Martin Mienkina. System Application Engineer

Transcription

1 Martin Mienkina System Application Engineer June 2012 Freescale, the Freescale logo, AltiVec, C-5, CodeTEST, CodeWarrior, ColdFire, ColdFire+, C-Ware, the Energy Efficient Solutions logo, Kinetis, mobilegt, PowerQUICC, Processor Expert, QorIQ, Qorivva, StarCore, Symphony and VortiQa are trademarks of Freescale Semiconductor, Inc., Reg. U.S. Pat. & Tm. Off. Airfast, BeeKit, BeeStack, CoreNet, Flexis, MagniV, MXC, Platform in a Package, QorIQ Qonverge, QUICC Engine, Ready Play, SafeAssure, the SafeAssure logo, SMAROS, TurboLink, Vybrid and Xtrinsic are trademarks of Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners Freescale Semiconductor, Inc..

2 Understand electricity meter metrology requirements Become familiar with Freescale metering specific reference designs and algorithm offerings Tutorials covered in this presentation: Rogowski Coil Digital Integrator Explicit RMS Converter MK30 ADC16-PGA Measurement Chain 2

3 Introduction to Energy Meters Energy Metering Tutorial International Standards Current Sensor Interfaces Energy Calculation An Effective Metering System on Chip Based Solutions Enablement Summary The first specimen of the AC kilowatt-hour meter produced on the basis of Hungarian Ottó Bláthy's patent and named after him was presented by the Ganz Works at the Frankfurt Fair in the autumn of 1889, and the first induction kilowatt-hour meter was already marketed by the factory at the end of the same year. These were the first alternating-current watt meters, known by the name of Bláthy-meters. Source: 3

4 Electricity Meter Types Electromechanical meters Limited accuracy Manual reading Contains moving parts (aluminum ring) Electronic meters MCUs, DSPs and ASICs based Accurate measurement Enhanced security Equipped with AMR No moving parts 4 Measured Quantities Active, reactive, apparent energy Active, reactive, apparent power RMS, peak values (voltage/current) Line frequency Power factor Temperature Measurement Types Single phase Common in EU residential meters One voltage and one current measurement Use of shunt resistors prevail due to low system cost Dual phase Common in US residential meters Two voltage and two current measurement Use of current transformers and Rogowski coils prevail Three phase Used in industrial and commercial meters Three voltage and three current measurement Use of current transformers and Rogowski coils prevail

5 IEC/EN compliant meters are typically quoted as class p: n(m)a Where n is the basic current m is the maximum current p is the accuracy class A(2%), B(1%), C(0.5%) i.e. Class C: 5(60)A Standardized values Voltage (V) Frequency rating 50 Hz Basic current (A) Maximum current - multiples of basic current Active power P [W], reactive power Q [VAR] and apparent power S [VA] should be measured Factors impacting accuracy Mains frequency Load power factor Harmonics in voltage and current Temperature DC current Limited power consumption in current circuit IEC/EN meter types dominate in Europe, Turkey, Russia, India and China. 5

6 Dynamic accuracy required Common static power meter Class B: (5)60A requirements Class Ireference [A] Imaximum [A] Itransitional [A] Itr = Iref/10 I minimum [A] Imin=0.5*Itr Accuracy [%] <imin; Itr> Accuracy [%] <Itr; Imax> A ,5 0,25 2,5 2 B ,5 0,25 1,5 1 C ,5 0,15 1 0,5 Dynamic range out of table Imax : Imin = 60 : 0,25 accuracy required for class B meter 1,5% (err) ADC requirements: e 240 : 1 (DR) ENOB ln DR / err ln(2) ln 240/ ln(2) 13,96 1 Imin Itr Imax 6

7 ANSI C12.20 compliant meters are typically quoted as Class p%: r(m) Where r is the reference current m is the maximum current p is the accuracy class 0.5%, 0.2% i.e. Class 0.2: 30(200)A Standardized values Voltage (V) Frequency rating 60 Hz Current classes (reference amperes) 2(0.25)-10(2.5)-20(2.5)-100(15)-200(30)- 320(50) (A) Starting current classes 0.001, 0.01, 0.01, 0.05, 0.1, 0.16 (A) Active power P [W], reactive power Q [VAR] and apparent power S [VA] should be measured Factors impacting accuracy ANSI C12.20 meters are used in Mains frequency US, Canada, Brazil and Mexico. Load power factor Harmonics in voltage and current Temperature DC current Limited power consumption in current circuit 7

