Design and implementation of the ELIQ Smart

Design and implementation of the ELIQ Smart
Design
and
implementation
of
the
Smart-plug system device
Master of Science Thesis in the Master Degree Programme
Embedded Electronic System Design
Yixin Liu
Department of Computer Science and Engineering
CHALMERS UNIVERSITY OF TECHNOLOGY
Göteborg, Sweden, 2014
ELIQ
The Author grants to Chalmers University of Technology and University of
Gothenburg the non-exclusive right to publish the Work electronically and in a
non-commercial purpose make it accessible on the Internet.
The Author warrants that he/she is the author to the Work, and warrants that the Work
does not contain text, pictures or other material that violates copyright law.
The Author shall, when transferring the rights of the Work to a third party (for
example a publisher or a company), acknowledge the third party about this agreement.
If the Author has signed a copyright agreement with a third party regarding the Work,
the Author warrants hereby that he/she has obtained any necessary permission from
this third party to let Chalmers University of Technology and University of
Gothenburg store the Work electronically and make it accessible on the Internet.
Design and implementation of the Smart-plug device system design
Master of Science Thesis in the Master Degree Programme Embedded Electronic
System Design
Yixin Liu
© Yixin Liu, July 2014.
Examiner: Per Larsson-Edefors
Chalmers University of Technology
University of Gothenburg
Department of Computer Science and Engineering
SE-412 96 Göteborg
Sweden
Telephone + 46 (0)31-772 1000
[Cover:
an explanatory caption for the (possible) cover picture
with page reference to detailed information in this essay.]
Department of Computer Science and Engineering
Göteborg, Sweden July 2014
The ELIQ Smart-Plug Design
Abstract
The company Exibea AB plans to design and implement a new product called ELIQ Smart-plug
device to help the households control the energy consumption at home more conveniently and
intelligently. The expected function of the ELIQ Smart-plug device is to measure the power
consumption value of the connected electrical appliance, send the measured value to the ELIQ
Display Unit through RF wireless communication, switch on or switch off the power supply to the
connected electrical appliance automatically and repeat the received ELIQ protocol RF message
to work as a wireless repeater for other ELIQ devices in the company.
This master thesis explores the design process and the implementation method of the ELIQ
Smart-plug system device including both of the hardware part and software part in depth.
MSP430AFE233 is chosen to be the microcontroller unit used on the Smart-plug electrical board
to control the whole system. In order to calculate the power consumption value, the RMS (Root
Mean Square) values of both of the voltage and current on the connected appliance should be
measured firstly, then the power consumption value will be obtained easily by multiplying the
voltage RMS value and the current RMS value. The calculated power consumption value will be
encoded and combined with some other specified bits such as preamble bits and header bits
together into a data packet according to the ELIQ protocol. The data packet will be transmitted to
the RF module on the Smart-plug electrical board through SPI communication firstly, then the
wireless communication between the antenna of this RF module and the antenna of another RF
module on the ELIQ Display Unit board will be implemented. The power consumption value will
be displayed on the ELIQ Display Unit device to show the households the transient power
consumption of the connected electrical appliance. The Smart-plug device can also switch on or
switch off the power supply to the appliance automatically according to the measured power
consumption value to achieve the aim of energy-saving in a house. In addition, this Smart-plug
device system can repeat the received RF message from other ELIQ devices and then send the
same RF message to the ELIQ Display Unit or the ELIQ Hub to work as a wireless repeater in order
to extend the wireless transmission distance and amplify the power of the signal.
Furthermore, the comparison of selecting components for the Smart-plug electrical board and
some important attentions need to be considered in both of the hardware design and software
design are discussed in this master thesis.
i
Yixin Liu
May 26th, 2014
The ELIQ Smart-Plug Design
ii
Yixin Liu
May 26th, 2014
The ELIQ Smart-Plug Design
Acknowledgement
It is my pleasure to show my gratitude towards everyone who supported me during my master
thesis.
Firstly, I would like to thank the company Exibea AB for offering this master thesis opportunity
and especially Hakan Ludvigsson who was my supervisor at the company. He is very pleasant to
work with. He provided me a good working environment and helped a lot about the component
selection and producing the PCB board. Also I would give thanks to Olof Hartelius and Johan Salin
at the company who were available for answering my questions and shared many technical
knowledge and experience with me together.
I would like to acknowledge the helps from Per Larsson-Edefors, who was my supervisor and
examiner at Chalmers University of Technology. He gave me a lot of valuable advice, feedback
and comments about how to write this master thesis.
A special thanks to my parents and my girlfriend, who encouraged me a lot when I encountered
with some difficulties in this master thesis project and provided me great impetus to overcome
these difficulties.
Finally, I also appreciate for the effort that my master thesis opponent, Gongpei Cui made. His
criticism and questions opened my eyes for the weak points in my master thesis.
iii
Yixin Liu
May 26th, 2014
The ELIQ Smart-Plug Design
iv
Yixin Liu
May 26th, 2014
The ELIQ Smart-Plug Design
List of Abbreviations
AC – Alternating Current
ACLK – Auxiliary Clock
A/D – Analog to Digital
ADC – Analog to Digital Converter
BCSCTLx – Basic Clock System Control Register x
CPU – Central Processor Unit
CS – Chip Selection
DC – Direct Current
DCO – Digitally Controlled Oscillator
EDA – Electronic Design Automation
FIFO – First Input First Output
FSR – Full Scale Range
GDO – General Digital Output
GFSK – Gaussian Frequency-Shift Keying
GIO – General Input/Output
GO – General Output
GUI – Graphic User Interface
I/O – Input or Output
ISR – Interrupt Service Routine
ISM – Industrial, Scientific and Medical
JTAG – Joint Test Action Group
LED – Light Emitting Diode
MCLK – Main System Clock
MCU – Microcontroller Unit
MSP – Mixed Signal Processor
NPN transistor – Negative-Positive-Negative Transistor
OSR – Oversampling Ratio
PCB – Printed Circuit Board
PWM – Pulse Width Modulation
PxDIR – Px Direct Register
PxIE – Px Interrupt Enable Register
PxIES – Px Interrupt Edge Select Register
PxIFG – Px Interrupt Flag Register
PxIN – Px Input Register
PxOUT – Px Output Register
PxREN – Px Pullup/Pulldown Resistor Register
RAM – Random Memory
RF – Radio Frequency
RISC –Reduced Instruction Set Computer
RMS – Root Mean Square
RX – Receive, Receive Mode
v
Yixin Liu
May 26th, 2014
The ELIQ Smart-Plug Design
SCLK – Serial Clock
SD24_A A/D Converter – 24-bit sigma-delta Analog to Digital Converter
SD24CTL – SD24 Control Register
SD24INTCTLx –SD24 Channel x Input Control Register
SD24CCTLx –SD24 Channel x Control Register
SD24MEMx –SD24 Channel x Conversion Memory Register
SI – Serial Input
SMCLK – Sub Main System Clock
SMD – Surface Mounted Device
SO – Serial Output
SPI – Serial Peripheral Interface
SPST – Single Pole Single Throw
SRAM – Static Random Memory
TACCRx – Timer_A Capture/Compare Register x
TACCTLx – Timer_A Capture/Compare Control Register x
TACTL – Timer_A Control Register
TAR – Timer_A Counter Register
TX – Transmit, Transmit Mode
UART –Universal Asynchronous Receiver/Transmitter
USART –Universal Synchronous/Asynchronous Receiver/Transmitter
USB – Universal Serial Bus
VCO – Voltage Controlled Oscillator
WDT – Watchdog Timer
WWF – World Wide Fund for Nature
XTAL – Crystal
vi
Yixin Liu
May 26th, 2014
The ELIQ Smart-Plug Design
Table of Contents
1. Introduction ............................................................................................................ 1
1.1 Background ......................................................................................................................... 1
1.2 Problem statement.............................................................................................................. 1
1.3 Specification ........................................................................................................................ 2
1.4 Purpose ............................................................................................................................... 2
1.5 Subcomponents .................................................................................................................. 2
1.6 Project procedure................................................................................................................ 3
1.7 Scope ................................................................................................................................... 3
1.8 Limitations ........................................................................................................................... 3
1.9 Thesis outline ...................................................................................................................... 3
2. Theory ..................................................................................................................... 5
2.1 Basic design principle .......................................................................................................... 5
2.2 Selection of MCU................................................................................................................. 6
2.3 Introduction of MSP430AFE233 .......................................................................................... 7
2.3.1 Description ........................................................................................................ 7
2.3.2 Features ............................................................................................................. 7
2.3.3 Digital I/O ports ................................................................................................. 8
2.3.3.1 Digital I/O ports input register PxIN .................................................................. 9
2.3.3.2 Digital I/O ports output register PxOUT ............................................................ 9
2.3.3.3 Digital I/O ports direction register PxDIR .......................................................... 9
2.3.3.4 Digital I/O ports pullup/pulldown resistor enable register PxREN .................... 9
2.3.3.5 Digital I/O ports interrupts .............................................................................. 10
2.3.4 Basic clock module .......................................................................................... 10
2.3.4.1 Basic clock module structure .......................................................................... 10
2.3.4.2 Basic clock module registers ........................................................................... 11
2.3.5 Timer_A ........................................................................................................... 12
2.3.5.1 Timer_A structure ........................................................................................... 12
2.3.5.2 Timer_A registers ............................................................................................ 13
2.3.6 SD24_A ............................................................................................................ 15
2.3.6.1 SD24_A structure ............................................................................................ 15
2.3.6.2 SD24_A registers ............................................................................................. 16
2.3.6.3 Analog-to-digital conversion in SD24_A .......................................................... 18
2.3.6.4 Digital filter in SD24_A .................................................................................... 19
3.
Materials ............................................................................................................... 21
3.1 Hardware components ...................................................................................................... 21
3.1.1 MCU ................................................................................................................... 21
3.1.2 RF module .......................................................................................................... 21
3.1.3 AC/DC Converter ................................................................................................ 25
3.1.4 Relay ................................................................................................................... 25
3.1.5 NPN transistor .................................................................................................... 26
3.1.6 Crystal ................................................................................................................. 26
3.1.7 Other components ............................................................................................. 27
vii
Yixin Liu
May 26th, 2014
The ELIQ Smart-Plug Design
3.2 Software tools ................................................................................................................... 28
3.2.1 Altium Designer 6 ............................................................................................... 28
3.2.2 Switcher CAD Ⅲ ................................................................................................ 28
3.2.3 IAR Embedded Workbench ................................................................................ 28
3.2.4 Lite Fet-Pro430 MSP430 flash programmer ....................................................... 28
4.
Hardware design ................................................................................................... 30
4.1 Power supply circuit design ............................................................................................... 30
4.2 Voltage sampling circuit design ......................................................................................... 31
4.3 Current sampling circuit design......................................................................................... 32
4.4 Switching on/off control unit circuit design ...................................................................... 35
4.5 Communication between MCU and RF module ................................................................ 37
5. Software design methodology .............................................................................. 39
5.1 Power consumption value calculation .............................................................................. 40
5.1.1 Peripherals initiation in MSP430AFE233 ............................................................ 40
5.1.1.1 Clock system initiation .................................................................................... 40
5.1.1.2 I/O ports initiation ........................................................................................... 41
5.1.1.3 SD24_A A/D converter initiation ..................................................................... 43
5.1.1.3.1 Configuring sampling frequency .................................................................. 43
5.1.1.3.2 Configuring analog-to-digital conversion parameters .................................. 44
5.1.1.4 Timer_A initiation............................................................................................ 47
5.1.2 Sampling voltage and current value in one second ........................................... 49
5.1.3 RMS value calculation for voltage and current .................................................. 50
5.1.4 Power consumption value calculation ............................................................... 52
5.2 Control switch on or off status of connected appliance ................................................... 52
5.3 SPI communication to RF module ..................................................................................... 53
5.4 Wireless communication ................................................................................................... 54
5.5 Working as a repeater ....................................................................................................... 55
6. Visual support ....................................................................................................... 57
7. Results ................................................................................................................... 59
8. Discussion ............................................................................................................. 60
8.1 Appliance with small power consumption cannot be measured accurately .................... 60
8.2 Sampling resistor with resistance tolerance 1% still influences the accuracy of
measurement .......................................................................................................................... 61
8.3 Component selection for the control unit......................................................................... 62
8.4 A/D converter output need load ....................................................................................... 63
8.5 Acquiring relay components ............................................................................................. 63
8.6 Width and thickness of the copper wire on the PCB board .............................................. 65
9.
Conclusion ............................................................................................................. 66
9.1 Summary ........................................................................................................................... 66
9.2 Future work ....................................................................................................................... 66
Bibliography ................................................................................................................. 68
viii
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
1. Introduction
1.1 Background
Over the last decade, energy saving has been a hot topic across the whole world. Households
always find it difficult to plan and handle their energy consumption at home. Unnecessary use of
energy is not only an environmental issue but also affects the economy negatively for the
consumers.
The company Exibea AB was founded in 2008 in Gothenburg, Sweden, with a focus on providing
products to help the consumers to control and reduce their energy consumption. Now the
company is planning to design and implement a new product called ELIQ Smart-plug device. This
device is expected to detect the energy use and display the measured value on the o pa y’s
flagship products ELIQ Energy Display Unit or ELIQ Energy Online to provide feedback to the
consumers [1][2]. The Smart-plug device will be inserted between the power outlet and any
household appliance to measure the energy use and switch on/off the power supply to the
connected appliance, which can be also controlled by the user from the ELIQ Energy Online web
GUI or the remote mobile phone, allowing the user to set up power on/off scheduling based on
the current electricity price, room temperature or other parameters to help facilitating a more
energy-conservative lifestyle.
1.2 Problem statement
Energy consumption in the world is on the rise, leading to a non-sustainable society in the long
run. Lack of energy is a serious challenge for human beings in the future. Today the government
in almost every country formulates a strategy to reduce the energy consumption in every field of
the society. It is not only a matter for governments, but also boils down to a matter for every
citizen of Earth: We all have the responsibility to save the electric energy in daily life in order to
provide an ecological and sustainable society [3] for the coming generations.
Although electricity does not create any pollution where it is used, there may be carbon dioxide
emission and other environmentally negative substances coming out at the production site. The
WWF (World Wide Fund for Nature) states that if the carbon dioxide emission is kept on the
same level as today, the global temperature will increase two degrees Celsius, which will lead to
the severe global warming crisis and then a series of natural disasters in the world [4]. Wind
power and hydropower almost do not produce any carbon dioxide emission, however, they can
still create other negative environmental problems such as the disrupted water flows.
In order to create an ecological society, the electricity energy consumption of households is an
important concern. Household energy consumption accounts for about 25% of the total energy
consumption in the world [5]. One of the effective ways to decrease electricity energy
consumption in households is by increasing consumers’ aware ess as people at least in the past
have been careless about lowering electricity energy consumption even though lowering
1
Design and implementation of the ELIQ Smart-plug device system
electricity energy consumption could reflect well on their household budget.
Taking steps towards achieving the ecological and smart home, the new product ELIQ Smart-plug
device which the company Exibea AB plans to develop is expected to manage home electricity
energy consumption. This Smart-plug device system can provide functions such as measurement
and visualization of the home energy consumption and automatic switch on/off control for home
electrical appliances. In addition, it can be also connected to the network of a smart home to
achieve efficient ecological operation.
Since the envisioned Smart-plug device is a new concept, this master thesis project work is
concerned with conceiving, implementing, integrating and evaluating a Smart-plug device based
on the different off-the-shelf subcomponents.
1.3 Specification
The ELIQ Smart-plug device should provide the following functions:

Measure the power consumption value of the connected appliances and send the value
once every 15 seconds to the ELIQ Energy Display Unit or the ELIQ Energy Online web
GUI through the ELIQ Hub to display to the households with using the RF-module in
868 MHz ISM band

Switch on/off power supply to the connected appliances according to the measured
power consumption value or by the user from the ELIQ Energy Online web GUI and
indicate on/off status

Repeat received ELIQ protocol RF communication, making the device work as a wireless
extender or repeater for other ELIQ devices
1.4 Purpose
With the measurement and visualization of the home energy consumption by using this
Smart-plug device system, the households can know the energy consumption value in a room
instantly and directly then can manage the energy use of the house easily and conveniently. The
automatic switch on/off control for the home appliances can achieve the automatic reduction of
the household energy consumption. The hope is that with the help of this Smart-plug device
system, household energy consumption can be lowered significantly, leading to not only better
household economy but also to a sustainable society.
