,000,000: Constant MC: Meter Constant, unit is imp/KWh, imp/Kvarh or imp/KVA Example: Calculation of PL_constant Assume: Meter Constant MC = 3200 Thus: PL_constant = 450,000,000,000 / 3200= 140,625,000 (Hex is 8614C68H) so the registers are set as below: PLconstH[31H] = 0861H PLconstL[32H] = 4C68H M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 19 4.2.3 Metering Method Configuration (MMode0) The M90E32AS can be used in difference systems and metering modes, which can be configured by the MMode0[33H] register. Address 33H Name MMode0 Bit15 ~ Bit0 Bit15 Bit14 Bit13 Bit12 Bit11 Bit10 Bit9 Bit8 - - - Freq60Hz HPFOff didtEn - 3P3W Bit7 Bit6 Bit5 Bit4 Bit3 Bit2 Bit1 Bit0 CF2varh - - ABSEnQ ABSEnP EnPA EnPB EnPC 1. Freq60Hz: grid operating line reference frequency 0: 50Hz (default) 1: 60Hz The M90E32AS is applicable in 50 Hz or 60 Hz power grid. The M90E32AS uses different calculation parameters in data processing according to different grid frequency. To improve the accuracy of measurement and metering, please set this control bit according to the real power grid frequency. 2. HPFOff: HPF enable control bit 0: enable HPF (default) 1: disable HPF Besides measuring the voltage/current RMS in 50Hz or 60Hz (AC) power grid, the M90E32AS can also measure the mean current value of DC condition. HPF should be disabled when using DC measurement functions. 3. didtEn: enable integrator for didt current sensor 0: disable integrator for didt current sensor; use CT sampling for current channel (default) 1: enable integrator for didt current sensor; use Rogowski coil sampling for current channel The M90E32AS supports sampling over CT or Rogowski coil. Please set this control bit according to the real current sampling means. Note that different sampling circuit should be adopted when using Rogowski coil. 4. 3P3W: connection type for three-phase energy meter 0: 3P4W connection (default) 1: 3P3W connection The M90E32AS uses different phase sequence judgment for different connection. Please set this control bit according to the real connection type. 5. CF2varh: CF2 pin source configuration 0: apparent energy 1: reactive energy (default) 6. ABSEnQ: configure the calculation method of total (all-phase-sum) reactive energy and power 0: total reactive energy equals to all-phase reactive energy arithmetic sum (default) 1: total reactive energy equals to all-phase reactive energy absolute sum 7. ABSEnP: configure the calculation method of total (all-phase-sum) active energy and power 0: total active energy equals to all-phase active energy arithmetic sum (default) 1: total active energy equals to all-phase active energy absolute sum 8. EnPA: this bit configures whether Phase A is counted into the all-phase sum energy/power (P/Q/S) 0: Corresponding Phase A not counted into the all-phase sum energy/power (P/Q/S) 1: Corresponding Phase A to be counted into the all-phase sum energy/power (P/Q/S) (default) 9. EnPB: this bit configures whether Phase B is counted into the all-phase sum energy/power (P/Q/S) 0: Corresponding Phase B not counted into the all-phase sum energy/power (P/Q/S) 1: Corresponding Phase B to be counted into the all-phase sum energy/power (P/Q/S) (default) 10.EnPC: this bit configures whether Phase C is counted into the all-phase sum energy/power (P/Q/S) 0: Corresponding Phase C not counted into the all-phase sum energy/power (P/Q/S) 1: Corresponding Phase C to be counted into the all-phase sum energy/power (P/Q/S) (default) Application note: Common configuration of MMode0 1. 3P4W, grid frequency 50Hz, MMode0 = 0087H 2. 