8 Dynamic accuracy required ANSI C12.20 power meter Class 0.2: (30)200 requirements Dynamic range out of table Imax : Imin@ref = 200 : : 1 (DR) accuracy required for class 0.2 meter 0.2% (err) ADC requirements: Current in Amperes Maximum error deviation [%] Current Class Accuracy class Conditions Starting current (1) Imin 1 2 ±1.0 ±0.4 (2) Imin@ref ±0.5 ±0.2 (3) 3 6 ±0.5 ±0.2 (4) ±0.5 ±0.2 (5) ±0.5 ±0.2 (6) ±0.5 ±0.2 (7) ±0.5 ±0.2 (8) ±0.5 ±0.2 (9) ±0.5 ±0.2 (10) Imax ±0.5 ±0.2 e ENOB ln DR / err ln(2) ln 66.6/ ln(2) Imin Imin@ref Imax 8

9 Introduction to Energy Meters Energy Metering Tutorial International Standards Current Sensor Interfaces Energy Calculation An Effective Metering System on Chip Based Solutions Enablement Summary 9

10 L_OUT Isense Rsense L_INP N_INP N_OUT Voltage divider Mathematical description: I 110/220V Power supply SENSE R R VDD Vref Vadc - 2 Gain RSENSE R R Anti-aliasing Filters (-6dB per octave) Rf Rf Cf Cf Vref Gain Vadc ADC PDB Pros: Commonly used Simple to design Inexpensive No magnetic effects Cons: Self-heating due to power dissipation Parasitic inductance introduces phase shift at low power factors Non-isolated I SENSE (A) R SENSE ( ) U SENSE (V P-P ) uV uV 60* 42.4mV * Power losses 60A Voltage drop across shunt resistor is proportional to the amplitude of the current and frequency. 10

11 L_OUT Isense L_INP N_INP N_OUT Voltage divider 110/220V Power supply Turn Ratio=1:N CT 1 N VDD Rb Anti-aliasing Filters (-6dB per octave) Rf Rf Vref R R Cf Cf Gain Vadc ADC PDB Pros: Provides electrical isolation Current in secondary is proportional to current in primary Preferred for poly-phase meters Output voltage scaled to ADC input signal range Cons: CT introduces phase error from 0.1º to 7.0º Phase shift depending on current and temperature Load must never be disconnected from secondary winding Iron core can saturate at current level beyond its rated current or at a large DC Sensitive to magnetic tampering Expensive Mathematical description: ISENSE Vref (Vadc - ) 2 Gain * Rb * N Courtesy of VACUUMSCHMELZE I SENSE (A) Ratio U SENSE (V P-P ) uV :2500 Rb= mV V 11

12 L_OUT I RS L_INP N_INP N_OUT Mathematical description: V Voltage divider R K R * Fr * I R Vadc where: 110/220V Power supply RC Vref 2 V RS VDD GAIN * V R Vref I R = rated primary current Fr = frequency of sinusoidal waveform K R = rotated transformer constant Rf Rf R R Cf Cf Anti-aliasing Filters (-6dB per octave) Gain Vadc ADC PDB Pros: Provides electrical isolation Capable of handling high current Low temperature drift Linear phase response No DC or high current saturation Immune to magnetic tampering Cons: Integration adds to extra circuitry (software load) Interference (far field) pickup - limited by design or shielding Courtesy of PULSE ELECTRONICS I SENSE (A) F r ( /Hz) U SENSE (V P-P ) uV uV (PA3202NL) mV The output voltage of the Rogowski coil is proportional to the timedifferentiation (di/dt) of the current. 12

13 Magnitude error Phase error (degrees) Magnitude response (db) Phase (degrees) Digital Integrator Block Diagram: Frequency & Phase Characteristics*: Hz Explanation: HPF1: The first high pass filter to prevent digital integrator overflow (used to remove offset and low frequency drift from input signal) Integrator: The numerical integrator (use of the Bilinear approximation is generally preferred over forward/backward Euler methods due to zero group delay at higher frequencies > 100 Hz) HPF2: The second high pass filter required to remove offset from the integrator output signal x Hz 1.3Hz 0.9Hz Hz Frequency (Hz) 0.5Hz 0 0.1Hz Frequency (Hz) Hz 0.9Hz 0.5Hz 0.1Hz 0.1Hz 0.5Hz 0.9Hz 1.3Hz Frequency (Hz) 1.7Hz Frequency (Hz) * Characteristics represents numerical integrator build of 1st order high pass. Butterworth filters designed and analog filter designed for fcut = 0.1, 0.5, 0.9, 1.3 and 1.7 Hz and sampling rate fs=3000 Hz. 13

14 Summary Small amplitude errors for filters cut off frequencies below 1.7Hz Phase error increases rapidly at higher filter cut-off frequencies (above 0.1Hz) The lower the filter cut-off frequency, the more sluggish the algorithm startup (initial offset removal) HPF cut-off frequency (Hz) Amplitude error (%)* Phase error (deg)* Energy UPF (%) Energy PF=0.5 (%) ** ** ** ** ** * Amplitude and phase error determined for 50Hz input waveform. ** Fine compensation using delaying ADC conversion is possible and widely used. 14