1.5 Subcomponents
The subcomponents the company has access to in this project are resistors, capacitors, diodes,
relay, switching power supply device, ferrite beads, transient voltage suppressor, NPN transistor,
crystal, fuse, varistor, LED, MCU, RF module, PCB board, plug and so on. The PCB board will be
manufactured by the PCB factory according to the designed circuit drawn by the specified EDA
software tool then the electrical subcomponents will be soldered on the PCB board. The
Smart-plug board will be connected to a plug to be inserted between the power outlet and the
2
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
household appliance. The tools multimeter, soldering iron and soldering tin will be used to
implement integrating these different off-the-shelf subcomponents into one testing product.
1.6 Project procedure
The electrical board of the Smart-plug device should be designed and implemented for testing
before the products in mass production, including both hardware design and software design. For
the hardware part, the electrical circuit design including the prototyping design and the selection
of the MCU and other suitable off-the-shelf components should be completed firstly. Then the
schematic and the PCB layout will be done by the specialized EDA tool. After the PCB board is
produced by the factory, soldering the components on the PCB board will be the following step.
The software task is to make programming for the MCU on the Smart-plug board to make the
Smart-plug device achieve the expected targets successfully.
1.7 Scope
The knowledge of this master thesis is related to the fields of analog circuit, digital circuit, A/D
mixed signal system, microcontroller unit, C code programming and wireless communication.
1.8 Limitations
There are some parts of this master thesis project left out due to the time reason. For instance,
the power supply circuit that drives the MCU, the RF module and the relay will not be designed
within this master thesis project work. Instead, a switching power supply device which can
transfer the 230V AC input from the mains to the 3.3V DC output to be the working supply
voltage is bought as the off-the shelf subcomponent. Since this temporary switching power
supply device is quite expensive and bulky, the final product in mass production will need a
different power solution on the Smart-plug electrical board to reduce cost and size. Secondly, the
wireless communication between the Smart-plug board and the ELIQ Energy Unit board or the
ELIQ Energy Hub does not need to be considered since this procedure is already done by the
company Exibea AB according to the company’s own RF protocol. In addition, on the ELIQ Energy
Unit board, the procedure of decoding received data packet and displaying the value on the LCD
display needs not to be mentioned as it has been also completed by the company.
Furthermore, the size, the exterior, the package, the manufacture, the process technology and
the arts and crafts of the final product have also not been considered as the aim of this master
thesis project is just designing a prototype.
1.9 Thesis outline
Chapter 2 introduces the theory of the basic design principle firstly, then the selection of the
suitable type of MCU and the basic knowledge of the selected MCU are described in detail. That
is vital for the readers to understand the following electrical circuits design and configurations in
software. Then, chapter 3 introduces the materials used in this project, including both of the
3
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
hardware tools and software tools. In chapter 4, the specific methodology of the hardware
design is elaborated. The software methodology design including the software design process
flowchart, the system setup initiation and the implementation of some special applications are
described step by step in chapter 5. Some important graphs such as the schematic of the whole
electrical circuits inside the Smart-plug device, the PCB layout of the Smart-plug electrical board
and the real photo of this electrical board are presented to be the visual support for this master
thesis in chapter 6. Chapter 7 will show the results of this project to the readers. In chapter 8,
some issues such as the comparison of suitable selection for the control unit component or some
conditions need be considered during the design will be discussed. Finally, in chapter 9, the
findings and insights from this project will be reviewed overall.
4
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
2. Theory
2.1 Basic design principle
The basic design principle diagram of the Smart-plug device system is shown in Figure 2.1 below.
L
Smart-plug device
N
MCU
re
Resistor
divider
A/D
converter
MCU
R0
RF
RF
module
module
ELIQ
Display Unit
A/D
converter
Load
Figure 2.1 The diagram of the basic design of the Smart-plug system
As can be seen from Figure 2.1 above clearly, L means the live wire and N means the neutral wire
from the household mains respectively. As the Smart-plug device needs to measure the power
consumption value of the electrical appliance, the value of the current on the live wire and
neutral wire of the circuit inside the smart-plug can be measured firstly in order to calculate the
power consumption. R0 is a resistor with small resistance value set on the neutral wire of the
circuit as the sampling resistor. The voltage value on R0 can be measured by connecting both
sides of the resistor to the A/D converter module of the MCU. If the voltage on R0 is known, then
the current value on the electrical appliance is easy to obtain by dividing the R0 resistance value.
Load means the load resistor of the electrical appliance connected to the Smart-plug device. The
voltage on the electrical appliance can also be measured by the A/D converter module of the
MCU. The resistor divider circuitry needs to be added between Live, neutral wire and the MCU
since the A/D converter of the MCU can only be convert small voltage values.
After sampling signal values during a specific time interval, both of the current effective value
and voltage effective value on the electrical appliance can be calculated using the sampled value
captured from the A/D converter, then the power consumption value can be calculated by
multiplying the effective voltage value and the effective current value. The power consumption
value will be transmitted to the ELIQ Display Unit via the respective RF module on both of the
5
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Smart-plug board and the ELIQ Energy Display Unit board.
In European countries, the voltage from the household mains is 230V AC. Since this Smart-plug
device is designed mainly for the market of the European countries, the design of the circuit is
based on the mains voltage equals to 230V AC. As the effective voltage value 230V from the
mains in European countries is almost highest among that across the world, the hardware circuit
design discussed in the later chapters will be also suitable for use in other regions outside
Europe.
2.2 Selection of MCU
A series of considerations are necessary to choose the type of MCU. As the product size limitation
and thermal regulation needs to be considered, the Texas Instruments MSP430 series
ultra-low-power microcontrollers are the first choice for the company since these types of MCU
possess the features of small size and ultra-low-power [6].
Initially, the MSP430F21x2 family MCUs, the MSP430G2x33 family MCUs and the MSP430AFE2x3
family MCUs have been considered to be the selections of the control MCU on the Smart-plug
electrical board in this project. Since the voltage from the household mains is alternating from
outside, the voltage on the sampling resistor is also alternating. The A/D converter in many types
of MCU such as the MSP430F21x2 family MCUs and the MSP430G2x33 family MCUs can only
detect the positive voltage value and the reference voltage on the A/D module is also positive
value as same as the supply voltage. If choosing these types of MCU, the voltage polarity
switching circuit need be added between the sampling resistor and MCU in order to do the
voltage bipolar-unipolar conversion. The MSP430AFE2x3 family MCUs have the different
analog-to-digital converters called SD24_A modules which are the multichannel 24-bit
sigma-delta analog-to-digital converters system can be input a pair of analog values. The
differential value of the pair will be processed by the SD24_A A/D converter. Therefore, if
choosing the MSP430AFE2x3 family MCUs, the AC signals can be interfaced directly without the
need of voltage bipolar-unipolar conversion.
The analog-to-digital converters in the MSP430F21x2 family MCUs and the MSP430G2x33 family
MCUs can only transfer the analog decimal data to the 10 bits digital binary data, however, the
MSP430AFE2x3 family MCUs with the SD24_A modules can transfer the analog decimal data to
the 16 bits digital binary data. That means the MSP430AFE2x3 family MCUs can implement the
analog-to-digital conversion much more accurately than the MSP430F21x2 family MCUs and the
MSP430G2x33 family MCUs with extremely higher resolution and smaller interval between each
adjacent transferred digital binary data.
Thus, the MSP430AFE2x3 family MCUs are the better selections for this project. Among the
different types of MCUs in the MSP430AFE2x3 family, in the overall consideration of the price,
the size, the power dissipation, the flash memory size, the SRAM memory size, the number of
ADC channels and the resolution of the ADC converter, the type MSP430AFE233 is chosen to be
the working microcontroller unit on the Smart-plug electrical board.
6
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
2.3 Introduction of MSP430AFE233
2.3.1 Description
MSP430AFE233 is a type of mixed signal ultra-low-power microcontroller which consists of
several devices featuring different sets of peripherals targeted for various applications [6]. The
architecture is optimized to achieve extended battery life in portable measurement applications.
The CPU inside this MCU is a 12MHz 16-bit RISC CPU including sixteen 16-bit registers and
constant generators that contribute to maximum code efficiency [6]. Three independent 24-bit
sigma-delta A/D converters, one 16-bit timer, one 16-bit hardware multiplier, one USART
communication interface, one watchdog timer and eleven I/O pins are also integrated in this
device.
2.3.2 Features
There are the basic features of MSP430AFE233 [6]:
 Low supply voltage range: 1.8 V to 3.6 V
 Ultra-low power consumption
-Active mode: 220 µA at 1 MHz, 2.2 V
-Standby mode: 0.5 µA
-Off mode (RAM retention): 0.1 µA
 Five power-saving modes
 Ultra-fast wake-up from standby mode in less than 1 µs
 16-Bit RISC Architecture, up to 12-MHz system clock
 Basic clock module configurations
-Internal frequencies up to 12 MHz
-Internal low-power low-frequency (LF) oscillator
-High-frequency (HF) crystal up to 16 MHz
-Resonator
-External digital clock source
 Basic clock module configurations
 Three 24-bit Sigma-Delta Analog-to-Digital (A/D) converters
with differential programmable gain amplifier inputs
 16-bit Timer_A with three capture/compare registers
 Serial communication interface (USART), Asynchronous UART or
Synchronous SPI selectable by software
 16-bit hardware multiplier
 Serial onboard programming, no external programming voltage needed
programmable code protection by security fuse
 On-chip emulation module
7
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 2.2 below shows the functional diagram of MCU MSP430AFE233.
Figure 2.2 The functional diagram of MSP430AFE233 [7]
Figure 2.3 below shows the pin designation of MCU MSP430AFE233. The MSP430AFE233 has 24
pins totally.
Figure 2.3 The pin designation of MSP430AFE233 [7]
2.3.3 Digital I/O ports
MSP430AFE233 has two I/O ports, P1 and P2. P1 has 8 pins, P1.0 to P1.7, P2 has 3 pins, P2.0,
P2.6 and P2.7. Each I/O pin can be individually configured in software for input or output
direction and each I/O line can be individually read or written [8].
8
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
The digital I/O ports have interrupt capability and each interrupt for the P1 and P2 I/O lines can
be enabled individually [8]. The P1 and P2 I/O lines can be configured to be provided an interrupt
by the rising edge or falling edge of an input signal [8]. Each I/O port pin contains the
independent input and output registers and the individually configurable pullup or pulldown
resistors.
2.3.3.1
Digital I/O ports input register PxIN
Each bit in each PxIN register reflects the value of the input signal at the corresponding I/O pin
when the pin is configured as I/O function [8]. When the bit is set to be 0 , the input signal to
the corresponding I/O pin is low; when the bit is set to be 1 , the input signal to the
corresponding I/O pin is high.
2.3.3.2
Digital I/O ports output register PxOUT
If the pullup/pulldown resistor of the corresponding pin is disabled by the PxREN register, the pin
is configured as I/O function, and the bit in each PxOUT register will reflect the value of the
output signal at the corresponding I/O [8]. When the bit is set to be 0 , the output signal from
the corresponding I/O pin is low; when the bit is set to be 1 , the output signal from the
corresponding I/O pin is high.
If the pullup/pulldown resistor of the corresponding pin is enabled by the PxREN register, the
corresponding bit in the PxOUT register will select whether the pullup resistor or pulldown
resistor is used [8]. When the bit is set to be 0 , the pin is pulled down; when the bit is set to be
1 , the pin is pulled up.
2.3.3.3
Digital I/O ports direction register PxDIR
Each bit in each PxDIR register will select the direction of the corresponding I/O pin [8]. When the
bit is set to be 0 , the corresponding pin is set to be input direction; when the bit is set to be 1 ,
the corresponding pin is set to be output direction.
2.3.3.4
Digital I/O ports pullup/pulldown resistor enable
register PxREN
Each bit in each PxREN register can enable or disable the pullup/pulldown resistor of the
corresponding I/O pin [8]. When the bit is set to be 0 , the pullup/pulldown resistor of the
corresponding I/O pin is disabled; when the bit is set to be 1 , the pullup/pulldown resistor of
the corresponding I/O pin is enabled, in this case, the corresponding bit in the PxOUT register will
select whether the pullup resistor or pulldown resistor is used.
9
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
2.3.3.5
Digital I/O ports interrupts
Each pin in ports P1 and P2 have interrupt capability, which is configured with PxIFG, PxIE and
PxIES registers individually and independently [8]. PxIFG is the Px interrupt flag register, PxIE is
the interrupt enable register and PxIES is the interrupt edge select register. All of the P1 pins
source the single interrupt vector and all of the P2 pins source a different single interrupt vector
[8].
Each PxIE bit can enable the associated PxIFG bit which is the interrupt flag for its corresponding
I/O pin. When one PxIE bit of the corresponding P1 port pin or P2 port pin is set to be 0 , that
means the I/O interrupt is disabled at thin pin; when one PxIE bit of the corresponding P1 port
pin or P2 port pin is set to be 1 , that means the I/O interrupt is enabled at thin pin.
Each PxIES bit will select which interrupt edge is used to control the PxIFGx flag of the
corresponding I/O port pin. When one PxIES bit of the corresponding P1 port pin or P2 port pin is
set to be 0 , the associated PxIFGx flag is set with a low-to-high transition; when one PxIES bit of
the corresponding P1 port pin or P2 port pin is set to be 1 , the associated PxIFGx flag is set with
a high-to-low transition [8]. It is important to notice that only transitions, not static levels, cause
the interrupts.
Each PxIFGx bit is set when the selected input signal edge occurs at the pin. All of the PxIFGx
interrupt flags request an interrupt when their corresponding PxIE bit and the GIE bit are set at
the same time [8]. Each PxIFG flag must be reset with software. In addition, software can also set
each PxIFG flag to generate a software initiated interrupt [8]. When one PxIFGx bit of the
corresponding P1 port pin or P2 port pin is set to be 0 , that means there is no interrupt
pending at this port pin; when one PxIFGx bit of the corresponding P1 port pin or P2 port pin is
set to be 1 , that means there is an interrupt pending at this port pin.
2.3.4 Basic clock module
2.3.4.1 Basic clock module structure
Figure 2.4 below shows the internal hardware structure of the basic clock module inside MCU
MSP430AFE233.
10
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 2.4 The internal structure of the basic clock module in MSP430AFE233 [8]
2.3.4.2 Basic clock module registers
There are a series of registers in the basic clock module. The initial state of the internal circuits of
the basic clock module needs to be set by configuring each bit of these registers firstly to
determine the working clock for the task executed in the MCU when the system starts working.
In BCSCTL1, the basic clock system register 1, the XT2OFF bit which occupies bit 7 is used to
choose whether the XT2 oscillator is turned on or turned off. If XT2OFF bit equals to 0 , XT2
oscillator is turned on; if XT2OFF bit equals to 1 , XT2 oscillator is turned off. Since XT2CLK needs
to be used for MCLK and SMCLK, the XT2 oscillator should be turned on by clearing the XT2OFF
bit.
Table 2.1 below shows the function of each bit in the BCSCTL1 register.
7
6
XT2OFF
XTS
5
4
3
DIVAx
2
1
0
RSELx
Table 2.1 The function of each bit in the BCSCTL1 register [8]
In BCSCTL2, the basic clock system register 2, the SELMx bits which occupy bit 7 and bit 6 are
used to select the clock source for the MCLK. The SELS bit which occupies bit 3 is used to select
clock source for the SMCLK. The DIVMx bits which occupy bit 5 and bit 4 will determine the value
11
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
of the divider ratio for the MCLK. Similarly, the DIVSx bits which occupy bit 2 and bit 1 will
determine the value of the divider ratio for the SMCLK. Table 2.2 below shows the function of
each bit in the BCSCTL2 register.
7
6
5
SELMx
4
3
DIVMx
2
SELS
1
DIVSx
0
DCOR
Table 2.2 The function of each bit in the BCSCTL2 register [8]
In BCSCTL3, the basic clock system register 3, the XT2Sx bits which occupy bit 7 and bit 6 are used
to select the frequency range for the XT2 clock. Table 2.3 below shows the function of each bit in
the BCSCTL3 register.
7
6
5
XT2Sx
4
3
LFXT1Sx
2
XCAPx
1
0
XT2OF
LFXT1OF
Table 2.3 The function of each bit in the BCSCTL3 register [8]
2.3.5 Timer_A
In the MCU MSP430AFE233, Timer_A is a 16-bit timer/counter with multiple capture/compare
registers. The 16-bit timer/counter can be chosen to work in four different operating modes.
Timer_A has a selectable clock source and configurable PWM outputs. It also has extensive
interrupt capabilities. The interrupts can be generated from the overflow condition of the
counter or from the capture/compare registers.
2.3.5.1
Timer_A structure
Figure 2.5 below shows the internal hardware structure of the Timer_A module inside MCU
MSP430AFE233.
12
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 2.5 The internal structure of Timer_A in MSP430AFE233 [8]
2.3.5.2
Timer_A registers
There are also some registers in the Timer_A module. Each bit of these registers needs to be
configured firstly to give the initial state of the circuits inside the Timer_A module when the
system starts working.