3P3W, grid frequency 50Hz, MMode0 = 0185H 20 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 4.2.4 PGA Gain Configuration (MMode1) The MMode1 register is used to configure PGA gain of ADC sampling channel, making chips applicable to meter designs of different current specifications. Address Name Bit15 ~ Bit0 Bit15 Bit14 - 34H MMode1 Bit7 Bit13 Bit12 PGA_GAIN (V3) Bit6 Bit5 - Bit4 PGA_GAIN (I3) Bit11 Bit10 PGA_GAIN (V2) Bit3 Bit9 Bit8 PGA_GAIN (V1) Bit2 PGA_GAIN (I2) Bit1 Bit0 PGA_GAIN (I1) PGA_GAIN (V1~V3, I1~I4): analog PGA gain for seven ADC channels 00: 1X (default) 01: 2X 10: 4X 11: N/A Application note: Configuration principle of PGA gain 1. Ensure that the ADC channel analog input signal should be within the dynamic range of 0~720mVrms 2. Configure PGA gain to be the maximum value within the whole dynamic range 4.2.5 Offset Calibration of Voltage/ Current/ Power In application, the input signal is often influenced by the interference signal. This interference will enter data processing module through ADC and high-pass filter, not only producing errors to the voltage/current RMS and power calculation, but also affecting accuracy of the energy metering. The M90E32AS provides offset calibration function to voltage, current and power, reducing the influence of the interference signal to measurement/metering accuracy. Every phase’s voltage/current offset calibration should be proceeded individually. Take phase A for example, the signal source is: Ub=Uc=Un, Ua=0, Ia=0. The calibration flow of voltage/current offset is as below: 1. Read measurement registers (32 bits). It is suggested to read several times to get the average value; 2. Right shift the 32-bit data by 7 bits (ignore the lowest 7 bits); 3. Invert all bits and add 1 (2’s complement); 4. Write the lower 16-bit result to the offset register Every phase’s power offset calibration should be proceeded individually. Take phase A for example, the signal source is: Ua=Ub=Uc=Un, Ia=0. The calibration flow of power offset is as below: 1. Read measurement registers (32 bits). It is suggested to read several times to get the average value; 2. Invert all bits and add 1 (2’s complement); 3. Write the lower 16-bits result to the offset register The corresponding offset register and measurement value registers are shown as below: M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 21 Offset Registers Address Voltage Current All-wave Power fundamental power Measurement Value Registers Register Name Address Register Name Address Register Name 63H UoffsetA 0D9H UrmsA 0E9H UrmsALSB 67H UoffsetB 0DAH UrmsB 0EAH UrmsBLSB 6BH UoffsetC 0DBH UrmsC 0EBH UrmsCLSB 64H IoffsetA 0DDH IrmsA 0EDH IrmsALSB 68H IoffsetB 0DEH IrmsB 0EEH IrmsBLSB 6CH IoffsetC 0DFH IrmsC 0EFH IrmsCLSB 6EH IoffsetN 0D8H IrmsN1 - - 41H PoffsetA 0B1H PmeanA 0C1H PmeanALSB 42H QoffsetA 0B5H QmeanA 0C5H QmeanALSB 43H PoffsetB 0B2H PmeanB 0C2H PmeanBLSB 44H QoffsetB 0B6H QmeanB 0C6H QmeanBLSB 45H PoffsetC 0B3H PmeanC 0C3H PmeanCLSB 46H QoffsetC 0B7H QmeanC 0C7H QmeanCLSB 51H PoffsetAF 0D1H PmeanAF 0E1H PmeanAFLSB 52H PoffsetBF 0D2H PmeanBF 0E2H PmeanBFLSB 53H PoffsetCF 0D3H PmeanCF 0E3H PmeanCFLSB 4.2.6 Voltage/ Current Measurement Calibration Measurement calibration means the calibration of voltage rms (Urms) gain and current rms (Irms) gain. Measurement calibration is the premise of energy metering calibration. 1. Voltage/current offset (Uoffset/Ioffset) calibration: For calibration method, please refer to 4.2.5 Offset Calibration of Voltage/ Current/ Power. No need of calibration if the voltage/current offset is very small in general application. 2 Voltage/current gain calibration: The three phases’ calibration can be proceeded simultaneously. The signal source is: Ua=Ub=Uc=Un, Ia=Ib=Ic=In(Ib). The calibration method is as below: a. Read voltage/current value of the external reference meter, and also read the voltage/current measurement value from chip registers; b. Calculate the voltage/current gain; Voltage Gain = Current Gain = reference voltage value Voltage measurement value x 32768 reference current value current measurement value x 32768 c. Write the result to the corresponding voltage/current gain registers Note: voltage/current gain calibration is not necessarily proceeded when gain register is the default value. That is, when the first calibration result is not ideal, there is no need to reset the gain register to the default value. Calibration can be performed again based on the current value. The formula is as below: New Voltage Gain = New Current Gain = reference voltage value voltage measurement value reference current value current measurement value x existing voltage gain x existing current gain The corresponding voltage/current gain register