15 Introduction to Energy Meters Energy Metering Tutorial International Standards Current Sensor Interfaces Energy Calculation An Effective Metering System on Chip Based Solutions Enablement Summary 15

16 Active Energy Term Definition: The electrical energy produced, flowing or supplied by an electric circuit during a time interval, being the integral with respect to time of instantaneous Active Power, measured in units of Watt-hours or standard multiples thereof. Equation: W 0 u( t) i( t) dt Reactive Energy Term Definition: The integral with respect to time of the product of voltage and current and the sine of the phase angle between them. Its is measured in units of volt-amperes reactive (VAR) and standard multiples thereof. Equation: VAR 0 u( t 90 ) i( t) dt RMS Voltage & Current Term Definition: In electrical engineering the root mean square (RMS) or effective value of a current is by definition such that the heating effect is the same for equal values of alternating or direct-current. Equation: rms 1 T T 0 2 value ( t) dt Active Power Term Definition: The product of voltage and the in-phase component of alternating current measured in units of watts and standard multiples thereof, Reactive Power Term Definition: Means the product of voltage and current and the sine of the phase angle between them, measured in units of volt-amperes reactive 16 that is: 1000 Watts = 1 kw, 1000 kw = 1 MW T Equation: (VAR) and standard 1 multiples thereof. T Equation: Q T 0 u( t 90 ) i( t) dt VAR Q* T Apparent Power Term Definition: In alternating-current power transmission and distribution apparent power is the product of the RMS voltage and amperage. Equation: P S S 1 T 0 u RMS i RMS Q u( t) i( t) dt 2 P W P* T u t S P Q

17 Term Definition: The electrical energy produced, flowing or supplied by an electric circuit during a time interval, being the integral with respect to time of instantaneous Active Power, measured in units of Watt-hours or standard multiples thereof. Equation: W u( t) i( t) dt 0 Processing/Calculation Steps: 1. Measure phase voltage u(t) and current i(t) samples 2. Remove offset from phase voltage measurements u(t) 3. Integrate product of phase voltage u(t) and current i(t) samples Model: Example: u(t)=urms*sqrt(2)*sin(2*pi*freq*t+alpha) i(t)=irms*sqrt(2)*sin(2*pi*freq*t+0) URMS=230V, IRMS=10A, freq=50hz, alpha=0deg. W=? Matlab Calculus: Ud = sym('u_max*sin(2*pi*f*t+alpha)') Id = sym('i_max*sin(2*pi*f*t+0)') W = int(ud*id,0,'t') subs(w,{'u_max','i_max','f','t','alpha'},{230*sqrt(2),10*sqrt(2),50,0.02,0})/3600 W = [watt-hours] Model Simulation: Active Energy Block accumulates phase voltage and phase current multiple. 17

18 Term Definition: Means the integral with respect to time of the product of voltage and current and the sine of the phase angle between them. Its is measured in units of voltamperes reactive (VAR) and standard multiples thereof. Equation: VAR 0 Processing/Calculation Steps: u( t 90 ) i( t) dt 1. Measure phase voltage u(t) and current i(t) samples 2. Remove offset from phase voltage measurements u(t) 3. Shift phase voltage u(t) by 90º. 4. Integrate product of shifted phase voltage u(t-90º) and current i(t) samples Model: Example: u(t)=urms*sqrt(2)*sin(2*pi*freq*t+alpha) i(t)=irms*sqrt(2)*sin(2*pi*freq*t+0) URMS=230V, IRMS=10A, freq=50hz, alpha=90deg. VAR=? Matlab Calculus: Ud = sym('u_max*sin(2*pi*f*t+alpha/180*pi)') Id = sym('i_max*sin(2*pi*f*t+0)') VAR = int(diff(ud/sym('-2*pi*f'))*id,0,'t') subs(var,{'u_max','i_max','f','t','alpha'},{230*sqrt(2),10*sqrt(2),50,0.2,90})/3600 VAR = [VAR-hours] Model Simulation: Reactive Energy Block accumulates phase voltage and phase current multiple. The phase voltage is shifted by 90º using 90 degree phase shift block. 18