Table 2.4 below shows the function of each bit in TACTL, the Timer_A control register. The
TASSELx bits which occupy bit 9 and bit 8 will select the Timer_A clock source from TACLK, ACLK,
SMCLK or INCLK. The IDx bits which occupy bit 7 and bit 6 will select the ratio of the divider to
divide the frequency of the input clock signal from the Timer_A clock source. The MCx bits which
occupy bit 5 and bit 4 will be used to choose which operating mode for the Timer_A to work with
from the up mode, the continuous mode or the up/down mode. The TACLR bit which is bit 2 of
the TACTL register is used to clear both of the count value in the TAR register and the ratio value
of the timer clock divider. The TAIE bit which is bit 1 of the TACTL register is used to enable the
Timer_A interrupt. The TAIFG bit which is bit 0 of the TACTL register is the Timer_A interrupt flag
bit.
13
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
15
14
13
12
11
10
9
unused
7
6
5
IDx
8
TASSELx
4
MCx
3
2
1
0
unused
TACLR
TAIE
TAIFG
Table 2.4 The function of each bit in the TACTL register [8]
Table 2.5 below shows the function of each bit in TAR, the Timer_A register. The 16-bit count
value of Timer_A is stored into the TAR register, which occupies every bit of the TAR register.
15
14
13
12
11
10
9
8
3
2
1
0
TARx
7
6
5
4
TARx
Table 2.5 The function of each bit in the TAR register [8]
Table 2.6 below shows the function of each bit in TACCRx, the Timer_A capture/compare register
x. Whether the compare mode or the capture mode is used can be chosen by the CAP bit in the
TACCTLx register. If choosing the compare mode, the data stored into the TACCRx register will be
compared with the count value in the Timer_A register. If choosing the capture mode, the count
value in the Timer_A register will be copied into the TACCRx register when a capture is
performed.
15
14
13
12
11
10
9
8
3
2
1
0
TACCRx
7
6
5
4
TACCRx
Table 2.6 The function of each bit in the TACCRx register [8]
Table 2.7 below shows the function of each bit in TACCTLx, the capture/compare control register
x. The CAP bit which occupies bit 8 is used to choose whether the compare mode or the capture
mode is employed for the TACCRx register. The CCIE bit which occupies bit 4 is to enable the
capture/compare mode. The CCIFG bit which is bit 0 is the capture/compare interrupt flag bit.
15
14
13
CMx
7
12
11
10
9
8
SCS
SCCI
unused
CAP
4
3
2
1
0
CCIE
CCI
OUT
COV
CCIFG
CCISx
6
5
OUTMODx
Table 2.7 The function of each bit in the TACCTLx register [8]
14
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
2.3.6 SD24_A
2.3.6.1
SD24_A structure
The MSP430AFE233 has three multichannel 24-bit sigma-delta analog-to-digital converters.
Figure 2.6 below shows the internal structure of the multichannel SD24_A analog-to-digital
converters inside the MCU MSP430AFE233. There are eight fully differential multiplexed analog
input pairs in each converter channel. Among the eight fully differential analog input pairs, A0 to
A5 are used to accept the external input analog signal. A5 is used to measure the working supply
voltage for the MCU with the resistive divider circuitry and A6 is used to measure the room
temperature by connecting a temperature sensor inside the MCU. The analog input module is
followed by the second-order oversampling sigma-delta modulator and the digital decimation
filter. The input analog differential value will be converted to a 1 bit data stream by the single-bit
comparator with the modulator frequency fM which can reach 1.1MHz inside the second-order
sigma-delta modulator.
Figure 2.6 The internal structure of the multichannel SD24_A analog-to-digital converters [8]
15
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
2.3.6.2 SD24_A registers
There are a series of registers inside the SD24_A module. These registers also need be configured
in software before the task starts.
Table 2.8 below shows the function of each bit in SD24CTL, the SD24 control register. The
SD24SSELx bits which occupy bit 5 and bit 4 will select the clock source from MCLK, SMCLK, ACLK
or external TACLK. Both of the SD24XDIVx bits which occupy bit 11 and bit 10 and the SD24DIVx
bits which occupy bit 7 and bit 6 can divide the clock to generate the modulating frequency f
for the sigma-delta modulator together. The SD24REFON bit which occupies bit 2 is used to select
whether the on-chip reference voltage 1.2 V is turned on or turned off for each SD24_A channel
by setting SD24REFON = 1 or SD24REFON = 0 .
15
14
13
12
11
reserved
7
6
9
SD24XDIVx
5
SD24DIVx
10
4
SD24SSELx
8
SD24LP
3
2
1
0
SD24VMIDON
SD24REFON
SD24OVIE
reserved
Table 2.8 The function of each bit in the SD24CTL register [8]
Table 2.9 below shows the function of each bit in SD24INTCTLx, the SD24 channel x input control
register. The SD24INCHx bits which occupy bit 2, bit 1 and bit 0 will select which differential input
pair to be used in the corresponding SD24_A channel. The SD24GAINx bits which occupy bit 5, bit
4 and bit 3 are used to choose the gain value of the preamplifier for the differential input AC
signal from 1, 2, 4, 8, 16 or 32 before the differential input AC signal coming into the 2nd
sigma-delta modulator. The SD24INTDLYx bits which occupy bit 7 and bit 6 are used to choose
how many samples delay for the first interrupt after the analog-to digital conversion starts. The
software variable will read the converted value from the SD24MEMx register in the SD24
interrupt service routine every time the SD24_A interrupt triggers. It needs to consume some
time to get the digital result from the analog-to-digital conversion, when the first interrupt
generates immediately just after the analog-to-digital conversion starts, the software variable will
read the empty value from the SD24MEMx register if the first digital result has not been got from
the analog-to-digital conversion. Thus, adding this interrupt delay between the analog-to digital
conversion starts and the first interrupt generates can avoid the data conversion confusion and
make the software work more stably.
7
6
SD24INTDLYx
5
4
3
SD24GAINx
2
1
0
SD24INCHx
Table 2.9 The function of each bit in the SD24INTCTLx register [8]
Table 2.10 below shows the function of each bit in SD24CCTLx, the SD24 channel x control
register. The SD24UNI bit which occupies bit 12 is to select whether the bipolar mode or unipolar
16
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
mode is used to accept the input AC signal. If choosing the unipolar mode, the analog input pair
in the SD24_A can only be entered into the positive signal, while the bipolar mode can deal with
both of the positive signal and negative signal. The SD24XOSR bit which occupies bit 11 and the
SD24OSRx bits which occupy bit 9 and bit 8 will decide the oversampling ratio together to get the
sampling frequency f from the modulating frequency f for the decimation digital filter. The
SD24LSBACC bit which occupies bit 6 is used to choose whether the 16 most significant bits or
the 16 least significant bits of the output conversion value from the decimation digital filter will
come into the SD24MEMx register. The SD24LSBTOG bit which occupies bit 7 is used to choose
whether the SD24LSBACC bit will be toggled or not be toggled each time the SD24MEMx register
is read by software. The SD24IFG bit which occupies bit 2 is the SD24_A interrupt flag bit, which
is set every time the new conversion result data is written into the SD24MEMx register, that
means the SD24_A interrupt will be triggered every time the converted data from the digital filter
output is stored into the SD24MEMx register. This bit will be cleared automatically after the data
in the SD24MEMx register is read by the CPU. Nevertheless, it is better that SD24IFG is cleared by
using the software code every time the SD24MEMx register is read by the CPU to avoid the data
confusion conflict in the small time interval. The SD24GRP bit which occupies bit 0 is used to
choose whether the corresponding channel works independently in single channel mode or the
corresponding channel is grouped with higher channel to work together. The SD24SNGL bit which
occupies bit 10 is used to select whether the single data conversion mode or the continuous data
conversion mode is selected. If choosing single conversion mode, the SD24SC bit will be cleared
automatically after one analog-to-digital conversion is completed then the conversion procedure
will be stopped immediately, while the conversion will continue until the SD24SC bit is cleared by
software in the continuous conversion mode. After the analog-to-digital conversion in the
channel is stopped, the channel will be powered down and the corresponding digital filter will be
turned off. Since the value in the SD24MEMx register can change when the analog-to-digital
conversion is stopped, it is better that the conversion data is read by the SD24MEMx before the
SD24SC bit is cleared to avoid reading an invalid result. The SD24BUFx bits which occupy bit 14
and bit 13 are used to choose whether the high impedance input buffer is enabled or disabled in
the SD24_A analog-to-digital converter. The SD24OVIFG bit which occupies bit 5 is the SD24_A
overflow interrupt flag to indicate whether the SD24_A overflow interrupt is pending or not
pending in the channel. The SD24DF bit which is bit 4 will choose whether the SD24_A conversion
data is in offset binary format or in two’s complement format by setting SD24DF = 0 or SD24DF =
1. The SD24IE bit which is bit 3 will enable the SD24_A interrupt. The SD24SC bit which is bit 1
will start the analog-to-digital conversion.
15
reserved
14
13
SD24BUFx
12
11
10
9
SD24UNI
SD24XOSR
SD24SNGL
8
SD24OSRx
7
6
5
4
3
2
1
0
SD24LSBTOG
SD24LSBACC
SD24OVIFG
SD24DF
SD24IE
SD24IFG
SD24SC
SD24GRP
Table 2.10 The function of each bit in the SD24CCTLx register [8]
Figure 2.7 below shows the single channel operation by using the continuous conversion mode. It
can be seen clearly that the conversion will continue until the SD24SC bit is cleared by software.
17
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 2.7 The diagram of the single channel operation using the continuous conversion mode [8]
Table 2.11 below shows the function of each bit in the SD24MEMx, the SD24_A channel x
conversion memory register. This register will store the 16-bit converted digital data result from
output of the digital filter. According to the value of the SD24LSBACC bit in the SD24CCTLx
register, the 16 most significant bits or the 16 least significant bits of the conversion result data
from the decimation digital filter output will be stored in the SD24MEMx register.
15
14
13
12
11
10
9
8
2
1
0
conversion result
7
6
5
4
3
conversion result
Table 2.11 The function of each bit in the SD24MEMx register [8]
2.3.6.3 Analog-to-digital conversion in SD24_A
According to the different oversampling ratios set in the decimation filter, the output value from
the filter ranges from 15 bits to 30 bits. When setting the oversampling ratio equal to 256 and
SD24LSBACC = 1 , the output value of the digital decimation filter is 24 bits and the 16 most
significant bits will come into the SD24_A conversion memory register. In the programmable gain
amplifier, the value of gain can be selected from 1, 2, 4, 8, 16 or 32. The relation of the full scale
input voltage range and the reference voltage is:
V
=
A
⁄
(2.1)
A
The resolution is :
resolution ∆=
n−
(2.2)
Figure 2.8 below shows the relationship of the digital output value and the input analog voltage
value using the bipolar mode and 2’s complement data conversion mode.
18
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 2.8 The relationship of the digital output value and the input analog voltage value
The converted digital value is:
i
converted digital value =
ia
a
a a
i
∆
i
ai
= calculated analog value
(2.3)
2.3.6.4 Digital filter in SD24_A
The digital filter will process the digital 1 bit data stream from the second-order sigma-delta
modulator with using a SINC comb filter.
The transfer function in the Z-Domain is:
×
H z =
− Z−
(2.4)
−Z−
The transfer function in the frequency domain is:
H f = [
i
(
i
×π×
(π ×
f
)
fM
f
)
fM
] = [
×
i (
×π×
i (π ×
f
)
fM
f
)
fM
]
The oversampling ratio (OSR) is the ratio of the modulator frequency f
frequency f , thus OSR =
M
(2.5)
to the sampling
. Figure 2.9 below shows the frequency response of the digital
filter in the SD24 A/D converter. The first filter notch is at f =
M
, as can be seen from Figure
2.9 below, every notch frequency is a multiple of the modulator frequency f .
19
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 2.9 The frequency response of the digital filter in the SD24_A A/D converter [8]
The decimation of the digital bit-stream is completed in the digital filter in each enabled SD24_A
analog-to-digital channel. Then the digital filter can output the new conversion results to the
corresponding SD24MEMx register at the sampling frequency f .
20
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
3. Materials
3.1 Hardware components
3.1.1 MCU
MSP430AFE233 is selected as the microcontroller unit to be used in the Smart-plug electrical
board. MSP430AFE233 has been described in detail in chapter 2.
3.1.2 RF module
The RF module with the chip CC1101 is chosen to transmit and receive the RF message in this
Smart-plug system as a series of advantageous the chip CC1101 possesses, such as small package
size, just 4×4 mm QFN and cheap price. The chip CC1101 is a low-cost sub-1 GHz transceiver
designed for low-power wireless applications [9]. The circuit inside the chip CC1101 is mainly
intended for the ISM (Industrial, Scientific and Medical) frequency bands at 315MHz, 433MHz,
868MHz and 915MHz. In this project, the 868MHz ISM band is used for the wireless
communication between the antennas of the RF modules in different ELIQ devices. Figure 3.1
below shows a photo of the RF module used in this project.
Figure 3.1 The photo of the RF module used in this project
This RF transceiver is integrated with a highly configurable baseband modem. The modem
supports various modulation formats and has a configurable data rate up to 600 kbps [9]. The
chip CC1101 can provide extensive hardware support for packet handling, data buffering, burst
transmissions, clear channel assessment, link quality indication and wake-on-radio [9].
The main operating parameters and the 64-byte transmit/receive FIFOs of CC1101 can be
controlled via an SPI interface [9]. In a typical system, the CC1101 will be used together with a
microcontroller and some other passive components.
Figure 3.2 below shows the schematic of the chip CC1101 and the peripherals circuits comprised
with the additional passive components such as resistors, inductors, capacitors and the crystal
using the 868MHz ISM band.
21
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 3.2 The schematic of the chip CC1101 and peripheral circuits on the RF module [9]
Figure 3.3 below shows the top view of the pin-out of the chip CC1101.
Figure 3.3 The top view of the pin-out of the chip CC1101 [9]
22
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Table 3.1 below shows the functional description of each pin of the chip CC1101.
Table 3.1 The functional description of each pin of the chip CC1101 [9]
23
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 3.4 below shows the internal hardware structure inside the chip CC1101.
Figure 3.4 The internal hardware structure inside the chip CC1101 [9]
The digital interface of this type of RF module has 6 port pins—SCLK, SO (GDO1), SI, CS, GDO0
and GDO2. These port pins will connect to the corresponding pins on the MCU respectively. The
CSn pin is the serial configuration interface to be used for the chip selection. The SCLK pin is the
serial configuration interface to be used for the clock input. The SI pin is the serial configuration
interface to be used for the data input. The chip CC1101 has two dedicated configurable pins, the
GDO0 pin and the GDO2 pin and one shared pin, the GDO1/SO pin that can output internal
status for the software [9]. All of these pins can be used to generate interrupts of the MCU. The
GDO1 pin is shared with the SO pin together. By selecting any other of the programming options,
the GDO1/SO pin will become a generic pin [9]. But when the CSn pin is low, the GDO1/SO pin
will always work as a normal SO pin which is the serial configuration interface to be used for the
data output. The GDO2 pin is the digital output pin for general use such as to be the test signal
interface, the FIFO status signal interface, the indicator of clearing channel, the clock output, the
serial output of RX data and the serial input of TX data [9]. The GDO0 pin is also the digital output
pin for general use, the function of the GDO0 pin is similar as the GDO2 pin except one difference
that the GDO2 pin can only be the serial output of RX data but not the serial input of TX data.
When the RF module is used to transmit the data to another device through the wireless
communication, the data input from the SI pin of the RF module will go into the TX FIFO firstly.
Then after processing the data in the packet handler and interleaver, the data will be modulated
in the modulator followed by the frequency systhesizer. The frequency synthesizer includes a
complete on-chip LC VCO and a 90 degree phase shifter for generating the signal to the PA
(power amplifier). Then after amplifying the power of the signal in the PA, the signal will go to
the positive RF output port pin and negative RF output port pin to the antenna for the wireless
communication.
When the RF module is used to receive the RF signals from another device through the wireless
communication, the received positive RF signal and negative RF signal from the wireless
communication will be amplified by the LNA (low noise amplifier) firstly. Then the positive RF
signal and negative RF signal will be down-converted in quadrature (I and Q) with the
24
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
intermediate frequency in the down-conversion mixers. At intermediate frequency, the I/Q
signals will be transferred to the digital signal by the analog-to-digital converters inside the chip
CC1101. Then after the demodulation in the demodulator and the bit synchronization in the
packet handler, the data will go to the digital interface to be transmitted to the MCU through the
SPI communication.
Figure 3.5 below shows the format of the data packet for the chip CC1101.