and measurement value registers are shown as below: 22 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 Gain Register Voltage Current Address Register Name Measurement Value Registers Address Register Name Address Register Name 61H UgainA 0D9H UrmsA 0E9H UrmsALSB 65H UgainB 0DAH UrmsB 0EAH UrmsBLSB 69H UgainC 0DBH UrmsC 0EBH UrmsCLSB 62H IgainA 0DDH IrmsA 0EDH IrmsALSB 66H IgainB 0DEH IrmsB 0EEH IrmsBLSB 6AH IgainC 0DFH IrmsC 0EFH IrmsCLSB Application Note: 1. Voltage rms is unsigned and the minimum unit 1LSB of the UrmsA/UrmsB/UrmsC registers is 0.01V. Only the higher 8 bits of the UrmsALSB/ UrmsBLSB/UrmsCLSB registers are valid, the lower 8 bits are always 0, and 1LSB is 0.01/ 256 V. 2. Current rms is unsigned and the minimum unit 1LSB of the IrmsA/IrmsB/IrmsC registers is 0.001A; Only the higher 8 bits of the IrmsALSB/ IrmsBLSB/IrmsCLSB registers are valid, the lower 8 bits are always 0, and 1LSB is 0.001/256 A. 3. Power is signed complement. Whole 32-bit register value should be read before caculation in power value caculation. 1LSB is 0.00032 (W/var/ VA). Example: Voltage gain calibration Assume: The initial value of phase A voltage gain register UgainA is 8000H (32768) Reference meter output voltage is 220.00V Voltage rms register readout UrmsA = 3039H (12345) The higher 8 bits of voltage LSB register readout UrmsALSB = 43H (67) Thus: voltage measured value = (UrmsA x 0.01) + (UrmsALSB x 0.01 / 256) = (12345 x 0.01) + (67 x 0.01 / 256) =123.453 V voltage gain = 220.00 / 123.453 x 32768 = 58395 = 0E41BH So the register can be set to: UgainA = 0E41BH M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 23 4.2.7 Energy Metering Calibration Only active energy is required for energy calibration. There is no need to calibrate reactive energy, the accuracy of which is guaranteed by chip design. Metering calibration flow is gain first then phase angle. Active energy pulse output (CF1) should be connected to the pulse input port of the calibration bench during calibration. Energy metering should be calibrated at In (Ib). 1. Power offset (Poffset/Qoffset) calibration For calibration method please refer to 4.2.5 Offset Calibration of Voltage/ Current/ Power. No need of calibration if the power offset is very small in general application. 2. Gain calibration Every phase’s gain calibration should be proceeded individually. Take phase A for example, the signal source is: Ua=Ub=Uc=Un, IA=In(Ib), Ib=Ic=0, PF=1.0. The calibration method is as below: a. Read the energy error value ε from calibration bench; b. Calculate the gain; Gain = Complement ary ( -ε x 215 ) 1+ ε c. Write the result to the corresponding gain registers. 3. Phase angle calibration. Take phase A for example, the signal source is: Ua=Ub=Uc=Un, Ia=In(Ib), Ib=Ic=0, PF=0.5L.The calibration method is as below: a. Read the energy error value εp from calibration bench; b. Calculate the phase angle error; AngleError = ε p × Gphase Gphase is a constant. When grid frequency is 50Hz, Gphase=3763.739. When grid frequency is 60Hz, Gphase=3136.449 c. Write the result to the corresponding phase angle error registers. The phase angle registers are signed and MSB of 1 indicates a negative value. The corresponding gain register and phase angle registers are shown as below: Address Register Name 47H GainA 48H PhiA Phase A Phase B Phase C 49H GainB 4AH PhiB 4BH GainC 4CH PhiC Example: Energy gain and phase angle calculation The condition is that power factor PF=1.0, current is Ib, energy error ε is -13.78%, so: -ε/(1+ε)=0.159823707, gain = int(0.159823707* 2^15)=5237.10=1475H Write 1475H to the gain register. After gain calibration, energy error εP is 0.95% in the condition of PF=0.5L, current is Ib and frequency is 50Hz, so: phase angle = εP*3763.739 =0.0095*3763.739=35.75553=24H; Write 24H to the phase angle register. 24 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 4.2.8 Fundamental Energy Metering Calibration For fundamental energy metering calibration, only gain and offset calibration is needed. There is no need to calibrate phase angle. Fundamental energy pulse output (CF3) should be connected to the pulse input port of the calibration bench during calibration. Only sigle-point calibration in In(Ib) is needed for fundamental energy metering calibration. Fundamental energy metering calibration is similar to energy metering calibration. 