19 Term Definition: In electrical engineering the root mean square (RMS) or effective value of a current is by definition such that the heating effect is the same for equal values of alternating or direct-current. In mathematics, the RMS is known as the quadratic mean, is a statistical measure of the magnitude of a varying quantity. It is especially useful when variations are positive and negative, e.g., sinusoids. Equation: rms 1 T T 0 2 value ( t) dt Processing/Calculation Steps: 1. Measure phase voltage u(t) and current i(t) samples 2. Remove offset from phase voltage measurements u(t) 3. Remove offset from phase current measurements i(t) 4. Calculate RMS values of the phase voltage URMS(t) and current IRMS(t). Model: Example: u(t)=urms*sqrt(2)*sin(2*pi*freq*t+alpha)+offset i(t)=irms*sqrt(2)*sin(2*pi*freq*t+0) IRMS=10A, URMS=230V, freq=50hz, alpha=0deg, offset=0v. Matlab Calculus: Ud = sym('u_max*sin(2*pi*f*t+alpha/180*pi)+offset') Id = sym('i_max*sin(2*pi*f*t+0)') UdRMS=sympow(symmul('1/T', int(ud*ud,0,'t')),1/2) IdRMS=sympow(symmul('1/T', int(id*id,0,'t')),1/2) subs(udrms,{'u_max','f','t','alpha','offset'},{230*sqrt(2),50,0.1,0,0}) UdRMS= [V] subs(idrms,{'i_max','f','t'},{10*sqrt(2),50,0.1}) IdRMS= 10.0 [A] Model Simulation: RMS Voltage & Current Averager are based on the Explicit RMS Converter *, therefore their output response is smoothed due to implied low pass filtering (IIR filter, first order, fcut=2hz). 19

20 Term Definition: The product of voltage and the in-phase component of alternating current measured in units of watts and standard multiples thereof, that is: 1000 Watts = 1 kw, 1000 kw = 1 MW Equation: P 1 T T 0 u( t) i( t) dt W P* T Processing/Calculation Steps: 1. Measure phase voltage u(t) and current i(t) samples 2. Remove offset from phase voltage measurements u(t) 3. Accumulate instantaneous product of phase voltage u(t) and current i(t) per fixed time period. 4. Divide accumulated sum by time period if not equal to 1s Model: Example: u(t)=urms*sqrt(2)*sin(2*pi*freq*t+alpha) i(t)=irms*sqrt(2)*sin(2*pi*freq*t+0) URMS=230V, IRMS=10A, freq=50hz, alpha=0deg. P=? Matlab Calculus: Ud = sym('u_max*sin(2*pi*f*t+alpha)') Id = sym('i_max*sin(2*pi*f*t+0)') P = symmul('1/t',int(symmul(ud,id),0,'t')) subs(p,{'u_max','i_max','f','t','alpha'},{230*sqrt(2),10*sqrt(2),50,1.0,0}) P = [watts] Model Simulation: Active Power Averager is also based on the Explicit RMS Converter*, therefore their output response is smoothed due to implied low pass filtering (IIR filter, first order, fcut=2hz). 20

21 Term Definition: The product of voltage and current and the sine of the phase angle between them, measured in units of voltamperes reactive (VAR) and standard multiples thereof. 1 T Equation: Q T u( t 90 ) i( t) dt 0 Processing/Calculation Steps: VAR Q* T 1. Measure phase voltage u(t) and current i(t) samples 2. Remove offset from phase voltage measurements u(t) 3. Shift phase voltage u(t) by 90º 4. Accumulate instantaneous products of shifted phase voltage u(t-90º) and current i(t) per fixed time period 5. Divide accumulated sum by time period if period doesn t equal 1s Model: Example: u(t)=urms*sqrt(2)*sin(2*pi*freq*t+alpha) i(t)=irms*sqrt(2)*sin(2*pi*freq*t+0) URMS=230V, IRMS=10A, freq=50hz, alpha=90deg. Q=? Matlab Calculus: Ud = sym('u_max*sin(2*pi*f*t+alpha/180*pi)') Id = sym('i_max*sin(2*pi*f*t+0)') Q = symmul('1/t',int(diff(ud/sym('-2*pi*f'))*id,0,'t')) subs(q,{'u_max','i_max','f','t','alpha'},{230*sqrt(2),10*sqrt(2),50,1.0,90})ud = sym('u_max*sin(2*pi*f*t+alpha)') Q = [VAR] Model Simulation: Reactive Power Averager is also based on the Explicit RMS Converter*, therefore their output response is smoothed due to implied low pass filtering (IIR filter, first order, fcut=2hz). The phase voltage is shifted by 90º using 90 degree phase shift block. 21

22 Term Definition: In alternating-current power transmission and distribution, apparent power is the product of the RMS voltage and amperage. Equation: S S u RMS i RMS Q 2 P 2 u t Processing/Calculation Steps: P 1. Measure phase voltage u(t) and current i(t) samples 2. Remove offset from phase voltage measurements u(t) 3. Remove offset from phase current measurements i(t) 4. Calculate RMS values of the phase voltage URMS(t) and current IRMS(t) 5. Calculate apparent power by multiplying URMS(t)*IRMS(t) Model: S Q Example: u(t)=urms*sqrt(2)*sin(2*pi*freq*t+alpha) i(t)=irms*sqrt(2)*sin(2*pi*freq*t+0) URMS=230V, IRMS=10A, freq=50hz, alpha=45deg. S=? Matlab Calculus: Ud = sym('u_max*sin(2*pi*f*t+alpha/180*pi)') Id = sym('i_max*sin(2*pi*f*t+0)') UdRMS=sympow(symmul('1/T', int(ud*ud,0,'t')),1/2) IdRMS=sympow(symmul('1/T', int(id*id,0,'t')),1/2) S=UdRMS*IdRMS subs(s,{'u_max','i_max','f','t','alpha'},{230*sqrt(2),10*sqrt(2),50,1.0,45}) S = [VA] Model Simulation: Apparent Power Block calculates apparent power using arithmetic approach from the RMS voltage and current samples. Performance of this block is therefore dependent on the performances of the RMS voltage and RMS current calculations. The RMS voltage and RMS current are based on the Explicit RMS Converters, therefore their output response is smoothed by low-pass filtering (IIR filter, first order, fcut=2hz). As URMS and IRMS contain all harmonic information, the apparent power computed by arithmetic approach is a Total Apparent Power. 22