Figure 3.5 The format of the data packet for the chip CC1101 [9]
The preamble pattern is an alternating sequence of ones and zeros (10101010… [9]. When
enabling TX, the modulator will start transmitting the preamble bytes firstly. After a number of
preamble bytes have been transmitted, the modulator will send the sync word and then the data
from the TX FIFO if the data is available. If the TX FIFO is empty, the modulator will continue to
send the preamble bytes until the first byte is written to the TX FIFO, then the modulator will
send the sync word and the data [9].
3.1.3 AC/DC Converter
In this project, the switching power device RECOM RAC02-3.3SC is chosen as the AC/DC converter
to transfer the high voltage AC signal to the low voltage DC signal to give the low working supply
voltage for both of the MCU and the RF module. It has a wide input voltage ranges from 80V AC
to 264V AC and a stable output voltage 3.3V DC [10]. The output rated power is 2 Watt so the
maximum output current is 600 mA. The use of this component is to convert 230 VAC from live
wire and neutral wire to 3.3V DC for the MCU and RF module. The RECOM RAC02-3.3SC has 4
pins, two of which are input pins connected to the live wire and neutral wire respectively, while
the other two pins are output pins connected to the pins DVCC and VDD on both of the MCU and
the RF module.
3.1.4 Relay
The electromagnetic relay HKE V6-S-DC3V is used to control the switch on or switch off status of
the electrical appliance. The structure inside this type of relay is SPST, single pole single throw,
which can be seen clearly in Figure 3.6 below, the pole is either normally open or normally closed.
Figure 3.7 shows the footprint of the relay HKE V6-S-DC3V. The coil rated voltage is 3V and the
coil resistance is 45Ωso the rated coil power is 200mW. The relay HKE V6-S-DC3V has four
terminals, two of which will be connected in the live wire, there is a mechanical switch device
which is set between these two terminals to operate the large current via the live wire. The small
25
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
current goes through the coil resistor which is set between the other two terminals will generate
the electromagnetic force to operate the latch on the mechanical switch device. When the
voltage on the coil achieves at the coil pick-up voltage 2.25V, the mechanical switch device will be
closed under the electromagnetic force from the coil then the electrical appliance will be
connected to the household mains; when the voltage on the coil decreases below 0.3V, the
mechanical switch device will be opened then the electrical appliance will be isolated from the
household mains.
Figure 3.6 The internal hardware connection diagram of the relay HKE V6-S-DC3V [10]
Figure 3.7 The footprint of the relay HKE V6-S-DC3V [10]
3.1.5 NPN transistor
The transistor BC847C is chosen as a control component in the control unit circuit to switch on or
switch off the power supply to the electrical appliance. It is a 3-pin NPN transistor whose
termination type is SMD and the footprint is SOT-23. The collector-emitter breakdown voltage is
45V and the collector-emitter saturation voltage is 200mV. The maximum value of the continuous
collector current I is 100mA. The power dissipation of the NPN transistor BC847C is 250mW
[11].
3.1.6 Crystal
All of the I/O ports in the MSP430AFE233 MCU have multiplex functions besides general purpose
input or output pins function. P2.6 and P2.7 can be interfaced with the external crystal oscillator.
P2.6 is the input terminal of external crystal and P2.7 is the output terminal of external crystal.
The MULTICOMP-HC49U-3.2768-20-30-60-12-ATF-CRYSTAL is used as the external crystal to
connect P2.6 and P2.7 of the MCU in this project. The frequency of this type crystal is 3.2768MHz.
It can output 3.2768MHz frequency with extremely high stability for the MCU to be selected as
the clock source for the ACLK, MCLK or SMCLK inside the MCU. The number of pins of the
MULTICOMP-HC49U-3.2768-20-30-60-12-ATF-CRYSTAL is 2 and the crystal mounting type of it is
26
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
through hole [12].
3.1.7 Other components
Table 3.2 below shows the other components used in this Smart-plug device system.
Name of the
component
Type
Termination type
Footprint
Quantity
Ferrite beads
BLM21BD121SN1D
SMD
0805
2
Varistor
B72214S0271K101
Axial Leaded
1
( 275VAC , 4500A )
TVS (Transient voltage
SMAJ6.0CA
SMD
SCHURTER – FUSE
SMD
0805
1
suppressor)
Fuse
1
( 250VAC, 1A )
LED
MULTICOMP - LED, HIGH
SMD
0805
1
SMD
1206
3
RED
Resistors
kΩ
Resistor
. kΩ
SMD
0805
1
Resistor
. kΩ
SMD
0805
1
Resistors
Ω
SMD
0805
1
Resistors
Ω
SMD
0805
1
Resistors
Ω
SMD
0805
2
Resistors
Ω
SMD
0805
1
. Ω
SMD
1206
1
Resistor
Metal film Resistor
.
LOB5 R010 FLF
Axial Leaded
1
Ω
Capacitor (330uF)
SMD
7243
1
Capacitors (10uF)
SMD
0805
2
Capacitors (0.1uF)
SMD
0805
3
Capacitors (15nF)
SMD
1206
2
Capacitors (47pF)
SMD
0805
4
Capacitors (12pF)
SMD
0805
2
SMD
0805
5
Diodes
IN4148
Table 3.2 The overview of the other components used in this project
27
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
3.2 Software tools
3.2.1 Altium Designer 6
Altium Designer 6 is an electronic design automation tool to plot the printed circuit board with
related libraries [13]. This EDA tool is developed by the company Altium Limited in Australia. It
has two major parts: schematic design and PCB design. In the schematic design part, the
electrical circuit schematic is drawn by setting design rules, search and selection of the electrical
components from the component library, placement of components and connecting the
components [13]. In the PCB design part, the printed circuit board graph file will be completed by
drawing footprints of the components, choosing footprints of the components from the
component footprint library, placement of footprints, manual or automatic trace routing in
different layers, signal integrity analysis and manufacturing files generation with support for
Gerber formats [13].
3.2.2 Switcher CAD Ⅲ
Switcher CAD Ⅲ is a computer aid design software produced by the company Linear Technology
Corporation [14]. It is a suitable simulation tool to implement the circuit feasibility analysis. The
simulation of the designed electrical circuit is executed in software by the simulator with a
schematic module and waveform viewer [14]. The data results can be seen clearly in the
waveform viewer after running the schematic design.
3.2.3 IAR Embedded Workbench
IAR Embedded Workbench is a high performance C/C++ compiler and debugger tool suite for the
8 bits, 16 bits or 32 bits microcontrollers [15]. It has an integrated development environment
including a compiler, an assembler, a linker and a debugger with user-friendly interface [16]. In
addition, IAR Embedded Workbench provides an uninterrupted workflow and a single toolbox in
which all of the components can integrate seamlessly. The objective code can be generated fast
and the execution time for developing low power applications is short in this advanced and
powerful software tool.
IAR Embedded Workbench supports a large number of different processors including MSP430
microcontroller families. In this project, IAR Embedded Workbench for TI MSP430 Kickstart
evaluation edition is used to debug and compile the C programming code for MSP430AFE233.
3.2.4 Lite Fet-Pro430 MSP430 flash programmer
The Lite Fet-Pro430 MSP430 flash programmer which is developed by Elprotronic Inc is used to
download the program into the MCU via JTAG interface. It is a software package to operate with
28
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
existing programming adapters such as MSP430-FET (using parallel port), MSP-FET430UIF (using
USB port) or Olimex' MSP430-JTAG [17]. In this project, MSP-FET430UIF using USB port is chosen
as the programming adapter.
29
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
4. Hardware design
As outlined in the basic design principles in chapter 2, both of the voltage and current on the
electrical appliance need be measured firstly in order to calculate the power consumption of the
appliance. If selecting unipolar mode in the SD24CCTLx register, the input voltage can only be the
positive value. Since the basic prototype is designed without the need of level shifter circuitry to
do the bipolar-unipolar conversion, the bipolar mode need be selected by setting SD24UNI = 0
in the SD24CCTLx register. The on-chip reference voltage 1.2V is turned on by setting SD24REFON
= 1 in the SD24CTL register to implement the analog-to-digital conversion.
4.1 Power supply circuit design
Figure 4.1 below shows the schematic of the power supply circuit inside the Smart-plug device.
Figure 4.1 The schematic of the power supply circuit of the Smart-plug system
Since the working voltage of the MCU ranges from 1.8V to 3.6V but the voltage from the mains
between live wire and neutral wire is 230V AC. An AC/DC converter needs to be added to transfer
the high AC voltage to the low DC voltage for both of the MCU and RF module. RECOM
RAC02-3.3SC is chosen as the AC/DC converter to supply low DC voltage. It has a wide input
voltage ranges from 80 V AC to 264V AC and a stable output voltage 3.3V DC. The output power is
2 Watt so the output current can reach 600 mA to give enough drive current for normal
operation of both of the MCU and the RF module. The RECOM RAC02-3.3SC has 4 pins, two of
which are input pins VACin(L) and VACin(N) connected to the live wire and neutral wire
respectively, other two pins are output pins Vout+ and Vout-.
There is a fuse SCHURTER – FUSE (250V AC, 1A) whose rated current is 1A set between the live
wire and the VACin(L) pin of the AC/DC converter before the 230V AC signal on the live wire going
into the AC/DC converter. When the current through the fuse from the mains to the VACin(L) pin
exceeds 1A, the metal wire inside this fuse will melt to prevent the input ports of the AC/DC
converter from being broken by the large current resulted from the short circuits, overloading,
mismatched loads, or device failure.
30
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
In addition, there is a varistor B72214S0271K101 (275V AC , 4500A) set between the VACin(L) pin
and the VACin(N) pin of the AC/DC converter as equivalent as between the live wire and neutral
wire from the mains outside. The varistor is an electronic component with a nonlinear current–
voltage characteristic as similar as a diode. The feature of the varistor is to conduct significantly
increased current when voltage is excessive. When the voltage on the varistor increases more
than a specific value, the resistance of the varistor will decrease a lot to be similar as the
conductive state to shunt the current which is created by the high voltage away. In this case, the
varistor can protect the AC/DC converter against excessive transient surge voltage.
The Vout+ pin is connected to the DVCC pin on the MCU and the VDD pin on the RF module. It is
also connected to the AVCC pin on the other side of the MCU via a 10Ω resistor. The Vout- pin is
connected to the DVSS pin, the AVSS pin on the MCU, the GND pin on the RF module and also the
board ground. In this design, the supply voltage for both of the MCU and the RF module is 3.3V
DC. The polarity capacitor C2 and the common capacitor C3 are set in parallel between the DVCC
and the board ground to remove the possible residual AC noise waves from the output pins of
the AC/DC converter. As similar as C2 and C3, the polarity capacitor C13 and the common
capacitor C12 are also set in parallel between the AVCC and the board ground to remove the
possible residual AC noise waves.
4.2 Voltage sampling circuit design
Since the effective value of the voltage from the household mains is not stable 230V AC
consistently but can fluctuate between 210V and 250V, it is necessary to measure the real voltage
value on the electrical appliance at the moment. Figure 4.2 below shows the schematic of the
designed voltage sampling circuit which is used to measure the effective voltage value on the
electrical appliance inside the Smart-plug device.
Figure 4.2 The schematic of the voltage sampling circuit of the Smart-plug system
31
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
As can be seen from Figure 4.2 above, the channel 0 of the SD24_A converters system is used as
input for the instantaneous voltage samples which is divided by the dividing resistor circuit, the
board ground is set to connect the neutral wire. The bipolar mode is selected and the internal
reference voltage is 1.2V, thus the full scale input voltage is ±0.6V. Since the A/D converter of the
MCU can only convert a small voltage value, the resistor divider circuitry needs to be added
between live neutral wire and the MCU. For the resistors, there should be 10% of maximum
voltage as margin. In order to verify the input voltage value does not exceed the critical value of
the A/D converter module to break the MCU, the input voltage should range from -0.54V to 0.54V.
Since the AC signal from the household mains may reach 250V AC, which triggers the peak value
is
× √
=
.
V, the divisor value of the voltage dividing circuit should satisfy that:
Divisor ≥
.
.
≈
(4.1)
That means the divisor value of the voltage dividing circuit should be at least 655. According to
this analysis, three identical 330kΩ resistors and one 1.5kΩ resistor are chosen to comprise the
dividing circuitry as Figure 4.2 above shows. The more resistors are deployed, the lower voltage
each resistor bears. In view of the thermal noise of resistors with high resistance, three identical
resistors R5, R6 and R7 are employed. Thus, the divisor value in the design is:
Divisor =
+
+
.
+ .
=
(4.2)
With this voltage dividing circuit, the input voltage to the A/D converter module of the MCU can
be kept between -0.54V and 0.54V to verify ensure a safe operation range for the MCU chip.
The resistor R9 and capacitor C6, as similar as the resistor R10 and the capacitor C7, constitute
the RC low pass filter circuit. Since there may be some high frequency flooding waves left
combined with the DC voltage signal together, the aliasing may occur in the circuit. The RC low
pass filter can remove the interrupts from these high frequency flooding waves.
4.3 Current sampling circuit design
Figure 4.3 below shows the schematic of the designed current sampling circuit which is used to
measure the effective current value on the electrical appliance inside the Smart-plug device.
32
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 4.3 The schematic of the current sampling circuit of the Smart-plug system
As can be seen from Figure 4.3 above, channel 1, another channel of the SD24_A
analog-to-digital converters system in MSP430AFE233, is used to capture the current value on the
instantaneous electrical appliance. R16 is the sampling resistor with extremely small resistance
value set in the main circuit. Since the current value on the electrical appliance needs to be
measured in order to calculate the power consumption, the voltage value on the sampling
resistor can be measured firstly. The Smart-plug device is designed to measure the electrical
appliance with the power consumption range from 2W to 3000W, that means the maximum
effective current value on the live wire and neutral wire can reach:
≈
A
(4.3)
As the input voltage to the SD24_A analog-to-digital converter should range from -0.54V to 0.54V,
so the maximum resistance value of the sampling resistor is:
.
√ ×
A
≈ .
Ω
(4.4)
If choosing the sampling resistor as 0.029Ω, the power consumption on the sampling resistor is:
× .
≈ . W
(4.5)
This power consumption value is quite large, which will make the Smart-plug device too hot, thus
a resistor with smaller resistance value should be chosen here as the sampling resistor. In order
to simplify the power value calculation, the resistor with a resistance value 0.01Ωis selected to
be the sampling resistor. In this case, if the 3000w electrical appliance is connected to the
Smart-plug device, the power consumption value on the sampling resistor is:
33
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
× .
= .
W
(4.6)
This power consumption value is suitable for the product.
The maximum voltage value on the sampling resistor can reach:
×√ × .
≈ .
V
(4.7)
This voltage value is inside the safe range as the input voltage value to the SD24_A
analog-to-digital converter module of MSP430AFE233.
Considering the thermal margin for the application, LOB5 R010 FLF, the 0.01Ωcurrent sense
resistor, with the rated power 5W, resistor tolerance ±1% and the metal film resistor element
type is chosen to be the sampling resistor in the Smart-plug device.
As can be seen in Figure 4.3 above, there are four IN4148 diodes set before the voltage signal on
the sampling resistor going to the MCU. Two of the four diodes are set in parallel with the
sampling resistor and the other two of the four diodes are set in the same way. The use of these
four IN4148 diodes is to clamp the input voltage to the analog inputs of the A/D converter inside
the MCU in a small voltage range. There may be some large spikes or surges in the live wire and
neutral wire from the mains, triggering the voltage value on the sampling resistor to exceed the
safe range of the analog inputs largely to break the MCU. Since the minimum threshold forward
voltage of the diode IN4148 is about 0.6V DC, if connecting the four diodes as Figure 4.3 above
shows, the voltage value of the two points on both of the two sides of the sampling resistor will
be clamped in a range from -0.6V to 3.9V. If the voltage value on either point of two sides of the
sampling resistor exceeds this range, one of the four diodes will face a breakdown to make the
current flow through the diode instantly in order to prevent the MCU from being damaged.
There are also a transient voltage suppressor SMAJ6.0CA which is called D7 in the figure 4.3 and a
burden resistor R15 set in parallel with the sampling resistor. The transient voltage suppressor
can protect the MCU from being damaged by the transient overvoltage in the circuit. Although
the 0.01Ωresistor is chosen as the sampling resistor and the maximum input voltage value to the
SD24_A analog-to-digital converter module of the MCU is 0.184V less than the full scale input
voltage 0.6V, there still exist some voltage spikes which may exceed the safe voltage range of the
MCU in the signal. The SMAJ6.0CA is a type of bidirectional transient voltage suppressor to deal
with the AC overvoltage signal. The circuit inside the bidirectional transient voltage suppressor
can be equivalent to two mutually opposing avalanche diodes in series with one another. When
the voltage on the device is lower than the avalanche breakdown voltage, the device will
represents high impedance. When the voltage on the device exceeds the avalanche breakdown
voltage, the impedance will decrease a lot instantly then the excess current will be shunted by
the device. In addition, all of the overvoltages above the breakdown voltage will be suppressed
less than the value of clamping voltage. The transient voltage suppression diode can respond to
the overvoltages much faster than other common over-voltage protection components such as
the varistor and absorb a large amount of the transient energy internally. When the overvoltage
goes away, the transient voltage suppressor can reset automatically.