1. Fundamental power offset (PoffsetxF) calibration For calibration method please refer to 4.2.5 Offset Calibration of Voltage/ Current/ Power. No need of calibration if the power offset is very small in general application. 2. Fundamental energy gain calibration Every phase’s fundamental energy calibration should be proceeded individually. Take phase A for example, the signal source is: Ua=Ub=Uc=Un, Ia=In(Ib), Ib=Ic=0, PF=1.0. The calibration method is as below: a. Read the error value ε from the external reference meter; b. Calculate the gain; -ε x 215 ) Gain = Complementary ( 1+ ε c. Write the result to the corresponding gain registers. The corresponding fundamental energy gain registers are shown as below: Address Register Name 54H PGainAF 55H PGainBF 56H PGainCF Fundamental Energy Calibration Startup Fundamental Power Offset Calibration Fundamental Energy Gain Calibration End Figure-6 Fundamental Energy Metering Calibration Flow M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 25 4.3 Calibration in Partial Measurement Mode The same measurement modules are used in Partial measurement mode as Normal mode, so no need to do special calibration for Partial measurement mode. 4.4 Calibration in Detection Mode Current detection is realized by low power consumption comparators. The comparator outputs low level when the external current is lower than the configured threshold; The comparator outputs high level when the external current is higher than the configured threshold, as shown in Figure-7. Current Input Current Threshold IRQ Output DAC Figure-7 Current Detection Principle 4.4.1 Current Detection Module Configuration Current detection module can be enabled in Normal mode to facilitate the current detection threshold calibration. Six current threshold comparators can be configured for the current detection module to detect positive and negative current of three phases. These six threshold comparators can be enabled and disabled by the control bits, as shown below: Address 10H Register Name DetectCtrl Bit15 ~ Bit0 Bit15 Bit14 Bit13 Bit12 Bit11 Bit10 Bit9 Bit8 - - - - - - - - Bit7 Bit6 Bit5 Bit4 Bit3 Bit2 Bit1 Bit0 - DetCalEn PDN3 PDN2 PDN1 PDP3 PDP2 PDP1 1. DetCalEn: Enable detector calibration in Normal mode; 0: Detector disable (default) 1: Detectors enabled 2. PDN3/2/1: Control bits for negative detector of channel 3/2/1; 0: Detector enable (default) 1: Detector disable 3. PDP3/2/1: Control bits for positive detector of channel 3/2/1; 0: Detector enable (default) 1: Detector disable Each of the six current threshold comparators has its own register configuration as shown below: Address Register Name 11H 12H 13H DetectTh1 DetectTh2 DetectTh3 Bit15 ~ Bit0 Bit15 Bit14 Bit13 Bit12 Bit11 Bit10 Bit9 Bit8 Bit3 Bit2 Bit1 Bit0 CalCodeN Bit7 Bit6 Bit5 Bit4 CalCodeP 1. CalCodeN: negative detector threshold, 8-bit width. 8’b0000-0000 corresponds to minimum threshold Vc=-1.2mV=-0.85mVrms 8’b1111-1111 corresponds to maximum threshold Vc=9mV=6.35mVrms 2. CalcodeP: positive detector threshold, definition is the same as CalCodeN. 26 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 4.4.2 Current Detection Threshold Calibration Because of the low power consumption consideration and the manufacturing process, the current detection threshold is different from different chips. Therefore, calibration is needed due to the offset of chip’s DAC output (less than ±5mVrms). The threshold current range is 2mVrms ~ 4mVrms within which the current detection module (low power consumption comparator) can detect accurately. It is recommended to proceed system design according to current detection threshold of 3mVrms. Dichotomy is suggested in current detection threshold calibration. The recommended calibration flow is as shown in Figure8. Reference source outputs current signal that needs detection (such as 5%Ib) Th_max = 100H Th_min = 00H Th_temp = 80H Th_min = Th_temp Th_temp = (Th_max – Th_min) / 2 