23 u 2 u sqrt(avg(u 2 )) sqrt(avg(u sqrt(avg(u 2 )) 2 )) avg(u 2 ) avg(u 2 ) avg(u 2 ) u u PGA Input Signal Range: LPF i i 2 avg(i 2 ) u X X 2 X 2 Equation: LPF LPF u 2 avg(u 2 ) Input Signal: Umax*sin(2*pi*f*t) avg(i 2 ) P avg(u 2 ) time(s) U RMS avg( u 2 ) u 2 u time(s) time(s) time(s) Step 1: Umax 2 *sin(2*pi*f*t) u time(s) time(s) time(s) 0.6 Step 2: avg(umax 2 *sin(2*pi*f*t)) Source: Charles Kitchin, Lew Counts, RMS to DC Conversion Application Guide, 2 nd Edition, Analogue Devices, time(s) time(s) time(s) time(s) time(s) time(s) Step 3: sqrt(avg(umax 2 *sin(2*pi*f*t)))

24 Summary Frequently used for RMS current, RMS voltage, active power and reactive power calculations Square root with 16-bit output dynamic range was found sufficient for representing non billing quantities (all above) Executes in less than 100 instruction cycles on ARM Cortex- M4 (compiled by IAR EW for ARM, full optimization, code inlining, the additional LPF {1st order} added to smooth square root output) 24

25 Introduction to Energy Meters Energy Metering Tutorial International Standards Current Sensor Interfaces Energy Calculation An Effective Metering System on Chip Based Solutions Enablement Summary 25

26 Cost optimized Large choice Industry standard cores Analog integration Lowest 32-bit power Single Phase 8-bit Three phase Flow LH/LL64 (S08,16-bit,8K-64K) GW/LL64 (S08,16-bit,32K-64K) MG64 (China) MZ60 (China) Coming Next 32-bit Apps Cont. 32-bit Int. AFE 32-bit Ext. AFE EM256 (CT) (CF,256K) EM128 (SR) (CF,128K) Kinetis : K10-K30 (M4,64K-1M,crypto) K30/K10/K12 (AFE,Welmec, M4) (256K-1M) K30/K10/K12 (Welmec, M4,) (128K-1M) Next Gen Met+App Next Gen Single Chip Next-Gen Met Flow NPI KA/LA8 LG32 Ext. AFE 32-bit AFE Smart AFE

27 Common System IP Common Analog IP Common Digital IP Development Tools K70 Family 512 KB-1 MB, pin K60 Family 256 KB-1 MB, pin K50 Family KB, pin K40 Family KB, pin K30 Family KB, pin K20 Family 32 KB-1 MB, pin K10 Family 32 KB-1 MB, pin 32-bit ARM Cortex-M4 Core w/ DSP Instructions Next Generation Flash Memory High Reliability, Fast Access FlexMemory w/ EEPROM capability SRAM Memory Protection Unit Low Voltage, Low Power Multiple Operating Modes, Clock Gating (1.71V-3.6V with 5V tolerant I/O) DMA -40 to 105 C 16-bit ADC Programmable Gain Amplifiers 12-bit DAC High-speed Comparators Low-power Touch Sensing CRC I 2 C SAI (I 2 S) UART/SPI Programmable Delay Block External Bus Interface Motor Control Timers esdhc RTC Bundled IDE w/ Processor Expert Bundled OS USB, TCP/IP, Security Modular Tower H/ware Development System Application Software Stacks, Peripheral Drivers & App. Libraries (Motor Control, HMI, USB) Broad 3rd party ecosystem 27

28 For Segment LCD Applications Flexible, low power LCD interface Segment LCD Blink mode lowers average power Segment fail detect prevents erroneous readouts and reduces LCD test cost Front/back plane reassignment provides pin-out flexibility and allows configuration changes in firmware Diverse communications suite A multitude of serial interfaces, with UART support for ISO7816 SIM/Smart Cards and IrDA interfaces Dual CAN for industrial network bridging System reliability and safety Hardware cyclic redundancy check safeguards memory contents and communication data Memory Protection Unit increase software reliability Independently-clocked watchdog prevents code runaway for fail-safe applications e.g. IEC60730 Hardware and software compatibility Common packages and peripherals across families enable rapid feature growth with minimal hardware and software disruption Built-in voltage regulator (K40) 5V regulator input with 3.3V regulated output 3.3V regulated output can power MCU and also external components (source current up to 120mA) K30 Family Block Diagram Up to 100MHz ARM Cortex-M4 Core DSP 1.25DMIPs/MHz Memory Protection Unit Up to 512KB Flash Memory Up to 128KB SRAM FlexMemory: up to 4KB EEPROM or 256KBFlash FlexBus 16-ch DMA Timers 3x FlexTimers Carrier Modulator Timer, Programmable Interrupt Timer, Low Power Timer, Programmable Delay Block CrossBar Switch (XBS) Peripheral Bus Controllers (x2) Analog 2x 16-bit SAR ADC 2x 12-bit DAC 3x High Speed Comparators 2x Programmable Gain Amplifiers Internal Voltage Ref (1.2V) Secure Digital Host Controller Communications 2x I2C 6x UART 3x DSPI 2x CAN IIS 32-bit Cyclic Redundancy Check Family USB OTG + DCD Segment LCD K30 - X K40 X X Clock Module: 2 Crystal inputs 2 internal oscillators PLL and FLL Segment LCD (8x40 /4x44) Low Power Xtrinsic Touch Sensing Interface Watchdog + External Watchdog Monitor Up to 98 GPIO Kinetis documentation: 28

29 Needs Cortex-M4 CPU Low Power Data Storage Flash Update User Interface Communication Memory Scalability Time Keeping Analog Modules Packages Kinetis Solution M4 core (DSP functionality) up to 100 MHz 298uA*/MHz Run Current Down to FlexMemory (Up to 4kB EEPROM, Dual Flash Bank). Fast Flash Programming (70us) Flash swap, security and protection. CRC32 for code check up Embedded LCD controller (integrated voltage generators) Up to 6 UARTS, 3 SPIs and 2 IIC. From 64 KB to 512 KB of Flash. From 16 KB to 128 KB of RAM RTC module with separate Vbat and Crystal compensation (Correction range ~ ±.12ppm.. ~ ±3900ppm) High Speed ADC with integrated PGA. 12-bit DAC. HSCMP. Built-in 1.2 V reference 33 ppm/c with dedicated output pin. 64LQFN, 64LQFP, 80LQFP, 81 BGA, 100LQFP, 104BGA, 144LQFP, 121BGA, 144BGA * Typical run mode current 72MHz core clock, 36MHz bus clock, and 24MHz flash clock. MCG configured for FEE mode. All peripheral clocks disabled. ** Typical very low-leakage stop mode 1 current at 3.0 V and 40 to 25 C 29

30 ADC0_DP1 ADC0_DM1 DAD1 ADC0 DAD0 PGA0_DP/ADC0_DP0/ADC1_DP3 PGA0_DM/ADC0_DM0/ADC1_DM3 PGA0 1x-64x DAD2 DAD3 DAD3 ADC1 PGA1_DP/ADC1_DP0/ADC0_DP3 PGA1_DM/ADC1_DM0/ADC0_DM3 PGA1 1x-64x DAD2 DAD0 ADC1_DP1 ADC1_DM1 DAD1 The Programmable Gain Amplifier (PGA) is designed to increase the dynamic range by amplifying low-amplitude signals before they are fed to the 16-bit ADC. 30

31 adc output [LSB] PGA Input Signal Range The formula for PGA differential input signal swing of the ADC16-PGA measurement chain is: V PPADC,DIF = min V x 0.2,V VREF V x 4 GAIN [V] where: V x is 700mV V VREF is 1.2V GAIN is PGA gain 1, 2, 4, 8, 16, 32 and 64. ADC16 digital output range is expressed as follows: ADC OUT = V PPADC,DIF 2 V VREF GAIN 2 N [LSB] where: N is number of bits for given ADC16 conversion mode. PGA Input Signal Range: Theoretical PGA GAIN V PP,DIF [mv] ADC OUT [LSB] x x x x x x x Gain=x64 ADC Measurement Use-Case: 1.2 V mv ADC16_DP 0.6 V PGA_DP 0.7V mv 1.0 V 0.2V 0V PGA ADC16 0V x V mv -1.0V 0.6 V PGA_DM mv 0.7V ADC16_DM 0.2V 0V ADC Digital Output (PGA Enabled, Gain=x64). x time [s] 31

32 Summary Kinetis K30 devices feature two 16-bit ADCs. Each ADC contains a differential input channel with PGA. Each ADC has two status and control registers as well as two result registers, thus up to four conversions can be initiated by hardware without software intervention. Up to two measurements can be trigger either simultaneously or with precisely defined delay with the help of the Programmable Delay Blocks the trigger setup is optimized for use in 1-phase electricity meters. 32

33 Secure, Prepaid MK30 with NFC 1Q12 33

34 Block diagram Photo More information: 34

35 Key Features: 5(60)A current range, nominal current is 5A, peak current is 60 A Four quadrant measurement 85V...264V, 50/60 Hz voltage range Active and reactive energy accuracy IEC Class B, 1% Line frequency measurement (for precision zero-cross detection) Cost-effective shunt resistor sensing circuit implementation Cost-effective bill of materials (BOM) Low-power modes effectively implemented, including the use of the built-in RTC LCD display, 4x31 segments including charge pump Values shown on the LCD: V,A,W,Var,VA, kwh, kvarh, cos, Hz, time, date Tamper detection via two tamper buttons (event is stored in FlexMemory) Built-in user push-button LEDs pulse outputs (kwh, kvarh) IEC1107 infrared hardware interface Optically isolated RS232 interface 2.4 GHz RF1322x interface (I 2 C Daughter Card) for connection to a ZigBee network MQX based for advanced markets Multiple advanced metering algorithms (FFT, filter-based method) 35

36 error[%] 2 Temp. drift of active energy error for MK30 1-ph meter (V bias tied on V REF ) C 0 C 22 C 45 C 70 C min. max current[a] According to EN , Class B 5(60)A, errors represents results at UPF 36

37 Where: Fs is sampling frequency More information: N is number of samples per voltage cycle k series of harmonics 1.. N/2-1 37

38 Conditions Implementation of FFT algorithm to ARM core (MK30X) Separation of real and imaginary parts of FFT signal to fix part (32-bits) and remainder part (16-bits) in total 48-bits (good precision for most of applications) Advantages of realization The same precision for both energies (active, reactive) in comparison to other computing methods depends only on AFE (HW) of the meter, not on software. Frequency analysis of input signal ability to compute total harmonic distortion Offset removal (0-harmonic is missed for power computing) Disadvantages of realization Adjustable sampling rate necessary to compensate for frequency changes Higher computational power - 10,2 MIPS for 6400 Hz sampling rate (CPUCLK = 48 MHz) 38

39 Demo purpose: To provide working solution for accuracy evaluation To specify schematic and BOM (low-cost) To verify accuracy Product Features and Specifications: Design Reference Manual (DRM122) AN FFT-based algorithm for metering applications Quick Start for MK30EMETERMQX MQX Reference Manual and MQX User s Guide 1322x Low-Power Node Reference Manual Device development tools: MK30EMETERMQX Device documentation: MK30X256 39

40 3-ph E Meter MK30 2Q12 40

41 2Q12 41

42 2Q12 Key Features: 5(100)A current range, nominal current is 5A, peak current is 100 A Four quadrant measurement 85V...264V, 50/60 Hz voltage range Active and reactive energy accuracy IEC Class B, 1% Line frequency measurement form all 3 phases (for precision zero-cross detection) Current transformers sensing circuit implementation Cost-effective bill of materials (BOM) Low-power modes effectively implemented, including the use of the built-in RTC LCD display, 4x44 segments including charge pump Values shown on the LCD: V,A,W,Var,VA, kwh, kvarh, cos j, Hz Tamper detection via ELECTRONIC TILT PMA8491Q sensor and tamper buttons (event is stored in FlexMemory) Built-in user push-button LEDs pulse outputs (kwh, kvarh) IEC1107 infrared hardware interface Optically isolated RS232 interface AMU and UMI connectors interface, could be use for (I 2 C Daughter Card) for connection to a ZigBee network Multiple advanced metering algorithms (FFT, filter-based method) 42

43 Introduction to Energy Meters Energy Metering Tutorial International Standards Sensor Interfaces Energy Calculation An Effective Metering System on Chip Based Solutions Enablement Summary 43

44 Special Edition Free. The following limitations apply - Unlimited assembly code Up to 32KB of C code for HC(S)08/RS08 derivatives Up to 64KB of C code for V1 ColdFire/ColdFire+ derivatives Up to 128KB of C code for V2-V4 ColdFire and Kinetis derivatives New Project Wizard Create a project in as few as six clicks Free Compiler up to 128KB! MCU Change Wizard Re-target to a new RS08, HCS08, ColdFire, ColdFire+ or Kinetis processor in as few as six clicks LiveView Allows registers, memory and global variables to be monitored without stopping the processor Processor Expert Creates tested, optimized initialization code and low-level drivers tuned to application needs and selected Freescale derivative Built-in knowledgebase immediately flags resource conflicts and incorrect settings, so errors are caught early in design cycle Processor Expert for Kinetis is fully integrated with MQX via RTOS adapter component Trace and profile support for on-chip trace buffer and real-time collection (external probe) Full debug support for low power modes Handles entry into and continuation of debug after exiting Low Power and Very Low Power modes. 44

45 Leading supplier of MCU development tools ANSI C/C++ compilers, Debuggers and Simulation Middleware components Extensive Device Database Directory of over 700 supported ARM MCUs Established support Phone, , Web and User Group Support Huge installed base 100K+ users world wide Best-in-class compilation tools Architecture-specific optimizations Smaller and faster code reduces system cost Products: MDK-STANDARD-FREESCALE: Supports any Kinetis Cortex-M4 series device, $745 (1year license with no code limits) MDK-Lite (32K Code Limited): $0 MDK-Standard (T) (1 Yr Lic): $1958 MDK-Basic (B) (256KB limit): $2695 MDK-Standard: $4895 MDK-Standard (F) (3 seats floating): $5874 MDK-Professional: $9995 MDK-Professional (F) (3 seats floating): $ MicroLib optimized C Libraries Superset of standard ARM C Library Optimized for embedded applications Additional support for the Cortex-M4 Supports Thumb2 Instruction sets including DSP and FPU CMSIS Signal Processing Library support

46 The most widely used C/C++ tool chain for ARM MCUs Professional and global technical support organization (10 offices) Reliable, powerful and easy to use Advanced trace debug functionality based on E and SWO Support for Kinetis 10/20/30/40/60/70, i.mx, ColdFire+ and ColdFire Freescale MQX RTOS integration Project examples for the Freescale Tower system Products: EWARM-CM-FSL: Supports any Kinetis Cortex-M4 series device, $2500 (1-year license), available now via Freescale Buy Direct e=ewarm-cm-fsl&fsrch=1&sr=2 EWARM-CM: Supports all ARM Cortex-M devices including Cortex-M4 EWARM-BL (BaseLine): Supports all ARM cores with a limit in code size of 256 KB EWARM: Supports all ARM cores Evaluation versions on to download 46

47 Free scalable, fully-featured and proven RTOS with 32-bit MCUs Full-featured and powerful BSPs incorporate tightly integrated RTOS, Middleware (USB, TCP/IP stacks), file system, and I/O drivers Designed for speed and size efficiency Market proven Available on Freescale processors for > 15 years Used in millions of products including medical and heavy industrial applications Simple and scalable As small as ~10KB for smallest implementation, or scale up to support sophisticated networking and threading Intuitive API & modular architecture enables straightforward fine-tuning of features Production source code provided Similar to other pay-for software OS $95K of free Software Software integration headache Integrated MQX Solution Stable Upgradable Easy to maintain 47

48 Application control and monitor Live graphs, variable watches, and graphical control page Real-time etpu operation monitor Supports: - HCS08, HC12, HCS12 and HCS12X BDM - 56F8000, 56F8100 and 56F8300 JTAG - SCI driver (FMASTERSCIDRV) for all platforms Download from: 48

49 8bit 16bit DSC 32bit - ColdFire 32bit Power Arch 32bit Kinetis Processor Modules ($39-$119) TWR-S08LL64 TWR-S08LH64 TWR-S08JE128 TWR-S08MM128 TWR-S08GW64 TWR-S08UNIV TWR-S08PT60 TWR-S12GN32 TWR-S12G128 TWR-56F8257 Serial Prototyping Wi-Fi Memory TWR-MCF51JE TWR-MCF51CN TWR-MCF51MM TWR-MCF51QM TWR-MCF5225X TWR-MCF5441X TWR-MCF51JF Sensors & Plug-Ins TWR-MPC5125 Displays TWR-K60N512 TWR-K40X256 TWR-K60D100M TWR-K40D100M TWR-K53N512 KWIKSTIK-K40 TWR-K20D50M Medical Peripheral Modules ($15 $149) TWR-SER TWR-SER2 TWR-PROTO TWR-WIFI-RS2101 TWR-WIFI-G1011MI TWR-WIFI-AR4100 Analog Audio Mesh Networking TWR-MEM TWR-SENSOR-PAK TWR-SENSOR-PAK-AUTO TWRPI-MMA6900 TWRPI-MPL115A TWR-LCD MED-EKG TWR-ADCDAC-LTC TWR-AUDIO-SGTL TWR-RF-SNAP 49

50 Introduction to Energy Meters Energy Metering Tutorial International Standards Sensor Interfaces Energy Calculation An Effective Metering System on Chip Based Solutions Enablement Summary 50

51 Freescale offers wide range of metering specific reference designs and application notes. Metering algorithms are tested according to EN (active energy) and IEC (reactive energy) as well as ANSI C12.20 standards. Performance and accuracy of all Kinetis MK30, ColdFire EM256, S08GW64 and S08LH64 reference designs have been thoroughly evaluated on metering test bench. The EMC testing is performed according to EN X basic standards taking into account regional specifics and preferences. 51

52 Facebook.com/Freescale Tag yourself in photos and upload your own! Tweeting? Please use hashtag #FTF2012 Session materials will be Look for announcements in the FTF Group on LinkedIn or follow Freescale on Twitter 52

53