34
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
There are two ferrite beads BLM21BD121SN1D in series with the anti-aliasing resistors R20 and
R21 respectively. When the signal through the ferrite bead BLM21BD121SN1D is low frequency,
there is almost not any impedance. However, if the ferrite bead BLM21BD121SN1D encounters
with high frequency signal, the resistance element of the ferrite bead will become dominant. As
this characteristic, it can be used to remove the high frequency noise and some spike disturbance
from the 50Hz AC signal.
As similar as the voltage sampling circuit, two RC low pass filter circuitries made up of the resistor
R13 and the capacitor C9, the resistor R14 and the capacitor C10 are also added before the signal
going into the analog input pairs of the MCU to avoid the signal aliasing in the current sampling
circuit.
4.4 Switching on/off control unit circuit design
The schematic of the control unit circuit which is used to switch on or off the power supply to the
connected appliance is shown in Figure 4.4 below.
Figure 4.4 The schematic of the control unit circuit
The electromagnetic relay HKE V6-S-DC3V is set in the live wire. As described in section 3.1.2,
there is a mechanical switch device to operate the large current via the live wire inside the
internal structure of this type of relay. The current goes through the coil resistor will generate the
electromagnetic force to operate the latch on the mechanical switch device. If the voltage on the
coil reaches the coil pick-up voltage 2.25V, the mechanical switch device will be closed under the
electromagnetic force from the coil then the electrical appliance will be connected to the
household mains; however, if the voltage on the coil decreases below 0.3V, the mechanical
35
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
switch device will be opened then the electrical appliance will be isolated from the household
mains. The I/O port pin P1.2 in the MCU is connected to the base pin of the NPN transistor. When
the P1.2 output is set to be 1 in software, the output voltage of the P1.2 is near 3.3V DC since
the power supply to the MCU is 3.3V DC. In this case, the voltage between the base pin and the
emitter pin will exceed the base-emitter saturation voltage of the BC847C. Then the circuit
between the collector pin and the emitter pin will conduct. Since the emitter pin is connected to
the board ground, the electric potential of the collector will be also near the board ground in this
case. Then there will be the electric potential difference between the point Vout+ of the A/D
converter and the collector of the BC847C, the current on the coil of the relay will increase
significantly. From the simulation waveform by using the software tool Switcher CAD Ⅲ, it can
be seen that the current on the coil is 64.8mA at this moment, so the voltage on the coil is:
. mA ×
Ω ≈
.
V
(4.8)
Now the voltage on the coil is more than the pick-up voltage 2.25V, thus the mechanical switch
device will be closed under the electromagnetic force from the coil, the electrical appliance will
be conducted as it is in the switch on status connected to the household mains.
When the P1.2 output is set to be 0 in software, the output voltage of the P1.2 is almost 0 as
the board ground. The circuit between the collector and the emitter will be isolated. There is no
voltage potential difference between the point Vout+ of the A/D converter and the collector of
the BC847C, so the current on the coil will be extremely small, almost 0, then the voltage on the
coil decreases below 0.3V, the mechanical switch device will be opened and the electrical
appliance will be isolated from the household mains.
P1.1 is connected to a LED and a 200Ω resistor. This pin is used to turn on or shut off the LED to
indicate the switch on or switch off status of the electrical appliance. When the electrical
appliance is switched off from the power supply by the relay, the output of P1.1 is 1 , that
means then the output voltage from this pin is close to 3.3V DC as the VCC and the LED will be
turned on. When the electrical appliance is switched on, the output of P1.1 is 0 , that means the
output voltage from this pin is close to 0 as the VDD, triggering the LED be shut off.
The diode IN4148 is set in parallel with the relay to prevent electric spark when the P1.2 output
value is toggled from 1 to 0 . As Figure 4.4 above shows, the positive pole of the IN4148 is
connected to the collector pin of the NPN transistor and the negative pole of the IN4148 is
connected to the Vout+ of the A/D converter. When the P1.2 output value is toggled from 1 to
0 , the NPN transistor will be toggled to the cutoff status from the saturation status
instantaneously, the circuit link between the collector pin and the emitter pin will be isolated. As
the coil is similar as a inductor and the effect of an inductor in a circuit is to oppose changes in
current through it by developing a voltage across it proportional to the rate of change of the
current, the instantaneous change of the current on the coil will generate the electromotive force
to make the coil work as a voltage source. There will still exist the voltage differential between
the point Vout+ of the A/D converter and the point collector pin of the NPN transistor, the
residual current will not disappear soon, so the electric spark may be generated on the coil. If
setting the diode IN4148 here, the coil and the diode IN4148 can compose a circuit loop, the
36
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
residual current can flow away through the diode IN4148 to prevent the coil from being damaged
by the electric spark.
4.5 Communication between MCU and RF
module
Figure 4.5 below shows the schematic of the designed circuit of the connection between the
MCU and the RF module.
Figure 4.5 The schematic of the communication circuit between MCU and RF module
The RF module has 8 port pins, 6 of them consist of the digital interface to connect the MCU and
other 2 port pins are VDD and GND connected to the Vout+ pin of the AC/DC converter and the
board ground respectively. During the 6 port pins of the digital interface, SI pin is connected to
the P1.5 (SIMO) pin of the MCU, SO pin is connected to the P1.6 (SOMI) pin of the MCU, SCLK pin
is connected to the P1.7 (UCLK) pin of the MCU, CS pin is connected to the P1.0 pin of the MCU,
these 4 pins will implement the 4-wire SPI communication together. The SCLK pin is the serial
configuration interface to be used for the clock input, the SI pin is the serial configuration
interface to be used for the data input, the SO pin is the serial configuration interface to be used
for the data output and the CS pin is the serial configuration interface to be used for the chip
selection.
The GO pin is connected to the P1.4 (URXD) pin of the MCU, the GIO pin is connected to the P2.0
pin of the MCU, meanwhile, the GO pin is connected to the GDO2 pin of the chip CC1101 and the
GIO pin is connected to the GDO0 pin of the chip CC1101 on the RF module. Both of the GDO2
pin and the GDO0 pin are the digital output pins for general use such as to be the test signal
interface, the FIFO status signal interface, the indicator of clearing channel and the clock output.
However, the GDO0 pin can be used as both of the serial output of RX data and the serial input of
TX data while the GDO2 pin can only be used as the serial output of RX data. From Figure 4.5
above, it can be noticed that there is a small empty pad outside the MCU on the PCB board
connected to the port pin P1.3 (UTXD) of the MCU. This small empty pad can be used to detect
37
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
whether the software code of the communication between the MCU and the RF module is
running successfully or not by writing some test code in software and using an oscilloscope as the
detection tool.
38
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
5. Software design methodology
Figure 5.1 below shows the flowchart of the software process. The details of the software
execution process will be explained specifically in the following sub chapters.
Reset
Peripherals initiation (Clock,
SD24_A, I/O pins, Timer, WDT)
No
SD24_A interrupts
3200 per second
3200 squared voltage and current value
accumulated? Wait for acknowledgement from
SD24_A interrupt
Read converted digital value of voltage and
current from SD24MEMx register every interrupt
Yes
Square every instantaneous voltage value
and current value
Acquire accumulation of squared voltage
and current value in one second from
SD24_A interrupt
Accumulate squared voltage value and
current value
Calculate RMS value of
voltage and current
3200 samples accumulated? One
second read 3200 samples
Yes
Calculate power
consumption value
Notify main function to process
accumulation of squared
voltage value and current value
Send packet data contains power
consumption value to RF module through
SPI communication
Return from SD24_A
interrupt
Figure 5.1 The flowchart of the implemented software
39
Yixin Liu
May 26th, 2014
No
Design and implementation of the ELIQ Smart-plug device system
5.1 Power consumption value calculation
5.1.1 Peripherals initiation in MSP430AFE233
In the MCU MSP430AFE233, the major peripherals are the SD24_A A/D converters, the clock
system, the timer and the watchdog timer. After a device RESET, the initial setup of these major
peripherals will be implemented in software firstly. The initialization routines include the setup of
the SD24_A analog-to-digital converters, clock system, general purpose input/output port pins,
timer and watchdog timer.
5.1.1.1 Clock system initiation
In the BCSCTL1 register, the XT2OFF bit is used to choose whether the XT2 oscillator is turned on
or turned off. Since XT2CLK needs to be used for MCLK and SMCLK, the XT2 oscillator should be
turned on by clearing the XT2OFF bit.
7
6
XT2OFF
XTS
5
4
3
2
DIVAx
1
0
RSELx
Table 5.1 The function of each bit in the BCSCTL1 register [8]
In software, the code to initialize the BCSCTL1 register is:
#define XT2OFF (0x80)
BCSCTL1 &= ~XT2OFF;
In the BCSCTL2 register, the SELMx bits will be used to select the clock source for the MCLK and
the SELS bit will be used to select clock source for the SMCLK. XT2CLK is used as the clock source
for MCLK and SMCLK and the XT2 oscillator has already been set present on chip, so the SELMx
bits = 1 0 and the SELS bit = 1 . Neither the MCLK nor the SMCLK needs to be divided, so the
DIVMx bits are equal to 0 0 to make the divider ratio for MCLK is 1. Similarly, the DIVSx bits are
equal to 0 0 to make the divider ratio for SMCLK is 1.
7
6
SELMx
5
4
3
DIVMx
SELS
2
1
DIVSx
0
DCOR
Table 5.2 The function of each bit in the BCSCTL2 register [8]
In software, the code to initialize the BCSCTL2 register is:
#define SELS (0x08)
#define SELM_2 (0x80)
BCSCTL2 |= SELS+SELM_2;
In the BCSCTL3 register, the XT2Sx bits are used to select the frequency range for the XT2 clock.
Since the frequency of the external crystal connected to the XT2 ports is 3.2768MHz, the XT2Sx
40
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
bits are equal to 1 0 to make the external crystal used on the electrical board is suitable for the
selected frequency range from 3MHz to 16MHz.
7
6
XT2Sx
5
4
3
LFXT1Sx
2
XCAPx
1
0
XT2OF
LFXT1OF
Table 5.3 The function of each bit in the BCSCTL3 register [8]
In software, the code to initialize the BCSCTL3 register is:
#define XT2S_2 (0x80)
#define LFXT1S_2 (0x20)
BCSCTL3 |= XT2S_2+LFXT1S_2;
5.1.1.2 I/O ports initiation
P1.0 is connected to the CS pin on the RF module. This pin is used to control the chip selection for
the Radio Frequency module. The chip selection is activated when the CS pin is low. In the P1DIR
register, bit 0 is set to be 1 since controlling the chip selection of the RF module is in the output
direction for the MCU. In the P1REN register, bit 0 is set to be 0 then the pullup/pulldown
resistor for this pin is disabled. Since the chip selection is activated when the CS pin is low, bit 0 is
set to be 1 in the P1OUT register to make the initial state of the chip selection function is
disabled.
As described in section 4.4, the port pin P1.1 is used to turn on or shut off the LED to indicate the
switch on or switch off status of the electrical appliance. In the P1DIR register, bit 0 is set to be 1
since controlling the LED is in the output direction for the MCU. In the P1REN register, bit 1 is set
to be 0 then the pullup/pulldown resistor for this pin is disabled. In the P1OUT register, bit 1 is
set to be 0 to shut off the LED in the beginning.
The port pin P1.2 is used to control the relay to switch on or switch off the power supply to the
electrical appliance, which has also been described in section 4.4 in detail. In the P1DIR register,
bit 2 is set to be 1 since controlling the NPN transistor is in the output direction for the MCU. In
the P1REN register, bit 2 is set to be 0 then the pullup/pulldown resistor for this pin is disabled.
In the P1OUT register, bit 2 is set to e 0 in software in the initial state, then the collector and
the emitter of the NPN transistor BC847C is isolated, there is negligibly small current on the coil
of the relay, so the mechanical switch device will be opened under the electromagnetic force
from the coil and the electrical appliance is switched off to the power supply in the beginning.
P1.3 is connected to a small empty pad outside the MCU on the PCB board. This pad is used to
detect the working state of the communication between the MCU and the RF module. In the
P1DIR register, bit 2 is set to be 1 since detecting the running state of the software by using an
oscilloscope is in the output direction for the MCU. In the P1REN register, bit 2 is set to be 0
then the pullup/pulldown resistor for this pin will be disabled. In the P1OUT register, bit 2 is set
to e 0 to give the initial state 0 for the output of the P1.3 port pin.
41
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
P1.4 is connected to the GO pin of the RF module. On the other side, GO pin is also connected to
the GDO2 pin of the CC1101 chip on the RF module. GO pin is the general output pin of the RF
module, which will send data to the port P1.4 of the MCU, so bit 4 in the P1DIR register is set to
be 0 to make the port pin P1.4 is in the input direction. In the P1REN register, bit 4 is set to be
1 to enable the pullup/pulldown register. In the P1OUT register, bit 4 is set to be 1 to ensure
the port P1.4 is pulled up.
P1.5 is also the SIMO0 pin for the MCU. It is connected to the SI pin of the RF module. Since MCU
is the master machine and RF module is the slave machine, the port pin P1.5 is used to send data
from the MCU to the RF module, then bit 5 in the P1DIR register is set to be 0 to ensure the
port pin P1.5 is in the output direction. In the P1REN register, if bit 5 is set to be 0 then the
pullup/pulldown resistor for this pin is disabled. In the P1OUT register, bit 5 is set to be 1 to
give the initial state of the output of P1.5 as high.
P1.6 is also the SOMI0 pin for the MCU. It is connected to the SO pin of the RF module. Since
MCU is the master machine and RF module is the slave machine, port pin P1.6 is used to receive
data from the RF module, then bit 6 in the P1DIR register is set to be 1 to ensure the port pin
P1.6 is in the input direction. In the P1REN register, bit 6 is set to be 1 to enable the
pullup/pulldown register. In the P1OUT register, bit 6 is set to be 1 to ensure the port pin P1.6 is
pulled up.
P1.7 is connected to the SCLK pin of the RF module. It is connected to the internal clock source
inside the MCU. This pin is used to output the clock signal to control the clock of the RF module
to receive data from the MCU, thus bit 7 in the P1DIR register is set to be 0 to ensure the port
pin P1.7 is in the output direction. In the P1REN register, bit 7 is set to be 0 then the
pullup/pulldown resistor for this pin is disabled. In the P1OUT register, bit 7 is set to be 1 to
give the initial state of the output of P1.7 as high.
P2.0 is connected to the GIO pin of the RF module. On the other side, GIO pin is connected to the
GDO0 pin of the CC1101 chip on the RF module. The GIO pin is the general input/output pin of
the RF module, which will send data to the port P2.0 of the MCU from the RF module, so the bit 0
in the P2DIR register is set to be 0 to ensure the port pin P2.0 is in the input direction. In the
P2REN register, the bit 0 is set to be 1 to enable the pullup/pulldown register. In the P2OUT
register, the bit 0 is set to be 1 to ensure the port P2.0 is pulled up. In this Smart-plug device
system, the I/O interrupt of the port pin P2.0 is used to control the function of repeating the RF
message. Fetching and processing the received data packet from other ELIQ devices via the
wireless communication will be implemented in the interrupt service routine of P2 port, thus the
bit 0 of P2IE register is set to be 1 to enable the I/O interrupt of P2.0 and bit 0 of P2IES register
is set to be 1 to ensure the interrupt edge of P2.0 is from high to low. This means that the
interrupt flag for the port pin P2.0 is set when the high-to-low transition occurs in the signal from
the GIO pin of the RF module to the P2.0 pin of the MCU, while the initial state of the bit 0 of the
P2IFG register is set to be 0 .
42
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
P2.6 and P2.7 are connected to the external crystal oscillator. P2.6 is the input terminal of the
external crystal and P2.7 is the output terminal of the external crystal. As described in Section
3.1.6, the MULTICOMP-HC49U-3.2768-20-30-60-12-ATF-CRYSTAL is used as the external crystal
for the MCU MSP430AFE233. The port pin P2.6 and the port pin P2.7 will receive the clock signal
from the external crystal, so both of bit 6 and bit 7 in the P2DIR register are set to be 0 to
ensure the port pin P2.6 and port pin P2.7 are in the input direction. In the P2REN register, both
of bit 6 and bit 7 are set to be 1 to enable the pullup/pulldown register. In the P2OUT register,
both of bit 6 and bit 7 are set to be 1 to ensure P2.6 and P2.7 are pulled up.