Th_max, Th_min, Th_temp: variable DetectThx: Threshold register Th_max = Th_temp Th_temp = (Th_max – Th_min) / 2 DetectThx = Th_temp N N Th_max – Th_temp == 1 ? Y IRQ0 output high level? N Th_temp – Th_min == 1 ? Y Y Current detection threshold = Th_temp Current detection threshold = Th_min End Figure-8 Current Detection Threshold Calibration Flow M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 27 5 FUNCTION REGISTERS CONFIGURATION 5.1 Startup Current Configuration The registers which related to startup current configuration is shown as below: Address Register Name Description 35H PStartTh All-phase Active Startup Power Threshold. 36H QStartTh All-phase Reactive Startup Power Threshold. 37H SStartTh All-phase Apparent Startup Power Threshold. 38H PPhaseTh Each-phase Active Startup Power Threshold. 39H QPhaseTh Each-phase Reactive Startup Power Threshold. 3AH SPhaseTh Each-phase Apparent Startup Power Threshold. 1. Due to system interference when current is 0, small signal may be gernerated in current sampling channel, producing a certain amount of energy and affecting the measurement and metering accuracy. To avoid this, the M90E32AS provides the each-phase startup power configuration/judgment function. 2. PPhaseTh, QPhaseTh and SPhaseTh are used to judge the startup power of each phase (A/B/C). Take active power for example, when a single phase input power is smaller than the configured PPhaseTh value, the input active power of that phase will be set to 0 by force, that means input to the next process is 0. Otherwise the signal will be streamed to the next process. Note that the threshold are configured separately to active(P), reactive(Q) and apparent (S). The compared value is (|P|+|Q|). 3. PStartTh, QStartTh and SStartTh are used to judge all-phase startup power. Take active power for example, when allphase-sum power is less than the configured PStartTh value, energy accumulation will not start. Otherwise energy accumulation will start. 4. Calculation methods of the two register groups are the same. The formula is as below: Register value = N / 0.00032, (N is the configured power threshold). Example: Startup Current Configuration Assume: meter voltage is 220V, current specification 5(100)A, active startup current is 0.1% considering the accuracy of current measurement in small-current state, it is recommended to configure the all-phase startup current threshold to be 50% of startup current (also can configure based the actual conditions). Assume the startup threshold of each-phase power is configured to be 10% of startup current. so: All-phase Active Startup Power Threshold = 3 x 5 x 0.1% x 50% x 220 = 1.65W Each-phase Active Startup Power Threshold = 5 x 0.1% x 10% x 220 = 0.11W register values are: PStartTh[35H] = 1.65 / 0.00032 = 5156 = 1424H PPhaseTh[38H] = 0.11 / 0.00032 = 344 = 158H 28 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 Power Threshold |P|+|Q|> PPhaseTh? A/B/C Phase Active Power from DSP 3 phases Total Active Energy Metering ABS > PStartTh? + 1 Phase Active Energy Metering 0 0 Power Threshold |P|+|Q|> QPhaseTh? Phase Reactive Power from DSP 0 0 Total Reactive Power 3 phases 0 1 Total Reactive Energy Metering ABS > QStartTh? + 1 1 0 Power Threshold |P|+|Q|> SPhaseTh? 0 0 Total Apparent Power 3 phases Phase Apparent Power from DSP Phase Reactive Energy Metering 0 0 A/B/C 0 1 1 0 A/B/C 0 Total Active Power ABS > SStartTh? + 0 1 Total (arithmetic sum) Apparent Energy Metering 1 1 0 Phase Apparent Energy Metering 0 0 0 Figure-9 Metering Startup Handling 5.2 Sag Function Sag detect function is provided in M90E32AS. The threshold of sag detection is configured through the SagTh register (08H). All three voltage phases use the same threshold. The threshold equation is as below: SagTh = Vth × 100 × 2 2 × Ugain / 32768 Vth: the voltage threshold to be configured; Ugain: the gain after calibration M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 29 6 COMPENSATION METHOD M90E32AS provids excellent metering accuracy over a dynamic range of 6000:1 and industrial temperature range. But in application of power meter, system metering accuracy would be infulenced by the performance difference of peripheral devices. So M90E32AS provides compensation function for system metering error considering the system application of power meter. M90E32AS provides three compensation functions based on different correspondences: 1. Current besed Compensation (per phase), it goes to Gain and Phi; 2. Frequency based compensation(all phases are the same), it goes to Phi; 3. Temperature based compensation (per phase), it goes to UGain 6.1 Current Segment Compensation Descriprion Delta-Gain (MeterErr) Delta-Gain Irms1 GainIrms2 30 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 Irms0 GainIrms1 Log(Irms) GainIrms0 Delta-Phi (MeterErr) Delta-Gain Irms1 Irms0 Log(Irms) Delta-Phi PhiIrms2 PhiIrms1 PhiIrms0 1. Irms0 and Irms1 are the current points of segments, through which the whole current range can be divided to three segments in calibration; gain and phi use the same current points in segment compensation; 2. GainIrms0, GainIrms1 and GainIrms2 are gain compensation coefficients for every current segment; Log(Irms) is the X axis, and delta-gain value when PF=1.0 is the Y axis value; 3. PhiIrms0, PhiIrms1 and PhiIrms2 are phi compensation coefficients for every current segment; Log(Irms) is the X axis, and delta- phi value when PF=1.0 is the Y axis value. Note: Log(X)=Log2(X)*6 in the following description, for example Log(2)=16 and Log(16)=64. M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 31 6.2 GainIrms Segment Compensation Example Per-phase compensation, take phase-A for example: MeterErr GainErr = 0.3% 5A GainIrms0 Irms0 0.5A Irms1 0.05A GainIrms1 50A Log(Irms) GainIrms2 GainErr = -0.5% 1. Irms0=5A register value is 5000, Log(Irms0) = Log2(5000)*16 = 197 (=0xC5), that is the LOGIrms0[0x20]register value should be set to 0xC5; 2. Irms1=0.5A register value is 500, Log(Irms1) = Log2(500)*16 = 143 (=0x8F), that is the LOGIrms1[0x21] register value should be 0x8F; 3. GainIrms0 X axis difference ΔX is: Log(50000) - Log(5000) = Log(50000/5000) = Log(10) = Log2(10)*16 = 53 Y axis difference ΔY is: 0.3% - 0 = 0.003, 19 Then Compensati onValue = −2 × ΔY 0.003 = −30 (Note: 219 is constant) = − 219 × ΔX 53 that is IGainAIrms01[0x26, Bit7:0] =-30 (=0xE2) (Note: complement); 4. GainIrms1 no compensation that is IGainAIrms01[0x26, Bit15:8]=0; 5. GainIrms2 difference ΔX is: Log(500) - Log(50) = Log(500/50) = Log(10) = Log2(10)*16 = 53, difference ΔY is: 0 - (-0.5%) = 0.005, 19 Then CompensationValue = −2 × ΔY 0.005 = − 219 × = −49 (Note: 219 is constant) 53 ΔX that is IGainAIrms2[0x27] =-49 (=0xCF)(Note: complement) 32 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 6.3 PhiIrms Segment Compensation Example Per-phase compensation, take phase-A for example: MeterErr MeterErr = 0.3% 0.5A Irms1 5A Irms0 0.05A PhiIrms0 50A PhiIrms1 Log(Irms) MeterErr = -0.2% PhiIrms2 MeterErr = -0.7% 1. Irms0=5A register value is 5000, Log(Irms0) = Log2(5000)*16 = 197 (=0xC5) that is LOGIrms0[0x20] register value should be set to 0xC5 2. Irms1=0.5A register vaue is 500, Log(Irms1) = Log2(500)*16 = 143 (=0x8F) that is LOGIrms1[0x21] register value should be set to 0x8F 3. PhiIrms0 X axis difference ΔX is: Log(50000) - Log(5000) = Log(50000/5000) = Log(10) = Log2(10)*16 = 53, Y axis differece ΔY:0.3% - 0 = 0.003, ΔY 0.003 8 8 then Compensati onValue = 3764 × 2 × ΔX = 3764 × 2 × 53 = 55 (Note: 3764 × 2 8 is constant), that is PhiAIrms01[0x24] Bit7:0=55 (=0x37) (Note: complement); 4. PhiIrms1 X axis differenceΔ is: Log(5000) - Log(500) = Log(5000/500) = Log(10) = Log2(10)*16 = 53 Y axis difference ΔY: 0 - (-0.2%) = 0.002, ΔY 0.002 8 8 then Compensati onValue = 3764 × 2 × ΔX = 3764 × 2 × 53 = 36 (Note: 3764 × 2 8 is constant), that is PhiAIrms01[0x24] Bit15:8=36 (=0x24) (Note: complement); 5. PhiIrms2 X axis difference ΔX is: Log(500) - Log(50) = Log(500/50) = Log(10) = Log2(10)*16 = 53, Y axix difference ΔY: (-0.2%) - (-0.7%) = 0.005, ΔY 0.005 8 8 then CompensationValue = 3764 × 2 × ΔX = 3764 × 2 × 53 = 91 (Note: 3764 × 2 8 is constant), that is PhiAIrms2[0x25] Bit7:0=91 (=0x5A) (Note: complement). M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 33 6.4 Frequency-Phase Compensation (PhiFreqComp) Example (all phases use the same compensation): 1. Assume reference frequency is 50Hz, that means the F0[0x22] register value is set to 5000 (=0x1388); 2. Assume offset is 0% when 50Hz(PF=0.5L) and 0.1% when 52Hz(PF=0.5L), 9 then Compensati onValue = -3764 × 2 × 0.1% - 0% = - 10 (Note: 3764 × 2 9 is constant), 5200 - 5000 That is PhiFreqComp[0x1C]=-10 (=0xF6)(Note: complement) 34 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 7 TEMPERATURE COMPENSATION The M90E32AS itself embodies good temperature characteristic. Considering that the external components might be affected by temperature in application, the M90E32AS also provides compensation function for external temperature drift. A series of special registers should be configured for temperature compensation. These registers are located in special addresses, and access to these registers should be strictly carried out as the following. 7.1 On-chip Temperature Sensor Configuration The M90E32AS provides a built-in temperature sensor. Due to the manufacturing process, the temperature sensor might be somewhat different for different chips. Therefore the on-chip temperature sensor should be configured before temperature compensation. The configuration method is as below: 1. Write 55AAH to address 2FFH 2. Write 5183H to address 216H 3. Write 01C1H to address 219H 4. Write 0000H to address 2FFH Read the Temp[0FCH] register directly to get the current temperature after configuration completed. Please note that, the temperature sensor will sense the temperature of the chip and it may have a few degrees of difference between the chip junction temperature and ambient temperature. M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 35 7.2 Temperature Compensation Based on Ugaint Example Test data before temperature compensation is as below: After linearization: The reference temperature (temperature in calibrating) is 25 ℃ , which means setting the T0[0x23] register value to 25 (=0x19) (complement), metering error is 0.0000%; The error at 85 ℃ point is 0.06% So the temperature coefficient is calculated as below: then Compensati onValue = -2 20 × 0.6% - 0% = - 105 85 - 25 (Note: 220 is constant), That is UGainTA[0x1A,Bit15:8]=-10 (=0xF6)(note: complement) 36 M90E32AS [Application Note] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 7.3 Temperature Compensation Based on Reference Voltage On-chip high-precision reference voltage is provided with excellent low temperature coefficient. But in application, what should be considered is the temperature drift of the whole system. Therefore, the M90E32AS specially provides temperature compensation based on reference voltage to minimize temperature drift caused by the on-board components. Note that, as voltage and current ADC sampling adopt the same reference voltage, compensation on the reference voltage will bring double effect on power and energy. Temperature compensation on reference voltage is proceeded with every 8 ℃ as a segment. In application, it is suggested to test on a small batch of components from the same lot to get the best temperature compensation coefficient. And then use this compensation coefficient as a fixed value to be written directly to register in production. The temperature compensation method is as below: 1. Write 55AAH to address 2FFH 2. Write the reference voltage coefficient of segment compensation to addresses 202H~209H 3. Write the curvature of segment compensation curve to address 201H 4. Write 0000H to address 2FFH In normal condition, reference voltage is 1200mV. The unit of reference voltage compensation is 0.020mV in this compensation method. The default value of the corresponding compensation registers is the ideal value in chip design. In application, only incremental adjustment is needed. Address Register Name 201H BGCurveK 202H BG_TEMP_P12 203H BG_TEMP_P34 204H BG_TEMP_P56 205H BG_TEMP_P78 206H BG_TEMP_N12 207H BG_TEMP_N34 208H BG_TEMP_N56 209H BG_TEMP_N78 Bit Read/Write Default Value Description 15:4 - - Reserved bit, readout value is 0 3:0 Read/Write 0 Reference voltage temperature compensation curve. The bit3 is assumed as sign bit, range is -8 to +7. Scaling factor = 1+ register value*1/8. So the scaling factor will be from 0, with step of 1/8, all the way to 1+7/8. 15:8 Read/Write 0 Compensation coefficient on the 25 ℃ temperature point 7:0 Read/Write 1 Compensation coefficient on the 17 ℃ temperature point 15:8 Read/Write 6 Compensation coefficient on the 41 ℃ temperature point 7:0 Read/Write 2 Compensation coefficient on the 33 ℃ temperature point 15:8 Read/Write 25 Compensation coefficient on the 57 ℃ temperature point 7:0 Read/Write 13 Compensation coefficient on the 49 ℃ temperature point 15:8 Read/Write 53 Compensation coefficient on the 73 ℃ temperature point 7:0 Read/Write 39 Compensation coefficient on the 65 ℃ temperature point 15:8 Read/Write 16 Compensation coefficient on the 1 ℃ temperature point 7:0 Read/Write 6 Compensation coefficient on the 9 ℃ temperature point 15:8 Read/Write 54 Compensation coefficient on the -15 ℃ temperature point 7:0 Read/Write 34 Compensation coefficient on the -7 ℃ temperature point 15:8 Read/Write 117 Compensation coefficient on the -31 ℃ temperature point 7:0 Read/Write 79 Compensation coefficient on the -23 ℃ temperature point 15:8 Read/Write 205 Compensation coefficient on the -47 ℃ temperature point 7:0 Read/Write 159 Compensation coefficient on the -39 ℃ temperature point M90E32AS [APPLICATION NOTE] Atmel-46103A-SE-M90E32AS-ApplicationNote_050514 37 Revision History Doc. Rev. Date 46103A 5/5/2014 Comments Initial release. X X X X Atmel Corporation 1600 Technology Drive, San Jose, CA 95110 USA T: (+1)(408) 441.0311 F: (+1)(408) 436.4200 | www.atmel.com © 2014 Atmel Corporation. All rights reserved. / Rev.: Atmel-46103A-SE-M90E32AS-ApplicationNote_050514. Atmel®, Atmel logo and combinations thereof, Enabling Unlimited Possibilities®, and others are registered trademarks or trademarks of Atmel Corporation or its subsidiaries. Other terms and product names may be trademarks of others. DISCLAIMER: The information in this document is provided in connection with Atmel products. No license, express or implied, by estoppel or otherwise, to any intellectual property right is granted by this document or in connection with the sale of Atmel products. EXCEPT AS SET FORTH IN THE ATMEL TERMS AND CONDITIONS OF SALES LOCATED ON THE ATMEL WEBSITE, ATMEL ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY WARRANTY RELATING TO ITS PRODUCTS INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT. IN NO EVENT SHALL ATMEL BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDENTAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS AND PROFITS, BUSINESS INTERRUPTION, OR LOSS OF INFORMATION) ARISING OUT OF THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF ATMEL HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Atmel makes no representations or warranties with respect to the accuracy or completeness of the contents of this document and reserves the right to make changes to specifications and products descriptions at any time without notice. Atmel does not make any commitment to update the information contained herein. Unless specifically provided otherwise, Atmel products are not suitable for, and shall not be used in, automotive applications. Atmel products are not intended, authorized, or warranted for use as components in applications intended to support or sustain life. SAFETY-CRITICAL, MILITARY, AND AUTOMOTIVE APPLICATIONS DISCLAIMER: Atmel products are not designed for and will not be used in connection with any applications where the failure of such products would reasonably be expected to result in significant personal injury or death (“Safety-Critical Applications”) without an Atmel officer's specific written consent. Safety-Critical Applications include, without limitation, life support devices and systems, equipment or systems for the operation of nuclear facilities and weapons systems. Atmel products are not designed nor intended for use in military or aerospace applications or environments unless specifically designated by Atmel as military-grade. Atmel products are not designed nor intended for use in automotive applications unless specifically designated by Atmel as automotive-grade.