In summary, the initialized value of every bit in P1DIR register, P1OUT register, P1REN register,
P2DIR register, P2OUT register, P2REN register, P2IES register, P2IFG register and P2IE register are
shown in Table 5.4 below:
register
initialized value
P1DIR
10101111
P1OUT
11110001
P1REN
01010000
P2DIR
00000000
P2OUT
11111111
P2REN
00111111
P2IES
00000001
P2IFG
00000000
P2IE
00000001
Table 5.4 The initialization of each bit of the I/O ports registers
In software, the code to configure the P1 I/O ports is:
P1DIR = 0xAF;
P1OUT = 0xF1;
P1REN =0x50;
P2DIR = 0x00;
P2OUT = 0xFF;
P2REN = 0x3F;
P2IES = 0x01;
P2IFG = 0x00;
P2IE = 0x01;
5.1.1.3 SD24_A A/D converter initiation
5.1.1.3.1 Configuring sampling frequency
The SD24_A A/D converter possesses a feature that the SD24_A interrupt is triggered every time
the converted data from the digital filter output is stored into the SD24MEMx register, that
means SD24_A A/D converter can generate the interrupt every sampling instant, which will be
quite convenient for the data sampling process.
43
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
The frequency of the mains voltage signal from outside is 50Hz if the mains voltage is stable.
According to the Nyquist Theorem, a signal waveform must be sampled at least twice as fast as
the frequency of the signal to reconstruct the original signal waveform accurately. However, the
more sampling points in one sine-wave period, the better original waveform can be retrieved,
similarly, if more data samples are captured in one sine-wave period of the signal, the RMS (Root
Mean Square) values of the voltage and current from the sampled instantaneous voltage values
and the sampled instantaneous current values can be calculated more accurately. The RMS values
of both of the voltage and current from the sampled instantaneous voltage values and the
sampled instantaneous current values are regarded to be equal to the effective voltage value and
the effective current value similarly in this project. It is designed to capture 3200 data samples in
one second to calculate the effective voltage value and the effective current value on the
connected electrical appliance, that means in one sine-wave period of the signal, 64 samples will
be captured, which is much larger than the number of sampling points to satisfy the Nyquist
Theorem.
The frequency of the clock from the external crystal is 3.2768MHz. In the SD24CTL register, the
SD24SSELx bits will select which one is the clock source from MCLK, SMCLK, ACLK or external
TACLK. Here SMCLK is chosen as the clock source for the SD24_A, so the SD24SSELx bits are equal
to 0 1 . Both of the SD24XDIVx bits and the SD24DIVx bits can divide the clock together to
generate the modulating frequency fM for the sigma-delta modulator. Here the SD24DIVx bits are
set to be equal to 1 0 and the SD24XDIVx bits are set to be equal to 0 0 to let the SMCLK be
divided by 4. So the modulator frequency fM to the 2nd order sigma-delta modulator in the
SD24_A is:
f
.
=
z
= .
MHz
(5.1)
OSR is the oversampling ratio value in the SD24_A. The sampling frequency fS is defined as
f =
M
. In SD24CCTLx register, the SD24XOSR bit and the SD24OSRx bits will select OSR to be
either 32, 64, 128, 256, 512 or 1024. In this project, OSR is designed to be 256, so the SD24XOSR
bit is set to be 0 and the SD24OSRx bits equal to 0 0 . Since the modulation frequency fM =
0.8192MHz, the sampling frequency fS in the SD24_A is:
f =
M
=
.
z
=
Hz
(5.2)
So the sampling frequency fS in the SD24_A is set to be 3200Hz, that means 3200 data samples
can be input to the SD24MEMx from the output of the digital filter in one second, also, it can be
read 3200 instantaneous voltage value and also 3200 instantaneous current value in one second
in software.
5.1.1.3.2 Configuring analog-to-digital conversion parameters
As described in section 5.1.1.3.1, in the SD24CTL register, the SD24SSELx bits are set to be 0 1 ,
the SD24DIVx bits are set to be 1 0 and the SD24XDIVx bits are set to be 0 0 . The SD24REFON
44
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
bit is used to select whether the on-chip reference voltage is turned on or turned off. In this
project, the on-chip reference voltage 1.2V is designed to be used in the analog-to-digital
conversion, so the SD24REFON bit is set to be 1 .
15
14
13
12
11
reserved
7
6
9
SD24XDIVx
5
SD24DIVx
10
4
SD24SSELx
8
SD24LP
3
2
1
0
SD24VMIDON
SD24REFON
SD24OVIE
reserved
Table 5.5 The function of each bit in the SD24CTL register [8]
In software, the code to initialize the SD24CTL register is:
#define SD24SSEL_1 (0x0010u)
#define SD24DIV_2 (0x0080u)
#define SD24REFON (0x0004u)
SD24CTL = SD24SSEL_1 | SD24DIV_2 |
SD24REFON;
As described in the basic design principle, two analog-to-digital channels need be used to
measure the effective voltage value and the effective current value on the connected electrical
appliance respectively. Thus, SD24_A channel 0 is designed to capture the voltage value and
thSD24_A channel 1 is designed to capture the current value.
In the SD24INTCTLx register, the SD24INCHx bits will select which differential input pair to be
used in the corresponding SD24_A channel. Since the differential input pair A0.0 is designed to
connect the input AC signal in channel 0 and the differential input pair A1.0 is designed to
connect the input AC signal in channel 1, both of the SD24INCH0 bits in the SD24INTCTL0 register
and the SD24INCH1 bits in the SD24INTCTL1 register are set to be 0 0 0 . The SD24GAINx bits are
used to choose the preamplifier gain value for the differential input AC signal from 1, 2, 4, 8, 16
or 32 before the differential input AC signal coming into the 2nd sigma-delta modulator. Here the
differential input AC signal need not be amplified in both of the channel 0 and channel 1 so both
of the SD24GAIN0 bits in the SD24INTCTL0 register and the SD24GAIN1 bits in the SD24INTCTL1
register are equal to 0 0 0 to keep the preamplifier gain as 1. The SD24INTDLYx bits are used to
choose how many samples delay for the first interrupt after the analog-to digital conversion
starts. This interrupt delay can avoid the data conversion confusion and make the software work
more stable. Here both of the SD24INTDLY0 bits in the SD24INTCTL0 register and the
SD24INTDLY1 bits in the SD24INTCTL1 register are chosen to be 0 0 to make the fourth sample
cause the first interrupt after the analog-to-digital conversion starts to avoid the data conversion
confusion as far as possible. Thus, all of the bits in the SD24INCTL0 register and the SD24INCTL1
register will be set to be 0 as the initial state.
7
6
SD24INTDLYx
5
4
3
SD24GAINx
2
1
0
SD24INCHx
Table 5.6 The function of each bit in the SD24INTCTLx register [8]
45
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
In software, the code to initialize the SD24INTCTLx register is:
#define SD24INCH_0 (0x0000u)
SD24INCTL0 = SD24INCH_0;
SD24INCTL1 = SD24INCH_0;
In the SD24CCTLx register, the SD24UNI bit is to select whether the bipolar mode or unipolar
mode is used to accept the input AC signal. If choosing the unipolar mode, the analog input pair
in the SD24_A can only be entered into the positive signal, while the bipolar mode can deal with
both the positive signal and negative signal. Since the voltage shift level circuit is not used
between the live wire, neutral wire and the MCU for both of the voltage sampling circuit and the
current sampling circuit, the bipolar mode is selected by setting the SD24UNI bit = 0 in both of
the SD24CCTL0 register and the SD24CCTL1 register. The SD24DF bit is used to choose whether
the data format of the converted digital signal is offset binary or two’s complement. Since the
converted digital signal is singed binary number from the positive or negative analog signal, two’s
complement data format is selected by setting the SD24DF bit = 1 in both of the SD24CCTL0
register and the SD24CCTL1 register. Since the SD24_A interrupt needs to be used when sampling
the data, the SD24IE bit is set to be 1 to enable the SD24_A interrupt in both of the SD24CCTL0
register and the SD24CCTL1 register. The SD24SC bit will start the analog-to-digital conversion on
the corresponding channel by setting this bit equal to 1 . The SD24LSBACC bit is set to be 0
then the 16 most significant bits of the conversion data from the decimation digital filter output
will come into the SD24MEMx register. The SD24LSBTOG bit in both of the SD24CCTL0 register
and the SD24CCTL1 register are set to be 0 then the SD24LSBACC bit in the SD24CCTL0 register
or in the SD24CCTL1 register will not be toggled each time the SD24MEM0 register or the
SD24MEM1 register is read by software to keep the SD24MEM0 register or SD24MEM1 register
store the 16 most significant bits of the conversion data from the digital filter output. The
SD24GRP bit is set to be 0 to make both of channel 0 and channel 1 work independently in
single channel mode since both of the channel 0 and channel 1 don’t need be grouped with
higher channel to work together in this project. The SD24SNGL bit is set to be 0 in both of the
SD24CCTL0 register and the SD24CCTL1 register to make both of channel 0 and channel 1 work in
the continuous data conversion mode then the data conversion will continue until the SD24SC bit
is cleared by software. Figure 5.2 below shows the single channel operation by selecting the
continuous conversion mode. There is no need to use the high impedance input buffer to deal
with the input signal, so the high impedance input buffer mode is disabled in both of the channel
0 and channel 1 by setting both of the SD24BUF0 bits in the SD24CCTL0 register and SD24BUF1
bits in the SD24CCTL1 register equal to 0 0 .
15
reserved
14
13
SD24BUFx
12
11
10
9
SD24UNI
SD24XOSR
SD24SNGL
8
SD24OSRx
7
6
5
4
3
2
1
0
SD24LSBTOG
SD24LSBACC
SD24OVIFG
SD24DF
SD24IE
SD24IFG
SD24SC
SD24GRP
Table 5.7 The function of each bit in the SD24CCTLx register [8]
46
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 5.2 The diagram of the single channel operation using continuous conversion mode [8]
In software, the code to initialize the SD24CCTLx register is:
#define
#define
#define
#define
SD24OSR_256 (0x0000u)
SD24DF (0x0010u)
SD24SC (0x0002u)
SD24IE (0x0008u)
SD24CCTL0 = SD24OSR_256 | SD24DF | SD24SC | SD24IE;
SD24CCTL1 = SD24OSR_256 | SD24DF | SD24SC | SD24IE;
5.1.1.4 Timer_A initiation
In the TACTL register, the TASSELx bits are set to be 1 0 to choose SMCLK as the Timer_A clock
source. The IDx bits are set to be 0 0 to determine the ratio of the divider is 1. The MCx bits are
set to be 1 0 to make the Timer_A work in the continuous mode. The TACLR bit is set to be 1
to clear the count value in the TAR register and the ratio value of the timer clock divider. The TAIE
bit is set to be 1 to enable the Timer_A interrupt and the TAIFG interrupt request.
15
14
13
12
11
10
9
unused
7
6
5
IDx
8
TASSELx
4
MCx
3
2
1
0
unused
TACLR
TAIE
TAIFG
Table 5.8 The function of each bit in the TACTL register [8]
The count value of Timer_A will be stored into the TAR register.
15
14
13
12
11
10
9
8
3
2
1
0
TARx
7
6
5
4
TARx
Table 5.9 The function of each bit in the TAR register [8]
In the TACCTL0 register, the CAP bit is set to be 0 to choose the compare mode for the TACCRx
register. The CCIE bit is set to be 1 to enable the capture/compare interrupt and the CCIFG
interrupt request.
47
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
15
14
13
CMx
7
12
11
10
9
8
SCS
SCCI
unused
CAP
4
3
2
1
0
CCIE
CCI
OUT
COV
CCIFG
CCISx
6
5
OUTMODx
Table 5.10 The function of each bit in the TACCTL0 register [8]
Since the CAP bit in the TACCTL0 register is set to be 0 , the TACCR0 register will work in the
compare mode. In software, the decimal data 32768 will be stored into the TACCR0 register
initially. This value will compare with the count value in the Timer_A register.
15
14
13
12
11
10
9
8
3
2
1
0
TACCRx
7
6
5
4
TACCRx
Table 5.11 The function of each bit in the TACCR0 register [8]
In the continuous mode, the timer counts up from 0 to 0FFFFh in the TAR register repeatedly. The
decimal value 32768 stored into the TACCR0 register is the independent time interval. Each time
the timer counts up to the value stored in the TACCR0 register, the TACCTL0 register interrupt is
generated. In the interrupt service routine, the next time interval value which also equals to
decimal value 32768 will be added to the value stored in the TACCR0 register. Then if the count
value in the TAR register achieves at the value stored in the TACCR0 register again, the TACCTL0
register interrupt is generated again. Since the SMLK is selected as the Timer_A clock source, that
means the clock frequency for the timer is 3.2768MHz, the time interval between the two
adjacent TACCTL0 register interrupts is:
Time interval =
.
×
×
=
s=
ms
(5.3)
In software, the code to initialize the Timer_A is:
#define TASSEL1
(0x0200u)
#define TACLR
(0x0004u)
#define CCIE
(0x0010u)
#define UEP_CYCLES_PER_RTC_INTERRUPT
#define MC1
(0x0020u)
#define TAIE
(0x0002u)
32768
TACTL = TASSEL1 + TACLR;
// Set clock source and clear TAR (SMCLK, divide by 1)
TACCTL0 = CCIE;
// CCR0 interrupt enabled
TACCTL1 = CCIE;
// CCR1 interrupt enabled
TACCR0 = UEP_CYCLES_PER_RTC_INTERRUPT;
// Set compare register 0
TACTL |= MC1 + TAIE;
// Start Timer_A in continues mode and enable interrupt
48
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
5.1.2 Sampling voltage and current value in one second
The SD24_A analog-to-digital converter has a fully differential input, the maximum full scale input
voltage range is:
±V
= ±
⁄
A
(5.4)
A
The on-chip reference voltage 1.2V is turned on in the SD24_A module, the bipolar mode is
selected and the gain of the preamplifier is selected to be 1, so the maximum input voltage range
now is:
±V
= ±
. ⁄
=± . V
(5.5)
The resolution is :
resolution ∆=
+
.
=
≈
.
mv
(5.6)
The converted digital value is:
converted digital value =
i
ia
a
a a
i
∆
i
ai
= calculated analog value
(5.7)
The output from the digital filter to the SD24MEMx register in the SD24_A is a signed binary
value. The simultaneous instantaneous voltage and current samples are captured by the two
independent SD24_A A/D converters at a sampling rate of 3200Hz. Track of the number of
samples that are present in one second is kept and used to obtain the RMS values for both of the
voltage and current respectively. Here the RMS values of both of the voltage and current are
regarded to be equal to the effective voltage value and the effective current value similarly. As
described in section 4.3, the effective current value on the electrical appliance will be inferred
from the measured effective voltage value on the sampling resistor. Channel 0 of the SD24_A
converters system is used to capture the instantaneous voltage samples which are divided by the
dividing resistor circuit and channel 1 of the SD24_A converters system is used to capture the
instantaneous voltage value on the sampling resistor to infer the instantaneous current value on
the electrical appliance. The input voltage samples in both of the channel 0 and channel 1 of the
SD24_A converters system are collected by using the SD24 interrupts. The function deals with the
timing critical events in software. The SD24 interrupt is generated once every
second. In
the SD24 interrupt service routine, a variable called adc_buffer used as the global data buffer will
read the data value stored in the SD24MEMx register firstly. Then the data will be squared by
using the multiplier to get the squared value. After that step, the squared value will be stored in a
dedicated 48-bit register since the squared value may exceed the range of the 16-bit signed
binary data. When the SD24 interrupt is generated next time and the function goes into the SD24
interrupt service routine again, the new squared value will be accumulated on the original
squared value in the dedicated 48-bit register. As the sampling frequency is 3200Hz, that means
3200 data samples of the input voltage signal in both of the channel 0 and channel 1 can be
49
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
captured in one second. After 3200 squared value of samples have been accumulated in the
dedicated 48-bit register, the calculation of the RMS values of voltage and current in one second
will start. A variable in software will be used as the sample counter to keep track of how many
samples have been accumulated over the frame period. When the value of the sample counter
variable reaches 3200, the main function will be notified to process the accumulation value of
both of the squared instantaneous voltage value and the squared instantaneous current value
from the SD24_A interrupt service routine. Then the software process will be returned from the
SD24_A interrupt service routine.
The formulas to calculate the accumulated squared values of the input voltage signal on both of
the channel 0 and channel 1 are shown below:
���
�
��
�
ℎ�
= ∑
�
=
ℎ�
(5.8)
���
�
��
�
ℎ�
= ∑
�
=
ℎ�
(5.9)
5.1.3 RMS value calculation for voltage and current
After reset and the peripherals initiation, the main function will wait for the acknowledgment
that 3200 squared instantaneous voltage values and 3200 squared instantaneous current values
have been accumulated in the dedicated registers in one second from the interrupt service
routine, then the RMS values of voltage and current in one second can be calculated from these
3200 data samples.
The formulas to calculate the RMS values of both of the voltage and current from the value read
from the SD24MEM0 register of channel 0 and the value read from the SD24MEM1 register of
channel 1 in software are shown below:
=
�
=
�
× √
∑
�
=
��
× √
∑
�
=
��
�
�
� �
�ℎ�
�
(5.10)
�
�
� �
�ℎ�
�
(5.11)
�
�
Where
is the scaling factor for the voltage and � is the scaling factor for the current.
Sample count is the number of samples in one second, which should equal to 3200 under these
configurations as long as the frequency of the mains is stable. The value of
and � can be
inferred backward according to the conversion from the input voltage signal to the analog inputs
of the SD24_A to the value written into the SD24MEMx register in both of the channel 0 and
channel 1 of the SD24_A A/D converters system. The relationship of the mains voltage value and
the value written into the SD24MEM0 register in channel 0 and the relationship of the current
50
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
value on the electrical appliance and the value written into the SD24MEM1 register in channel 1
are:
�
�
ℎ�
=
�
�
ℎ�
=
� ��
(5.12)
.
��
�� �� ∗
.
.
(5.13)
According to the formulas of the relationship between the mains voltage value and the value
written into the SD24MEM0 register in channel 0 and the relationship between the current value
on the electrical appliance and the value written into the SD24MEM1 register in the channel 1
above, the value of the scaling factors
and � can be inferred backward as the formulas
shown below:
=
�
=
.
.
×
×
(5.14)
.
(5.15)
In software, dividing
can be implemented by shifting the value � bits to the right. The
calculation of the RMS value for both of the voltage and current on the electrical appliance is
slow since the calculation steps such as multiplying, square, extraction of a root and shifting bits
are time consuming functions. The time consumed in the complex calculations may exceed the
sampling time interval, which may lead to the data conflict and calculation confusion in the
software process. However, if the calculation of the RMS value is divided into two parts
implemented in the SD24_A interrupt service routine and the main function respectively, the
data conflict can be conflicted in software as far as possible. The calculation of the square and the
accumulation of 3200 data samples are implemented in the SD24_A interrupt service routine,
then the extraction of the accumulation value and multiplying the scaling factors will be
implemented in the main function. The different variables are used to be the dedicated 48-bit
registers to store the accumulation of both of 3200 squared instantaneous voltage values and
3200 squared instantaneous current values. In this way, the software process can be prevented
from being congested and confused.
The reason using 3200 data samples in one second to calculate the RMS values of both of the
voltage and current but not using 64 data samples in one sine-wave period to calculate the RMS
values is that the deviation influence from the noises or spikes in the signal for the calculation
can be avoided as far as possible if using many more sampled points to implement the calculation.
Although the RC low pass filter circuit is added before the signal going into the analog input pairs
of the SD24_A A/D converter of the MCU, there is still residual noise in the signal. If the data
sample points deviated from the estimated value so far away in the signal are happened to be
captured by the SD24_A A/D converter, it will influence the accuracy of the calculated RMS values
results. The less sampled points to implement the calculation of the RMS values, the more
deviation of the accuracy of the RMS values results from the influence of these deviated points. If
only using 64 samples in one sine-wave period to calculate the RMS values, these points will
51
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
influence the results significantly. Thus, 3200 data samples are decided to be used in one second
to implement the calculation of the RMS values of both of the voltage and current. In addition,
this solution can also reduce the deviation of the calculated RMS values when the frequency of
the mains voltage is not stable at 50Hz. For example, if the frequency of the main voltage is a
little higher than the expected frequency, just sampling 64 points in one sine-wave period may
lead to some points in the positive region of the next sine-wave period being captured by the
SD24_A A/D converter. Then the RMS values result may be deviating from the influence of these
points in the positive region of the next sine-wave period. However, if capturing 3200 sampled
points in one second, that means capturing 3200 sampled points in 50 sine-wave periods, even if
the frequency of the mains voltage signal is not stable at 50Hz, for example, the frequency of the
main voltage is a little higher than the excepted frequency and some points in the positive region
of the 51st sine-wave period, the calculated RMS values result will almost not be changed
anything since there are 3200 data samples used to calculate the RMS values and the number of
the points in the positive region of the 51st sine-wave period just occupies a very small
percentage of the total number of the sampled points used to calculate the RMS values result.
One second is still a quite short time interval compared to the time interval 15 seconds between
each RF message sent from the ELIQ Smart-plug device to the ELIQ Energy Display Unit, which
can satisfy the fast calculation to get the measured power consumption result for the RF
communication.
5.1.4 Power consumption value calculation
When the RMS values of both of the voltage and current on the electrical appliance in one
second are obtained, since the RMS values of both of the voltage and current are regarded to be
equal to the effective voltage value and the effective current, then the power consumption value
of the electrical appliance can be calculated by multiplying the RMS value of voltage and RMS
value of current. The formulas are shown below:
power comsumption value = effective voltage value × effective current value
��
��
=
× �
(5.16)
(5.17)
5.2 Control switch on or off status of
connected appliance
In the beginning, the output of the I/O port pin P1.2 is set to e 0 in software, the collector pin
and the emitter pin of the NPN transistor BC847C are isolated, there is negligibly small current on
the coil of the relay, so the mechanical switch device will be opened under the electromagnetic
force from the coil and the electrical appliance is switched off from the power supply. When
measurement of the power consumption starts, the output of P1.2 is toggled from 0 to 1 to
make the NPN transistor BC847C conduct, then the mechanical switch device will be closed
under the electromagnetic force from the coil and the electrical appliance is switched on to the
52
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
power supply. If the measured power consumption of the electrical appliance is less than 2500w,
the output of P1.2 is still 1 to keep the electrical appliance in switch on status. If the measured
power consumption of the electrical appliance is more than 2500w, then the P1.2 is toggled from
1 to 0 , turning off the NPN transistor BC847C, then the mechanical switch device will be
opened under the electromagnetic force from the coil and the electrical appliance is switched off
from the power supply.
5.3 SPI communication to RF module
The internal SPI port of the MCU is not used but the SPI communication is handled instead in
software. A 32-bit unsigned variable is used to store the calculated power consumption value. A
16-bit unsigned variable is used to store the step value. An 8-bit unsigned variable is used to
store the serial number and another 8-bit unsigned variable is used to store the battery status. All
of these four variables consist of a struct variable. The report message data will be stored into
this struct variable.
According to the data packet format of the chip CC1101 defined in the datasheet of CC1101,
which can be seen clearly in Figure 5.3 below, there are two bytes for which one byte is used to
store the length of the data packet and another byte is used to store the address of the message
data consisting of the message header for the report message data. The message data will be
added to a message buffer, toghther with the message header. Thus, in this message buffer, the
first byte indicates the length, the number of bytes of the message data to be sent, the second
byte indicates the destination address of the message data, then the payload data starts at the
third byte.
Figure 5.3 The format of the data packet for the chip CC1101 [9]
As can be seen from Figure 4.5, the schematic of the communication circuit between MCU and
RF module presented in section 4.5, there are 4-wire SPI compatible interfaces (SI, SO, SCLK and
CSn) on the chip CC1101 used to read and write the buffered data to implement the SPI
communication between the MCU and the chip CC1101 where the MCU is the master and the
chip CC1101 is the slave. All transfers on the SPI interface are done from most significant bit to
least significant bit of the data.
Since the CSn port pin is active with the low state, the CSn pin must be kept low during the
transfers on the SPI bus. If CSn goes high during transferring the data, the transfer will be
cancelled immediately. The message data will be transferred serially to the SI pin of the RF
module from the MOSI pin of the MCU.
53
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 5.4 below shows the flowchart of sending the value from the MCU to the RF module by
using SPI communication. The data message contains the measured power consumption value
will be encoded into the data packet firstly, then the data packet will be transferred to the RF
module via the SPI interface once every 15 seconds, the time interval of transferring the data
packet will be controlled by the Timer_A interrupt which is initiated in the beginning of the
software process.
Figure 5.4 The flowchart of sending the value from MCU to RF module using SPI communication
As can be seen from Figure 5.4 above, Timer_A interrupt is used to control the SPI
communication to the RF module. Since it is expected to send the message one time every 15
seconds in the project plan and the P2 interrupt service routine is called every 10ms, a variable
with the value 1500 is added in the interrupt service routine to control an if statement, the
function to send the packet data from the MCU to the RF module is written under this if
statement, then the function to send the packet data will be called once when the Timer_A
interrupt service routine is called 1500 times, thus, the time interval for the event to report is:
ms ×
=
s
(5.18)
That means the function to send the packet data from the MCU to the RF module is called once
every 15 seconds, so the measured power consumption value is sent to the ELIQ Display Unit
from the Smart-plug system through the RF communication once every 15 seconds.
5.4 Wireless communication
When the RF module is used to transmit the data to another device using the wireless
communication, the data input from the SI pin of the RF module will go into the TX FIFO firstly,
then the packet handler, the interleaver, the modulator, the frequency synthesizer and the power
amplifier, the signal will go to the positive RF output port pin and negative RF output port pin and
54
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
then the antenna for the wireless communication.
When the RF module is used to receive the RF signals from another device using the wireless
communication, the received positive RF signal and negative RF signal from the wireless
communication will go to the LNA (low noise amplifier) firstly, then the down-conversion mixers,
the analog-to-digital converters, the demodulator, the packet handler, and finally, the data
packet will go to the digital interface to be transferred to the MCU through the SPI
communication.
The 868 MHz ISM band is used for the wireless communication. GFSK(Gaussian shaped
frequency-shift keying) is chosen to be the modulation mode used in this project. GFSK is a type
of frequency shift keying modulation that uses a Gaussian filter to smooth the positive or
negative frequency deviations, which represents a binary 1 or 0 . A GFSK modulator is similar
to a FSK modulator, except that before the baseband waveform (levels −1 and +1) goes into the
FSK modulator, the signal is passed through a Gaussian filter to make the transitions smoother in
order to limit the spectral width of the signal. Gaussian filtering is a standard way for reducing
the spectral width of the signal and is called "pulse shaping" in this application. In this project,
the carrier frequency is 868MHz with a deviation of 20KHz, the data rate is 38.4KBaud, the digital
channel filter bandwidth is 100KHz and the output power is +10dBm.
5.5 Working as a repeater
This Smart-plug device can also repeat the received ELIQ protocol RF message for other ELIQ
devices to work as a wireless repeater. The P2 I/O interrupt is used to control the received RF
message from some transmitter devices such as an ELIQ sensor. After processing the received RF
message in software, the same RF message will be transmitted outside to the receiver devices
such as ELIQ Display unit or ELIQ wireless hub. With this function of repeating the message, the
power of the RF message signal can be amplified and the wireless communication distance can
be extended largely by using the Smart-plug device system.
As can be seen from Figure 4.5, the schematic of the communication circuit between the MCU
and the RF module presented in section 4.5, the port pin P2.0 of the MCU is connected to the
GIO pin of the RF module. The GDO0 pin of the chip CC1101 is attached to the GIO pin of the RF
module and the GDO0 signal from the chip CC1101 is used to indicate data packet received.
When CC1101 receives the RF message from other ELIQ devices via the wireless communication,
the high-to-low transition occurs in the GDO0 signal from the chip CC1101 to the GIO pin of the
RF module then to the port pin P2.0 of the MCU, the interrupt flag for P2.0 is set and the P2
interrupt is triggered, then fetching and processing the received RF message is implemented in
the P2 interrupt service routine. The received message data will be transferred serially to the
MISO pin of the MCU from the SO pin of the RF module. When the MCU receives the data packet
from the RF module, the data packet will be unfolded to remove the preamble bits and
synchronization words to get the useful data. Then the same data will be encoded and packeted
again to transfer to the RF module from the MCU immediately via the SPI interface, the message
data will still be transferred serially to the SI pin of the RF module from the MOSI pin of the MCU
as same as the analysis in section 5.3.
55
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 5.5 below shows the flowchart of repeating the RF message when the Smart-plug system
working as a repeater.
Figure 5.5 The flowchart of repeating the RF message when the Smart-plug system working as a repeater
56
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
6. Visual support
Figure 6.1 below shows the schematic of the whole circuits on the Smart-plug electrical board.
Figure 6.1 The schematic of the whole circuits of the Smart-plug system
Figure 6.2 below shows the PCB layout of the whole circuits on the Smart-plug electrical board.
Figure 6.2 The PCB layout of the whole circuits of the Smart-plug system
57
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Figure 6.3 below shows a real photo of the Smart-plug electrical board.
Figure 6.3 The real photo of the electrical board of the Smart-plug system
58
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
7. Results
This ELIQ Smart-plug device system has achieved its basic expected functions that measuring the
power consumption value of the connected electrical appliance and sending the data message
which contains the measured power consumption value of the electrical appliance to the ELIQ
Energy Display Unit via the RF wireless communication once every 15 seconds to display to the
households successfully. Table 7.1 below shows the testing results displayed on the ELIQ Display
Unit when measuring one 25W light and the different number of the same lights in parallel with
the power consumption value of one light is 185W.
Rated power value
of the light
Measurement
results
One 25W light
25W
37W
One 185W light
185W
180W
Two 185W lights in parallel
370W
362W
Three 185W lights in parallel
555W
540W
Four 185W lights in parallel
740W
716W
Five 185W lights in parallel
925W
895W
Six 185W lights in parallel
1110W
1050W
Table 7.1 The testing results of measuring the electrical appliances
with different power consumption value
When measuring two 185W lights in parallel, whose rated power value is 370W, the ELIQ Display
Unit can display the power consumption value 362W, in this case, the accuracy of the power
consumption measurement is:
≈
.
%
(7.1)
This is the best measurement accuracy among a series of tests above.
This ELIQ Smart-plug device system can also switch on or switch off the power supply to the
connected electrical appliance according to whether the measured power consumption value of
the connected appliance is less than or more than 2500W. In addition, this ELIQ Smart-plug
device system can work as a repeater to repeat the same RF message for other ELIQ devices in
order to amplify the power of the RF message signal and extend the RF wireless communication
distance largely.
59
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
8. Discussion
8.1 Appliance with small power consumption
cannot be measured accurately
As the tested results shown in the chapter 7 above, when measuring the 25W light, the ELIQ
Display Unit will display the power consumption value of the light 35W. It is observed that the
electrical appliance with small power consumption value which is less than 100W cannot be
measured accurately to achieve the expected range of the accuracy degree of the measured
result by trying several times of testing with different electrical appliances.
When measuring the electrical appliance with the power consumption value is about 25W, the
effective current value on the appliance is:
≈ .
A
(8.1)
The effective voltage value on the sampling resistor is:
.
A × .
Ω =
So the peak voltage value on the sampling resistor is:
.
Since the resolution is :
.
V
V × √ = .
resolution ∆=
+
=
.
(8.2)
V
≈
.
(8.3)
mV
(8.4)
The digital output value transferred from the peak voltage value on the sampling resistor to the
analog input pairs of the SD24_A A/D converter is:
.
.
≈
=
(8.5)
That means the number of intervals from the lowest peak point to the highest peak point for the
sine wave signal of the voltage on the sampling resistor is
× =
, there are only 168
steps from the lowest peak point to the highest peak point. Compared to
× =
×
=
steps from the lowest peak point to the highest peak point for the full scale voltage
value of the SD24_A A/D converter, 168 is quite small. Since 64 points need be sampled in one
sine-wave period of the voltage to the analog input pairs of the SD24_A, it cannot be
distinguished clearly for different sampling points to different steps if there are only 168 steps
from the lowest peak point to the highest peak point of the signal. In addition, if noise is present
in the signal, this will influence the calculation of the RMS value of the current value significantly
since the digital output value can change dramatically due to the total number of intervals from
the lowest peak point to the highest peak point being only 168. These factors will lead to the
60
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
circumstance that the electrical appliance with small power consumption value which is less than
100W cannot be measured accurately.
8.2 Sampling resistor with resistance tolerance
1% still influences the accuracy of
measurement
The sampling resistor chosen to be used on the Smart-plug device electrical board is LOB5 R010
FLF, the 0.01Ω metal film current sense resistor with a rated power of 5W and the resistor
tolerance is ±1%. As discussed in section 2.3.6.2, the SD24_A A/D converter has 16-bit resolution,
the number of the intervals for the full scale voltage to the analog inputs of the SD24_A A/D
converter is 65536 from the lowest peak point to the highest peak point of the sine wave signal.
That means the signal will be divided into 65536 steps from the lowest peak point to the highest
peak point, so if measuring a 100W electrical appliance, the number of intervals from the lowest
peak point to the highest peak point in the signal can be inferred from the following calculations:
The effective current value on the sampling resistor is:
≈ .
A
(8.6)
The effective value of the voltage to the analog input pairs of the SD24_A A/D converter is:
.
A × .
Ω =
.
V
(8.7)
The peak value of the voltage to the analog input pairs of the SD24_A A/D converter is:
.
V × √ = .
.
≈
V
(8.8)
The digital output value transferred from the peak voltage value on the sampling resistor to the
analog input pairs of the SD24_A A/D converter is:
.
=
(8.9)
The number of intervals from the lowest peak point to the highest peak point for the sine wave
signal of the voltage input to the analog input pairs of the SD24_A A/D converter is:
× =
(8.10)
That means if measuring the 100W electrical appliance with such small power consumption, the
signal from the lowest peak point to the highest peak point will be divided into 672 steps.
As the resistance tolerance of the sampling resistor chosen to be used in this project is 1%, the
deviation of the voltage on the sampling resistor can reach 1%. That means if measuring the
electrical appliance with 100W power consumption and the resistance deviation of the sampling
resistor is 1%, there are about 6 steps or 7 steps deviated of the digital output compared to the
61
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
accurate converted result after the analog-to-digital conversion. If measuring the electrical
appliance with larger power consumption value, there will be more steps deviated of the digital
output compared to the accurate result.
Thus, if choosing the sampling resistor with a resistance tolerance 1%, the result of the
calculation still cannot be accurate using the MCU with the 16 bits A/D converter, since when
transferring the analog decimal data to the digital binary data, there will be not any difference
among the 16 bits, 12 bits or 10 bits A/D converters in this case.
The bottomline is that in order for the Smart-plug device system to measure the power
consumption value of the connected appliance more accurately, a sampling resistor with much
lower resistance tolerance, such as 0.1% or 0.01%, must be used. However, the price of the
metal film current sense resistor with only 0.1% or 0.01% resistance tolerance is much more
expensive than the price of the metal film current sense resistor with 1% resistance tolerance, so
the tradeoff between the price and the measurement performance of the Smart-plug device
should be considered seriously when designing this product.
8.3 Component selection for the control unit
There are multiply choices of choosing the component to switch on or off the power supply to
the connected appliance. For instance, both of a relay and a triac can implement the function of
switching on or off the circuit. Table 8.1 below shows the comparison between the relay and the
triac. The size of the triac is much smaller than the relay which can be searched on the market
commonly, and only a very small current (which is about 10mA) can trigger the gate of the triac
to switch on or switch off the circuit, however, the relay need large current which is about 60mA
for the coil to make the voltage on the coil is more than the pick-up voltage value to switch on or
switch off the circuit. Since the maximum output current from one I/O port pin of the MCU
MSP430AFE233 is only about 10mA~15mA, that means the NPN transistor and some other
electronic components need be added in the circuit on the board to amplify the output current
from one I/O port pin of the MCU to achieve at enough current value for the relay to control the
switch on or off status of the circuit, while the NPN transistor circuits need not be used to
simplify the circuit and save the components and area of the board if choosing the triac as the
control unit component [18][19]. Although the triac has these advantageous compared to the
relay, the power dissipation of the triac is much higher than that of the relay when the current
through it achieves at large value. The power dissipation of the triac increases along with the
current through it increases, for example, at 18W, if the current through the triac is more than
10A [18]. In addition, the price of the triac is much more expensive a relay. In overall
considerations above, as the power dissipation of the whole Smart-plug device system and the
price of the components on the board are important factors, the relay is eventually chosen to be
control unit component to switch on or off the power supply to the appliance in the live wire. The
comparison of the relay and the triac is shown in Table 8.1 below.
62
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Relay
Triac
Large size
Cheap price
Low power dissipation (500mW)
Small size
Expensive price
High power dissipation (0 – 18W
according to current) [18]
Need large current through the coil to Small current can trigger gate (10mA)
control switch on/off (at least 60mA)
[18]
Need driver circuit with NPN transistor to Do not need driver circuit with NPN
amplify current
transistor
Table 8.1 The comparison between the relay and the triac
8.4 A/D converter output need load
The function of the A/D converter RAC02-3.3SC is to transfer the 230V AC signal from input to
3.3V DC signal to the output. In this Smart-plug electrical board, the A/D converter RAC02-3.3SC
is used to provide the supply working voltage to both of the MCU and the RF module. When
soldering the components on the Smart-plug electrical board, soldering the A/D converter
RAC02-3.3SC which is the power supply device on the board should be the first step because it is
necessary to detect whether the output voltage is really 3.3V DC to ensure that the output
voltage is safe for both of the MCU and the RF module. However, if just soldering the A/D
converter RAC02-3.3SC on the board without any load at the output and connecting the input of
the A/D converter to the 230V AC mains, the detected voltage value of the output displayed on
the multimeter is about 4.2V DC which is higher than the rated output voltage 3.3V DC. That is
because the minimum load for proper working of the A/D converter is 10% then the specification
of the A/D converter is valid. At no load condition, the converter will not be damaged but its
output will be higher than the rated output voltage. So it is necessary to connect some load such
as resistors on the output when detecting the output voltage of the A/D converter. After
soldering all of the components including the MCU and the RF module on the Smart-plug
electrical board, the output voltage is stable 3.3V DC when using the multimeter to detect as
there are enough load connected to the output of the A/D converter, so it will be safe supply
working voltage for both of the MCU and the RF module.
8.5 Acquiring relay components
The 3v relay is hard to find on the market while the 5V relay is easy to find on the market. In the
relay RECOM RAC02-05SC, the coil rated voltage is 5V and the coil resistance is 45Ωso the rated
coil power is 200mW. When the voltage value on the coil reaches the coil pick-up voltage value of
2.25V, the mechanical switch device will be closed under the electromagnetic force from the coil
then the electrical appliance will be connected to the household mains, so a LDO (low-dropout
regulator) needs to be used in the circuit if choosing the 5V relay.
63
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
The LDO can transfer the DC voltage signal in a range from the input to the stable DC voltage to
the output. If choosing the relay HKE V6-S-DC5V to control switching on or off power supply to
the connected appliance, the LDO which can transfer the 5V DC signal from input to the stable
3.3V DC signal to the output needs to be added to the circuit and the A/D converter should be
changed to the type with 5V DC output. The LDO HT7133 is a three terminal voltage regulator
implemented in CMOS technology with features like low power consumption, low voltage drop
and low temperature coefficient. It allows an input voltage up to 24V DC and it is available with
the fixed output voltage 3.3V DC. In this case, the type HT7133 is suitable to be chosen as the
LDO and the RECOM RAC02-05SC is chosen to be the A/D converter. Figure 8.1 below shows the
block diagram of the internal hardware circuit inside the LDO HT7133. Figure 8.2 below shows
the schematic of peripheral circuits for the LDO HT7133. There are one 10uF capacitor between
the Vin port and common GND, as well as one 10uF capacitor between the Vout port and
common GND to remove the possible residual AC voltage signal combined with the DC voltage
signal together.
Figure 8.1 The block diagram of the internal hardware circuit inside the LDO HT7133 [20]
Figure 8.2 The schematic of peripheral circuits for the LDO HT7133 [20]
The output pin Vout+ of the A/D converter RECOM RAC02-05SC will be connected to the relay
HKE V6-S-DC5V through a resistor since the voltage on the coil needs to exceed the pick-up
voltage 3.75V to control the switch device inside the relay. It will be also connected to the input
pin Vin of the LDO HT7133 to give the 5V DC signal for the LDO HT7133. Then the output pin Vout
of the LDO HT7133 will be connected to the VDD pin of both of the MCU and the RF module to
provide the 3.3V DC voltage signal as the supply working voltage for both of the MCU and the RF
module.
64
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
8.6 Width and thickness of the copper wire on
the PCB board
The ELIQ Smart-plug device can measure the power in an electrical appliance with the power
consumption value ranges from 2W to 3000W. That means the maximum value of the current
through the live wire and neutral wire can reach 13A, so the copper wire of the live line and the
neutral line should be wide enough and thick enough in order to make the current with such
large value flow through safely.
In this PCB layout design, the width of the copper wire of the live line and neutral line is set to be
3mm (118.11mil), the width of the copper wire of 3.3V DC signal is set to be 0.5mm (19.685mil),
and the width of the copper wire of other small signal is set to be 0.254mm (10mil). The
thickness of the copper wire is chosen to be 70um when manufacturing the PCB board.
65
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
9. Conclusion
9.1 Summary
The work progress of the project has achieved the initial goals, that is, the electrical appliance
with the power consumption in a range from 100W to 1000W can be measured accurately by
using this Smart-plug device system and the RF message which contains the measured power
consumption value of the electrical appliance can be sent to the ELIQ Energy Display Unit via the
RF wireless communication once every 15 seconds to display to the households. In addition, this
Smart-plug device system can work as a repeater to repeat the RF message for other ELIQ devices,
which can amplify the power of the RF message signal and extend the RF wireless communication
distance effectively.
MSP430AFE233 is a type of suitable microcontroller unit to be the core chip on the Smart-plug
device electrical board to implement the power consumption measurement of the connected
appliance, switching on or off the power supply to the connected appliance and handle
communication with the RF module. Besides the features of small size, ultra-low power
consumption and cheap price, the MCU MSP430AFE233 integrates the SD24_A A/D converter
which has a pair of analog inputs and can transfer the differential value of the analog voltage
signal from the analog input pairs to the digital signal directly. Thus, the circuits on the
Smart-plug electrical board can be significantly simplified without designing the level shifter
circuit to transfer the bipolar voltage value to the unipolar voltage value, the number of
components on the Smart-plug electrical board can be reduced and the area of the board can be
smaller. The SD24_A interrupt is used to control sampling 3200 points from the signal in one
second to calculate the RMS values of both of the voltage and current, the Timer_A interrupt is
used to control so that the RF message which contains the measured power consumption value
of the connected appliance is sent to the ELIQ Energy Display Unit every 15 seconds, the I/O
interrupt of P2 port is used to detect and receive the RF message from other ELIQ devices when
the Smart-plug device working as a repeater.
The RF module with the core chip CC1101 is used to implement the wireless communication
between different ELIQ devices. The Chip CC1101 is a low-cost sub-1 GHz transceiver designed
for very low-power wireless applications in the ISM band at 868MHz. The RF transceiver is
integrated with a highly configurable baseband modem. The chip CC1101 can be controlled by
the MCU via the SPI interface.
9.2 Future work
Now the RECOM RAC02-3.3SC AC/DC converter is used as the power supply for both of the MCU
and the RF module and controls the current on the coil of the relay by transferring the 230V AC
signal from mains to the 3.3V DC signal. However, the size of the RECOM RAC02-3.3SC AC/DC
converter is too large and the price of the RECOM RAC02-3.3SC AC/DC converter is too expensive,
about $16 per piece. If the design of the Smart-plug device is put in mass production, the whole
66
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
cost of the components on the Smart-plug device system must be considered and the area of the
Smart-plug device electrical board should be minimized. Thus, designing the power supply circuit
using resistors, capacitors, diodes and some other electronic components for the MCU, the RF
module and the relay is necessary to save both of the cost and the area of the Smart-plug device
electrical board. It is important to notice that the output current from the designed power supply
circuit should be sufficient enough to drive the MCU, the RF module and the coil of the relay. In
addition, the empty space between the different components on the ELIQ Smart-plug device
electrical board should be as small as possible to make the board layout of the product more
compact than the PCB layout designed in this project, which is also helpful to reduce the cost of
the materials and the size of the product.
The designed Smart-plug device system can measure the power consumption value of the
connected electrical appliance and send the RF message data which contains this measured
power consumption value to the ELIQ Energy Display Unit to let the households know the current
power consumption value of the electrical appliances in a house until now. Besides, this
Smart-plug device system is also expected to send the measured power consumption value to the
ELIQ Online through the ELIQ Hub, which is also implemented with the RF module which is
similar to the work already done. As far as the future work, it is expected to achieve the function
that using the intelligent cellphone to implement the remote monitor to control the switching on
or off status of the connected electrical appliance according to the power consumption value
read from ELIQ online. Furthermore, the Smart-plug device system is also planned to control the
switching on or off status of the connected electrical appliance automatically based on high
demand time or low demand time of the electricity utilization in a city by importing the
specialized algorithm in software. All of these features will bring lower cost and reduced more
energy consumption.
67
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
Bibliography
[1] Exibea AB, ELIQ Energy Display , Copyright 2014 Exibea AB,
http://eliq.se/en/products/eliq-energy-display/
[2] Exibea AB, ELIQ Energy Online , Copyright 2014 Exibea AB,
http://eliq.se/en/products/eliq-online/
[3] N.Saito, Ecological Home Network: An Overview , Published in Procedings of the IEEE ,
Volume: 101, Issue: 11, Publication Year: 2013, Page(s): 2428 – 2435
[4] Anna Berglund and Olof Hartelius, Development of an Energy Visualization Tool ,
Department of Product and Production Development, Chalmers University of Technology,
Gothenburg, Sweden, 2010, Page(s): 13 – 15
[5] Marusic.L, Skocir.P, Petric.A, Ho e-in-Palm — A mobile service for remote control of
household e ergy o su ptio , Telecommunications (ConTEL), Proceedings of the 2011 11th
International Conference on, Publication Year: 2011 , Page(s): 109 – 116
[6] Texas Instruments, MSP430AFE233 16-bit Ultra-Low-Power Microcontroller, 8KB Flash, 512B
RAM, 3 x SD24 , Copyright 1995-2004, Texas Instruments Incorporated,
http://www.ti.com/product/msp430afe233
[7] Texas Instruments, MSP430AFE2X3, MSP430AFE2X2, MSP430AFE2X1, Mixed Signal
Microcontroller , Literature number: SLAS701A, November 2010-Revised March 2011,
http://www.ti.com/general/docs/lit/getliterature.tsp?genericPartNumber=msp430afe233&fileTy
pe=pdf
[8] Texas Instruments, MSP430x2xx Family User's Guide , Literature number: SLAU144J,
December 2004-Revised July 2013,
http://www.ti.com/general/docs/lit/getliterature.tsp?baseLiteratureNumber=slau144&fileType=p
df
[9] Texas Instruments, CC1101 Low-Power Sub-1 GHz RF Transceiver , Literature number:
SWRS061I, Copyright 2014, Texas Instruments Incorporated,
www.ti.com/lit/ds/symlink/cc1101.pdf
[10] RECOM, Powerline AC/DC-Converter 1-2 Watt Single Output, RAC01-C, RAC02-C ,
Revised:January 2014, http://www.recomasia.com/pdf/Powerline-AC-DC/RAC01_02-SC.pdf
[11] Farnell Element14, NXP-BC847C-Transistor, NPN, 45v, SOT-23 , Copyright 2007-2014,
[Premier Franell plc],
http://se.farnell.com/nxp/bc847c/transistor-npn-45v-sot-23/dp/1081235
68
Yixin Liu
May 26th, 2014
Design and implementation of the ELIQ Smart-plug device system
[12] Farnell Element14, MULTICOMP-HC49U-3.2768-20-30-60-12-ATF-CRYSTAL, HC-49/U,
3.2768MHZ , Copyright 2007-2014, [Premier Franell plc],
http://se.farnell.com/multicomp/hc49u-3-2768-20-30-60-12-atf/crystal-hc-49-u-3-2768mhz/dp/1
666947
[13] Wikipedia-the free encyclopedia, Altiu Desig er , Last modified on June 11th 2014 at 8:53,
http://en.wikipedia.org/wiki/Altium_Designer
[14] Linear Technology, Design Simulation and Device Models , Copyright 2014,
http://www.linear.com/designtools/software/#LTspice
[15] IAR Systems, IAR Embedded Workbench , Copyright 1995-2014,
http://www.iar.com/Products/IAR-Embedded-Workbench/
[16] Wikipedia-the free encyclopedia, IAR Sytems , Last modified on April 17th 2014 at 8:53,
http://en.wikipedia.org/wiki/IAR_Systems
[17] FET-Pro430 User's Manual, http://www.elprotronic.com/files/Manual-FET-Pro430.pdf
[18] Wikipedia-the free encyclopedia, TRIAC , Last modified on May 8th 2014 at 14:50,
http://en.wikipedia.org/wiki/TRIAC#Triggering_in_Quadrant_I
[19] STMicroelectronics, BTA20, 20A SnubberlessTM Triacs , Copyright 2012,
http://www.farnell.com/datasheets/1689731.pdf
[20] HOLTEK, HT71xx, High Voltage Regulator , May 17th 1999, HOLTEK SEMICONDUCTOR INC,
http://pdf.datasheetcatalog.com/datasheets2/97/97210_1.pdf
69
Yixin Liu
May 26th, 2014
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising