Laboratory practical with the
C8051Fxxx microcontroller family
Gingl Zoltán, Mingesz Róbert
2014
A tananyag a TÁMOP-4.1.2.A/1-11/1-2011-0104 “A felsőfokú informatikai
oktatás minőségének fejlesztése, modernizációja” c. projekt keretében a Pannon
Egyetem és a Szegedi Tudományegyetem együttműködésében készült.
Laboratory practicals with the
C8051Fxxx microcontroller family
Authors:
Dr. Zoltán Gingl and Dr. Róbert Zoltán Mingesz
Keywords:
Microcontrollers, embedded programming, timers, counters, serial communication,
analogue-to-digital conversion, sensors.
Summary
The purpose of this book is to help the teaching of the applications of microcontrollers in
various projects. Several books and manuals are available [1-19]; this book contributes to
these by covering the knowledge needed to use the powerful C8051Fxxx family of
microcontrollers from Silicon Laboratories in practice. Our aim was to synthesise the most
useful information found in manuals, tutorials, datasheets, user forums, application notes,
electronic design notes and example code in a single book. Most chapters feature brief
application guidelines and troubleshooting based on our teaching and development
experience. This can be highly useful for students and for developers as well.
We believe that the brief discussion of the architecture, peripherals, analogue and digital
signal interfacing helps to understand how these can be used to build various applications.
We provide tested example code and recommended exercises and discuss several application
examples, including single-supply analogue signal conditioning, sensor interfacing and
microcontroller-host computer communication. In the last chapter, we show the schematic
and layout of an extension board that supports the use of the C8051F410DK development kit
and can also be modified for use with other target boards.
Up-to-date, high quality references were chosen that are provided by industry leading
companies [1–19]. Almost all of the references are available on-line on the companies’ web
pages.
1
TABLE OF CONTENTS
1
2
3
4
5
6
Introduction ........................................................................................................................ 5
1.1
Real-world signal processing and control ................................................................... 5
1.2
Microcontrollers .......................................................................................................... 6
1.3
Microcontroller core and integrated peripherals ........................................................ 7
1.4
Microcontroller classification .....................................................................................12
Architecture and properties of the C8051Fxxx microcontroller family.............................13
2.1
8051 microcontrollers ................................................................................................13
2.2
The C8051Fxxx microcontroller family ......................................................................13
2.3
The CIP-51 architecture ..............................................................................................14
Assembler and C programming ........................................................................................ 27
3.1
SDCC C compiler ....................................................................................................... 27
3.2
Interrupt programming in assembler ....................................................................... 29
3.3
Interrupt handling in C ............................................................................................. 30
3.4
Interrupt programming guidelines ........................................................................... 32
3.5
Using an integrated development environment and the associated tools ................ 33
3.6
Config Wizard ............................................................................................................ 34
Digital input and output; crossbar .................................................................................... 36
4.1
The I/O structure....................................................................................................... 36
4.2
Crossbar ..................................................................................................................... 38
4.3
Port I/O applications ................................................................................................. 39
4.4
Application guidelines ............................................................................................... 47
4.5
Troubleshooting ........................................................................................................ 48
4.6
Exercises .................................................................................................................... 48
Timers and counters ......................................................................................................... 50
5.1
Timer 0 and Timer 1 .................................................................................................. 50
5.2
Timer 2, Timer 3 and Timer 4 ................................................................................... 52
5.3
Timer applications ..................................................................................................... 54
5.4
Application guidelines ............................................................................................... 58
5.5
Troubleshooting ........................................................................................................ 59
5.6
Exercises .................................................................................................................... 60
Programmable counter array .............................................................................................61
6.1
Edge-triggered capture mode .....................................................................................61
2
7
8
9
6.2
Software timer and high-speed output mode ............................................................ 62
6.3
Frequency output mode............................................................................................. 63
6.4
8-bit and 16-bit PWM modes .................................................................................... 64
6.5
Application guidelines ............................................................................................... 66
6.6
Troubleshooting ........................................................................................................ 67
6.7
Exercises .................................................................................................................... 67
Serial communication peripherals .................................................................................... 69
7.1
UART ......................................................................................................................... 69
7.2
SPI.............................................................................................................................. 74
7.3
SMBus ........................................................................................................................ 78
7.4
C standard I/O redirection ........................................................................................ 82
7.5
Exercises .................................................................................................................... 83
Analogue peripherals ........................................................................................................ 84
8.1
Comparators .............................................................................................................. 84
8.2
Voltage reference ....................................................................................................... 87
8.3
ADC ............................................................................................................................ 89
8.4
DAC ............................................................................................................................ 96
8.5
Temperature sensor ................................................................................................... 98
8.6
Exercises .................................................................................................................... 98
Sensor interfacing ........................................................................................................... 100
9.1
Voltage output sensors ............................................................................................ 100
9.2
Current output sensors ............................................................................................ 102
9.3
Resistive sensors ...................................................................................................... 103
9.4
Exercises .................................................................................................................. 105
10 Real-time clock ............................................................................................................... 107
10.2
11
Exercises .................................................................................................................. 109
Watchdog and power supply monitor.............................................................................. 110
11.1
The watchdog timer .................................................................................................. 110
11.2
Supply monitor ......................................................................................................... 110
11.3
Exercises ................................................................................................................... 111
12 Low-power and micropower applications ....................................................................... 112
12.1
Low-power modes .................................................................................................... 112
12.2
Clock speed tuning ................................................................................................... 112
12.3
Peripheral power consumption ................................................................................ 113
3
12.4
Supply voltage........................................................................................................... 113
12.5
Exercises ................................................................................................................... 115
13
USB, wired and wireless communications ...................................................................... 116
13.1
USB-UART interfaces ............................................................................................... 116
13.2
Wireless communication possibilities ...................................................................... 118
13.3
Exercises ................................................................................................................... 119
14 Development kit .............................................................................................................. 120
14.1
The C8051F410 development kit ............................................................................. 120
14.2
Extension board....................................................................................................... 120
15
Acknowledgements ......................................................................................................... 124
16 References ........................................................................................................................125
4
Introduction
1
1.1
Introduction
Real-world signal processing and control
It is a typical aim to construct machines to make life more comfortable and more economical.
From simple mechanical machines to advanced electronic devices such as smart phones the
range is really wide. The most efficient devices are based on electronics, sophisticated signal
processing and modern software.
In order to allow processing, real signals must be converted into another format that can be
processed and the result should be used for intervention, as shown in Figure 1.1.
Sensing
Real
system
Processing
Acting
Figure 1.1. General real-world interaction.
The same principle is used in machines in general (Figure 1.2).
External
signals
Signal
conversion
Impact
Signal
conversion
Machine
processing
Figure 1.2. Machine – real world interaction.
The most efficient devices use analogue and digital electronics and run software to process
information. Many of today’s devices are small, battery-operated and incredibly efficient.
Again, a good example is the smart phone that integrates telephony, camera, wireless
communication, computer, sensors, GPS and many more in a handful of electronics.
The detailed block diagram of such an electronic device is shown in Figure 1.3. Sensors
convert several physical signals (displacement, force, pressure, acceleration, temperature,
light intensity, etc.) to signals that can be handled by electronics (voltage, current, resistance,
capacitance, inductance). The output of sensors is converted to voltage in the proper range (a
few volts) that can be easily used in processing. The analogue-to-digital converter translates
this voltage to integer numbers for digital processing. A similar principle is applied in the
reverse transformations.
5
Introduction
External
signals
Sensor
Signal
conditioning
analog electrical
signals
Impact
Actuator
A/D
converter
digital electrical signals
(binary values)
Signal
conditioning
Processor
and
software
D/A
converter
Figure 1.3. Electronic device – real world interaction.
Several analogue and digital integrated circuits have been developed to support the
manufacture of electronic devices. One of the most compact and most efficient components is
the microcontroller.
1.2
Microcontrollers
The microcontroller unit (MCU) is a small but powerful digital building block, a single-chip
microcomputer. It contains everything required for operation; very few external components
are needed – sometimes only supply decoupling capacitors. Of course, the device must be
powered, typically from a single supply voltage that ranges from 1.8 V to 5 V. Sometimes even
a coin cell battery suffices.
The microcontroller has several peripherals to sense real-world signals and initiate realworld events, and has a processor core to run software. It is a very flexible, powerful and
compact electronic component. Since most of the processing is done by the software, the
same hardware can be used for several applications; the performance can be upgraded easily
by replacing the software only.
There is a very wide range of microcontrollers on the market from sizes of 2 mm × 2 mm and
from a power consumption of 30 W to a speed of several hundred MHz.
Most modern microcontrollers incorporate comparators, analogue-to-digital and digital-toanalogue converters and temperature sensors – therefore, they are often called mixed-signal
(both analogue and digital) microcontrollers.
Figure 1.4 illustrates some typical components of a modern mixed-signal microcontroller; the
details will be given in the next chapter.
6
Introduction
ANALOGUE PERIPHERALS
DIGITAL PERIPHERALS
VOLTAGE
REFERENCE
TIMER
COUNTER
PORT INPUT
AND OUTPUT
D/A
CONVERTER
TEMPERATURE
SENSOR
COUNTER
ARRAY
UART
CAPACITANCE
TO DIGITAL
VOLTAGE
REGULATOR
CRC
CALCULATION
SMBUS/I2C
MULTIPLY/
ACCUMULATE
SPI BUS
USB/WIRELESS
CAN/LIN BUS
MUX
A/D
CONVERTER
PGA
PROCESSOR SUPPORT PERIPHERLS
POWER ON
RESET
SUPPLY
MONITOR
WATCHDOG
TIMER
OSCILLATOR
AND PLL
REAL TIME
CLOCK
MICROCONTROLLER CORE
PROCESSOR
CORE
INTERRUPT
HANDLER
MEMORY
RAM/FLASH
DMA
DEBUG
INTERFACE
Figure 1.4. Microcontroller components.
1.3
Microcontroller core and integrated peripherals
The microcontroller core is based on a processor with its typical components including an
arithmetic logic unit (ALU) and several registers. The architecture may follow the CISC (like
the 8051 family) or, more probably, the RISC principles (for example, the PIC, AVR and ARM
microcontrollers) in today’s popular microcontrollers.
Most of the devices use separate memory for the data and for the program; that is, they have
Harvard architecture. This fits well the need for non-volatile program memory and at the
same time it prevents code corruption and provides even faster execution in some cases. The
word length of the two kinds of memory can also be different. Microcontrollers may use
Neumann or Harvard architecture, or the user can even configure the memory usage (for
example, in the case of the ARM Cortex-M3 32-bit microcontroller family).
All modern microcontrollers have volatile (SRAM) memory and non-volatile,
reprogrammable flash memory. The flash memory contains the code, so no external
integrated circuits are needed. The flash memory can be reprogrammed by special
programming devices using a few (from 2 to 6-8) pins of the microcontroller (in-circuit
programming, JTAG) or can even be overwritten by the microcontroller itself in some cases.
Additional separate flash or EEPROM may also be integrated to support non-volatile data
storage (configuration data, calibration data, statistical data, etc.). The flash memory can be
7
Introduction
rewritten about 100000 times, and the typical data retention time is longer than 20 years.
The flash memory can be protected, i.e., the code can be prevented from being read by the
user.
If the on-chip memory is not enough for a certain application, the developer can choose
microcontrollers with an external memory interface that support the connection of static
RAM or other memories of various sizes. Note also that this interface may support the use of
‘memory mapped’ peripherals including A/D converters, D/A converters, FIFO memories,
etc.
1.3.1
Processor support
In the following the most typical processor support peripherals will be described briefly.
Power on reset (POR) generator. After switching the power on, the supply voltage may
rise a bit slowly due to the fact that the supply decoupling and filtering capacitors must be
charged and the supply current is limited. At the same time, the digital circuitry needs a
certain minimum supply voltage for proper operation, so the start-up of the microcontroller
must be delayed until the supply voltage reaches the safe operating level. Having detected the
crossing of this level, the POR generates an additional short delay (in the range from below
1 ms to about 100 ms) and finally releases the reset line.
Power supply monitor (Brown-out detector). In some cases, the supply voltage may go
below the safe operating level even during operation (for example, when sudden heavy
current loading occurs). This may result in erroneous code execution, therefore the supply
monitor circuit will generate a reset in this case. Note that this feature can be disabled by the
programmer, although the use of the supply monitor is strongly recommended.
Low-dropout (LDO) regulator. Some microcontrollers have separate voltage levels for
their core and digital input and output ports. Integrated voltage regulators can provide stable
and sometimes even programmable supply voltage from the input supply voltage. Lowdropout regulators need only a slightly (roughly about 100 mV) higher input supply voltage
than their output voltage.
Watchdog timer (WDT). Even properly powered processors can fall into infinite loops or
get disturbed by electromagnetic or conducted interference (for example, in the case of
lightning or power line transients), which may cause serious problems in several applications
(motor control, heating control, healthcare devices, etc.). The watchdog timer refresh register
needs to be written within a certain amount of time (that can be typically programmed from
tens of milliseconds to several seconds); otherwise, a reset will be generated. If the code
writes to the register, the timer will be restarted and no reset will be generated. If the
processor code execution fails, this will not occur and a reset will be initiated. The best
practice is to always use the watchdog timer except in code development phase or in simple
test projects. The watchdog timer is enabled automatically upon reset in quality
microcontrollers.
Oscillator, PLL. All processors need a clock signal to schedule instruction execution.
Modern microcontrollers have on-chip oscillators but also support the use of external quartz
crystals or external clock signals. Optional phase-locked loop (PLL) clock multipliers often
combined with clock dividers allow the generation of a wide range of higher processor clock
frequencies. Typically, on-chip oscillators have an accuracy of 1%-20%, while the precision of
8
Introduction
crystal oscillators can fall below 0.01%. The developer can choose the solution that suits the
particular application best.
Debug interface. This interface is used by the integrated development environment to
download code to the flash memory. Memory upload is also supported and the developer can
program the security bits to protect the code from being uploaded. The debug port allows
single stepping, supports breakpoints and can track the content of variables, memory and
peripheral registers. The debug interface makes code development and testing easy and it is
an essential tool for all modern microcontrollers. The most commonly used interface
standard is called JTAG (Joint Test Action Group, IEEE 1149.1 Standard Test Access Port and
Boundary-Scan Architecture).
1.3.2 Digital peripherals
Digital peripherals include the digital input/output pin drivers and internal digital circuits
related to timing, communication and computation acceleration.
General-purpose input/output (GPIO), port input/output (Port I/O). The
processor reads from and writes to memory and all on-chip peripherals using the
bidirectional data bus. Some processors may also incorporate a direct memory access (DMA)
controller, which transfers data between memory and a peripheral without processor
intervention. The data is valid only for a short duration in order to free the bus for other
transactions, so the general purpose output requires latches that can keep the data until the
code writes new data to it. The output of these latches is connected to the pins of the chip and
can drive LEDs and provide logic output signals for external digital circuit inputs. These
signals are mostly arranged in 8-bit groups to form a byte. The pins can also be configured as
digital inputs that can be read any time by the microcontroller. This way buttons, switches
and digital signals can be connected as well with the help of internal or external pull-up
resistors.
Timer/Counter modules. Microcontrollers are designed to control electronic equipment
for household, automotive, industrial, test and measurement applications; therefore, timing,
event counting, periodic event generation and time duration measurement are important. All
microcontrollers contain 8-, 16- or 32-bit counters that can be configured as timers (when an
oscillator drives the counter) or as counters, when the rising or the falling edge of an external
signal increments the counter. Timers also provide timing for serial communication
peripherals, A/D converters and D/A converters.
Programmable Counter Array (PCA). The PCA contains a simple free-running counter
that is driven by an oscillator. There are several (from 3 to 6) independent compare/capture
registers that can be used to latch the counter value upon an event (a change in a digital input
signal). These registers can also hold data to be compared with the counter value and to
generate an event when a match occurs. The PCA can be used to measure pulse width, period
or frequency, to generate pulse width modulated (PWM) signals and special logic signal
patterns, periodic interrupts and even more.
Real-Time clock (RTC). In order to measure the real time or synchronise events to it, a
dedicated precise oscillator and an associated 32 to 48-bit counter is provided in some
microcontrollers. The oscillator typically uses 32768-Hz tuning fork crystals and a very low
power oscillator. Practically, a clock is integrated into the microcontroller that can be
powered from a button battery and can run even if the processor is not powered. Besides
9
Introduction
measuring real time, this peripheral can serve to wake the microcontroller up at a certain
time – in other words, to provide alarm function.
Computing support (MAC, CRC, AES). Some microcontrollers contain computation and
digital signal processing acceleration hardware. For example, 8-bit microcontrollers can have
a multiply and accumulate (MAC) unit that can multiply and add 16-bit data in a few clock
cycles. This can be used efficiently in digital filtering and to compute fast Fourier transforms
(FFT). Cyclic redundancy check (CRC) is frequently used to check data integrity in
communications and the Advanced Encryption Standard (AES) algorithm is also supported
by some microcontrollers.
1.3.3 Communication
Universal Asynchronous Receiver/Transmitter (UART). This serial (one bit at a
time) communication interface uses one wire to send and another wire to receive bits of a
byte. The sender and receiver must have a closely matched time base that determines the
duration of transmitting a bit, since no timing synchronisation is provided. Every transaction
is initiated by sending a start bit, followed by the data bits. The receiver detects the start bit
and can then decode the data bits by sampling the signal at evenly spaced time instants. The
UART interface is used for low wire count inter-processor communications, host computer
communication via USB-UART interfaces, infrared communications and device-to-device
communications. In most cases, it is a two-device bus; the use of more devices introduces
hardware and software overheads.
Serial Peripheral Interface (SPI). The SPI interface is typically used for high-speed
communication with off-chip peripherals including analogue-to-digital converters, digital-toanalogue converters, digital output sensors and other processors. Two wires are used to carry
data bits in two directions and one wire for a clock signal that synchronises the timing
between the communicating devices. A rising or a falling edge of this signal indicates the
beginning of the transmission of each bit. A fourth signal may also be used to provide a frame
for the communication. If this signal is inactive, the other signals are ignored, which can be
used to connect multiple devices on the same bus and select one for which communication is
enabled.
Inter-Integrated Circuit (IIC or I2C) and System Management Bus (SMBus). This
medium-speed interface is specially developed for communication between a host
microcontroller and several peripheral chips (memories, data converters, sensors or other
processors) on the same printed circuit board or within equipment over only two wires. One
wire carries data in both directions, while the other is used to provide a frame (start and stop
conditions) for the transaction and to synchronise the transmission of the bits through clock
pulses.
Controller Area Network (CAN), Local Interconnect Network (LIN). These serial
interfaces are only available on some microcontrollers that target automotive or other
industrial applications. Most of the protocol is implemented in hardware.
Universal Serial Bus (USB). The USB is the most popular and innovative interface for
connecting peripherals to personal computers or tablets. Some microcontrollers have built-in
slave (and rarely host) USB ports. This allows direct connection to the USB port; however,
the programmer should know the most important parts of the USB protocol and a driver is
typically required on the host computer.
10
Introduction
Wireless communication peripherals. Wireless communication is becoming a more
and more popular interface between small devices, since it supports very flexible location and
networking options and no wires are required. There are microcontrollers with integrated
wireless transmitters and receivers (transceivers) with several frequency options and a
number of wireless protocols can be implemented by software. Bluetooth, ZigBee and the
open-source TinyOS system are among the most widely used platforms.
1.3.4 Analogue peripherals
Several microcontrollers have analogue parts to handle analogue signals even without
external analogue circuitry. This makes microcontrollers even more compact: a single
microcontroller and only a few external components can implement a complete solution for a
real-world application that requires the monitoring of signals and the controlling of
processes. Microcontrollers that have a significant analogue part and can therefore handle
both digital and analogue signals are often called mixed-signal microcontrollers or analogue
microcontrollers.
Comparator. Comparators have two analogue voltage inputs and a digital output. Their
output is logical high if the voltage connected to their positive input is higher than the voltage
at their negative input. They may also have hysteresis to reduce potential noise-induced
switching.
Analogue-to-Digital Converter (ADC). Analogue voltages can be translated into the
digital domain using ADCs. The output is an integer number with a various number of bits. A
resolution of 10 bits is the most typical, but precision microcontrollers can have 12-, 16- or
even 24-bit ADCs. Note that the accuracy is normally less than the resolution; therefore, the
datasheet should always be consulted to obtain reliable information.
Digital-to-Analogue Converter (DAC). DACs output analogue signals proportional to
the integer number at their input. The resolution range includes 8, 10, 12 or 16 bits. The
output signal can be voltage or current.
Voltage reference (VREF). All data converters (ADCs, DACs) need a reference voltage
that serves as an etalon of conversion. The input range of the ADCs and the output range of
voltage output DACs are both determined by Vref; in most cases it is between 0 and Vref. The
internal reference voltage can be switched off to support the use of more precise external
reference voltage circuits.
Capacitance-to-Digital Converter (CDC). One of the most popular modern user
interfaces is based on touch sensing that effectively replaces mechanical buttons, which have
limited reliability and lifetime. The change in a capacitance is measured, which change
depends on the proximity of the finger of the user from the sensing pad. The capacitance is
digitised and the data can be used for evaluation.
Analogue Multiplexer (MUX). Monitoring multiple analogue signals is often needed in
real-world applications. This can be supported by a network of switches, called an analogue
multiplexer, that connects one of the signals to the input of the ADC at a time. After the
conversion of a signal, the next signal can be selected. Since conversion only takes a short
time, this means a quasi-simultaneous conversion if the signals change only slowly. However,
the different signals are measured at slightly different time instants, which should be
considered anyway.
11
Introduction
Programmable Gain Amplifier (PGA). Some microcontrollers have preamplifiers before
their integrated ADCs to support voltage range extension. The preamplifiers can have
software-programmable gains of 0.5, 1, 2, 4, 8, 16, 32, 64, 128. Single-ended and differential
input PGAs are both available.
Temperature sensor. Most mixed-signal microcontrollers include diode-based
temperature sensors that can be connected to the input of the internal ADC using the
analogue multiplexer. The on-chip sensor outputs a voltage that has linear dependence on the
chip temperature. The accuracy of the sensor is roughly about 3 °C. It can be used to protect
the device from overheating or to estimate the ambient temperature if the power dissipation
of the microcontroller is low enough so that we can neglect self-heating.
1.4
Microcontroller classification
Depending on the different features and according to target applications microcontrollers can
be broken down into the following categories:
General-purpose microcontrollers have common digital peripherals including timers,
GPIO or UART. Their typical clock frequency is around 10 MHz.
Low power microcontrollers can operate at lower clock frequencies, from 1 MHz down
to tuning fork crystal frequency of 32768 Hz or even below. At 1 MHz the supply current is
well below 1 mA, and supply sensitivity is close to 200 A/MHz. During power down state the
supply current can fall below 1 A.
Precision mixed-signal microcontrollers incorporate 12-bit or higher resolution ADCs
and DACs. Sigma-delta ADCs can even have a resolution of 24-bits and a PGA can provide
software programmable gains in the range of 1 to 128.
High-speed microcontrollers execute most of their instructions within a single clock
cycle and can operate at frequencies from about 25 MHz to several hundred MHz.
According to the bus width there are 8-bit, 16-bit and 32-bit microcontroller families.
8-bit microcontrollers typically consume less power, while 32-bit microcontrollers have more
processing power.
Industrial and automotive microcontrollers operate at a full industrial temperature
range of -40 °C to 85 °C. The internal peripherals have stricter specifications to provide
additional reliability under various conditions, and accuracy of the internal oscillator is better
than 1% over the full operating temperature range. These microcontrollers typically have
industrial or automotive communication peripherals like CAN buses or LIN buses.
Secure microcontrollers are used in security-sensitive applications including electronic
banking and payment, application protection, communication and more. These
microcontrollers offer protection of code and data, prevent reverse engineering, tampering,
data monitoring and physical attacks. Hardware cryptographic modules, random number
generators, fast data and code encryption are implemented to support secure applications.
12
Architecture and properties of the C8051Fxxx microcontroller family
2
Architecture and properties of the C8051Fxxx microcontroller
family
C8051Fxxx microcontrollers developed by Silicon Laboratories [1, 2] are among the most
powerful modern derivatives of the popular MCS-8051 MCU [2] introduced by Intel. A short
summary of these devices follows.
2.1
8051 microcontrollers
The 8051 or MCS-51 family of 8-bit Harvard architecture microcontrollers were developed by
Intel in the eighties for embedded applications. Their easily upgradable architecture proved
successful, became a standard for many manufacturers and several derivatives are still
popular on the market due to their ease of use and carefully designed peripheral handling.
The 8051 family can be easily programmed. There are many free and professional
development tools, so the 8051 microcontrollers can be used by practiced experts, lecturers,
students and hobbyists at the same time. Many source code examples are available to solve
various problems and the manufacturers provide very useful application notes, knowledge
base and user forums.
Manufacturers include Silicon Laboratories, Maxim/Dallas, Analog Devices, Atmel and NXP
(formerly Philips).
Very wide ranges of speed, code and data memory size, analogue and digital peripherals,
power requirement are provided by the C8051Fxxx family developed by Silicon Laboratories.
The 1 MIPS peak performance of the original 8051 microcontroller has been upgraded up to
100 MIPS peak speed and the integrated flash memory, debug interface and very rich set of
analogue and digital peripherals make the C8051Fxxx family a good choice for various
applications.
The chips can have sizes of 2 mm × 2 mm (10 pins) to 16 mm × 16 mm (100 pins).
2.2
The C8051Fxxx microcontroller family
The maximum clock frequency of the C8051Fxxx microcontrollers is in the range of 25 MHz
to 100 MHz. Slower clock speeds are allowed, practically down to DC, so no minimum is
specified. The frequency of the internal oscillator is programmable, so the user can choose
low power operation at low frequencies, while higher processing speeds can be achieved at
the expense of higher power consumption. For example, the C8051F410 processor can be
operated at 50 MHz, when the core supply current is about 15 mA, while at 32 kHz the device
draws less than 20 A from the supply rail, allowing long lasting operation from a battery.
The size of on-chip flash memory available varies from 2 kbyte to 128 kbyte, while the
internal RAM can store 256 to 8448 bytes of data. The flash memory contains the code and
may also be written by the code to support non-volatile data storage.
The C8051Fxxx microcontrollers can have up to six 16-bit timers and a programmable
counter array with 6 independent channels. Some devices include a real-time clock with
battery backup power option.
Communication peripherals include UARTs, I2C/SMBus, SPI, USB, CAN, LIN serial
interfaces and the parallel external memory interface that also supports the connection of fast
external ADCs, DACs and more.
13
Architecture and properties of the C8051Fxxx microcontroller family
From 6 to 64 GPIO pins are available with configurable output driving options (open-drain
with or without internal pull-up and push-pull mode).
The C8051Fxxx family provides high-performance analogue peripherals. ADC resolutions
from 10 to 12 bits with sample rates from 100 kHz to 200 kHz are common, while the
C8051F06x devices incorporate two independent 1-MHz 16-bit ADCs, and the C8051F35x
microcontroller has an 8-channel 24-bit ADC with programmable-gain amplifier to resolve
sub-V signals. Some devices have a 32-channel multiplexer before their ADCs, and DACs
with resolutions of 8 to 12 bits are also available. The list of analogue peripherals may also
include up to 3 comparators with programmable response time and hysteresis.
The company provides several development tools including a free integrated development
environment that supports the use of the popular open-source Small Device C Compiler
(SDCC). A configuration wizard application helps much in configuring the peripherals
properly by generating even the source code (assembly or C).
Hardware development platforms are also available. There are simple and full-featured
development kits for almost all C8051Fxxx processors.
2.3
The CIP-51 architecture
The Silicon Laboratories C8051Fxxx microcontrollers have the so-called CIP-51 architecture
[6]. The simplified block diagram is shown in Figure 2.1.
PSW
TMP1
ACC
RAM ADDRESS
ALU
RAM
TMP2
DATA POINTER
PROGRAM COUNTER
PROGRAM
ADDRESS
REGISTER
PC INCREMENTER
BUFFER
FLASH/PGM MEMORY
INTERNAL BUS
INSTRUCTION
REGISTER
STACK POINTER
INTERRUPT
CONTROLLER
MEMORY
INTERFACE
SFR INTERFACE
I/O PORTS
PERIPHERALS
Figure 2.1. A simplified CIP-51 architecture.
The architecture is closely matched with the original 8051 architecture developed by Intel;
code compatibility is provided. The main improvements include much faster instruction
execution, integrated flash memory and larger integrated RAM.
In the following the main features of the architecture will be discussed.
2.3.1 Registers
The following table summarises the 8-bit registers, with short descriptions and the reset
values [2]. The registers can be used in several instructions. The accumulator (A or ACC)
14
Architecture and properties of the C8051Fxxx microcontroller family
holds the result of arithmetic and logic operations and the program status word (PSW), and
contains several flags modified by operations. Additional registers support indirect
addressing and stack handling. Instructions typically execute faster when the operands are
registers.
Register
A, ACC
B
R0.–R7
PSW
Description
accumulator, ALU result
0
general-purpose register and register for multiplication and
division
0
general-purpose registers, R0 and R1 are also used in indirect
addressing
0
Bit 7: CY
carry bit (set by addition or subtraction, ADDC,
SUBB)
0
Bit 6: AC
auxiliary carry bit (at 3rd bit, used in 4-bit
arithmetics)
0
Bit 5: F0
user flag
0
Bit 4: RS1
Bit 3: RS0
DPH, DPL
SP
Reset value
R0–R7 at
00: 0x00
R0–R7 at
01: 0x08
R0–R7 at
10: 0x10
R0–R7 at
11: 0x18
0
0
Bit 2: OV
overflow (set by instructions MUL, DIV, ADD,
SUBB)
0
Bit 1: F1
user flag
0
Bit 0: PAR
parity bit: 1 if sum of bits in A is 1
0
DPTR, data pointer, used in 16-bit indirect code or RAM
addressing
0
stack pointer, modified by subroutine and interrupt routine calls or
push/pop instructions
7
2.3.2 Special function registers
The special function registers (SFRs) are used to access the peripherals and some registers.
For example, ACC is the same as A (accumulator); therefore, it can be accessed as an SFR or
as a register. This allows the accumulator to be used in some instructions when registers
cannot be used (like push and pop, see later)
SFRs can be accessed by direct addressing instructions, where the address falls in the range
of 0x80–0xFF. Therefore, SFRs can be thought of as memory-mapped registers; the program
can read or write their content as if they were in the RAM.
The following table shows the standard 8051 SFR registers.
15
Architecture and properties of the C8051Fxxx microcontroller family
Address
0
1
2
3
4
5
TL1
TH1
6
7
0xF8
0xF0
B
0xE8
0xE0
ACC
0xD8
0xD0
PSW
0xC8
0xC0
0xB8
IP
0xB0
P3
0xA8
IE
0xA0
P2
0x98
SCON
0x90
P1
0x88
TCON
TMOD
TL0
TH0
0x80
P0
SP
DPL
DPH
SBUF
Note that SFRs in column 0 are bit addressable.
The SFRs listed in the table are the following (some of them will be discussed in the next
chapters):
P0, P1, P2 and P3 are the port input/output SFRs that are associated with the pins of
the microcontroller. For example, the byte written to P0 determines the logic signal
on the 8 pins corresponding to P0. The programmer must be careful: for example,
writing 1 to P0 sets the least significant bit but will clear all the other 7 bits. Since the
P0 register is bit addressable, a single bit can be written or read without affecting the
other bits. For example, setting P0.0 sets the least significant bit only; all the other
bits remain unchanged. Bit addressing is also useful for accessing a single bit of the
status and other registers where the individual bits have special meanings.
ACC and B provide SFR access to the accumulator and to the B register.
PSW is the program status word. Its individual bits are accessible using bit
addressing. For example, PSW.7 is the carry bit.
SP is the stack pointer.
DPL and DPH are the low- and high-order bytes of the data pointer DPTR.
IE and IP are the interrupt enable and priority registers. Their individual bits are
accessible using bit addressing.
TCON, TMOD, TL0, TH0, TL1 and TH1 are used to access and control the Timer 0
and Timer 1 peripherals.
16
Architecture and properties of the C8051Fxxx microcontroller family
SCON and SBUF are associated with the serial port communication peripheral.
2.3.3 Memory structure
8051 processors have Harvard architecture [2]; they have separate memory for code and
data. The code memory can store constant data, so it can be used as a read-only data
memory. Two types of RAM are available: internal and external. The internal RAM size is 256
bytes, while the external RAM is addressed by a 16-bit pointer, so the maximum size is
64 kbyte.
Figure 2.2 shows the internal RAM structure. The first 128 bytes (from 0x00-0x7F) can be
accessed by direct or indirect addressing. The general-purpose registers occupy 8 bytes at the
location defined by the RS0 and RS1 bits of the PSW register. The 16-byte space at address
0x20-0x2F is bit addressable, so 128 individual bit variables can be used here.
0x00-0xFF
SFR
0x80-0xFF
(direct)
0x80-0xFF
(indirect)
0x00-0x7F
0x30-0x7F
BIT
ADDRESSABLE
0x00-0x7F
(direct,
indirect)
RS1,RS0=11
RS1,RS0=10
RS1,RS0=01
RS1,RS0=00
0x20-0x2F
0x18-0x1F
0x10-0x17
0x08-0x0F
0x00-0x07
Figure 2.2. Internal memory structure of CIP-51 microcontrollers.
The SFR registers are mapped to the upper 128 bytes of the address space. SFRs are accessed
by direct addressing; otherwise, the upper 128 bytes of the internal RAM can be used. Note
that since stack handling is based on indirect addressing by the stack pointer, the upper 128
bytes of RAM can also be used as stack space. Upon reset, the stack pointer has the value of 7
and increases from there. However, it is best to set the initial value of the stack pointer (SP)
to the first free location of data memory, just above the variables. In this case, all free
memory is available as stack.
The external RAM (XRAM) was originally provided by SRAM chips, but modern C8051Fxxx
processors integrate a certain amount (up to 8192 bytes) of this kind of RAM. XRAM memory
can only be accessed by 16-bit indirect addressing using the DPTR pointer (DPH and DPL
registers).
XRAM at 0x00-0xFF can also be accessed by 8-bit indirect addressing using either the R0 or
the R1 register.
Since off-chip memory can be slower than the on-chip memory, the control timing (data
setup and hold time, write/read pulse width, etc.) can be set by dedicated SFR registers.
17
Architecture and properties of the C8051Fxxx microcontroller family
Figure 2.3 shows the XRAM arrangement in C8051Fxxx processors. The processor can be
configured to access the on-chip memory only, the off-chip memory only or on-chip only if it
is available and off-chip otherwise. The 8-bit addressable space can also be moved to another
256-byte page. Note that not all C8051Fxxx processors support off-chip memory.
ON-CHIP
0x0100(16-bit indirect)
0x0000-0x00FF
(8-bit indirect)
OFF-CHIP
0x0100-0xFFFF
(16-bit indirect)
0x0000-0x00FF
(8-bit indirect)
Figure 2.3. External memory structure of CIP-51 microcontrollers.
2.3.4 Addressing modes
Data can be accessed in different ways depending on its location (register, memory or code)
and on the so-called addressing mode. The following table summarises the four possible
addressing modes and shows examples.
Addressing
mode
MNEMONIC
example
Description
register
MOV A, B
A = B, copy the content of B to A
immediate
constant
MOV A, #10
A = 10 (value), copy the value 10 to A
direct
MOV A, 10
MOV A, P0
A = byte in internal RAM at address 10
A = bits at port P0 (SFR access)
indirect
MOV A, @R0
MOVX A,@DPTR
A = byte in internal RAM at address pointed to by R0
A = byte in external RAM at address pointed to by DPTR
2.3.5 Instructions
A brief summary of the available instructions are given in the following [2]. Instructions are
classified into groups and tables summarise their function and the flags affected by them.
18
Architecture and properties of the C8051Fxxx microcontroller family
2.3.5.1 Arithmetic operations
MNEMONIC
OPERATION
ADDRESSING
FLAGS
(C-style syntax)
DIR
IND
REG
IMM
CY
AC
OV
P
ADD A, byte
A=A+byte
ADDC A, byte
A=A+byte+C
SUBB A, byte
A=A–byte–C
INC A
A=A+1
INC byte
byte=byte+1
INC DPTR
DPTR=DPTR+1
DEC A
A=A–1
DEC byte
byte=byte–1
MUL AB
A=(B*A) % 256
B=(B*A) / 256
only A and B
0
DIV AB
A=integer part of A/B
B=remainder of A/B
only A and B
0
DA A
Decimal Adjust
only A
only DPTR
only A
2.3.5.2 Logic operations
MNEMONIC
OPERATION
(C-style syntax)
ADDRESSING
FLAG
DIR
IND
REG
IMM
P
ANL A,byte
A=A & byte
ANL byte,A
byte=byte & A
ANL byte,#const
byte=byte & const
ORL A,byte
A=A | byte
ORL byte,A
byte=byte | A
ORL byte,#const
byte=byte | const
XRL A,byte
A=A ^ byte
XRL byte,A
byte=byte ^ A
XRL byte,#const
byte=byte ^ const
19
Architecture and properties of the C8051Fxxx microcontroller family
2.3.5.3 Accumulator manipulation
MNEMONIC
OPERATION
(C-style syntax)
ADDRESSING
FLAGS
CY
AC
OV
P
CRL A
A=0
only A
CPL A
A = ~A
only A
RL A
Rotate A left by 1 bit
A = A << 1
only A
RLC A
Rotate A left through Carry
A = (A << 1) + C
C = bit 7 of the original value of A
only A
RR A
Rotate A right by 1 bit
A = A >> 1
only A
RRC A
Rotate A right through Carry
A = (A >> 1) + (C << 7)
C = bit 7 of the original value of A
only A
SWAP A
Swap nibbles of A
only A
2.3.5.4 Bit-variable operations
MNEMONIC
OPERATION (C-style syntax)
ANL C,bit
C = C && bit
ANL C,/bit
C = C && !bit
ORL C,bit
C = C || bit
ORL C,/bit
C = C || !bit
MOV C,bit
C = bit
MOV bit,C
bit = C
CLR C
C=0
CLR bit
bit = 0
SETB C
C=1
SETB bit
bit = 1
CPL C
C = !C
CPL bit
bit = !bit
20
Architecture and properties of the C8051Fxxx microcontroller family
2.3.5.5 Data move operations
MNEMONIC
OPERATION
(C-style syntax)
ADDRESSING
DIR
IND
REG
IMM
MOV A,byte
A = byte
MOV byte,A
byte = A
MOV byte1, byte2
byte1 = byte2
MOV DPTR,#const16
DPTR = 16-bit immediate constant
PUSH byte
SP = SP+1
RAM[SP]= byte
POP byte
byte = RAM[SP]
SP = SP-1
XCH A,byte
exchange the content of A and byte
XCHD A,@Ri
exchange low nibbles of A and
RAM[Ri]
2.3.5.6 External and code memory access
MNEMONIC
OPERATION (C-style syntax)
MOVX A,@Ri
A = XRAM[Ri]
MOVX @Ri,A
XRAM[Ri]= A
MOVX A,@DPTR
A = XRAM[DPTR]
MOVX @DPTR,A
XRAM[DPTR] = A
MOVC A,@A+DPTR
A = CODE[A+DPTR]
MOVC A,@A+PC
A = CODE[A+PC]
21
Architecture and properties of the C8051Fxxx microcontroller family
2.3.5.7 Jump and subroutine call
MNEMONIC
OPERATION (C-style syntax)
JMP address
Jump to address
PC = address
JMP @A+DPTR
Jump to A+DPTR
PC = A+DPTR
ACALL address
Call subroutine at 11-bit <address>
PC = PC+2
SP = SP+1
RAM[SP] = PC lower order byte
SP = SP+1
RAM[SP] = PC higher order byte
PC = address
LCALL address
Call subroutine at 16-bit address
PC = PC+3
SP = SP+1
RAM[SP]= PC lower order byte
SP = SP+1
RAM[SP] = PC higher order byte
PC = address
2.3.5.8 Return from subroutines and interrupts
MNEMONIC
OPERATION (C-style syntax)
RET
Return from subroutine
PC = RAM[SP]*256 + RAM[SP-1]
SP = SP-2
RETI
Return from interrupt
PC = RAM[SP]*256 + RAM[SP-1]
SP = SP-2
restore the interrupt logic to accept further interrupts
NOP
No operation
22
Architecture and properties of the C8051Fxxx microcontroller family
2.3.5.9 Conditional jumps
Note that if a conditional jump occurs, the program counter is updated as PC=PC+address,
where the address is an 8-bit two’s complement number in the range of -128 to 127.
MNEMONIC
OPERATION
ADDRESSING
DIR
IND
REG
JZ address
Jump if A = 0
only A
JNZ address
Jump if A !=0
only A
DJNZ byte, address
Decrement and jump if not zero
CJNE A,byte, address
Jump if A != byte
CJNE byte,#const, address
Jump if byte != const
JC address
Jump if C = 1
JNC address
Jump if C = 0
JB bit, address
Jump if bit = 1
JNB bit, address
Jump if bit = 0
JBC bit, address
Jump if bit = 1; CLR bit
IMM
2.3.6 Instruction timing and coding
The CIP-51 architecture executes most of the operations in 1 or 2 system clock cycles.
Depending on the specific device, the system clock can have maximum frequencies from
25 MHZ to 100 MHz; therefore, the fastest instruction execution time can be as low as 10 ns.
The following table shows the distribution of the cycle time for the available instructions.
Note that processors operating at clock frequencies above 25 MHz may use pipelining
(prefetching instructions into a fast buffer) due to flash code memory access time limitations.
This means that the processor may stall for a few clock cycles in some cases (for example,
when a jump or a branching occurs).
cycles
1
2
2/4
3
3/5
4
5
4/6
6
8
instructions
26
50
5
10
7
5
2
1
2
1
The CISC architecture of the 8051 processors allows instructions to be coded using 1, 2 or 3
bytes. The first byte is associated with the type of the instruction, while the remaining one or
two identify the operands. A few examples are shown in the next table.
instruction
1. byte
2. byte
ADD A, Rn
0010 1nnn
ADD A, #10
0010 0100
0000 1010
ANL 15,#10
0101 0011
0000 1111
DIV AB
1000 0100
3. byte
cycles
1
2
0000 1010
3
8
23
Architecture and properties of the C8051Fxxx microcontroller family
JZ address
0110 0000
relative address
2/4
2.3.7 Interrupt handler
Event handling is one of the most important aspects of embedded programming. Events can
be generated by peripherals such as timers, communication ports, and analogue-to-digital
converters and also by changes of external signals. In the 8051 environment, events can
generate interrupts, which can be serviced by subprograms. If an event occurs, a flag is set
(which can even be polled by software) and an associated interrupt routine is called if
enabled. The interrupt mechanism is visualised in Figure 2.4. When the event occurs, the
system detects this within the system clock cycle time t (the reciprocal of the system clock
frequency). Upon completion of the currently running instruction (which can take from 1 to 8
cycles; see the previous chapter), an LCALL instruction is executed and the program jumps
to the interrupt service routine. After processing, a RETI instruction is executed to return to
the main program and restore the interrupt logic to accept further interrupts. One can easily
see that the time elapsed from the event to the execution of the first instruction of the
interrupt handler requires a minimum latency time and has some uncertainty as well.
t
t
RETI (6t)
Instruction
#3
Interrupt
handler
Instruction
#2
Detect (t)
Instruction
#1
t
LCALL (5t)
t
main program paused
latency: 7t -19t
IRQ
Figure 2.4. Interrupt mechanism. The interrupt latency time varies from 7 to 19
system clock periods.
It is very important to keep this in mind, since in a real-time application it can cause
problems. For example, if a periodic interrupt is used to generate a square wave, this causes
some fluctuation of the switching times, which should be considered, especially when
switching times are short. For example, if a 100-kHz square wave is to be generated by a
timer interrupt routine, the routine must be called 200000 times per second to change the
signal state at every 5 s. At a system clock frequency of 25 MHz, the clock period is 40 ns, so
the latency time can vary from 740 ns to 1940 ns, resulting in an uncertainty of (197)40 ns=480 ns. This can cause a maximum error of 9.6% in the 5-s switching time.
The main program can be interrupted at any time, even during a task requiring multiple
instructions. This means that all temporary variables and register content modified by the
24
Architecture and properties of the C8051Fxxx microcontroller family
interrupt service routine must be saved at the beginning of the interrupt handling routine
and must be restored upon return to the main program. Also note that the peripheral state or
the input/output can also be changed during interrupt handling, which also needs careful
attention.
If an interrupt service routine is running, another request can only be serviced if it has higher
priority. Only two priority levels are provided, so no further interrupts can be serviced. The
priority of the interrupts is defined by the bits of the IP, EIP1 and EIP2 registers.
Correspondingly, there are only two priority levels: normal and high. If more interrupts are
detected simultaneously, the higher priority interrupt will be serviced first. Since the
interrupt flag set by the event can be cleared only when the associated interrupt routine is
called, no interrupts are lost if multiple requests are detected or the request occurs during the
servicing of another one. Of course, if a request is generated two or more times without
servicing, only the last request can be serviced.
Interrupt sources are associated with a number that also defines priority (lower number
means higher priority). The execution address of the interrupt routines is fixed and only 8
bytes are available up to the next address. Therefore, longer routines are located elsewhere
and only a jump to that space is needed here.
A few interrupt flags are automatically cleared by the hardware when the service routine is
called; all others must be cleared by the software – otherwise, the request will remain active
and will be serviced continuously.
Interrupts can be individually enabled and disabled using the bits of the IE, EIE1 and EIE2
registers. IE.7 (which can also be accessed as the SFR bit EA) is a global enable bit. Note that
if an interrupt is enabled, it must have an interrupt handler code; otherwise, the processor
can go into an uncertain state.
Reset
0x0000
-
/INT0 external
0x0003
0
IE.0
Timer 0 overflow
0x000B
1
/INT1 external
0x0013
Timer 1 overflow
Source
Flag
Priority bit
Execution
Address
Number
Enable bit
The interrupt sources available on C8051F410 processors are listed in the following table [6].
name
Cleared by
hardware
-
yes
IP.0
IE0
yes
IE.1
IP.1
TF0
yes
2
IE.2
IP.2
IE1
yes
0x001B
3
IE.3
IP.3
TF1
no
UART
0x0023
4
IE.4
IP.4
RI, TI
no
Timer 2 overflow
0x002B
5
IE.5
IP.5
TF2H, TF2L
no
SPI0
0x0033
6
IE.6
IP.6
SPIF, WCOL,MODF,
RXOVRN
no
SMB0
0x003B
7
EIE1.0
EIP1.0
SI
no
smaRTClock
0x0043
8
EIE1.1
EIP1.1
ALRM, OSCFAIL
no
25
Architecture and properties of the C8051Fxxx microcontroller family
ADC0 Window
Comparator
0x004B
9
EIE1.2
EIP1.2
ADC0 End of
Conversion
0x0053
10
EIE1.3
Programmable
Counter Array
0x005B
11
Comparator 0
0x0063
Comparator 1
AD0WINT
no
EIP1.3
AD0INT
no
EIE1.4
EIP1.4
CF, CCFn (up to six
flags)
no
12
EIE1.5
EIP1.5
CP0FIF, CP0RIF
no
0x006B
13
EIE1.6
EIP1.6
CP1FIF, CP1RIF
no
Timer 3 overflow
0x0073
14
EIE1.7
EIP1.7
TF3H, TF3L
no
Voltage regulator
dropout
0x007B
15
EIE2.0
EIP2.0
-
no
Port match
0x0083
16
EIE2.1
EIP2.1
-
no
26
Assembler and C programming
3
Assembler and C programming
Programming 8051 microcontrollers requires special attention due to limited processing
power, small memory space and the direct access of peripherals. No operating system is used
in most cases; therefore, the programmer must take care of everything that the
microcontroller does. The programmer must have extensive knowledge about the hardware,
including memory types, instructions, SFRs, the interrupt handler and digital and analogue
peripherals.
Simple programs can be written in assembler, but C is recommended for general-purpose
code development. Although code optimisations are done by the C compiler, some fragments
of code can be further enhanced by mixing assembler and C. C compilers allow inserting
assembly code in C and C and assembly code can work on the same variables. C programmers
can write efficient embedded code only if they know assembler as well.
3.1
SDCC C compiler
There are many 8051 C Compilers on the market. The most popular professional compiler is
the KEIL C51 [3] and there exists an open-source alternative called Small Device C Compiler
(SDCC) [4]. The free availability, good quality and the detailed documentation of SDCC make
it an ideal tool to use in education. Here only the most important additions to C are
mentioned that are needed to use the features of the 8051 processor.
Variables can be placed in different memory types; for this purpose, the compiler supports
the declaration of storage classes:
__data unsigned char x;
// internal RAM
__xdata unsigned char x; // external RAM
__idata unsigned char x; // internal indirectly addressable RAM
__pdata unsigned char x; // 8-bit addressed external RAM
__code unsigned char x=3; // constant in code memory
__bit b;
// bit addressable RAM
__sfr __at 0x80 P0;
// SFR byte
__sbit __at 0xD7 CARRY;
// SFR bit
__xdata __at (0x4000) unsigned char x[16];
// external RAM, absolute address
__code __at (0x7f00) char Msg[] = "Message"; // code memory, absolute address
__bit __at (0x80) GPIO_0;
// bit, absolute address
Inserting assembly into C can be done using the __asm and __endasm directives:
unsigned char x;
__asm
clr a
mov R0,#0
mov R1,#0x80
mov a,R2
mov _x,a
jz
L1
mov R0,#0
L1:
mov R1,#1
__endasm;
//
/*
//
//
//
//
//
//
beginning of assembly code fragment
C style comment */
P0, C++ style comment
C style hexadecimal constant
copy the content of R2 register to accumulator
accessing x declared in C
use of a label
clear register R0
// load 1 into register R1
// end of assembly code fragment
The variable types are listed in the following table.
27
Assembler and C programming
type
width
(bits)
default
signed range
unsigned range
__bit
1
unsigned
-
0,1
char
8
signed
-128–127
0–255
short
16
signed
-32768–32767
0–65535
int
16
signed
-32768–32767
0–65535
long
32
signed
-2147483648
+2147483647
0–4294967296
float
IEEE754
32
signed
pointer
8-24
generic
1.175494351 ∙ 10-38,
3.402823466 ∙ 10+38
Most of the variable types are the same as in standard C, but due to the limited resources,
there are some exceptions. For example, the SDCC compiler allows defining bit variables
using the __bit keyword. The variable can be placed in the bit addressable memory space,
optimising memory usage. Floating-point arithmetic is supported; however, only single
precision 4-byte wide float type variables can be used. This is fine in most embedded
applications due to its 6-7 digits of precision. Double precision is not available, because it
would take a long execution time and significantly longer code.
Generic pointers are rather special, since the 8051 microcontroller uses several different
memory types. The 3-byte wide generic pointer defines the address in two bytes and the
memory type (internal RAM, external RAM or code memory) on the third byte. Of course, the
programmer can declare a pointer that points explicitly to an internal memory location. This
pointer is stored in a single byte since only 256 different locations are possible.
Microcontroller programming often requires the manipulation of bits. Here are two simple
examples:
x = x & ~(1 << 3);
x = x | (1 << 3);
// clearing a bit
// setting a bit
Working with integer numbers that are not 8, 16 or 32 bits long is also common. Left or right
shifting may be required, but care must be taken concerning signed and unsigned numbers,
since the behaviour of the shift operator is different for signed and unsigned numbers. For
example, to handle a 2s complement 12-bit number (the four most significant bits are in
ADCH and the eight least significant bits are in ADCL), one may use the following code:
short x; // define a signed 16-bit integer variable
x = (ADCH << 12) + (ADCL << 4);
// left justified
x = ((signed short)(ADCH << 12) >> 4) + ADCL; // right justified
In most cases, unsigned integers are used for the data of the peripherals (such as counter
value or ADC value). The programmer should always declare the variable as unsigned if it
contains an unsigned number. However, the use of negative constants can help in some
cases, especially when calculating the value used in timer programming (see Chapter 5):
unsigned short x; // define an unsigned 16-bit integer variable
x = -100; // this is equivalent to 65536-100, i.e. 65436
28
Assembler and C programming
Note that 65536 cannot be represented by an unsigned short variable and long arithmetic
would take more time and longer code.
3.2
Interrupt programming in assembler
A simple assembler interrupt handler example code is listed below. At the beginning, the
registers in use are pushed onto the stack and restored at the end of the routine. The
interrupt pending flag (in this example RI) is cleared. Note the use of assembler-style
comments.
push
push
clr
mov
add
mov
pop
ACC
PSW
RI
A,SBUF
A, #1
P0,A
PSW
pop ACC
reti
;
;
;
;
;
;
;
;
;
;
ACC (SFR access of A) to the stack
status register to the stack
clear interrupt flag
A is changed here
A and PSW are changed here
copy the content of the accumulator to port P0
PSW restored here
reverse order!
ACC (A) is restored here
return to the main program
If the R registers are used, they must be saved and then restored as well. However, the 8-byte
register bank can be moved to four memory locations; therefore, the interrupt routine can
use one bank while the main code uses another bank.
push PSW
; status register to the stack
mov PSW,#8 ; use register bank #1
; use R registers here
pop PSW
; PSW and the register bank selection is restored here
The following complete assembler code illustrates the use of a timer interrupt to make an
LED blink. The system clock after reset for the C8051F410 processor is 191406 Hz, and its 16bit Timer 2 runs with 1/12 of this rate by default: 191406/12 Hz 15950 Hz. Since the
interrupt occurs when the 16-bit timer overflows, 15950 steps are needed to reach 216=65536
in order to wait 1 second before overflow. Therefore, the initial value of the timer should be
set to 65536-15950 = 49586 = 0xC1B2, and this value will be reloaded upon overflow
automatically. This way, a periodic interrupt will be generated every second. Note that the
detailed description of the peripherals can be found in the following chapters.
$include (C8051F410.INC) ; load the definitions used for the C8051F410 MCU
LED
EQU
P0.2
CSEG at 0000h
jmp Main
ORG 002Bh
anl TMR2CN,#07Fh
cpl LED
reti
Main:
anl PCA0MD, #0BFh
mov PCA0MD, #000h
mov XBR1,
#040h
mov TMR2RLL, #0B2h
mov TMR2RLH, #0C1h
; the LED is connected to bit 2 of port 0.
; reset, jump to the label ‘Main’
;
;
;
;
Timer 2 interrupt location
clear interrupt flag
complement LED
return from interrupt
;
;
;
;
;
switch watchdog off
switch watchdog off
enable the crossbar to allow input and output
set the Timer 2 reload register (low and high bytes)
to provide 1-Hz interrupt rate
29
Assembler and C programming
mov
mov
mov
mov
jmp
END
3.3
TMR2L,
TMR2H,
TMR2CN,
IE,
$
#0B2h
#0C1h
#004h
#0A0h
;
;
;
;
;
Timer 2 counter initial value
is the same as the reload value
Start Timer 2 now
enable global interrupts and Timer 2 interrupt
repeat forever, interrupt routine will blink the LED
Interrupt handling in C
The SDCC C compiler for the 8051 family of processors supports interrupt programming. If a
function is intended to be an interrupt handler, it must be declared accordingly. The
programmer should use the __interrupt keyword and include the number of the interrupt to
identify which interrupt will be handled. For example:
void InterruptHandler(void) __interrupt 5
or using predefined constants
void InterruptHandler(void) __interrupt INT_TIMER2
This code defines an interrupt handler (no return value or input arguments can be defined)
for the Timer 2 interrupt that is numbered as 5. The location of the R0—R7 register bank can
also be defined:
void InterruptHandler(void) __interrupt 5 __using 1
where the number following __using keyword defines which one of the four possible register
banks are used (the default is 0). This can help the compiler to generate faster code since
saving/restoring the registers is not necessarily needed.
An interrupt handler can use local variables, which are initialised in any execution of the
routine. A simple example is the use of temporary variables. However, in some cases a
variable must retain its value after exiting from the interrupt handler routine. For example, if
the code must count how many interrupts are generated, a counter value must be
incremented each time the interrupt routine is called. This variable can be declared as a
global variable just at the beginning of the code, but if it is used in the interrupt routine only,
it is best to hide the variable from other parts of the code. In this case, the variable should be
declared in the interrupt handler routine using the static keyword. The following example
code toggles the state of an LED upon every hundredth Timer 2 overflow interrupt request.
The static variable named ‘counter’ counts how many requests are detected, and if this
number reaches 100, the LED is toggled. Since the counter value is only used in the interrupt
routine, it can be declared within the scope of the routine. Note that the initialisation of the
variable is done only when the program starts.
/*************************************************************************
Timer 2 interrupt handler
**************************************************************************/
void IntHandler(void) __interrupt 5
{
static unsigned int counter = 0; // will be initialised only once!
TMR2CN&=~0x80; // or TF2H=0, clear interrupt pending flag
counter++;
// increment the value of the counter variable
if (counter == 100) // the routine has been called 100 times
{
counter = 0;
// reset counter
LED = !LED;
// complement LED
}
}
30
Assembler and C programming
If a variable is used both in an interrupt routine and in other parts of the program then it
must be declared as volatile:
// define the global variable that is used both in the interrupt routine
// and in the main program
volatile unsigned char counter;
/*************************************************************************
Timer 2 interrupt handler
**************************************************************************/
void IntHandler(void) __interrupt 5
{
TMR2CN&=~0x80; // or TF2H=0, clear interrupt pending flag
counter++;
// increment the value of the counter variable
}
The volatile keyword tells the compiler that the variable can be changed at any time, so it
cannot be assumed to remain unchanged in a sequence of a few lines of computations. It is a
typical error to get unexpected results due to missing volatile declarations.
There are several considerations to be kept in mind if both interrupts and regular code work
on the same data. The main program may work on data in several assembler instructions that
are hidden from the programmer. For example, checking the value of a 16-bit number needs
several assembler instructions: the higher- and lower-order bytes must be separately checked
– it is not an atomic operation. If an interrupt routine changes the value of this integer
during this process, an unexpected error can occur. Therefore, non-atomic operations must
be protected.
One solution is the use of critical blocks as shown below:
volatile unsigned short x; // variable declaration
.
.
/*************************************************************************
Interrupt handler routine
**************************************************************************/
void IntHandler(void) __interrupt 5
{
TF2H=0; // clear interrupt pending flag
x++;
// increment the value of x
}
.
.
.
__critical // define the critical block
{
if (x>1024) DoSomething(); // here x cannot be changed by IntHandler()
}
At the beginning of the critical block, the compiler disables interrupts (saves then clears EA)
and at the end re-enables them if necessary (restores the value of EA); therefore, no interrupt
can be executed within the critical block. Note that this may cause extra interrupt latency and
even missing interrupts if the critical block needs too much time to complete.
It is also possible to protect a variable from being modified by the interrupt routine by
introducing a user flag as illustrated in the following example.
In the main code:
31
Assembler and C programming
volatile __bit protect_x;
volatile unsigned short x;
// flag variable
// this variable can be changed in the
// interrupt handler routine
protect_x=1;
if (x>10) DoSomething();
protect_x=0;
// switch protection on
// here x cannot be changed
// switch protection off
In the interrupt handler:
if (!protect_x)
// allow changes only if protect_x is 0
{
x=(ADC0H << 8) | ADC0L; // do the change of x
}
Note that some events generate the same interrupt. For example, interrupt 6 corresponds
both to serial port receive and transmit. Therefore, the interrupt handler must check if the RI
or the TI interrupt flag is set and execute the code accordingly.
3.4
Interrupt programming guidelines
Interrupt programming is rather difficult, as there are many potential pitfalls. Response time,
latency, processing time, variable and memory content, priority, peripheral status,
simultaneous requests and a lot more are all to be considered carefully. Debugging is not easy
due to the complexity and the differences between real-time versus single stepping operating
modes. Here are some guidelines to follow to reduce the probability of unexpected behaviour.
If the interrupt handler and the rest of the code work on the same data,
synchronisation must be carefully designed.
Atomic and non-atomic operations should be identified. Temporary results of nonatomic operations on data must be protected from being modified by an interrupt
service routine.
Before enabling the interrupt, the corresponding peripheral must be configured and
the variables used must be initialised.
An enabled interrupt must have an interrupt handler routine.
The stack size must be set to have enough space for saving and restoring variables and
for subprogram calls. It is best to set the initial value of the stack pointer (SP) to the
first free location of data memory, just above the variables. In this case, all free
memory is available as stack.
Interrupt routines should take as short a time as possible, and only the most
important processing that cannot be done by the main program should be performed
here.
Too frequent interrupt calls can slow the processor down; too frequent multiple
concurrent interrupt requests can cause a failure to service certain requests.
Multiple interrupts can generate extra interrupt latency time – another reason to keep
interrupt service routine execution time as short as possible.
Interrupt pending flags must be cleared in the interrupt service routine.
Interrupt priorities must be taken into account. Priority of critical interrupts requiring
fast response must be set to high.
Consider using different register banks for interrupts.
Several interrupt flags can be associated with the same interrupt routine; therefore,
the routine must handle all of them.
32
Assembler and C programming
3.5
Do not mix event handling by polling the interrupt pending flag with event handling
by an interrupt service routine. Choose between the two possibilities.
Do not use library routines in interrupt routines. They can be slow and can be nonreentrant. Only reentrant functions can be called while one instance is already
running. On the other hand, reentrant functions are slower and need more resources.
For example, floating-point arithmetic and 16-bit and 32-bit integer multiplication,
division and modulus operations use non-reentrant support functions. See the SDCC
manual how to overcome this limitation.
Using an integrated development environment and the associated tools
Silicon Laboratories provides a free integrated development environment (IDE) and several
other software tools to support code development [2].
Many different compilers can be integrated with the IDE, including the open-source and free
SDCC C compiler. In the Tool Chain Integration menu item the compiler can be selected.
Projects can be created and header and C source files or libraries can be added to the project
as usual in IDEs.
The IDE handles the USB debug adapter that connects the PC to the target microcontroller.
The adapter allows the downloading of the compiled code and also provides debug functions.
After compilation, the code can be downloaded.
Breakpoints can be defined, so after starting the code, real-time execution will automatically
stop when a breakpoint is found. This means that the real system can be monitored and no
simulation is performed. During debugging both the assembly and C code can be viewed.
Another very useful feature is the watch window, which can show the actual content of
several variables. Besides this, there are many debug windows available to view and even
change the contents of the registers, memories and peripherals of the 8051.
Single stepping, full speed execution, run-to-cursor execution are all possible. An example
screenshot of the IDE can be seen in Figure 3.1. Solid red circles indicate breakpoints, while
the blue bar shows the current source line being executed. Keyword highlighting is also
provided.
On the right the peripheral watch windows – programmable counter array (PCA) window;
the 8051 register (including the program counter, PC and accumulator, ACC) window; the
disassembly windows and the variable watch window (which shows the INT0counter and the
PCAcounter) – can be seen. Red colour indicates recently changed values.
33
Assembler and C programming
Figure 3.1. Debugging in the Silicon Laboratories IDE.
Note that peripherals are stopped if the program is paused in the debugger, and after a
single-step operation the code is halted again. This means that all peripherals are stopped at
this point. This must be kept in mind, because full-speed execution might differ significantly.
Examples are given below.
3.6
Assume that the voltage reference is switched on in a program line and in the next
line an analogue-to-digital conversion is initiated. Since the voltage reference needs a
settling time of a few milliseconds, at full-speed execution the A/D conversion value
will be invalid (switching the reference on needs only a few clock cycles), while in
single stepping mode there is enough time for the voltage reference to settle.
Sending a byte over a serial port is initiated by writing the SBUF register, for
example: SBUF=0xAA;. Since the individual bits of SBUF are transferred at a certain
rate (at each overflow of a timer), several clock cycles are needed to complete the
transfer. Therefore, if the user places a breakpoint after SBUF loading (that is a twocycle instruction) or performs single stepping, the data will not be transferred,
because the timers will be halted.
Config Wizard
The C8051Fxxx processors have a very rich set of peripherals that are configured with many
SFRs — typically each has independent configuration bits. Therefore, it would be very hard to
read the datasheet and set the individual bits of these SFRs accordingly, and the probability
of making an error would be rather high. The Config Wizard 2 free graphical user interface
development tool helps to configure the processor and its peripherals very efficiently. After
34
Assembler and C programming
choosing the processor, it is possible to configure any of its peripherals by dialogue boxes and
the corresponding source code will be generated in C or assembly format. This code can be
copied into the user code. Figures 3.2 and 3.3 show two examples: the Port input/output and
the Timer configuration dialogue boxes.
Figure 3.2. Port input/output configuration dialogue box.
Figure 3.3. Timer configuration dialogue box.
35
Digital input and output; crossbar
4
Digital input and output; crossbar
Microcontrollers provide ports for general-purpose digital input and output signals [6]. The
ports are organised in 8-bit groups (named P0, P1, etc.), but bits can also be accessed
individually (for example, P1.3 in assembler or P1_3 in SDCC to access the third bit of port
P1). All port bits are associated with the pins of the package of the chip and can be configured
as input or as output, and have several operating modes. Some port pins can also be
configured in analogue mode.
4.1
The I/O structure
The port structure can be seen in Figure 4.1.
SET
DATA BUS
D
WRITE
Q
PORT OUT
C
CLEAR
DIGITAL
PERIPHERAL
PORT
DRIVER
PORT IN
READ
READ-MODIFY-WRITE
ANALOGUE
PERIPHERAL
ANALOGUE I/O
Figure 4.1. I/O port structure.
Writing to a port (for example, MOV P0, #1 in assembler or P0=1 in SDCC) means writing
the data into a D-latch that is connected to a port pad via the port driver. During reading
from a port, the port pad is connected to the internal data bus. Note that read-modify-write
instructions (for example, INC P1, ANL P1,#1 in assembler and P1++, P1&=1 in C) do not
read the state of the external signal itself, but rather use the output of the D-latch instead to
guarantee consistent operation.
The simplified schematic of the port driver is shown in Figure 4.2. A complementary
transistor pair can pull the line down to GND or up to Vdd (power supply of the driver stage)
and a third transistor can switch on a weak pull-up. The port input uses a Schmitt trigger to
guarantee valid logic levels for slowly changing or noisy signals. In order to use the analogue
mode, all transistors must be switched off and the input Schmitt trigger must also be
disabled.
36
Digital input and output; crossbar
PORT DRIVER
WEAK PULL-UP
Vdd
OR
ANALOGUE MODE
PUSH-PULL MODE
Rp
Vdd
AND
Vdd
PORT OUT
OR
OUT ENABLE
PORT IN
ANALOGUE IN
Figure 4.2. Simplified schematic of the I/O port driver. Bold indicates internal
I/O signals.
4.1.1
Port input
The port can be configured as digital input by switching off the output drivers. Therefore, it is
important to write logic 1 to the corresponding port bit, otherwise the transistor connected
to the ground will short-circuit the port pin to the ground. Push-pull mode must be disabled.
The weak pull-up (Rp, roughly 100 kΩ, actually a weak P-channel FET) can be enabled or
disabled globally for all port pins. When disabled, the leakage current is typically 10 nA at
room temperature and is guaranteed to be less than 1 A. This must be considered in
analogue mode, whereas digital signals will not typically be affected by this small current that
is matched with the specifications of other CMOS devices. The input capacitance is close to
5 pF and the diodes protect the internal circuitry against electrostatic discharge (ESD). The
simplified equivalent schematic is shown in Figure 4.3.
Rp
Vdd
Vdd
Vdd
IL
Figure 4.3. Port input configurations. On the left, the digital input with weak
pull-up is shown. On the right, the digital input with no pull-up and the
analogue input can be seen. The typical IL leakage current is about 10 nA but
can be as high as 1 A.
Note that the diodes protect the inputs from electrostatic discharge (ESD) and from over- or
undervoltage, but the current cannot exceed the specifications given in the absolute
37
Digital input and output; crossbar
maximum ratings section in the datasheet. Since the supply voltage is less than 5 V, some
ports provide 5 V tolerant inputs. In this case, the diode connected to Vdd is missing.
4.1.2 Port output
The ports can operate either as open-drain or as push-pull outputs.
In open-drain mode, the weak pull-up can be switched on (reset default) or off. Logic 1 state
can only be set by the large value (roughly 100 kΩ) pull-up resistor; therefore, the port
cannot be loaded and signal transition from 0 to 1 will be rather slow, since the external
capacitances can only be charged through the resistor. For example, a loading of 50 pF will
reduce the rise time (90% of the final value) to about 10 s. External pull-up resistors (down
to about 1 kΩ) can make the switching faster and can source more current to the load at the
expense of a larger quiescent current when 0 is written to the port bit. Some peripherals
(such as I2C) require open drain mode.
In push-pull mode the output drive strength is symmetric, and the port can sink and source
large currents and guarantee fast switching from 0-to-1 and from 1-to-0. Therefore, it is
strongly recommended to use push-pull mode for the output in most applications, especially
for communication peripherals, in order to avoid data corruption.
Vdd
Rp
Vdd
PORT OUT
PORT OUT
Figure 4.4. Open-drain and push-pull output modes.
4.2
Crossbar
After reset, the ports are not connected to the core and all peripherals are idle. Port pins can
be associated with the port latches or with the enabled peripherals, which can output or input
signals (see Figure 4.5). The priority crossbar provides a flexible way to connect the internal
peripherals and port latches to the port pins. If it is enabled, the port pins are accessible. If a
peripheral is used, its signals are associated with port pins. Peripherals are numbered and the
port pins are associated in this order with the enabled peripherals. For example, if peripheral
#1 is enabled with two signals and peripheral #5 is enabled with three signals, peripheral #1
will be connected to the first two pins (P0.0 and P0.1), while peripheral #5 will be associated
with the next three pins (P0.2, P0.3 and P0.4). The state of these pins cannot be modified by
writing to the port latches but their state can be monitored by reading the corresponding port
bit. The push-pull or open-drain settings can still be set by firmware.
38
Digital input and output; crossbar
PERIPHERAL #1
CROSSBAR
P0.0
PORT
CELL
PERIPHERAL #2
P0.7
PERIPHERAL #3
P1.0
PORT
CELL
P0
P1.7
P1
Figure 4.5. The crossbar assigns peripherals and port latches to port pins.
Crossbar settings and peripheral configuration are typically done just after reset. However, it
is possible to reconfigure the system during execution. In this case, the crossbar must be
disabled first, then can the changes be made before re-enabling the crossbar. Note that
during this process the port pins may exhibit transitions, which must be tolerated by the
system.
4.3
Port I/O applications
In this chapter port input and output application examples will be shown. The port I/O
allows realising various user interfaces and communication with external circuits.
4.3.1 Reading buttons and switches
R
C
Vdd
Vdd
R
Vdd
R
Vdd
R
One of the simplest and most common digital input types is the state of a button or a switch.
Figure 4.6 shows four ways of connecting buttons to the port pins. The most popular
connection uses a pull-up resistor and a grounded button, as can be seen in the two
configurations on the left. If the button is pressed, the corresponding logic value is 0. A
capacitor is sometimes used to eliminate bouncing and to reduce noise. Positive logic can be
realised by swapping the resistor and the button: in this case logic 1 is obtained when the
button is pressed.
Figure 4.6. Four ways of connecting a button or switch to the port pins. The two
configurations on the left represent negative logic, while the other two
correspond to positive logic. The value of R is typically 10 kΩ.
39
Digital input and output; crossbar
Figure 4.7 shows that if the internal weak pull-ups (Rp) are switched on, the external pull-up
resistor can be eliminated. Although the external capacitor may cause voltage at the input to
change slowly, the internal Schmitt trigger ensures reliable operation.
C8051Fxxx
Vdd
Vdd
Rp
Rp
Vdd
C8051Fxxx
Vdd
C
Figure 4.7.
The following, very simple source code shows an example of reading the state of a button
connected to the third bit of port P0:
#define BUTTON_ON (!P0_3) // define an alias to access
// the port bit of the button
/*************************************************************************
The main function
**************************************************************************/
void main(void)
{
while (1) // infinite loop; the microcontroller never stops
{
if (BUTTON_ON)
// if the button is pressed
{
while (BUTTON_ON); // wait while button is released
DoShortProcess();
// some process to be completed
}
}
}
There are many problems associated with the code above. For example, potential bouncing is
not handled and during process execution button pressings are lost.
An improved code uses a timer interrupt to detect button pressing only if the button is
pressed for a period of at least 100 ms:
#define BUTTON_ON (!P0_3) //
//
#define BUTTON_ON_TICK 10 //
//
define an alias to access
the port bit of the button
number of ticks to be counted
defines the minimum time for button detection
volatile bit ButtonPressed; // variable to indicate if the button
// has been pressed
/*************************************************************************
Timer interrupt handler routine
**************************************************************************/
void TMRHandler __interrupt TMRVECTOR // 10-ms period
{
// static variables retain their values upon exiting the function
40
Digital input and output; crossbar
static bit buttonstate=0; // this bit stores the state of the button
static bit detected=0;
// set if button pressing is detected
static unsigned char counter=0; // counter for ticks
if (ButtonPressed)
// button pressing not yet handled
return;
// nothing to do
if (BUTTON_ON)
// button is in a pressed state
{
if (buttonstate)
// it has already been pressed
{
counter++;
// increment time interval counter
if (counter == BUTTON_ON_TICK) // if enough time has elapsed
detected=1;
// pressing detected
}
buttonstate=1;
// save button state
}
else
// it is in a released state
{
buttonstate=0;
// save button state for the next function call
counter=0;
// reset time interval counter
if (detected)
{
ButtonPressed=1; // notify main program
detected=0;
// reset; end of detection; enable next detection
}
}
}
// in the main function:
…
if (ButtonPressed)
{
DoSomething();
ButtonPressed=0;
}
// button pressing has been detected
// execute a process
// clear flag to enable further detections
4.3.2 Reading a keyboard
A more advanced user input interface is the keyboard. Keys are arranged in columns and
rows. Columns and rows have associated wires, which are connected to each other if a key is
pressed. Figure 4.8 shows how to interface the keyboard to the microcontroller. The wires of
the rows are connected to port pins configured as inputs (Pn.3 to Pn.6), while the columns
are driven by port pins configured as outputs (Pn.0 to Pn.2). Note that the optional pull-up
resistors may be used on the inputs (Pn.3 to Pn.6).
41
Digital input and output; crossbar
Pn.0
Pn.1
MCU
Pn.2
Pn.3
Pn.4
Pn.5
Pn.6
1
2
3
4
5
6
7
8
9
*
0
#
Figure 4.8. Connecting a keyboard to the microcontroller port.
The microcontroller typically scans all keys to determine which key is pressed. To do so, one
of the column wires must be pulled down (by clearing one of the Pn.0, Pn.1 or Pn.2 outputs
while the others are at logic 1) and check which row wire is at logic low. This procedure must
be performed on all columns to determine which keys are pressed. In most cases, the
algorithm can be stopped if a key if found to be pressed and no further keys need to be
checked.
4.3.3 Driving LEDs
LEDs are the simplest indicators that can inform the user about logic values. They can be
connected in negative or positive logic, i.e., they can be lit by writing either logic low or logic
high to the corresponding port bit. Figure 4.9 shows all connections. Open-drain output
mode can only be used if the anode is connected to the supply, while push-pull mode can be
used in both connections.
Vdd
Vdd
Vdd
C8051Fxxx
C8051Fxxx
Vdd
Vdd
R
Rp
R
C8051Fxxx
PORT OUT
R
PORT OUT
PORT OUT
Figure 4.9. An LED can be connected between a port pin and the supply or
ground via a series resistor that sets the current. The push-pull configuration is
needed to drive an LED whose cathode is grounded. Note that the push-pull
mode can be used in both cases.
The current setting resistor should be selected to provide enough light intensity, but keep in
mind that output current of the port is limited and if many LEDs are driven, the total current
sourced or sunk can be too large. Values from 330 Ω to 1 kΩ are typical. External drivers or
transistors can be used to overcome this limitation.
42
Digital input and output; crossbar
4.3.4 Driving 7-segment displays
The 7-segment display contains 7 LEDs to display a decimal digit and one LED to represent
an optional decimal point if multiple displays are used. The anodes or cathodes of the LEDs
are connected to support positive or negative logic. Figure 4.10 shows the common-anode
version associated with the negative-logic mode, which allows the port output to be
configured either in open drain or in push-pull mode.
G
F Vdd A
B
Vdd
A
A
C
D
E
F
G
DP
R
R
R
R
R
R
E
R
G
Pn.0
Pn.1
Pn.2
Pn.3
Pn.4
Pn.5
Pn.6
Pn.7
C
D
E
B
B
R
F
DP
D Vdd C
DP
Figure 4.10. An 8-bit port can drive a 7-segment display.
The following table shows the port bits and the port byte to be written to display a specific
digit.
Digit
G
F
E
D
C
B
A
PORT
0
1
2
3
4
5
6
7
8
9
0
0
1
1
1
1
1
0
1
1
1
0
0
0
1
1
1
0
1
1
1
0
1
0
0
0
1
0
1
0
1
0
1
1
0
1
1
0
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
0
0
1
1
1
1
0
1
1
0
1
1
1
1
1
0x3F
0x06
0x5B
0x4F
0x66
0x6E
0x7E
0x07
0x7F
0x6F
Note that multiple 7-segment displays can be connected to the same port provided that only
one is enabled at a time. For example, the extension board (see the Appendix) has two 7segment displays connected to Port 2 and one bit (P1.3) selects the display whose common
cathode will be connected to Vdd. Therefore, only one display can be active at a time, but if
they are toggled quickly enough, the user will see both displays working with different
numbers. Of course, the brightness will be halved in this case.
4.3.5 Driving alphanumeric LCD displays
A much more powerful popular user interface is the alphanumeric liquid crystal display
(LCD). The microcontroller can communicate with its integrated processor over a special 8-
43
Digital input and output; crossbar
bit parallel interface that can also be configured as a 4-bit interface. A detailed description
can be found in the datasheet of the HD44780 or a compatible processor [16].
GND
VDD
Vo
RS
R/W
E
DB0
DB1
DB2
DB3
DB4
DB5
DB6
DB7
BLA/C
BLC/A
The LCD display can communicate with the processor over its parallel interface. The 8-bit
bidirectional bus can be connected to one port. This port can be configured in open-drain
mode, but in this case external pull-up resistors may be needed (3 kΩ-10 kΩ) to ensure the
short rise time of the signals. Alternatively, the port can be configured in push-pull mode and
should be changed to open drain only for read operations. The 3 control lines are driven by
the port bits of the microcontroller; push-pull mode is strongly recommended. The R/W line
selects between read (R/W=1) and write (R/W=0) operations. RS selects which one of the
two register sets are written or read. If RS is logic high, then the display RAM is accessed and
characters can be written to the display; otherwise instructions to the LCD display can be
sent (for example clearing the display). A pulse on E reads or writes the data. Vdd (5 V in
most cases) and GND are the supply lines. The voltage input V0 is used to set the contrast of
the display. If the LCD display has internal backlighting LEDs, pins 15 and 16 can be used to
power these. The anode and cathode can be connected in both ways; the datasheet must be
consulted to determine the proper connection. Figure 4.11 shows the connector pinout of the
LCD display.
1
16
Figure 4.11. Pinout of the standard LCD display connector.
Instruction
data
comment
B7
B6
B5
B4
B3
B2
B1
B0
Clear
display
0
0
0
0
0
0
0
1
Clears the whole display
Return
home
0
0
0
0
0
0
1
-
Reset cursor and entry mode
Entry
mode set
0
0
0
0
0
1
ID
S
Entry mode:
ID=1:cursor right
S=1:entire display shift
Display
on/off
0
0
0
0
1
D
C
B
D=1 display on,
C=1 cursor on, B=1 cursor blinks
44
Digital input and output; crossbar
Cursor or
display
shift
0
0
0
1
SC
RL
-
-
SC=1:display, 0:cursor
RL=1:right, 0: left shift
Function
set
0
0
1
DL
N
F
-
-
DL=1:8-bit, 0:4-bites mode
N=1:2 sor, 0:1 sor
F=1:5x10, 0:5x8 pixels
Set
CGRAM
address
0
1
Address of writing to the character
generator RAM, defining characters
Set
DDRAM
address
1
0
Address of writing to a specific
location of the display, cursor
positioning
When writing to or reading from the display, certain timing conditions must be met (see
Figure 4.12). The display is a rather slow external peripheral, and the microcontroller code
must be written with this taken into account.
READ TIMING
WRITE TIMING
RS
R/W
RS
Tas
>40ns
Tah
>10ns
R/W
Tpw
>230ns
E
Tas
>40ns
Tpw
>230ns
Tah
>10ns
E
Tdsw
>40ns
DATA
IN
Tddr
>160ns
Th
>10ns
VALID
TcycE >500ns
DATA
OUT
Th
>5ns
VALID
TcycE >500ns
Figure 4.12. Time diagram of write and read operations (for details see the
HD44780 datasheet [16]).
The following example code introduces a few functions to initialise the display and to write to
the display.
#define
#define
#define
#define
LCD_RS
LCD_RW
LCD_E
LCD_PORT
P0_5
P0_6
P0_7
P1
//
//
//
//
RS is the register select input
specifies read or write
enable line serves as write or read pulse
the data bits are connected to port P1
unsigned char line_address[4];
// this array holds the address of the
// first character in a row
/*************************************************************************
LCD initialisation function
Input parameters are the number of rows and columns
**************************************************************************/
void LCD_Init(unsigned char rows, unsigned char columns)
{
unsigned char i;
45
Digital input and output; crossbar
line_address[0]=0;
// initial address of the first row
line_address[1]=0x40; // initial address of the second row
line_address[2]= columns;
// initial address of the third row
line_address[3]=0x40+ columns; // initial address of the fourth row
LCD_RW=0;
// assume write operations as default
LCD_E=0;
// the E line should be inactive
LCD_RS=0;
// register select must be 0 to send commands
Delay_ms(50);
// special initialisation sequence after 50 ms of delay
LCD_DATA=0x30;
// 8-bit mode is selected
LCD_PulseE();
// generate pulse on the E line
Delay_ms(5);
// wait for approximately 5 ms
LCD_PulseE();
// generate pulse on the E line
Delay_ms(1);
// wait for approximately 1 ms
LCD_PulseE();
// generate pulse on the E line
LCD_Write(0x38); // set 8-bit mode, 2 lines
LCD_Write(0x08); // set display off
LCD_Write(0x01); // execute display clear function
LCD_Write(0x06); // set entry mode: increment cursor
LCD_Write(0x0C); // switch display on; no cursor; no blinking selected
}
/*************************************************************************
Pulses the E line to initiate a write to or read from the LCD
**************************************************************************/
void LCD_PulseE(void)
{
__asm
// wait for about 1 µs
mov R7,#7
// system clock frequency in MHz divided by 4
L1: djnz R7,L1
// loop to wait for the specified time
__endasm;
LCD_E=1;
// set E
__asm
// wait for about 1 µs
mov R7,#7
// system clock frequency in MHz divided by 4
L2: djnz R7,L2
// loop to wait for the specified time
__endasm;
LCD_E=0;
// clear E
}
/*************************************************************************
Writes a byte to the LCD
**************************************************************************/
void LCD_Write(unsigned char a)
{
LCD_DATA=a;
// set the data bus according to the value of a
LCD_PulseE();
// generate pulse on the E line
Delay_ms(2);
// wait 2 ms here or, alternatively, check busy flag
}
/*************************************************************************
Clears the entire LCD display
**************************************************************************/
void LCD_Clear(void)
{
LCD_RS=0;
// register select must be 0 to send commands
LCD_Write(1);
// write command to LCD
LCD_RS=1;
// register select default value is 1 (display RAM)
}
46
Digital input and output; crossbar
/*************************************************************************
Moves the cursor to the specified location
**************************************************************************/
void LCD_MoveTo(unsigned char line, unsigned char pos)
{
LCD_RS=0;
// register select must be 0 to send commands
LCD_Write(0x80 | (line_address[line]+pos)); // select position of char
LCD_RS=1;
// register select default value is 1 (display RAM)
}
/*************************************************************************
Redirects the standard C output to the LCD
**************************************************************************/
void putchar(char c) // redefined standard output function
{
LCD_Write(c);
// send the character to the LCD
}
LCD_MoveTo(0,10);
printf("Hello");
// first line, 10th position
// write to the display
4.3.6 Driving relays and motors
Microcontrollers must sometimes control higher power devices such as motors, stepper
motors, valves or high power LEDs. Since the output can source and sink only a few
milliamperes, external drivers are required for heavy loads. A simple solution is to use
bipolar or MOS transistors connected to port pins configured as push-pull, as shown in
Figure 4.13. Inductive loads – coils or motors – can cause very high voltage spikes during
turn-off; therefore, a protection diode is used across the two terminals of such a load.
V+
V+
BC817
500mA
10k
R
IRLML2502
4A
MCU
MCU
R
BC817
500mA
V+
10k
V+
IRLML2502
4A
Figure 4.13. Higher power loads can be driven by external transistors.
4.4
Application guidelines
The crossbar must be enabled to connect the port bits to the pins.
Pins used as digital inputs should be configured as open-drain and logic high must be
written to the corresponding port bits.
After reset, the port latches are set to 1.
Pins used as digital outputs should be configured as push-pull.
47
Digital input and output; crossbar
4.5
The sink current (open-drain or push-pull mode) or source current (push-pull mode
only) must not exceed the datasheet specifications (about 3 mA-4 mA). The total
current of all output pins should be limited to meet the specifications.
An LED can be connected to an output via a series current limiting resistor (330 Ω–
1 kΩ). If using open drain output, the cathode must be connected to the supply
voltage.
Buttons can be connected between the input and GND. An optional pull-up resistor of
about 10 kΩ can be used. Parallel capacitors may help to reduce switching noise.
Troubleshooting
Problem:
Cannot read from or write to the port pin; unexpected output or input values
experienced.
Possible reasons:
4.6
The crossbar is disabled. Enable the crossbar.
Writing 1 to a port bit does not mean that reading it returns 1, since the voltage can be
pulled low by an external circuit.
Push-pull is not set for an output. Open-drain output cannot source current.
Be sure that for input the port pin is configured as open-drain and 1 is written to the
port bit.
External short-circuit on the pin can be present. Check the voltage with a voltmeter.
Improper logic voltage level may be present on the pin.
The output drive current limit may be violated; too high a load may be present on the
pin.
A peripheral is associated with the port pin by the crossbar. In this case the pin state
can only be driven by the peripheral.
Short glitches (temporary pulses) on a signal can corrupt reading.
Exercises
Write a program that reads the state of a button and toggles the state of an LED on
each pressing. Successive button pressings within 200 ms should not be detected.
Continuously pressing a button should be detected as a single pressing.
Connect four LEDs and a button to the microcontroller and write code that
illuminates only one LED and switches to the next LED if the button is pressed. The
first LED should follow the last one.
Write code that displays incremented numbers from 0 to 9 in a cyclic manner upon
each pressing of a button.
Write a program that detects the first button pressing for a longer time and halves
the detection time for additional detections if the button is continuously pressed.
Change the code to reduce detection period if the button is continuously pressed.
Write code to read which key is pressed on a 3 × 4 keyboard matrix. Flash an LED as
many times as the number corresponding to the key pressed.
Write code to display the number of button pressings on an LCD display.
48
Digital input and output; crossbar
Connect a unipolar stepper motor to four pins of the microcontroller. Write a
program that energises only one coil at a time and switches to the next coil ten times
per second.
Modify the program to switch between rotating clockwise and anticlockwise
directions when another button is pressed.
49
Timers and counters
5
Timers and counters
The basis of the timers integrated into microcontrollers is a binary counter. It realises a timer
function if it is driven by an oscillator. It can be used to count events corresponding to falling
or rising edges of an external logic signal.
5.1
Timer 0 and Timer 1
C8051Fxxx processors contain enhanced versions of the standard 16-bit 8051 timers Timer 0
and Timer 1 [6]. Their clock input can be configured as shown in Figure 5.1. SYSCLK is the
system clock; EXT OSC represents an external oscillator. In timer mode, the TCLK signal
driving the timer clock input is derived from these sources, while in counter mode the T0
input is used. The TR0 SFR bit enables the timer and the /INT0 input can be used to gate the
timer depending on the state of the IN0PL and GATE0 SFR bits. The timer/counter values
can be accessed via the SFRs TH0 and TL0 as well as TH1 and TL1, representing the higher
and lower order bytes of Timer 0 and Timer 1, respectively. If the timer overflows, a flag is set
(TF0 for Timer 0 and TF1 for Timer 1) and an interrupt can be generated if enabled.
SYSCLK/12
SYSCLK/4
SYSCLK/48
CLK0
EXT OSC/8
SYSCLK
T0
/INT0
IN0PL
AND
TR0
XOR
OR
GATE0
TCLK0
AND
Figure 5.1. Timer input clock configuring circuit. Labels in bold indicate external
signals and configuration bits are in italic.
Timer 0 and 1 have four modes of operation. In mode 0, they are operated as 13-bit timers;
only the lower 5 bits of the lower order byte of the timers are used (see Figure 5.2).
MODE 0
TCLK0
TL0, 5 LSBs
TH0
TF0
MODE 1
TCLK0
TL0
TH0
TF0
Figure 5.2. Timer 0 in mode 0 (13-bit timer) and in mode 1 (16-bit timer). Each
falling edge on TCLK0 increments the timer.
In mode 2, only the lower 8 bits are used for counting and the higher order byte is used as a
starting value upon overflow. This is called auto reload mode and is useful for generating
programmable periodic events. The block diagram of this mode is illustrated in Figure 5.3.
50
Timers and counters
TL0
TCLK0
TF0
TH0
Figure 5.3. Timer 0 in auto reload mode. Upon overflow, the initial value is
loaded to TL0 from TH0.
The time that passes from the initial state (when the counter value TL0 is equal to TH0) to the
overflow is
T (256 TH 0) tTCLK 0 ,
(5.1)
where tTCLK0 is the period of the input clock of the timer.
The frequency of the periodic overflows can be given by
f f TCLK 0
1
,
256 TH 0
(5.2)
where fTCLK0 is the frequency of the input clock of the timer
If the initial value of the counter (TL0) is less than the reload value, an overflow must occur
before generating the overflows with the desired rate. The mechanism of the auto reload
mode is illustrated in Figure 5.4.
t
TL0
0
1
TF0
0
0
255
252
253
254
255
252
253
254
255
252
253
0
1
1
1
1
1
1
0
0
1
1
CLR TF0
Figure 5.4. Timer 0 operation in auto reload mode.
Mode 3 is rarely used; here, the two 8-bit parts of Timer 0 are used as two 8-bit timers as
shown in Figure 5.5. Timer 1 is inactive in this mode.
51
Timers and counters
SYSCLK/12
SYSCLK/4
SYSCLK/48
CLK0
EXT OSC/8
TH0
TF1
TL0
TF0
SYSCLK
T0
AND
/INT0
TR0
XOR
IN0PL
AND
OR
GATE0
TCLK0
Figure 5.5. In mode 3, two 8-bit timers can be used. Labels in bold indicate
external signals and configuration bits are in italic.
5.2
Timer 2, Timer 3 and Timer 4
C8051Fxxx processors have 2, 3 or 4 additional 16-bit timers with various features [6].
The timers of the C8051F410 can be operated as 16-bit auto reload timers or as dual 8-bit
auto reload timers.
SYSCLK
SYSCLK/12
AND
EXT OSC/8
TCLKn
TMRnL
TMRnH
TMRnRLL
TMRnRLH
TFnH
TRn
Figure 5.6. Timers 2, 3 and 4 in 16-bit auto reload mode. The label in italic is
used for the timer enable bit.
Figure 5.6. shows the block diagram of the auto reload operation, while the operation itself is
illustrated in Figure 5.7.
TMR2
=TMR2H*256
+TMR2L
TMR2RL
=TMR2RLH*256
+TMR2RLL
TMR2
TMR2RL
TMR2RL+1
TF2H ( interrupt)
0
t
0
Nt
65534
65535
TMR2RL
TMR2RL+1
0
0
1
1
Nt
TMR2RL
65534
65535
TMR2RL
TMR2RL+1
0
0
1
1
Hardware sets TF2H
generates interrupt
Software clears
TF2H
Figure 5.7. Timer operation in 16-bit auto reload mode.
52
Timers and counters
Using Timer 2 the frequency of the periodic overflows is given by
f f TCLK 2
1
1
f TCLK 2
;
65536 TMR2RL
65536 256 TMR2RLH TMR2RLL
(5.3)
therefore, if the desired frequency is known, the value of the reload registers can be
calculated:
TMR2RL 65536
f TCLK 2
,
f
TMR2RL
TMR2RLH
,
256
TMR2RLL TMR2RL mod 256
(5.4)
The following example code illustrates the calculation of the reload value. The desired period
is given in s units; the timer input clock must be entered in Hz units.
/*************************************************************************
steps = period/dt
dt=1/timer clock
steps = timer clock*period
reload value = 65536-steps
**************************************************************************/
unsigned long period;
unsigned long tmrclk;
unsigned short tmrrl;
// in µs
// timer clock in Hz
// reload value
tmrclk = 24500000;
// timer clock frequency is 24.5 MHz
period = 100;
// 100 µs
tmrrl = -period*tmrclk/1000000L; // means 65536-period*tmrclk/1000000L
Note that it is also possible to configure the timer as two 8-bit auto reload timers; see
Figure 5.8.
SYSCLK
SYSCLK/12
AND
EXT OSC/8
TMRnH
TFnH
TRn
TMRnRLH
SYSCLK
TMRnL
TFnL
TMRnRLL
Figure 5.8. Timers 2, 3 and 4 in C8051F410 can be configured as two 8-bit auto
reload timers. The label in italic is used for the timer enable bit.
An enhanced timer is available in some C8051Fxxx processors, including, for example, the
C8051F120 100-MHz microcontrollers. As illustrated in Figure 5.9, this timer can count up or
down, and can be clocked by an external signal Tn. The external signal TnEX can be used to
53
Timers and counters
latch the counter value into the reload registers, which allows the accurate detection of the
time instant of an event or a series of events.
SYSCLK/12
SYSCLK/4
SYSCLK
0xFF
0xFF
TMRnL
TMRnH
EXT OSC/8
Tn
AND
TFn
TRn
TnEX
EXENn
D Q
AND
OR
TMRnRLL
TMRnRLH
C Q
Tn
EXFn
Figure 5.9. Timers 2, 3 and 4 in C8051F120 can count up or down, toggle an
output signal and capture the timer value. Labels in bold indicate external
signals and italic is used for configuration bits.
5.3
Timer applications
Timers can be used in many applications. They are employed to generate periodic interrupts,
to provide programmable clock frequency for serial communication ports and also to
generate events at certain time instants and to measure the time elapsed between events.
Some examples are given below.
5.3.1 Delay generation
Timers can be used to generate a desired amount of time delay. The time required to reach
the overflow state from the initial value of Timer 0 (TH0, TL0) can be given by the following
formula:
T 65536 TH 0 256 TL 0 tTCLK 0 ,
(5.5)
where tTCLK0 is the period of the input clock of the timer. The code example below
implements a function that waits for a period equal to stepstSYSCLK.
/*************************************************************************
Waits for a specified number of Timer 0 steps
**************************************************************************/
void Delay(unsigned short steps)
{
TMOD=(TMOD & 0xF0) | 0x01; // 16-bit timer mode
CKCON=CKCON | 0x04;
// Timer 0 clock is SYSCLK
TH0=-steps >> 8;
// 65536-steps, higher-order byte
TL0=-steps;
// 65536-steps, lower-order byte
TF0=0;
// clear timer overflow flag
TR0=1;
// run Timer 0
while (!TF0);
// wait for Timer 0 to overflow
TR0=0;
// stop Timer 0
}
54
Timers and counters
5.3.2 Generating periodic interrupts
Periodic interrupt generation is a common application. Timer auto reload mode is one
option. The following C8051F410 code is the C version of the code given in Chapter 3.2:
/*************************************************************************
Timer 2 interrupt handler routine
**************************************************************************/
void IntHandler(void) __interrupt 5 // define Timer 2 interrupt handler
{
TF2H=0;
// clear interrupt pending flag
LED = !LED; // complement LED; flashing rate is half
// of the Timer 2 overlow rate
}
/*************************************************************************
The main function
**************************************************************************/
void main(void)
{
PCA0MD &= 0x40h; // switch watchdog off
PCA0MD = 0x00h; // switch watchdog off
XBR1
= 0x40h; // enable the crossbar to allow input and output
TMR2RLL = 0xB2h; // set the Timer 2 reload register (low and high bytes)
TMR2RLH = 0xC1h; // to provide 1-Hz interrupt rate
TMR2L
= 0xB2h; // Timer 2 counter initial value
TMR2H
= 0xC1h; // is the same as the reload value
TMR2CN = 0x04; // start Timer 2 now
IE
= 0xA0; // enable global interrupts and Timer 2 interrupt
while (1);
// infinite loop
}
Timer 0 and Timer 1 provide 8-bit auto reload mode; therefore, only higher frequencies can
be generated. In the example below, the program sets the initial value upon overflow. Note
that due to the latency time it is not as accurate as the hardware auto reload mode.
TMOD=(TMOD & 0xF0) | 0x01; // 16-bit timer
TR0=1;
// run timer
IE=0x82;
// enable global & timer0 interrupts
/*************************************************************************
Timer 0 interrupt handler routine
**************************************************************************/
void Timer0Handler(void) __interrupt 1 // define Timer 0 interrupt handler
{
TR0=0;
// stop timer
TH0=-steps >> 8; // initial value is 65536-steps (higher-order byte)
TL0=-steps;
// 65536-steps (lower-order byte)
TR0=1;
// restart timer
…
// perform the required operation
}
5.3.3 Software extended counter
The 16-bit counter can easily be extended by software. For example, if a 24-bit counter is
needed, an 8-bit variable can be added to represent the most significant 8-bits while the
hardware timer provides the least significant 16-bits. Upon overflow of the timer the variable
is incremented as shown in Figure 5.10.
55
Timers and counters
TMR (HW)
TF
65534
65535
0
1
0
0
1
1
65534
65535
0
1
0
0
1
1
COUNTER (SW)
INT+SW
DELAY
0
1
Figure 5.10. 24-bit timer operation emulated by software.
/*************************************************************************
Timer interrupt handler routine
**************************************************************************/
void TimerIRQ(void) __interrupt TIMER_VECTOR // define interrupt handler
{
static unsigned char counter=0; // static variable retains value
counter = (counter+1) % countermax; // increment and implement overflow
if (!counter) Process();
// if counter returns to zero
// (overflows)
}
5.3.4 Pulse width measurement
Timers can be used to measure event timing. Figure 5.11 shows the time diagram of a possible
pulse width measurement. The /INT0 external signal is used to gate the timer: the timer
counts while this signal is high. In order to set up properly, the code waits first for /INT0 to
go low, then enables the timer. After this, the code should wait for the next falling edge of
/INT0, which identifies the end of the pulse. Note that the /INT0 state or a falling edge can
set the IE0 flag, which can be polled or used to generate an interrupt. /INT0 must be enabled
using the crossbar.
T0 CLK
TH0,TL0
0
0
1
K-2 K-1
K
Kt
/INT0
TR0
Counting
waiting while high
waiting for falling edge
Figure 5.11. Time diagram of the pulse width measurement of the /INT0 signal.
56
Timers and counters
TR0=0;
TH0=TL0=0
TMOD=0x09;
IT0=0;
IE0=0;
while (!IE0);
IT0=1;
IE0=0;
TR0=1;
while (!IE0);
TR0=0;
//
//
//
//
//
//
//
//
//
//
//
stop timer
clear timer
T0: 16-bit gated timer mode
set level-triggered /INT0 mode (IE0 is high if /INT0 is low)
clear INT0 flag
wait for input to go down
set edge-triggered /INT0 mode to detect falling edge
clear INT0 flag
enable timer
wait for end of pulse
stop timer
The result of the measurement is in the registers TH0 and TL0.
Note that since part of the detection is done by software, the accuracy can be affected by
accidental interrupts, which halt the main code for a while.
5.3.5 Frequency measurement
The time diagram of a possible frequency measurement algorithm can be seen in Figure 5.12.
The timer is enabled for a given amount of time (for example 1 s) and the counter counts the
external signal falling edges. Therefore, the frequency is the counter value divided by the
running time.
T0
TH0, TL0
TR0
0
1
2
3
4
5
6
Counting
Figure 5.12. Time diagram of the frequency measurement.
In the following example, Timer 1 is used to set TR0 high for a given amount of time, while
Timer 0 counts the pulses.
TCON=0;
// stop Timer 0 and Timer 1
TMOD=0x15;
// Timer 0: 16-bit counter; Timer 1: timer
TH0=0;
// initialise counter of Timer 0 (high byte)
TL0=0;
// initialise counter of Timer 0 (low byte)
TH1=-steps >> 8; // 65536-steps (high byte)
TL1=-steps;
// 65536-steps (low byte)
TF1=0;
// clear Timer 1 overflow flag
TCON=0x50;
// run both timers (both TR0 and TR1 are set)
while (!TF1);
// wait for Timer 1 overflow (Timer 0 counts during this
time)
TCON=0;
// stop both timers
The result of the measurement is in the registers TH0 and TL0.
Note that since part of the detection is done by software, the accuracy can be affected by
accidental interrupts, which halt the main code for a while.
57
Timers and counters
5.3.6 Period measurement
Period measurement means that the time of one or more periods is measured. If the period is
short, it is better to measure the time of multiple periods. One timer can be used to count the
periods; its initial value must be set to 65536 minus the number of periods to be counted. The
other timer is driven by a clock source and runs while the first counter is counting. Therefore,
the time of one period is the clock period of the second timer multiplied by the clock period
and divided by the number of pulses counted.
N periods (L=65536-N)
T0
TH0, TL0
L
L
L+1
65534 65535
0
2
K-3 K-2 K-1
K
TF0
T1 CLK
TH1, TL1
0
0
1
Kt
TR0, TR1
Counting
Figure 5.13. Time diagram of the period measurement.
TCON=0;
TMOD=0x15;
TH1=0;
TL1=0;
TH0=-N>> 8;
TL0=-N;
TF0=0;
TCON=0x50;
while (!TF0);
TCON=0;
//
//
//
//
//
//
//
//
//
//
stop timers
T0: 16-bit counter; T1: timer dTCLK1 period
timer1 initial value (high byte)
timer1 initial value (low byte)
65536-N
N events to TF0=1
clear timer 0 flag
run both timers
wait for N events
stop both timers
The result of the measurement is in the registers TH1 and TL1. The period is:
T
TH1 256 TL1 tTCLK1
N
,
(5.6)
Note that since part of the detection is done by software, the accuracy can be affected by
accidental interrupts, which halt the main code for a while.
5.4
Application guidelines
The timer input clock must be configured first. Choose a frequency value that allows
the desired rate to be accurately set. If the clock frequency is too high, the overflow
rate cannot be set low enough or longer time intervals cannot be measured. If it is too
low, the accuracy of the timing can be low.
58
Timers and counters
5.5
Verify the settings by calculating the timing using the timer SFR values.
Set the desired operating mode.
After proper setup, enable the timer.
Enable the timer interrupt if needed. Do not forget to clear the interrupt pending flag
in the service routine.
The timer interrupt must not be enabled if no service routine is defined. This is the
case when the timer is used to generate a periodic signal for other communication or
other peripherals (UART, PCA, etc.)
Keep in mind that using the timer for multiple purposes simultaneously needs special
attention and is a potential source of problems.
Do not read 16-bit timer values during operation, since the high and low bytes cannot
be read simultaneously and thus they may not correspond to the same timer value.
Troubleshooting
Problem:
The timer is not running or unexpected timing occurs.
Possible reasons:
The timer is not enabled.
The timer is not configured in the proper mode.
The input clock is not configured properly.
Timer 0 and Timer 1 may be in gated mode and the gate signal may be inactive.
The SFR values are miscalculated or not properly written.
Problem:
No timer interrupt occurs or the interrupt rate is not as expected.
Possible reasons:
The timer is not enabled.
The associated interrupt is not enabled.
The interrupt flag is not cleared, so the interrupt is generated continuously. Most of
the processor power is taken in this case.
Execution of other interrupt service routines can cause a delay of the timer interrupt.
The service routine may take longer than the time between two overflows; the
overflow rate is too high.
The timer is used for multiple purposes simultaneously and the settings are different.
Problem:
Unexpected frequency, period or pulse width of an external signal is experienced.
Possible reasons:
The crossbar is not configured properly to connect the external signal to the timer.
The timer settings (like input clock or mode) are improperly set or miscalculated.
The resolution of the time measurement is too low; for example, short periods are
measured by only a few timer increments.
The external signal is noisy; oscillations occur at transitions.
59
Timers and counters
5.6
The software-dependent part of the measurement code is delayed by an unexpected
interrupt.
Exercises
Write a function that waits for a specified number of milliseconds given by the
argument of the function. Use Timer 0 to implement the function.
Write a program that flashes an LED at 1 Hz. Use a button to set the flashing rate:
on each pressing of the button, the rate should be doubled, but at 16 Hz the pressing
of the button should reset the rate to 1 Hz.
Write a program that emulates a pulse width modulated signal. An LED should be
flashed at a rate of 100 Hz and the on time should be set by button pressings from
1 ms to 9 ms in a cyclic manner.
Write a program that generates an output signal of 100 Hz using Timer 2. Measure
the period of this signal using the other timers.
Connect a 555 timer circuit based 100 Hz oscillator output to the microcontroller.
Measure the frequency of this signal using timers.
Measure the pulse width of button pressings using timers.
60
Programmable counter array
6
Programmable counter array
The programmable counter array (PCA) contains a simple 16-bit free-running counter, which
is driven by a periodic clock signal [6, 17]. There are several (from 3 to 6) independent
compare/capture registers, which can be used to latch the counter value upon an event
(change in a digital input signal). These registers can also hold data to be compared with the
counter value and to generate an event when a match occurs. The corresponding flag (CCFn)
is set upon these events, while the CF flag is set when the main counter overflows. All of these
events can generate interrupts; however, the same interrupt routine is called, so the flags
must be checked to identify the source of the interrupt. The structure of the PCA is shown in
Figure 6.1.
SFR BUS
SYSCLK/12
SYSCLK/4
SNAPSHOT
REGISTER
PCA0L READ
PCA0H
CF
TIMER0
EXT CLK IN
PCA0L
SYSCLK
EXT OSC/8
RTC OSC/8
PCA0CPL0 PCA0CPH0
CCF0
PCA0CPL1 PCA0CPH1
CCF1
Figure 6.1. The main counter (PCA0L and PCA0H) can be driven from different
clock sources. There are up to six compare/capture registers (PCA0CPLn and
PCA0CPHn). The label in bold indicates an external signal.
Depending on the operating mode, several useful functions can be implemented.
In the following, only one of the six compare/capture registers is shown, and the names of the
corresponding SFRs are appended with an n that identifies one of the six possible registers.
Note that all compare/capture registers can be associated with an external input/output
signal named CEXn (n = 0, 1, …) via the crossbar. The function of this signal is determined by
the operation mode.
6.1
Edge-triggered capture mode
The edge-triggered capture mode uses an external signal CEXn to latch the value of the
counter into one of the capture registers (PCA0CPLn or PCA0CPHn). This can happen on
rising or falling transitions, or on both. The CCFn flag is set and an interrupt can be
generated, if enabled. Figure 6.2 shows the block diagram of the edge-triggered capture
mode.
61
Programmable counter array
PCA0CPLn PCA0CPHn
CEXn
CCFn
PCA0L
PCA clock
PCA0H
Figure 6.2. Edge-triggered capture mode. The label in bold indicates an external
signal.
CAPTURE
REGISTER=K
CAPTURE
REGISTER=L
PCA CLK
PCA0
0
1
2
K-1
K
K+1
L-1
L
L+1
CEXn
(L-K)t
INTERRUPT
INTERRUPT
Figure 6.3. Time diagram of the edge-triggered capture mode.
6.2
Software timer and high-speed output mode
The PCA0CPLn and PCA0CPLh registers can be compared to the actual value of the main
counter, setting the CCFn flag and generating an interrupt if a match occurs. The software
timer and the high-speed output mode are practically the same, except that in the output
mode the CEXn output is toggled upon each match event. Figure 6.4 shows the block
diagram of these modes. Note that a write to PCA0CPLn disables the comparator, while
writing PCA0CPHn enables it. This ensures that both the low and the high byte of the
capture/compare register are valid when the comparator is enabled. The programmer must
take it into account, so PCA0CPLn must be written first and then should the value of
PCA0CPHn be set. Changing only PCA0CPLn stops the operation.
RESET
Write to
PCA0CPLn
OR
PCA0CPLn PCA0CPHn
0
Write to
PCA0CPHn
PCA clock
1
D Q
C Q
16-bit COMPARATOR
PCA0L
CEXn
CCFn
PCA0H
Figure 6.4. Software timer and high-speed output mode. The label in bold
indicates an external signal.
62
Programmable counter array
Software updates compare register
COMPARE REGISTER=K
COMPARE REGISTER=L
PCA CLK
PCA0
0
1
2
K-1
K
K+1
L-1
L
L+1
CEXn
(L-K)t
Figure 6.5. Time diagram of the software timer and of the high-speed output
mode.
6.3
Frequency output mode
Frequency output mode (Figure 6.6) can be used to output a periodic square wave. Only the 8
least significant bits of the counter are compared to PCA0CPLn and upon a match the output
is toggled and the PCA0CPLn is incremented by the value stored in PCA0CPHn. Of course,
PCA0CPLn will overflow at a certain time but it does not affect the operation.
PCA0CPLn
Enable
8-bit
comparator
PCA clock
PCA0L
8-bit adder
PCA0CPHn
D Q
C Q
CEXn
Figure 6.6. Frequency output mode.
f
f PCA
2 PCA0CPHn
(6.1)
Figure 6.7 shows a sample time diagram when the output frequency is fPCA/6.
63
Programmable counter array
PCA0CPHn=3
PCACLK
PCA0L
PCA0CPLn
255
0
1
2
3
4
5
6
7
8
9
10
0
3
3
3
6
6
6
9
9
9
12
12
CEXn
Figure 6.7. Time diagram of the frequency output mode when the frequency is
1/6 of the PCA input frequency.
6.4
8-bit and 16-bit PWM modes
One of the most useful modes is the generation of pulse width modulated (PWM) signals.
Since a single digital signal can only have two different values, its applications in control are
strictly limited. Using PWM signals, this limitation can be significantly reduced.
A PWM signal is a periodic pulse train whose pulse width can be varied. If the frequency of
this signal is high enough, it can be used as a fine control of slow systems. Typical
applications include motor control, temperature control and light control, where the driven
system cannot follow fast changes and thus only the average of the signal will be effective.
This average is proportional to the duty cycle of the PWM signal.
The PCA module supports 8- and 16-bit PWM modes. In the 8-bit mode, only the 8 least
significant bits of the counter are used. When the value is equal to PCA0CPLn, the output
signal is set, and at the overflow of PCA0L, it will be reset; see Figure 6.8. This way, the
signal is low for PCA0CPLn steps and high for 256-PCA0CPLn steps. This can be changed
by writing a new value to PCA0CPHn, which will take effect only upon the overflow of
PCA0L, ensuring reliable changes. The frequency of the PWM signal is fPCA/256.
PCA0CPHn
PCA0CPLn
Enable
PCA clock
8-bit
comparator
S Q
PCA0L
OVERFLOW
CEXn
R
Figure 6.8. 8-bit PWM mode. The label in bold indicates an external signal.
64
Programmable counter array
PCA CLK
PCA0L
255
0
1
K-1
K
K+1
254 255
0
1
CEXn
Kt
(256-K)t
256t
Figure 6.9. Time diagram of the 8-bit PWM mode.
The 16-bit PWM operation is similar, but here all 16 bits of the counter as well as compare
registers are used; see Figure 6.10 and Figure 6.11.
PCA0CPLn
Enable
PCA0CPHn
S Q
16-bit comparator
CEXn
R
PCA0L
PCA clock
OVERFLOW
PCA0H
Figure 6.10. 16-bit PWM mode. The label in bold indicates an external signal.
1
0
65535
65534
K+1
K
K-1
1
0
PCA0
65535
PCA CLK
CEXn
Kt
(65536-K)t
65536t
Figure 6.11. Time diagram of the 16-bit PWM mode.
6.4.1 PWM DAC
The PWM signal can also be used to generate analogue voltages if the signal is filtered with a
low-pass filter. This way, a digital-to-analogue converter can be emulated. A simple firstorder filtering is shown in Figure 6.12. The ripple of the signal depends on the filter and on
the frequency of the PWN signal. If the ripple allowed at the PWM frequency is given, the
filter corner frequency 1/(2RC) can be determined.
65
Programmable counter array
VIO
R
C
V
VIO
CEXn
Figure 6.12. The PWM signal can be low-pass filtered to approximate a DC
voltage with a low-ripple signal.
A simple estimation can be made assuming a 50% duty cycle, which is the worst case. If V is
small, the capacitor charging current is nearly constant; therefore, V can be approximated
by the following formula
T VIO
T
V 2 2
C
RC 2
I
V
T
;
VIO 4 RC
(6.2)
(6.3)
therefore, choosing
RC
T V IO
4 V
(6.4)
will keep the ripple under the desired limit.
Note that the precision of the output signal is limited by the precision of the VIO supply
voltage. The supply voltage tolerance is not strict; 10% is typical. If higher accuracy is
required, external circuitry should be used.
6.5
Application guidelines
The input clock must be configured first. Choose a frequency value that allows the
desired timing to be accurately set. If the clock frequency is too high, longer timing
can be impossible. If it is too low, the accuracy of the timing can be low.
Verify the settings by calculating the timing using the PCA SFR values.
Configure the compare/capture modules in the required mode.
After proper setup, enable the PCA counter and the modules used.
Enable the PCA interrupt if needed. Do not forget to clear the interrupt pending flag
in the service routine. Note that all modules generate the same interrupt; therefore,
all possible requests must be handled within a single service routine and the
corresponding pending flag must be cleared.
Keep in mind that if the watchdog timer module is used, the PCA input clock cannot
be changed while the watchdog timer is enabled.
The 16-bit PCA counter value can be safely read by reading the lower-order byte
(PCA0L) first.
Always write the lower-order byte of the compare/capture registers (PCA0CPLn)
first. Even if the higher-order byte is not changed, it must be written to re-enable the
PCA comparator, which is disabled by writing to the lower-order byte. If only the
higher-order byte needs updating, writing to the lower-order byte is not required.
66
Programmable counter array
6.6
Troubleshooting
Problem:
The PCA is not running or unexpected timing occurs.
Possible reasons:
The PCA is not enabled.
The PCA modules are not configured in the proper mode.
The input clock is not configured properly.
The input clock source is missing – for example, the external signal is missing or
Timer 0 overflows do not occur.
The SRF values are miscalculated or not properly written.
The higher-order byte of the capture/compare register is not written, so the PCA
comparator can permanently remain in a disabled state.
Problem:
No PCA interrupt occurs or the interrupt rate is not as expected.
Possible reasons:
The PCA or the modules are not enabled.
The associated interrupt is not enabled.
The interrupt flag is not cleared; therefore, the interrupt is generated continuously.
Execution of other interrupt service routines can cause delay of the timer interrupt.
The service routine may take longer than the time between two PCA interrupt
requests; the interrupt request rate is too high.
Problem:
Unexpected frequency, period or pulse width of an external signal is experienced.
Possible reasons:
6.7
The crossbar is not configured properly to connect the external signal to the PCA.
The PCA settings (such as input clock or mode) are improperly set or miscalculated.
The resolution of the time measurement is too low; for example, short periods are
measured by only a few timer increments.
The external signal is noisy, and oscillations occur at transitions.
The PCA interrupt requests are generated faster than the interrupt service routine can
handle them.
Exercises
Write code that can measure the width of a button pressing pulse using the edgetriggered capture mode. Solve the problem both with polling and with interrupt
techniques
Write code that generates a signal that is toggled with the following timing: 1 ms,
2 ms, 4 ms and 8 ms, and repeats this sequence infinitely. Use the high-speed output
mode to implement the code. Check the result on an oscilloscope.
67
Programmable counter array
Write a program that drives an LED with a PWM signal at a rate of 1000 Hz, with
the pulse width set by button pressings from 10% to 90% by steps of 10% in a cyclic
manner.
Generate a 1-kHz PWM signal and pass it through a simple RC filter that reduces the
ripple to 1%. Check the result with an oscilloscope.
68
Serial communication peripherals
7
Serial communication peripherals
Today’s electronic equipment is optimised for small size, low cost and reliability. Board space
must be kept small and the wiring of the printed circuit board must be simple. Integrated
circuits can be smaller if their pin count is small. Reliability is also improved with a lower
number of contacts and a simpler design.
Since microcontrollers often communicate with other components, the above-mentioned
requirements can be supported by serial interfaces that use only a few pins and wires to
connect the devices. Microcontrollers provide several kinds of serial ports where the idea is to
exchange bytes as bit streams, with one bit transferred at a time.
7.1
UART
One of the most popular serial interfaces is the so-called universal asynchronous
receiver/transmitter (UART), developed with the aim of communicating with distant devices,
using circuits that are typically on a separated printed circuit board [6]. Depending on the
distance, a longer cable may be used to connect the devices, in which case a driver/receiver –
aka transceiver – circuit is needed (for example, RS232 and RS485 transceivers). The data
are sent over a single wire in one direction. The communicating devices have a transmit
output (TX) and a receiver input (RX) pin. These must be cross-connected, i.e. the TX pin of
one device should be connected to the RX pin of the other and vice versa. The interface is
symmetrical: any side can send data asynchronously. If the TX pin can be disabled, even a
single wire can be used for bidirectional data transfer. Sometimes one-directional data
transfer is sufficient. Figure 7.1 summarises the connection possibilities.
FULL DUPLEX
C
SIMPLEX
Device
C
Device
TX
TX
TX
TX
RX
RX
RX
RX
HALF DUPLEX
C
SIMPLEX
Device
C
Device
TX
TX
TX
TX
RX
RX
RX
RX
Figure 7.1. Typical UART interconnections.
69
Serial communication peripherals
Vdd
V+
VOLTAGE
CONVERTER
TX
R OUT
VR IN
RS232 SIGNALS
RX
T IN
T OUT
RS232 TRANSCEIVER
MAX202, MAX3232
C
Figure 7.2. RS-232 transceivers allow the use of longer cables between the
communicating devices.
The idle state of the signal is logic high; the transmission is started by setting the signal low
for a given amount of time t. After sending this ‘start bit’, the data bits will be sent and each
bit will be placed on the wire for time t. The transfer is terminated by a so-called stop bit,
which is logic high for a duration of at least t. The transmitter and receiver must have the
same timing; they detect the start bit and then sample the signal to determine the value of the
bits. Sometimes a ninth bit is sent, which can be a parity bit or can be used in multiprocessor
communication to mark the byte as a control or address word rather than data. Figure 7.3
shows the time diagram of the data transfer.
t
t
t
t
t
t
t
t
t
t
START
BIT
B0
B1
B2
B3
B4
B5
B6
B7
STOP
BIT
bit sampling
Figure 7.3. Time diagram of the data transfer.
The 1/t bit rate (called the baud rate) is generated by timer overflows in two different ways
depending on which processor is used:
baud rate = timer overflow rate / 16, (for example, C8051F120 UART 0)
baud rate = Timer 1 overflow rate / 2 (for example, C8051F410 or C8051F120
UART 1).
or
Using Timer 1 in auto reload mode, the timer reload value can be determined in the following
way:
TH1=256-TCLK0/(16baud rate) (for example, C8051F120 UART 0),
70
Serial communication peripherals
or
TH1=256-TCLK0/(2baud rate) (for example, C8051F410 or C8051F120 UART1).
If one of Timers 2–4 is used, the 16-bit reload value is
TMRRL=65536- TCLK/(16baud rate).
The timers must be configured in auto reload mode and must be enabled. The associated
timer interrupts are not enabled.
Note that the transmitter and receiver baud rate cannot be exactly the same since they are
derived from different oscillators. Figure 7.4 shows a time diagram example of a 3%
mismatch.
START
BIT
B0
B1
B2
B3
B4
B5
B6
B7
STOP
BIT
1,03t
Figure 7.4. Time diagram of data transfer with 3% baud rate tolerance. The red
bits are not sampled properly.
Depending on the signal transition time, the allowed tolerance is different. A higher baud rate
or a longer transition time needs more strict matching [18]. It is strongly recommended to
configure outputs as push-pull to keep transition times as short as possible.
t/8
t
t
BIT
BIT
0,75t
0,25t
0,75t
BIT
BIT
Figure 7.5. Depending on the rise and fall time of the signals, the valid state can
be longer, which allows a less strict tolerance of the baud rate.
The following C8051F410
communication.
code
examples
illustrate
simple
polling-mode
UART
/*************************************************************************
UART initialisation function
**************************************************************************/
void UART_Init()
{
SCON0 = 0x10; // 8-bit, variable baud mode
TI=1;
// assume empty output buffer!
}
/*************************************************************************
UART input function, polling mode
71
Serial communication peripherals
**************************************************************************/
unsigned char UARTIn(void)
{
while (!RI); // wait for a byte
RI=0;
// clear UART receive flag
return SBUF; // return the byte
}
/*************************************************************************
UART output function, polling mode
**************************************************************************/
void UARTOut(char a)
{
while (!TI); // wait for end of previous transmission
TI=0;
// clear UART transmit flag
SBUF=a;
// transmit a byte, do not wait for end
// this will also trigger the transmission process
}
The next example shows the use of ring buffers to transmit and receive in interrupt mode.
#define BUFFERSIZE 8
// declare ring buffers for input and output queue
volatile unsigned char TxBuffer[BUFFERSIZE];
volatile unsigned char RxBuffer[BUFFERSIZE];
// TX buffer read and RX buffer write pointers
// used in the interrupt routine
volatile unsigned char TxReadPtr=0, RxWritePtr=0;
// TX buffer write and RX buffer read pointers
unsigned char TxWritePtr=0, RxReadPtr=0;
// Number of data available in the ring buffers
volatile unsigned char TxNumberOfData=0;
volatile unsigned char RxNumberOfData=0;
/*************************************************************************
UART interrupt routine
**************************************************************************/
void UARTInterrupt(void) __interrupt UART_VECTOR
{
if (RI)
// if byte has been received
{
RI=0; // clear UART receive flag
if (RxNumberOfData < BUFFERSIZE) // does it fit in the buffer
{
RxBuffer[RxWritePtr]=SBUF;
// save the byte into the buffer
RxWritePtr = (RxWritePtr+1) % BUFFERSIZE; // ring buffer indexing
RxNumberOfData++; // increment number of received bytes
}
}
if (TI)
// if byte has been transmitted
{
TI=0; // clear UART transmit flag
if (TxNumberOfData) // if there are still bytes to be sent
{
// this will also trigger the transmission process
SBUF=TxBuffer[TxReadPtr]; // send the byte
TxReadPtr = (TxReadPtr+1) % BUFFERSIZE; // ring buffer indexing
72
Serial communication peripherals
TxNumberOfData--; // decrement the number of bytes in the queue
}
}
}
/*************************************************************************
UART input function, interrupt mode
**************************************************************************/
unsigned char UARTIn(unsigned char *c)
{
if (RxNumberOfData) // if bytes are available
{
RxNumberOfData--; // decrement the number of available bytes
*c=RxBuffer[RxReadPtr]; // read a byte from the buffer and return it
RxReadPtr = (RxReadPtr+1) % BUFFERSIZE; // ring buffer indexing
return 0; // return 0 if successful
}
return 1; // no byte could be read from the buffer
}
/*************************************************************************
UART output function, interrupt mode
**************************************************************************/
unsigned char UARTOut(unsigned char c)
{
if (TxNumberOfData < BUFFERSIZE) // is there space in the transmit queue
{
TxNumberOfData++; // increment number of bytes in the transmit queue
TxBuffer[TxWritePtr]=c; // put the byte in the transmit queue
TxWritePtr = (TxWritePtr+1) % BUFFERSIZE; // ring buffer indexing
return 0; // return 0 if successful
}
return 1; // no byte could be placed into the transmit queue
}
7.1.1
Application guidelines
UART must be enabled on the crossbar and the TX output must be configured as
push-pull.
The baud rate should be set by configuring the associated timer overflow rate. The
maximum baud rate is SYSCLK/16; however, for transmission, the baud rate can be
SYSCLK/2 if the baud rate is equal to the timer overflow rate divided by 2.
The timer must be enabled but the timer interrupt must not.
The UART reception must be enabled.
Either polling or interrupt mode can be used but the two modes should not be used
simultaneously.
If interrupt mode is used, the UART interrupt must be enabled. The interrupt pending
flag must be cleared in the service routine. Note that both transmit and receive
interrupts invoke the same service routine, so both events should be handled.
The UART has a single-byte input buffer; therefore, the input data will be overwritten
by the next incoming byte of data if the buffer has not been read by the processor in
time.
73
Serial communication peripherals
In order to avoid data loss, some kind of handshaking can be implemented. For
example, a received byte can be sent back to confirm reception.
7.1.2 Troubleshooting
Problem:
The UART does not seem to send or receive data.
Possible reasons:
The UART is not enabled on the crossbar or the crossbar is not enabled.
The UART reception is not enabled.
The UART baud rate timer is not enabled.
The baud rate time is not configured properly.
Broken or short-circuited wires or links between the communicating devices.
Problem:
The data sent or received do not seem to be valid.
Possible reasons:
The baud rates of the communicating devices do not match (due to improper settings
or the limited accuracy of the internal oscillator, etc.).
The baud rate is too high (higher than SYSCLK/16).
The baud rate timer is used for another purpose and has been overwritten
accidentally.
The TX output signal is not configured as push-pull.
Problem:
Some bytes are missing during data transfer.
Possible reasons:
7.2
The receive buffer is not read in time by the processor before new data are received.
The data transfer is too fast.
SPI
Serial peripheral interface (SPI) is normally used to connect integrated circuits – sensors,
ADCs, DACs, other microcontrollers, etc. – on the same board in a master-slave fashion [6].
SPI uses one wire for outgoing data (master out – slave in, MOSI) and another for incoming
data (master in – slave out, MISO). A third wire (serial clock, SCK) driven by the master
synchronises the transfer by providing a clock signal that changes when a bit of the data is
available. Typically, the data are shifted out on one edge and can be read on the following
opposite transition; the polarity can be chosen. An optional active low fourth signal (often
called negated slave select, NSS) can be used to select the slave device, which ignores all
communication signals if this line is inactive. This is useful to provide a safe frame
(accidental pulses on the SCK line can corrupt data transmission) or to use multiple slave
devices on the same bus.
74
Serial communication peripherals
MISO
MISO
MISO
MISO
MOSI
MOSI
MOSI
MOSI
SCK
SCK
SCK
SCK
NSS
NSS
GPIO
Master
Slave
Master
Slave
MISO
MOSI
SCK
NSS
Slave
Figure 7.6. SPI master and slave connections.
The time diagram of a 3-wire transaction can be seen in Figure 7.7, while Figure 7.8
illustrates the use of the NSS signal. Note that there is no separate read operation; during a
read the MOSI wire is driven. The slave device typically ignores this byte, but the datasheet
must be consulted to ensure proper operation.
The SPI clock rate can be expressed as:
fSCK
SYSCLK
,
2 SPI0CKR 1
(7.1)
where SPI0CKR is an SFR that can be given by the following formula:
SPI0CKR
SYSCLK
1.
2 fSCK
(7.2)
75
Serial communication peripherals
SCK
POL=1
POL=0
MOSI
B7
B6
B5
B4
B3
B2
B1
B0
PHA=0
MISO
B7
MOSI
B7
B6
B6
B5
B5
B4
B4
B3
B3
B2
B2
B1
B1
B0
B0
PHA=1
MISO
B7
B6
B5
B4
B3
B2
B1
B0
Figure 7.7. Time diagram of an SPI transaction. The clock polarity and phase
can be programmed.
SCK
POL=1
POL=0
NSS
MOSI
B7
B6
B5
B4
B3
B2
B1
B0
PHA=0
MISO
B7
B6
B5
B4
B3
B2
B1
B0
Figure 7.8. Time diagram of a 4-wire SPI transaction.
The following is a polling-mode SPI example code.
/*************************************************************************
SPI output function, polling mode
**************************************************************************/
void SPIOut(unsigned char c)
{
SELECT = 0;
// activate the select signal (negative logic)
SPIF
= 0;
// clear the end of transmission flag
SPI0DAT = c;
// place the byte into the transmit register
// this will also trigger the transmission process
while (!SPIF); // wait until the end of transmission
SELECT = 1;
// deactivate the select signal (negative logic)
}
/*************************************************************************
SPI input function, polling mode
76
Serial communication peripherals
**************************************************************************/
unsigned char SPIIn(void)
{
SELECT = 0;
// activate the select signal (negative logic)
SPIF
= 0;
// clear the end of transmission flag
SPI0DAT = 0;
// dummy write starts SPI clocking and therefore reception
while (!SPIF); // wait until the end of reception (8 bits)
SELECT = 1;
// deactivate the select signal (negative logic)
return SPI0DAT; // return the received byte
}
7.2.1 Application guidelines
The SPI must be enabled on the crossbar and the outputs (MOSI and SCK in master
mode, MISO in slave mode, and the select signal if applicable) must be configured as
push-pull.
The SCK clock rate should be set by setting the dedicated system clock divider value
SPI0CKR. The maximum clock rate in master mode is the SYSCLK/2 or 12.5 MHz,
whichever is lower; in slave mode, it must be less than SYSCLK/10.
All parameters of the SPI port – clock phase, polarity, 3- or 4-wire mode and master
or slave mode – must be set according to the communication requirements.
Use either polling or interrupt mode, but not the two modes simultaneously.
If interrupt mode is used, the SPI interrupt must be enabled. The interrupt pending
flag must be cleared in the service routine. Note that several SPI events are handled
by the same service routine, so care should be taken to handle interrupts from all
possible sources in the routine.
The SPI has a single-byte input buffer; therefore, the input data will be overwritten in
slave mode by the next byte of incoming data if the buffer has not been read by the
processor in time.
7.2.2 Troubleshooting
Problem:
The SPI does not seem to send or receive data.
Possible reasons:
The SPI is not enabled on the crossbar or the crossbar is not enabled.
The SPI is not configured properly.
The select signal is used but inactive.
Broken or short circuited wires or links between the communicating devices.
Problem:
The data received or sent do not seem to be valid.
Possible reasons:
The setup (clock phase, polarity, etc.) of the communicating devices does not match
The clock rate is too high (higher than 12.5 MHZ or SYSCLK/10 in slave mode).
The output signals are not configured as push-pull.
The interface is not initialised properly; an accidental transaction has not finished.
77
Serial communication peripherals
7.3
SMBus
System management bus (SMBus) is a two-wire master-slave interface to connect multiple
masters (microcontrollers) and multiple slaves (digital output sensors, ADCs, DACs, flash
memory, etc.) on the same bus [6]. It is practically compatible with the I2C (Inter-integrated
circuit) bus; a few minor differences include timeout handling, clock speed and line driving
specifications. The communicating chips are typically found on the same printed circuit
board or at least in the same equipment. The bus is not designed to use long wires (more than
a few tens of centimetres).
Vdd=5V
Vdd=3V
Vdd=5V
Vdd=3V
Master 1
Master 2
Slave 1
Slave 2
R
Vdd=5V
R
One wire carries the data in both directions (serial data, SDA) and another (serial clock, SCL)
is supplied by a master and synchronises the communication devices by clock pulses
indicating valid bits on the bus.
SCL
SDA
Figure 7.9. Connecting devices to the two-wire SMBus.
The SMBus has open-drain output drivers and needs external pull-up resistors for proper
operation. Higher value resistors (about 10 kΩ or greater) lower the power consumption,
while smaller resistances (down to about 2 kΩ) provide higher speed. The open-drain
structure only allows both the master and the slave to set the signal logic low (pull-down); the
resistors ensure logic high when none of the devices pull the signal down.
FILTER
SCL
CROSSBAR
CLOCK CONTROL
SHIFT REGISTER
7 6 5 4 3 2 1 0
FILTER
SDA
SDA CONTROL
ACK
Figure 7.10. Block diagram of the SMBus peripheral of the microcontroller.
78
Serial communication peripherals
The bit rate can be set by Timer 0 and Timer 1 overflows and Timer 2 high or low byte
overflows according to the following formula:
fSCL
Timer Overflow Rate
.
3
(7.3)
R/W
7-bit address
and direction bit
D7
D0
8-bit data
STOP
A0
NACK
START
A6
ACK
SDA
SCL
The time diagram of a typical transaction can be seen in Figure 7.11. In the inactive state both
SDA and SCL are high. Data exchange is initiated by a start condition, which is the master
pulling the SDA low. Then the master sends 7 address bits to select a device and a direction
bit. This bit is logic high if the master reads from the slave and is logic low otherwise. After
this, the slave must pull the SDA line low to acknowledge (ACK) the request; otherwise, the
transaction will fail. Depending on the direction, the master or the slave then puts data on the
SDA line, while the master controls the timing by driving the SCL line. Each byte must be
acknowledged. After the last byte has been sent, either the master or the slave can send a notacknowledge (NACK, release the SDA wire) to stop the data transfer. The transfer is ended by
a stop condition: the master keeps SDA low while releasing the SCL to go high then releases
the SDA line to let the external resistor to pull the line high.
Figure 7.11. Time diagram of an SMBus transaction.
The communication is more complicated than in the case of UART or SPI. Two examples in
Figure 7.12 and 7.13 show read and write time diagrams and interrupt flag behaviour. It is
recommended to handle the transfer in an interrupt routine; however, polling the interrupt
bit (SI) can be easier to implement and understand in simple transfers.
79
Serial communication peripherals
MASTER
IRQ
IRQ
IRQ
IRQ
S
ADDR
W A
DATA
A
DATA
A P
S
ADDR
W A
DATA
A
DATA
A P
IRQ
IRQ
IRQ
SLAVE
Figure 7.12. SMBus write operation.
MASTER
IRQ
IRQ
IRQ
IRQ
S
ADDR
R A
DATA
A
DATA
N P
S
ADDR
R A
DATA
A
DATA
N P
IRQ
IRQ
IRQ
SLAVE
Figure 7.13. SMBus read operation.
A simple polling mode example code can be seen below.
/*************************************************************************
SMBus/I2C output function, polling mode
**************************************************************************/
void SMBusOut(unsigned char address, unsigned char c)
{
STO = 0;
// stop condition bit must be zero
STA = 1;
// start transfer
SI = 0;
// continue
while (!SI);
// wait for start to complete
STA = 0;
// manually clear STA
SMB0DAT = address << 1; // A6..A0 + write
80
Serial communication peripherals
SI = 0;
while (!SI);
if (!ACK)
{
STO = 1;
SI = 0;
return;
}
SMB0DAT = c;
SI = 0;
while (!SI);
STO = 1;
SI = 0;
// continue
// wait for completion
// if not acknowledged, stop
// stop condition bit
// generate stop condition
//
//
//
//
//
put data into shift register
continue
wait for completion
stop condition bit
generate stop condition
}
/*************************************************************************
SMBus/I2C input function, polling mode
**************************************************************************/
unsigned char SMBusIn(unsigned char address)
{
STO = 0;
// stop condition bit must be zero
STA = 1;
// start transfer
while (!SI);
// wait for start to complete
STA = 0;
// manually clear STA
SMB0DAT = (address << 1) | 1; // A6..A0 + read
SI = 0;
// continue
while (!SI);
// wait for completion
if (!ACK)
// if not acknowledged, stop
{
STO = 1;
// stop condition bit
SI = 0;
// generate stop condition
return;
}
ACK = 0;
// NACK, last byte
SI = 0;
// continue
while (!SI);
// wait for completion
STO = 1;
// stop condition bit
SI = 0;
// generate stop condition
return SMB0DAT;
}
7.3.1 Application guidelines
The SMBus must be enabled on the crossbar and the associated two pins must be
configured as open-drain.
The SMBus data rate should be set by configuring the associated timer overflow rate.
The maximum clock rate is SYSCLK/20. In most cases, rates close to the standard
100 kbit/s are used
Master or slave mode can be selected.
The associated timer must be enabled but the timer interrupt must not.
Use either polling or interrupt mode, but not the two modes simultaneously.
If interrupt mode is used, the SMBus interrupt must be enabled. The interrupt
pending flag must be cleared in the service routine. Note that several SMBus events
are handled by the same service routine, so care should be taken to handle interrupts
from all possible sources in the routine..
81
Serial communication peripherals
External pull-up resistors must be used. The typical range is from 2 kΩ to 10 kΩ.
Lower values enable higher speed; higher values lower the power consumption.
7.3.2 Troubleshooting
Problem:
The SMBus does not send and receive data or the data sent or received do not seem to
be valid.
Possible reasons:
7.4
The SMBus is not enabled on the crossbar or the crossbar is not enabled.
The SMBus is not configured properly.
The SMBus data rate timer is not enabled.
The data rate is not configured properly.
The device address is invalid or no device is present on the bus.
Broken or short-circuited wires or links between the communicating devices.
The device communication protocol is not followed properly.
The data rate (clock frequency) is too high and the clock pulse width is too narrow.
The devices are too far from each other, and if they are connected with a cable, it may
be too long.
C standard I/O redirection
The standard C input/output functions in the SDCC environment use the putchar and
getchar basic functions [4]. Therefore, it is possible to redirect the input and output to use a
serial port. The printf, scanf and other functions will use the port. A code example of using
the UART port can be seen below.
/*************************************************************************
Redirected standard C output function
**************************************************************************/
void putchar(char c)
{
UARTOut(c);
// UART, SPI, SMBus, etc.
}
/*************************************************************************
Redirected standard C input function
**************************************************************************/
char getchar(void)
{
return UARTIn(c); // UART, SPI, SMBus, etc.
}
…
printf("x=%d",x);
// example write operation
82
Serial communication peripherals
7.5
Exercises
Write code that sends back every received byte over the UART. The baud rate should
be 9600 bit/s; use a personal computer connected to the UART via a USB-UART
converter to the microcontroller board.
Upgrade the UART example code to use the RTS and CTS handshake lines.
Connect the MCP4141 digital potentiometer via the SPI port to the microcontroller.
Write code to set the potentiometer value. Check the result with a digital multimeter.
Connect the SST25VF020B 2-Mbit flash memory via the SPI port to the
microcontroller. Write code to fill the first 256 locations with the location index and
read back the data. Measure the SPI signals with an oscilloscope.
Read the temperature value using an LM75 sensor via the SMBus interface. Display
the value on the two 7-segment displays in degrees. Measure the SMBus signals with
an oscilloscope.
83
Analogue peripherals
8
Analogue peripherals
Modern microcontrollers have several built-in analogue peripherals – comparators, ADCs,
DACs and voltage references – to support very compact real-world signal processing. Some
sensors can be connected even without additional analogue circuitry.
In this section, the analogue peripherals of the C8051F410 microcontroller will be detailed
[6].
8.1
Comparators
Comparators are the simplest digitising components. They have a single logic output that is
logic high if the voltage on the positive input is greater than the voltage connected to the
negative input. Some hysteresis turns the comparator into a Schmitt trigger, making it less
sensitive to the noise on slowly changing signals, which can cause oscillations on the output
when the signal is close to the switching threshold.
Vp
Vp
Vp
Vn
Vp
Vn
Vn
Vn+h/2
Vn-h/2
t
OUT
t
OUT
Figure 8.1. Comparator and Schmitt trigger (comparator with hysteresis)
waveforms.
Note that comparators have some delay that varies from device to device and may also
depend on the settings.
Tdh
Tdl
Vn+h/2
Vn-h/2
t
OUT
Figure 8.2. Comparator delay.
Comparators in C8051F410 processors have analogue multiplexers at their inputs so the
signals can be chosen during program execution. The corresponding port pins must be
configured as analogue input using the crossbar. The output of the comparators can be
polled, used to generate an interrupt or connected to the pins of the microcontroller via the
crossbar. In the latter case, the output can be either synchronised to the internal system clock
or left as is (analogue mode).
84
Analogue peripherals
P0.0
CP0A
P2.4
P2.6
D Q
CP0
CROSSBAR
P0.2
C
P0.1
P0.3
OR
INTERRUPT
P2.5
P2.7
Figure 8.3. C8051F410 comparator circuit.
Comparators are useful for detecting levels of signals in applications such as heating control.
They can be applied to convert a sinusoidal signal to a logic signal in order to measure its
period; this way, the output of voltage-to-frequency converters can be digitised. They can also
make the inputs compatible with different logic levels.
The comparator is useful for implementing a certain kind of analogue-to-digital converter.
Figure 8.4 shows how it can be used to digitise the time constant RC of a resistor-capacitor
circuit. When the push-pull output port bit is switched to logic high, the capacitor will be
charged through R and the positive input is fixed at a fraction of the driving voltage. The time
t needed to charge the capacitor to reach this voltage can be measured by a timer. This time
can be obtained from the following derivation:
R1
t
1 exp
,
RC R1 R2
(8.1)
R1
t
exp
,
R1 R2
RC
(8.2)
R1 R2
t
exp
,
R2
RC
(8.3)
R R
t RC ln 1 2 .
R2
(8.4)
1
The logic high voltage level is not accurate, but the time does not depend on it, so the
accuracy is not degraded. There are several resistive and capacitive sensors whose signals can
be digitised using this method. Note that the input leakage current and the input capacitance
can affect accuracy if the charging current is low or if the capacitor value C is low.
85
Analogue peripherals
C
R2
PROCESSING
R1
R
PORT BIT
C8051F410
Figure 8.4. The comparator can be used to measure the time constant of the RC
circuit, thus R or C can be measured if one of them is known.
8.1.1 Application guidelines
The comparator input pins must be associated with port pins using the input
multiplexer.
The comparator input pins must be configured as analogue and must be skipped
using the crossbar.
The comparator input pins must not be left floating.
The source impedance of the two input signals must be low and balanced to avoid an
undesired voltage drop due to the input leakage current, and to reduce noise pickup.
The comparator raw analogue output or the output synchronised to the system clock
can be connected to a port pin using the crossbar.
The comparator response time and power should be selected.
The hysteresis should be configured. Noisier or slower signals typically need higher
hysteresis.
The input voltage range must be within the specifications; typically between ground
and supply.
The comparator interrupt can be generated on falling edge, on rising edge or on both
transitions. If used, it must be enabled and the interrupt pending flag must be cleared
in the interrupt service routine.
8.1.2 Troubleshooting
Problem:
The comparator does not seem to detect the polarity of the voltage between its input
terminals properly.
Possible reasons:
The comparator is not enabled.
The input multiplexer is not configured to assign the desired signals.
The signals are out of the valid input range.
The signals never cross each other; for example, one of the inputs is at GND while the
other is always positive.
The input signals are too fast; the polarity of the voltage between the input terminals
is changing too quickly.
An input is floating
86
Analogue peripherals
The source impedances of the two input signals are highly different; therefore, the
input leakage current causes an undesired voltage drop.
Problem:
The comparator interrupt is not generated.
Possible reasons:
The comparator is not configured properly,
The interrupt is not enabled or it is not configured properly.
Problem:
The comparator output changes randomly or irregularly.
Possible reasons:
8.2
The comparator is not configured properly.
One or both of the inputs are floating or the source impedance is too high.
The input signals are noisy or very slowly changing; the hysteresis is too small.
Voltage reference
Most of the analogue peripherals need a stable, clean and accurate voltage. Comparators may
need accurate voltage levels to which they can relate their input signals. Analogue-to-digital
converters compare the input voltage to a reference voltage and determine the ratio of these
voltages. Digital-to-analogue converters output a current or voltage that is an integer
multiple of a small quantity. Although some use the supply voltage as a reference, it is
absolutely not recommended, since the value is not stable enough: it depends on loading, it
can be rather noisy and its accuracy is very poor. One must always use a precise dedicated
reference voltage. Note that the accuracy of the reference determines the accuracy of all
circuits that use it.
Mixed-signal microcontrollers have integrated voltage references but can use external
references as well. Typically, the internal reference is not as accurate; some applications may
need better performance. Typical parameters are the following.
Absolute accuracy. Guaranteed minimum and maximum voltage limits. Sometimes the
error is given as percentage of the nominal voltage. Internal references have accuracies of 1%2%, while external references can have accuracies below 0.1%.
The C8051F410 value is 1.8%.
Temperature coefficient. The reference voltage is somewhat temperature dependent. The
value is given as ppm/K: parts-per-million change in the nominal voltage per degree. The
value is typically around 30 ppm/K; for external references it can be below 3 ppm/K.
The C8051F410 value is 35 ppm/K.
Load regulation. The reference voltage depends slightly on the loading current. It can be
given as voltage change per loading current or as ppm per loading current. A few ppm/A is
typical.
The C8051F410 value is 10 ppm/A.
87
Analogue peripherals
Power supply rejection. Changes in the value of supply voltage can cause small changes in
the reference voltage. The smaller the change, the better the rejection. The ratio of the voltage
reference change and the supply voltage change is given (mV/V or ppm/V).
The C8051F410 value is 2 mV/V.
Maximum loading current. Voltage references can drive external resistor dividers and
other circuitry. However, the loading current cannot exceed a certain value; otherwise, the
specifications are not guaranteed.
The C8051F410 value is 200 A.
Turn-on time. The voltage reference can be turned on and off, which allows the use of an
external reference and can help to optimise power management. Since the output voltage of
the reference must be decoupled with capacitors (100 nF and 4.7 F-10 F), it takes a
considerable amount of time – typically a few milliseconds – for the voltage to reach the final
accurate value. It is a typical mistake to enable the reference and do an analogue-to-digital
conversion without having waited for the reference settling time to pass. The error can be
very large in this case.
The C8051F410 value is about 7 ms using 100 nF and 4.7 F capacitors.
Figure 8.5 shows typical reference connections.
OPTIONAL
EXTERNAL
VREF
4u7
100nF
ADC, DAC
INTERNAL
VREF
C8051Fxxx
Figure 8.5. Both external and internal reference can be used. The decoupling
capacitors must be placed as close as possible to the reference pin.
8.2.1 Application guidelines
Decide whether to use internal or external reference. In both cases use decoupling
capacitors.
If internal reference is used, it must be enabled and the internal reference buffer
should also be enabled.
Do not overload the reference. It is a good practice to apply only loads well under the
maximum (below 10%). Use an external operational amplifier buffer to provide higher
current.
Consider the reference turn-on time. Use it only after the value is surely stabilised.
The reference port pin must be configured as analogue and must be skipped by the
crossbar.
8.2.2 Troubleshooting
The troubleshooting for references is given in the following sections.
88
Analogue peripherals
8.3
ADC
The analogue-to-digital converter outputs a b-bit binary number d that depends on the input
voltage as
V
d 2b 0.5 ,
Vref
(8.5)
where Vref is the voltage of the internal or external voltage reference [12]. The smallest voltage
change that can be detected, Vref/2b, is called the voltage of the least significant bit (VLSB) or
voltage resolution. There are also differential ADCs, which measure voltage difference
between their inputs and output a 2s complement binary number. Negative voltages are not
allowed, because they are out of the range of the supply voltage — microcontrollers do not
have a negative supply. This means that external signal conditioning is needed if the signal is
out of range, if it is too small or if it is not a low-impedance voltage.
Figure 8.6 shows the block diagram of the 12-bit ADC integrated into the C8051F410
microcontroller. This successive approximation register (SAR) ADC is a very common
architecture. Note that the conversion takes several steps (the number of bits plus 1), so a
start signal and a periodic clock (SAR clock) signal are needed for proper operation.
B11
B10
B9
B8
B7
B6
B5
B4
B3
B2
B1
B0
B11
B10
B9
B8
B7
B6
B5
B4
B3
B2
B1
B0
The analogue multiplexer allows 27 different signals to be digitised. If the analogue signal
(voltage) to be digitised is connected to a port pin, this pin must be configured as an analogue
input using the crossbar and it must also be ‘skipped’, i.e. the crossbar must not assign any
other peripherals to this pin. Note that the voltage reference pin (P1.2) must also be
configured as an analogue input and skipped regardless of whether external or internal
voltage reference is used.
P0.1
P2.6
P2.7
TEMP
VDD
ANALOGUE MULTIPLEXER
P0.0
INTERRUPT
(AD0INT)
12-bit A/D
converter
START
Vref
CONVERSION
ADC0LTH
ADC0H
Left justified
Right justified
ADC0L
ACCUMULATOR
1,4,8,16 samples
WINDOW
COMPARATOR
ADC0LTL
INTERRUPT
ADC0GTH ADC0GTL
Figure 8.6. The A/D converter circuit of the C8051F410.
The 12-bit result of the conversion can immediately go to the 16 bits of SFRs ADC0H and
ADC0L in left- or right-justified format. It is also possible to accumulate 1, 4, 8 or 16 samples
and transfer their sum to the ADC0H and the ADC0L. In this case, the data must be rightjustified, since the sum can take all 16 bits. A window comparator can set a flag or generate
an interrupt depending on whether the result is in a specified range.
89
Analogue peripherals
The conversion can be started in several ways, as shown in Figure 8.7. At the end of
conversion the AD0INT flag is set and an interrupt can be generated.
WRITE 1 TO AD0BUSY
TIMER 3 OVERFLOW
START CONVERSION
CNVSTR (P0.6)
TIMER 2 OVERFLOW
Figure 8.7. Start of conversion sources.
It is important to note that before conversion, the input signal is sampled by a capacitor Cs, as
shown in Figure 8.8.
TRACK
Rext
CONVERT
RMUX
IL
12-bit A/D
converter
Cs
C8051F410
Figure 8.8. Simplified schematic of the ADC input.
The capacitor must be charged to a voltage that is close enough to the input voltage. If the
deviation is less than half of the voltage resolution ½ VLSB = Vref/213, the error thus
introduced does not degrade the 12-bit resolution. Since the capacitor is charged through Rext
and the internal resistance RMUX=5 kΩ, the sampling time must be at least
Rext RMUX Cs ln213 Rext 5 k 12pF ln213 Rext 110 ns 540 ns ,
k
(8.6)
so at least a 540-ns signal tracking time is required. However, the datasheet specifies 1000 ns
as minimum due to the uncertainty of the values 5 kΩ and 12 pF. In order to ensure reliable
operation, it is best to keep a minimum of 1000 ns tracking time and add 200 ns for every
kΩ of output impedance of the signal source:
ttrack
Rext
200 ns 1000 ns ,
1 k
(8.7)
Even if a DC signal is measured, this minimum tracking (or sampling) time must be
guaranteed, because the sampling capacitor is discharged during conversion.
Note that the analogue input has a leakage current. It flows through Rext, therefore causes an
error voltage. For higher impedances, an operational amplifier is recommended to provide
low-impedance output.
A more common connection can be seen in Figure 8.9, when an external capacitor is placed
between the input and ground. This capacitor can remove high-frequency noise, charge the
sampling capacitor quickly and isolate the signal source from the current transients caused
90
Analogue peripherals
by the switched sampling capacitor. The latter is especially useful when the signal source is
the output of an operational amplifier. Cext is typically chosen to be much greater than the
sampling capacitance. For example, if it is 1000 times greater, it can charge the sampling
capacitor to 99.9% even without drawing current from the signal source. However, between
conversions the external capacitor must be recharged to represent the input voltage with a
specified accuracy.
TRACK
Rext
CONVERT
RMUX
Cext
IL
12-bit A/D
converter
Cs
C8051F410
Figure 8.9. Signal source connected to the input via a series resistor (Rext) and a
capacitor (Cext).
In order to make it clearer, an example follows. Let us assume that the voltage of the source is
Vin and the sampling frequency is fs. This means that in every conversion cycle the sampling
capacitor drains a charge of VinCs, so the average current flowing to the input is the charge
divided by the sampling period ts (which is equal to 1/fs):
I
Vin Cs
Vin Cs f s ,
ts
(8.8)
This current flows through Rext, therefore causes an average voltage drop on it:
V Rext Vin Cs f s;
(8.9)
therefore, the relative error can be estimated as
V
Rext Cs f s .
Vin
(8.10)
Thus at a given sample rate and relative error the Rext resistance is limited as
Rext
V 1
,
Vin Cs f s
(8.11)
or for a given Rext value the sample rate must be limited as
fs
V
1
.
Vin Cs R ext
(8.12)
For example, if 0.1% error is allowed, then at a sample rate of 10 kHz Rext must be less than
0.001/(1210-12104) Ω 8333 Ω.
In summary:
when no external capacitor is used, the minimum tracking time is given by Equation
8.7;
using an external capacitor much (about 1000 times) greater than the sampling
capacitor, the tracking time can be kept at its minimum, but the sample rate is limited
according to Equation 8.12.
91
Analogue peripherals
According to the above the external resistor and capacitor can only be used as an simple antialiasing filter if Equations 8.11 and 8.12 are satisfied. On the other hand, single pole filters do
not reduce higher frequency components properly. If the signal contains significant
components above fs/2, then an active anti-aliasing filter is preferred. A popular simple
second-order low-pass filter [11] is shown in Figure 8.10.
Rext
R1
R2
C1
Cext
C2
Figure 8.10. Sallen-Key second-order low-pass filter. Note that Rext and Cext are
not parts of the filter.
The transfer function of this filter can be given as:
A( )
1
,
1 i R1 R2 C2 2 R1R2C1C2
(8.13)
where is the angular frequency: =2f.
The general formula of a second-order low pass filter can be written as:
A( )
1
.
i 2
1
Q0 02
(8.14)
The values of Q and 0 can be obtained from tables or by using filter design software, while
the values of the resistors and capacitors can be determined.
Higher-order filters with better high-frequency rejection can be realised by cascading several
first- or second-order stages. Note that Rext and Cext are not parts of the filter: these
components are needed to isolate the output of the operational amplifier from the transient
load caused by the switched sampling capacitor.
The C8051F410 microcontroller offers several tracking options that are illustrated in Figure
8.11.
92
Analogue peripherals
CONVERT START
PRE
TRACKING TRACK
POST
TRACKING
IDLE
CONVERT
TRACK
CONVERT
DUAL
TRACKING TRACK TRACK
CONVERT
TRACK
IDLE
CONVERT
TRACK
TRACK
CONVERT
TRACK TRACK
CONVERT
13 ADC CLOCKS
Figure 8.11. Time diagram of the different tracking modes.
The safest mode is the dual tracking mode. The post-tracking mode can be used to save
power, since the ADC is in an idle state between conversions. The pre-tracking mode can help
to achieve the highest possible conversion rate, but one must be very careful, because a
minimum tracking time is not guaranteed. Therefore, the use of this mode is not
recommended.
The following simple example code illustrates ADC handling in polling mode.
P0MDIN
P1MDIN
P0SKIP
P1SKIP
REF0CN
ADC0CF
ADC0CN
=
=
=
=
=
=
=
0xFE;
0xFB;
0x01;
0x04;
0x13;
0x00;
0x80;
//
//
//
//
//
//
//
P0.0 analogue input
P1.2 analogue input (VREF)
skip P0.0 since it is an analogue input
skip P1.2 since it is an analogue input
enable internal VREF
191406-Hz ADC clock
enable ADC (conversion start: set AD0BUSY)
unsigned int GetADC(unsigned char channel)
{
ADC0MX = channel; // set the multiplexer
ADC0CN = 0x80;
// enable the ADC
AD0INT=0;
// clear the end of conversion flag
AD0BUSY=1;
// start A/D conversion
while (!AD0INT); // wait for end of conversion
AD0INT=0;
// clear the end of conversion flag
return (ADC0H << 8)+ADC0L; // return the result of the A/D conversion
}
A more efficient way is to read the converted data in an interrupt service routine. One
possible implementation can be seen below.
TMR2RLL
TMR2RLH
TMR2CN
P0MDIN
P1MDIN
P0SKIP
P1SKIP
REF0CN
ADC0CF
ADC0CN
EIE1
=
=
=
=
=
=
=
=
=
=
=
0x60;
0xFF;
0x04;
0xFE;
0xFB;
0x01;
0x04;
0x13;
0x00;
0x83;
0x08;
//
//
//
//
//
//
//
//
//
//
//
high byte of reload register for a 100-Hz overflow rate
low byte of reload register for a 100-Hz overflow rate
enable Timer 2
P0.0 analogue input
P1.2 analogue input (VREF)
skip P0.0 (input signal)
skip P1.2 (VREF)
enable internal VREF
191406-Hz ADC clock
enable ADC (conversion: TIMER 2)
enable ADC interrupt
93
Analogue peripherals
IE
= 0x80;
// enable interrupts
/*************************************************************************
ADC interrupt handler routine
**************************************************************************/
void ADC_interrupt(void) __interrupt ADC_VECTOR
{
AD0INT = 0; // clear the end of conversion flag
adc_data = (ADC0H << 8) | ADC0L; // save the result of the A/D conversion
}
8.3.1 Application guidelines
The ADC should be enabled for proper operation.
Select the event that starts the conversion: it can be the write signal to AD0BUSY SFR
bit, the overflow of Timer 2 or Timer 3 or a rising edge of an external signal (CNVSTR,
P0.6). If CNVSTR is used, the port pin must be skipped and configured as open-drain
and 1 must be written to the corresponding port bit.
If a timer overflow is used to start a conversion periodically, the timer must be
configured properly, and the timer interrupt should not be enabled.
Select the input signal by setting the multiplexer to the desired port pin. The pin must
be configured as analogue and must be skipped using the crossbar. If multiple signals
must be converted, then all associated pins must be configured as analogue and must
be skipped.
Select the desired ADC SAR conversion clock. Choose the highest frequency available
but not higher than the specified maximum (3 MHz for the C8051F410). The full
conversion takes 13 cycles plus the tracking time.
If the system clock frequency is low or low-power operation is needed, the use of burst
mode is recommended. In this mode, the ADC is operated from a high-speed clock
independent of the system clock and is only out of idle state during conversion.
Select the proper tracking mode and post-tracking time. Dual tracking mode is
preferred in most cases, since it guarantees a minimum tracking time. Post-tracking
mode can be used when low-power operation is needed.
Consider the output impedance Rext of the signal, the external capacitance Cext, the
internal resistance of the multiplexer and the value of the sampling capacitor to
estimate the minimum tracking time and maximum sample rate using Equations 8.7
and 8.11–8.12. The input leakage current flows through Rext, so it also causes an error.
An operational amplifier can be used if the impedance is high.
The voltage reference pin (P1.2) must be configured as analogue and must be skipped
using the crossbar. If the internal voltage reference is used, it must be enabled and the
internal reference buffer must be enabled. The internal bias generator must also be
enabled.
If the conversion is started by writing to the AD0BUSY bit, polling the AD0INT bit
can be used to wait for the end of conversion. The AD0INT bit must be cleared before
starting the conversion. In multichannel applications, the next channel must be
selected just after the end of conversion.
If the conversion is started by a timer overflow or by an external signal (CNVSTR),
then the end of conversion event should be handled by the ADC interrupt service
routine. In the routine, the AD0INT flag must be cleared and in multichannel
94
Analogue peripherals
applications, the next channel must be selected at the beginning of the routine to
allow the longest possible settling time.
The 12-bit ADC can be left- or right-justified. Multiple (4, 8 or 16) samples can be
accumulated and then their sum can be read. In this mode, the data must be rightjustified, because addition would cause an overflow otherwise. If 16 samples are
accumulated, the result will be a 16-bit word. Note that averaging may reduce the
noise in certain cases but does not improve accuracy. Taking 16 samples can reduce
the noise to one fourth.
8.3.2 Troubleshooting
Problem:
No A/D conversions can be detected.
Possible reasons:
The ADC is not enabled.
Interrupt mode is planned but the ADC interrupt is not enabled or the global
interrupt flag is disabled.
The start of conversion signal is missing or not configured properly. If timers are
used, they might not be enabled. The external start of conversion signal pulse can be
too narrow.
Problem:
The conversion result is not valid.
Possible reasons:
The port pin is not configured as an analogue input.
Due to improper multiplexer settings, the signal is not connected to the ADC input.
The voltage reference or the internal bias generator is not enabled.
Internal reference is used, but the internal reference buffer is not enabled.
The voltage reference is enabled just before starting a conversion. Note that the
voltage reference stabilisation time can be several milliseconds, which must be
allowed to pass before starting a conversion.
The voltage reference is overloaded, so it does not provide the proper value.
Polling mode is used and the data are read before the end of conversion. The AD0INT
flag might not be logic low before starting a conversion.
Improper integer data handling occurred. For example, left-justified or accumulated
data must be stored in an unsigned short.
The ADC SAR clock frequency is too high (>3 MHz) or too low.
The tracking time is too short. The signal output impedance might be too high, which
necessitates a longer tracking time; see Equation 8.7.
The signal output impedance is high, so the input leakage current causes significant
error.
The signal is out of the measurement range (0–Vref).
95
Analogue peripherals
8.4
DAC
The rich set of analogue peripherals of the C8051F410 includes two independent 12-bit
current output digital-to-analogue converters. The output current range (full-scale output
current, Imax) can be set as 0.25 mA, 0.5 mA, 1 mA or 2 mA.
Current output DACs are typically faster, but need external circuitry if voltage output is
needed. Even a simple resistor of value R suffices (see Figure 8.12), but the specified
compliance range must be met, and the output voltage must be below Vdd-1.2 V. Note also
that in such a configuration, the output impedance is R. For example, 1 V output range at
1 mA full-scale current requires R=1 kΩ.
12-bit D/A
converter
R
Vout
C8051F410
Figure 8.12. A resistor converts current to voltage for the C8051F410 current
output DACs. The output voltage is Vout=RImaxN/212.
The digital-to-analogue conversion can be initiated by simply writing to the DACs SRFs
(IDA0L first then IDA0H), but for the most accurate timing – needed for example for
waveform generation –, timer overflow or the transitions of an external signal can be used, as
shown in Figure 8.13.
IDA0H
TIMER 0 OVERFLOW
TIMER 1 OVERFLOW
LATCH
IDA0L
WRITE TO IDA0H
12-bit D/A
converter
TIMER 2 OVERFLOW
TIMER 3 OVERFLOW
CNVSTR
P0.6
OR
Figure 8.13. Sources that can control the DAC output update.
A simple example of using DAC 0 with output update upon writing to IDA0H:
P0MDIN = 0xFE; //
P0SKIP = 0x01; //
IDA0CN = 0xF2; //
//
P0.0 analogue mode
skip P0.0
enable DAC0, update by write to IDA0H
1 mA full scale, left-justified data
96
Analogue peripherals
IDA0L = 0;
// low byte of the DAC output register
// writing to the high byte of the DAC output register updates the DAC output:
IDA0H = 128;
// high byte of the DAC output register, half scale: 0.5mA
The following code updates the DAC 0 output at a rate of 100 Hz and generates a ramp
signal:
unsigned
P0MDIN =
P0SKIP =
IDA0CN =
short dac_data = 0;
0xFE;
// P0.0 analogue mode
0x01;
// skip P0.0
0xA2;
// enable DAC0, update: Timer 2 overflow
// 1 mA full scale, left justified data
TMR2RLL = 0x60; // high byte of the reload register for 100-Hz overflow rate
TMR2RLH = 0xFF; // low byte of the reload register for 100-Hz overflow rate
TMR2CN = 0x04; // enable Timer 2
IE
= 0xA0;
// enable Timer 2 a global interrupts
/*************************************************************************
Timer 2 interrupt handler routine
**************************************************************************/
void Timer2_interrupt(void) interrupt TIMER2_VECTOR
{
TF2H=0;
// clear interrupt pending flag
IDA0L=dac_data;
// set lower order byte
IDA0H=dac_data >> 8; // set higher order byte
dac_data++;
// increment the value for linear ramp
}
8.4.1 Application guidelines
The port pin associated with the DAC output should be set as analogue and must be
skipped using the crossbar.
Select the DAC output current range.
Select the DAC output update source: write to the DAC register, timer overflow, or
external signal (CNVSTR, P0.6). If CNVSTR is used, then it must be configured as
open-drain and must be skipped. If the update source is timer overflow or an external
signal, an associated interrupt service routine should set the next DAC value.
The DAC update can be precisely synchronised to ADC conversions. In this case, the
ADC interrupt routine must update the DAC registers and the ADC start of conversion
and the DAC update source must be the same (timer overflow or CNVSTR).
Enable the DAC and the internal bias generator.
If voltage output is needed, connect a resistor between the output and ground. The
output compliance range (Vdd–1.2 V) should not be exceeded, so at full-scale current
the voltage on the resistor must be within this range.
Select left or right justification. If only the 8 most significant bits are used, leftjustified mode should be selected, since only the higher-order byte is used in this case.
If used, the lower-order byte (IDA0L or IDA1L) must be written first. The DAC output
is updated if the higher order byte is written (IDA0H or IDA1H), unless the update
method is configured differently.
8.4.2 Troubleshooting
Problem:
97
Analogue peripherals
The DAC output is unchanged or invalid when writing to the DAC registers
Possible reasons:
8.5
The DAC or the bias generator is not enabled.
Only the lower-order byte (IDA0L) is written.
The output update source is a timer, but it is not configured properly or it is not
enabled.
If an interrupt routine must update the DAC output, the interrupt might not be
enabled.
The output is not configured as analogue or is not skipped. The port pin is used by
another peripheral or port latch.
The output compliance range is violated.
The output update rate is too high, whereas the DAC output needs a certain settling
time.
Temperature sensor
The C8051F410 microcontroller has an internal diode-based temperature sensor, whose
output voltage can be measured by the ADC. The voltage depends almost linearly on the chip
temperature:
V a T b,
(8.15)
where the value of a is 2.950 mV/⁰C0.073 mV/⁰C and b=900 mV17 mV.
The linearity error – the maximum deviation from the ideal linear dependence – is 0.2 ⁰C;
the overall accuracy is about 3 ⁰C.
Note that depending on the operating power the chip temperature can be significantly higher
than the ambient temperature. In a low-power application (below 5 mW), the chip
temperature is close to the temperature of the printed circuit board (approximately within
1 ⁰C–2 ⁰C).
The temperature sensor can be used to monitor the operating temperature of the chip and the
containing printed circuit board. In some very low power applications it can be used to
estimate the ambient temperature with an accuracy of about 3 ⁰C. Temperature changes can
be monitored more accurately.
8.6
Exercises
Measure the resistance of a resistor in the range of 1 kΩ to 99 kΩ using the
comparator as shown in Figure 8.4. Use C=100 nF, R1=27 kΩ and R2=47 kΩ. Display
the result in kΩ units on two 7-segment displays. Measure 1% precision resistors
(1 kΩ, 3.3 kΩ, 10 kΩ and 33 kΩ) and compare the results with the nominal values.
Measure the capacitance of a capacitor in the range of 1 nF to 99 nF using the
comparator as shown in Figure 8.4. Use R=100 kΩ, R1=27 kΩ and R2=47 kΩ.
Display the result in nF units on two 7-segment displays. Measure 1 nF, 3.3 nF, 10 nF
and 33 nF capacitors and compare the results with the nominal values. Also consider
the precision of the nominal values.
Try to measure the internal voltage reference turn-on time using the ADC.
98
Analogue peripherals
Read the position of the potentiometer with the ADC and display the value converted
into 0 to 99 on two 7-segment displays.
Write a program that generates a sinusoidal waveform using the DAC. One period
should contain 64 points; the output data rate defined by a timer overflow should be
100 kHz; use a 24.5-MHz system clock. Check the result with an oscilloscope.
Implement a voltage-to-frequency converter by measuring the voltage at the output
of the potentiometer and generate a logic signal whose frequency is a linear function
of this voltage. The frequency range should be 1 kHz to 10 kHz. Use the PCA
frequency output mode.
Write code that can delay a signal by 100 µs to 10000 µs in 100 µs units. Convert the
signal at a rate of 10 kHz and output the delayed signal on DAC0. Use a timer to
synchronise the ADC sample rate and DAC update rate. The delay value should be
set by the potentiometer. Check the result on an oscilloscope.
Measure the on-chip temperature using the internal temperature sensor. Measure
the on-chip temperature as a function of the system clock frequency.
Find a method to estimate how much higher the on-chip temperature is than the
temperature of the printed circuit board.
99
Sensor interfacing
9
Sensor interfacing
There are many different sensors that can be interfaced to mixed-signal microcontrollers. In
most cases, some external analogue signal conditioning circuitry is needed [11]. Since the
microcontroller is a single-supply device, level shifting is used to handle bipolar signals.
External active signal conditioning is typically based on single-supply operational amplifiers
that may need additional attention.
In the following the most important solutions are presented briefly.
9.1
Voltage output sensors
If the voltage to be measured is in the range of the ADC (0–Vref), then it can be connected
directly to one of the ADC inputs. If the voltage is unipolar but can exceed Vref, then a simple
resistive voltage divider can be used to reduce the voltage to match the input range. Figure
9.1 shows the above-mentioned connections.
12-bit A/D
converter
R
C
4u7
C
V
R2
V
12-bit A/D
converter
R1
INTERNAL
VREF
100nF
4u7
100nF
C8051F410
INTERNAL
VREF
C8051F410
Figure 9.1. Unipolar voltage output sensors can be connected directly or via a
voltage divider to the ADC input. On the left, the ADC input voltage VADC is equal
to V, while on the right it is R2/(R1+R2)V.
R1
R2
V
12-bit A/D
converter
R
C
Max 200uA
R4
R3
4u7
100nF
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.2. Small or large bipolar voltages can be measured using an
operational amplifier.
If the voltage is small or bipolar, then an operational amplifier can be used to convert the
signal int0 0–Vref. A general-purpose inverting circuit is shown in Figure 9.2. The output
voltage of this circuit is fed to the ADC and is equal to
100
Sensor interfacing
VADC
Vref R4
R3 R4
R2 R2
1
V
R
R1
1
(9.1)
One can see that this formula allows small and large voltages, since the signal amplification is
–R2/R1, so it can be less or greater than 1. At zero input signal, VADC should be close to Vref/2
for the optimal usage of the input range.
The instrumentation amplifier circuit containing three high-accuracy operational amplifiers
is very useful for handling small differential sensor signals where high input impedance (no
loading of the signal) is required [15]. The gain can be set by a single resistor Rg, and it has a
level-shifting input called reference or ground. Note that the supply range of the amplifier
limits the input signal range as well. Figure 9.3 shows the simplified schematic of the
instrumentation amplifier.
V1
Va
R
R
Rg
Rf
Vout
Rf
R
V2
R
V0
Vb
Figure 9.3. Instrumentation amplifier (IA) circuit. Integrated IAs contain the
parts drawn within the rectangle. Vout=G(V2-V1)+V0, where G=1+2Rf/Rg.
The instrumentation amplifier is also available as a single integrated circuit including the low
voltage AD623 amplifier, which is ideally suited to single-supply microcontroller
applications. Figure 9.4 shows a typical input signal conditioning circuit using an
instrumentation amplifier. Note that the operational amplifier is needed to ensure a lowimpedance drive to define the middle output voltage (in our example, Vref/2) of the
instrumentation amplifier.
101
Sensor interfacing
Rg
V
IA
R3
C
Vo
12-bit A/D
converter
Vref/2
Max 200uA
Rd
Rd
4u7
100nF
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.4. Small voltage differences can be measured by applying an
instrumentation amplifier. The ADC input voltage is VADC=V+Vref/2.
9.2
Current output sensors
Current-to-voltage conversion can be done by even a single resistor (Figure 9.5) if the current
is not too high (which would cause high power dissipation) or not too low (too high
impedance because of the high-value resistor). The resistor R must be chosen to get a voltage
equal to Vref when the maximum current flows.
A
R
C
12-bit A/D
converter
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.5. Current-to-voltage conversion using a resistor.
If the current is low, as in the case of a photodiode, a low input current operational amplifier
based current-to-voltage converter circuit should be used; see Figure 9.6. The feedback
resistor value determines the output voltage, IRf.
102
Sensor interfacing
C
Rf
R
C
4u7
100nF
12-bit A/D
converter
INTERNAL
VREF
C8051F410
Figure 9.6. Photodiode current-to-voltage conversion using an operational
amplifier.
Bipolar currents can be handled by simply shifting the zero-current output voltage to Vref/2,
as shown in Figure 9.7.
Rf
A
R
C
12-bit A/D
converter
Max 200uA
Rd
Rd
4u7
100nF
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.7. Bipolar current-to-voltage conversion. Here the ADC input voltage
VADC is equal to IRf+Vref/2. At zero current the voltage is equal to Vref/2.
9.3
Resistive sensors
Resistive sensors, such as thermistors and photoresistors, can output a voltage if they form a
resistive voltage divider with a resistor of known value (Figure 9.8). The input of the divider
is the reference voltage Vref. This circuit works in a ratiometric operation, since the ADC uses
the same reference voltage as the voltage divider, so the result of the conversion does not
depend on Vref.
103
Sensor interfacing
12-bit A/D
converter
C
Max 200uA
Rs
R
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.8. A voltage divider allows the measurement of Rs. VADC seen by the
ADC is equal to Rs/(R+Rs)Vref.
Potentiometric sensors can also be connected in a very similar manner, as shown in
Figure 9.9.
12-bit A/D
converter
C
αRs
(1-α)Rs
Max 200uA
Rs
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.9. Potentiometric sensors can be used as voltage dividers of Vref. The
ADC input voltage is VADC= Vref.
If the Vref loading were violated because of too small resistor values, the Vref voltage can be
buffered by an operational amplifier; see Figure 9.10.
12-bit A/D
converter
R
C
Rs
Max 200uA
1k
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.10. An operational amplifier buffer removes reference loading. VADC is
equal to Rs/(R+Rs)Vref.
104
Sensor interfacing
Pressure sensors, load cells and force sensors are typically based on a resistor bridge. The
bridge can be driven by a voltage and a small differential voltage between two terminals must
be measured by a high input impedance stage. The instrumentation amplifier is the ideal
choice in this case, because the gain can be set by a single resistor and the output can be
level-shifted by connecting a voltage – typically Vref/2 – to its reference input. Figure 9.11
shows a possible solution.
Rg
Vref/2
IA
R3
C
Vref/2
12-bit A/D
converter
Max 200uA
Rd
Rd
4u7
100nF
4u7
100nF
INTERNAL
VREF
C8051F410
Figure 9.11. Bridge sensors can be connected to the analogue input in
ratiometric configuration using an instrumentation amplifier. The ADC input
voltage is VADC=(GR/R+1)Vref/2.
9.3.1 Application guidelines
Always consider the following in voltage measurement:
o Voltage range. Unipolar or bipolar signal handling may be required.
o Output impedance of the source. If it is too high, then the tracking time can be
too short. Inverting amplifiers have quite a low input impedance.
9.3.2 Troubleshooting
Problem:
Cannot communicate with the real-time clock peripheral.
Possible reasons:
9.4
The interface is not opened properly. Only a reset can end the blocked state and
restore normal operation.
Exercises
Design a circuit that can convert the voltage range of -10 V–10 V to 0 V–2.5 V. Check
the transfer function using a circuit simulator.
Design a circuit to measure the supply voltage.
105
Sensor interfacing
Design a circuit to measure the supply current.
Connect a thermistor and a 10-kΩ resistor as a voltage divider of Vref to the ADC
input. Convert the input continuously and display the temperature on the two 7segment displays. The temperature in Celsius as a function of thermistor resistance
is given by the formula T=1/(1/298 °C+ln(R/R25)/B)-298 °C, where R25 is the value
of the thermistor resistance at 25 ⁰C (nominally 10 kΩ) and B can be found in the
datasheet of the thermistor (it is typically 4000 °C).
Replace the thermistor with a photoresistor and display the resistance in kΩ on the
two 7-segment displays.
106
Real-time clock
10 Real-time clock
The real-time clock circuit is a counter that is driven by an oscillator based on a tuning fork
crystal with a frequency of 32768 Hz (Figure 10.1). Crystals typically have an absolute
accuracy of about 10 ppm-20 ppm. For example, a 20-ppm accuracy means an error of about
1 minute in a month.
The real-time clock (called smaRTClock [6]) of the C8051F410 microcontroller uses a 47-bit
binary counter that is mapped to 6 bytes (RTC0–RTC5). The least significant bit of this sixbyte value is not used. This means that the four most significant bytes (RTC2–RTC5) can be
considered as a 32-bit counter driven by a 1-Hz source, so it is incremented in every second.
Typically the value corresponds to seconds elapsed from 0:00 January 1, 1970 and can be
converted to date format using the standard C library functions declared in time.h.
ALARM5
ALARM4
ALARM3
ALARM2
ALARM1
ALARM0
ALARM
IRQ
47-bit COMPARATOR
44444444333333333322222222221111111111000000000X
76543210987654321098765432109876543210987654321
RTC5
RTC4
RTC3
RTC2
RTC1
RTC0
CAPTURE5
CAPTURE4
CAPTURE3
CAPTURE2
CAPTURE1
CAPTURE0
32768Hz
OSC
Figure 10.1. Structure of the real-time clock.
CAPTURE0–CAPTURE5 registers are used to read and write the 47-bit counter. If it is to be
written first, the CAPTURE registers must be loaded, then a write operation must be
performed. A read transaction copies the current value to the CAPTURE registers for bytewise reading. The ALARM registers can define a certain value that is compared to the counter
and an interrupt can be generated if the values match. This can be used to wake the processor
up at a certain time.
Note that if no crystal is present, the counter can also be driven by an internal oscillator,
which has a selectable frequency of 20 kHz or 40 kHz. Due to its low accuracy, it cannot be
used to measure real time; however, it can be useful in very low power applications when
selected as the system clock.
Write or read operations of the smaRTClock registers can only be performed if the interface
is opened by sending a special keyword to enable these operations. Three SFRs are available
for operations: RTC0KEY, RTC0ADR and RTC0DAT. All internal registers can be accessed
through the RTC0ADR address and the RTC0DAT data registers (see the datasheet for
details).
107
Real-time clock
A simple example code of basic communication functions is listed below.
/*************************************************************************
Unlock the smaRTClock interface
**************************************************************************/
void OpenRTC()
{
RTC0KEY = 0xA5; // first this value must be written to the key register
RTC0KEY = 0xF1; // next this value must be written to the key register
}
/*************************************************************************
Write to the smaRTClock registers
**************************************************************************/
void WriteRTC(unsigned char Address, unsigned char Data)
{
while (RTC0ADR & 0x80); // wait while the smaRTClock is busy
RTC0ADR = Address;
// set the target address
RTC0DAT = Data;
// write the data into the register
}
/*************************************************************************
Read from the smaRTClock registers
**************************************************************************/
unsigned char ReadRTC(unsigned char Address)
{
while (RTC0ADR & 0x80); // wait while the smaRTClock is busy
RTC0ADR = Address;
// set the target address
RTC0ADR|=0x80;
// define a read operation
while (RTC0ADR & 0x80); // wait for the data
return RTC0DAT;
// return the data
}
The following code initialises and starts the smaRTClock in crystal oscillator mode:
OpenRTC();
WriteRTC(0x07,0xE0);
WriteRTC(0x06,RTC0CN_DEF);
for (i = 0; i < 3000; i++);
while ((ReadRTC(0x07) & 0x10)==0);
WriteRTC(0x07,0xC0);
//
//
//
//
//
crystal mode, auto gain, double bias
power on the oscillator
wait a bit for stabilisation
wait for oscillator OK
crystal mode, auto gain, single bias
Note that at the beginning, the bias current of the oscillator is doubled for enhanced
reliability and robustness. After stabilisation, the bias current can be reduced to normal to
save power but can also be left doubled if power consumption is not a concern.
10.1.1 Application guidelines
Unlock the interface of the smaRTClock peripheral by writing 0xA5 and 0xF1, in this
order, to RTC0KEY. If other codes are written or invalid read or write operations are
initiated, the interface will be disabled until a system reset occurs.
Select crystal oscillator mode for accurate real-time operation (an external 32768-Hz
crystal must be used).
Enable the smaRTClock crystal oscillator and wait for stabilisation by polling the
corresponding bit.
Write the initial value to the counter.
Enable the timer by setting the timer run control bit.
108
Real-time clock
If an alarm is needed, write the corresponding counter value to the alarm registers
and enable the alarm events. Provide the proper interrupt service code and enable the
real time clock interrupt.
The actual value of the counter can be set or read at any time.
Conversion between the counter value and real time (date, hour, minute, second) can
be done by the time and mktime functions defined in the standard C library (time.h).
10.1.2 Troubleshooting
Problem:
Cannot communicate with the real-time clock peripheral.
Possible reasons:
The interface is not opened properly. Only a reset can end the blocked state and
restore normal operation.
Problem:
The alarm interrupt is not generated.
Possible reasons:
The crystal oscillator is not running. This can be checked by reading the valid
oscillation bit.
Counting is not enabled.
The counter value or the alarm value is invalid.
The alarm events are not enabled or the alarm interrupt is not enabled.
10.2 Exercises
Using the alarm function of the smaRTClock, write code that generates an interrupt
in every second. Display the seconds on the 7-segment display.
Write code that reads the real-time clock value and converts it to date and time
using the standard C library functions.
109
Watchdog and power supply monitor
11 Watchdog and power supply monitor
In every real-world– commercial, industrial, automotive, etc. – application reliable operation
is probably the most important. Microcontrollers are widely used in such embedded
applications, and since they have rather complex structure, contain digital and analogue
hardware components, run software when powered on, they can be sensitive to both
hardware and software problems. Electromagnetic interference, spikes on the supply line,
software hang-ups due to core errors, unexpected values on peripherals, infinite loops and
unhandled exceptions are all potential sources that can permanently break normal operation.
Since it is impossible to prevent these from occurring, some methods have been developed to
safely return to normal operation.
11.1 The watchdog timer
One solution to avoid permanent hang-ups is the use of the so-called watchdog timer, which
can detect hang-ups and reset the processor to restart normal operation. Of course, the code
must be developed keeping this possibility in mind. The watchdog peripheral has an internal
timer, which measures time and can reset the processor if a timeout is occurred. The timer is
always restarted if the software writes to its dedicated register, so timeout will not happen if
the software notifies the watchdog timer periodically within the timeout period. If any hangup happens due to hardware or software failure, the processor will be reset within the defined
timeout value.
The C8051F410 processor uses the last PCA channel to implement the watchdog timer
function [6]. It is automatically enabled upon reset; therefore, the code must be developed
accordingly. During prototyping and practicing the watchdog timer can be disabled, but this
must be done at the beginning of the code, because otherwise a reset will be generated. Note
that the C compiler generates startup code, which is executed before calling the main
program and which can take longer than the default timeout period set after reset. For
example, since SDCC initialises the variables by default, if a large array is declared in the
external RAM space, the startup code may not be finished before a watchdog reset is
generated.
In order to prevent this situation, the startup code can be redefined:
/*************************************************************************
Startup code redefinition
**************************************************************************/
unsigned char _sdcc_external_startup ()
{
PCA0MD &= ~0x40; // disable watchdog timer
PCA0MD = 0x00;
// disable watchdog timer
VDM0CN = 0xA0;
return 1;
// enable VDD monitor
// 1: do not initialise variables
}
11.2 Supply monitor
The supply monitor generates a reset if the supply voltage falls below the safe level. Since
proper operation of digital circuits can only be guaranteed if the supply is within a certain
range, unexpected behaviour may happen if the voltage gets out of this range even for a short
110
Watchdog and power supply monitor
period. If the supply monitor is enabled, normal operation is restored by generating a reset in
such cases.
11.2.1 Application guidelines
Every final version of code should use the watchdog timer and the power supply
monitor to ensure reliable operation.
During testing or code development, the watchdog timer can be disabled. It must be
done at the beginning of the code to prevent undesired resets.
The watchdog timeout can be programmed. Prefer intervals short enough to ensure
quick recovery after a fault, yet long enough for the code to safely contact the
watchdog timer within the timeout period in normal operation.
The supply monitor should be switched on at the beginning of the code. After a few
microseconds allowed for stabilisation, it can be enabled.
11.2.2 Troubleshooting
Problem:
The code does not start or unexpected resets occur.
Possible reasons:
The watchdog timer is not disabled and not handled by the code.
The watchdog timer is not restarted in time. This can be due to too short a timeout,
improperly written code, time delay caused by interrupt routines or miscalculated
timings.
C compilers generate code executed before the main function, which can delay the
switching off of the watchdog timer in the main function. Most compilers allow the
redefining of the startup code (_sdcc_external_startup using the SDCC compiler),
which can help to prevent this.
11.3 Exercises
Write code that uses the watchdog timer with 1 s timeout. Try to simulate an infinite
loop and check the watchdog-generated reset.
111
Low-power and micropower applications
12 Low-power and micropower applications
In certain cases, the microcontroller operates from a low power-supply such as a battery,
solar cell or similar source. In this case, the power consumption must be kept as low as
possible to meet the supply specifications and to increase battery life and reliability at the
same time.
In most cases, the processor must perform operations only in a fraction of the time.
Therefore, keeping the power consumption low means keeping the active operating current
low and it is desirable to put the processor into an idle mode during the inactive state. Of
course, some event must be used to terminate this idle mode and to resume normal
operation.
12.1 Low-power modes
The C8051F410 processor has some low-power inactive states [6].
12.1.1 Idle mode
The processor can be placed in idle mode by setting the PCON.0 high. In this mode, program
execution is stopped and will be resumed if an enabled interrupt request occurs or a reset is
generated. The oscillator and the peripherals are not stopped in idle mode. The supply
current is reduced in idle mode: for example, the typical core supply current of 0.43 mA in
normal mode at a 1-MHz system clock will be reduced to 0.21 mA in idle mode.
12.1.2 Stop mode
A more efficient power saving mode can be realised by the stop mode. In this mode, the
internal oscillator, the core and all digital peripherals are stopped. The status of the analogue
peripherals is unaffected; they can be powered down by software before entering stop mode.
An internal or external reset is required to exit from stop mode. Therefore, program
execution will be restarted.
The power consumption can be very low in stop mode: the digital supply current can be as
low as 0.150 A
12.1.3 Suspend mode
Suspend mode is very similar to stop mode, but can be terminated by additional events
including port 0 or port 1 match to a specified bit pattern, the output of an enabled
comparator going low or real-time clock (smaRTClock) alarm or fail.
12.2 Clock speed tuning
The supply current depends on the system clock frequency in a roughly linear manner. For
example, below 15 MHz the supply current can be estimated as the actual system clock
frequency multiplied by 390 A/MHz.
This allows efficient power management even without entering idle, stop or suspend modes
and without stopping program execution. The system clock can be changed at any time, so it
can be kept low and it is only switched to a higher frequency when more processing power is
112
Low-power and micropower applications
needed. The average supply current depends on the ratio of the time spent in slower mode to
that in faster mode.
The internal clock generation module of C8051F410 processor provides several different
clock speeds. A 24.5-MHz internal oscillator serves as a base of the system clock generation;
this value can be divided by 1, 2, 4, 8, 16, 32, 64 and 128. This means that the system clock
can be as low as 191406 Hz for lowest power consumption and can be 24.5 Mhz for fastest
execution. If the internal clock multiplier is used, even a system clock of 49 MHz can be
generated.
Since the frequency of the internal oscillator is 24.5 Mhz regardless of the division used to
generate the system clock, its power consumption is constant, typically 200 A. In
conclusion, the supply current cannot be less than this value.
Note that the real-time clock (smaRTCclock) frequency can also be selected as system clock,
which allows very low supply current down to about 20 A. The 24.5-MHz internal oscillator
should be switched off to save its 200-A operating current.
12.3 Peripheral power consumption
In most low-power applications, some peripherals are used and of course they consume
power. Some considerations follow concerning the power requirements of different
components of the microcontroller.
Port pins typically drive external devices, so they may require significant current which must
be considered. For example, LEDs, pull-up resistors (like those used for SMBus) and external
circuitry load the ports. Note that for lowest-power operation even the internal weak pull-up
resistors should be disabled.
The input clock of digital peripherals (such as timers, the programmable counter array,
or communication peripherals) is derived from the system clock; therefore, their operating
current is reduced if the system clock is reduced. These peripherals require significantly less
power than the processor core.
Analogue peripherals need a certain bias current for proper operation, so they contribute
to the total supply current. Comparators can be configured in four different power modes.
Lower power can be realised at the expense of slower response. In order to reduce power
consumption, the ADC has a special burst mode. In this mode the ADC is powered only
during conversions and powered down between conversions. Therefore lowering the sample
rate lowers the power required as well. Current-output DACs definitely provide considerable
current, so if they are used, they contribute to the total supply current significantly.
12.4 Supply voltage
The supply current is roughly proportional to the supply voltage of the core and of the
peripherals. Since the total power dissipated by the system is equal to the supply current
multiplied by the supply voltage, it is very useful to reduce the supply voltage in order to
achieve low power consumption. For example, the typical supply current of the C8051F410 is
430 A at a 2.5-V core supply voltage, which is reduced to 300 A at 2.0 V. This means a
power consumption reduction from 1.1 mW to 0.6 mW.
113
Low-power and micropower applications
12.4.1 Application guidelines
The microcontroller power can be reduced using low-power modes when the core is
halted. Analogue peripherals must be switched off by software. Consider the wake-up
sources.
The supply voltage should be kept low for low-power operation.
Lower system clock frequency corresponds to lower supply current. Consider the
constant current of the internal 24.5-Mhz oscillator.
The system clock frequency can be changed during operation, but be careful: serial
data transfer, timer and even ADC operation can be seriously affected.
Minimise the loading on the port pins. Always take the current required by external
components into account.
Consider the supply current used by active digital and analogue peripherals. They
should be active only during the period they are required.
Use burst mode if the ADC is used. Keep in mind that the ADC SAR clock is derived
from a dedicated 24.5-Mhz oscillator.
Use low-power settings if comparators are used. Consider the reduced response time
of the comparators.
12.4.2 Troubleshooting
Problem:
The supply current is significantly greater than the value given in the datasheet.
Possible reasons:
The ports are loaded by external components.
The debug adapter is connected to the system. It is safest to remove it during supply
current measurement.
Some of the active peripherals are not considered.
Problem:
Invalid data are received during serial communication.
Possible reasons:
The system clock frequency is changed during data transfer or the transfer speed does
not match.
Problem:
The ADC data seem to be invalid.
Possible reasons:
The voltage reference or the ADC is powered up too close to the start of the
conversion. The time is too short for accurate settling of the voltage reference, which
can take several milliseconds.
If the ADC SAR clock is too low, the internal capacitors may lose charge during
conversion. Keep the ADC SAR clock as high as possible or use burst mode to avoid
this problem.
114
Low-power and micropower applications
12.5 Exercises
Write code that iterates the system clock frequency upon each pressing of a button
from 24.5 Mhz/128 to 24.5 MHz in a cyclic manner. Measure the digital supply
current as a function of the clock frequency. Consider any possible loads on the port
pins (including the debug adapter).
Write code that wakes up the microcontroller in every second from a suspend state
using the smaRTClock alarm function. The code must switch an LED on for 100 ms
then should go back to suspend mode.
Write code that wakes the microcontroller up from a suspend state if a button has
been pressed. The code must switch an LED on for 100 ms then should go back to
suspend mode. Use the port match event to detect button pressings and to terminate
the suspend state.
115
USB, wired and wireless communications
13 USB, wired and wireless communications
Most microcontrollers do not have communication interfaces that support direct connection
to personal computers or host computers. The most popular wired interface is the USB port,
which can even provide power supply for the connected peripheral. Devices can be wirelessly
connected via a Bluetooth module especially developed for low-power small device
applications.
There are microcontrollers with built-in USB interfaces or wireless communication modules,
but they only represent a fraction of the wide selection of microcontrollers with a rich set of
analogue and digital peripherals.
A more general solution is to use a USB-UART, Bluetooth or other wireless module
connected to the UART or similar port available on all microcontrollers. Somewhat more
space and at least two integrated circuits are required, but in this case, practically any
microcontroller can be used, which guarantees exceptional flexibility.
13.1 USB-UART interfaces
One of the most popular and most reliable USB-UART converters is the FT232R [19]. The
chip can be connected to the UART port and can handle the quite complicated USB protocol.
Only a few external capacitors are needed as power supply decoupling capacitors. The
FT232R chip supports full-speed USB communication (12 Mbit/s); however, baud rates are
limited to a maximum of 3 Mbit/s. Sending a byte means sending a start bit, 8 data bits and
1-2 stop bits, so the achievable throughput is somewhat below 300 kbyte/s. The FT232R
contains a 25-byte FIFO (first in-first out) buffer memory to avoid data loss at high data
rates.
Note that since downstream data must be directly received by the microcontroller from the
FT232R chip, the transmit FIFO of the FT232R cannot be used. Therefore, a software FIFO
must be implemented in the microcontroller code at high speed transfers. See the UART
interrupt mode examples in Chapter 7.2.
The host computer can communicate with the microcontroller via the native driver or via the
virtual COM port driver, which is easy to use even with a simple terminal software and easy
to program in C, C++, C#, Java, LabVIEW or Matlab.
Note that the virtual COM port mode has limited configuration possibilities. For example, the
so-called latency time cannot be set and its default value is 16 ms. This means that if the host
wants to send only a few bytes (at least less than the buffer size to trigger an USB transmit
transaction), then the latency time must elapse before sending the data. This can slow
communication down, so it is recommended to set the latency time to its minimum, 1 ms,
using the hardware configuration utility of the operating system.
116
USB, wired and wireless communications
5V
RTS
RTS
CTS
CTS
TX
TX
RX
RX
USB-UART
TRANSCEIVER
D+
D-
USB CONNECTOR
LDO REG
USB CONNECTOR
LDO REG
USB CABLE
C
FT232R
GND
Figure 13.1. Connecting a microcontroller to a USB port using the FT232R USBUART converter.
Figure 13.1 shows how the microcontroller can be connected to a USB port using the FT232R
USB-UART converter. The TX and RX are the UART port bits, while the RTS (ready-to-send)
and CTS (clear-to-send) on the microcontroller are provided by general-purpose port bits.
These lines are optional and can be used for handshaking – checking if data is available or if
the receiver is ready to accept data. Note that the USB port can even power the circuit; the
low dropout regulators (LDO REG) output the required supply voltage that is normally less
than 5 V.
The complete example schematic and board layout can be seen in Figure 13.2 and Figure
13.3.
Figure 13.2. Schematic of the C8051F410 microcontroller USB interface.
117
USB, wired and wireless communications
Figure 13.3. Component (red) and bottom (blue) side of the C8051F410
microcontroller USB interface printed circuit board.
All supply lines are decoupled with ceramic chip capacitors placed as close to the supply pins
as possible. The bottom side realises the required solid ground plane, and the signal ground
and USB grounds are connected at the microcontroller. This separates the sensitive analogue
circuitry of the microcontroller from the noisy ground return currents of the digital part. D1
and D2 are USB data line protection diodes, X1 is the debug port and JP1 is a connector for
port P1. This port can accept both analogue and digital input or output signals depending on
the configuration of the port P1.
Note that there are faster (USB 2.0) USB-UART interfaces, including the FT2232H, which
uses the same drivers on the computer side and therefore can be used to seamlessly upgrade
communication speed. However, the microcontroller bit rate is limited, so only high clock
frequency microcontrollers can benefit from this solution.
USB-to-parallel interfaces can also be used to transfer a whole byte at a time. This provides
the fastest communication at the expense of more complex circuitry and of the fact that much
more pins of the microcontroller must be used.
13.2 Wireless communication possibilities
There are small wireless modules that can be also connected to the microcontroller.
Bluetooth modules are widely available and have a standard SPP (serial port protocol) mode
to be driven directly from a UART port of a device. After setting up the module, it will be fully
transparent: a virtual COM port on the host computer can be used in the same way as for the
FT232R USB-UART converter or a regular COM port. In such cases, even smart phones can
be used to easily communicate with the microcontroller-based hardware unit.
118
USB, wired and wireless communications
13.3 Exercises
Write code that measures the state of the potentiometer and sends the data in text
format over the UART using a 9600-bit/s baud rate. Check the result with terminal
software using the virtual COM port.
Write code that measures the state of the potentiometer and sends the data in text
format over the BTM-112 Bluetooth module. Check the result with terminal software
using the virtual COM port.
Write code that measures the state of the in-chip temperature and sends the data in
text format over the BTM-112 Bluetooth module. Check the result with terminal
software running on a smart phone.
119
Development kit
14 Development kit
14.1 The C8051F410 development kit
The C8051F410 development kit is manufactured by Silicon Laboratories to support rapid
development and testing [7]. It can be used as a general-purpose platform to develop many
different microcontroller applications. The board is powered from a wall-plug adapter and
integrates LEDs, push buttons, a serial host interface, a potentiometer, a watch crystal, a
battery socket and even more. The complete description can be found in the user manual that
can be downloaded from the manufacturer’s pages.
The board has a two-row pin header connector that allows access of any port pin of the
microcontroller and supports the connection of various external circuitries. An extension
board with 6 additional LEDs, two 7-segment displays, a 3-pin general purpose analogue
sensor port and an LM75 temperature sensor is shown in the photo in Figure 15.1.
Figure 15.1. The C8051F410TB target board with the extension board. On the left
side, a thermistor connected to the general-purpose analogue input can be seen
The extension board is documented in the next chapter.
14.2 Extension board
The extension board is a powerful supplement to the C8051F410 development kit. It can be
used to practice many features of the microcontroller, while it also serves as a reference
design.
The six LEDs are driven from pins of port P0 and P1. The anodes of the LEDs are connected
to the positive supply, so both open-drain or push-pull mode can be used to light them. The
120
Development kit
current limiting resistors have a value of 1 kΩ, which ensures proper light intensity. The
LEDS are arranged on the board as two traffic lights and their colours are red, yellow and
green. This supports practicing several related applications. Note that the LEDs form a sixpoint rectangle (or circle), so, for example, stepper motor control can also be simulated and
visualised.
Two 7-segment displays are connected to port P2 via a buffer to reduce the total port current
of the microcontroller. Port bit P1.3 is used to select which 7-segment display is active. Both
displays cannot be used at the same time; however, this can be used to demonstrate how a
fast alternation of the displays can be applied to implement a simultaneous-looking display of
two digits. The display therefore can be used to count from 0 to 99, implement a second
counter or display a temperature in degrees, etc.
The U$3 and U$4 pin headers are only used to connect the ground and the positive supply to
the extension boards.
An LM75 I2C temperature sensor is connected to port pins P0.0 and P0.1. This supports the
measurement of external temperature and also allows the learning of the use of the
SMBus/I2C interface.
The 3-pin header labelled IN1 is a general-purpose analogue and sensor interface. The three
pins are connected to the system ground, the 5 V supply and a high impedance input of a railto-rail input and output operational amplifier. The output of this operational amplifier is
connected to pin P1.7 of the microcontroller via a voltage divider and the voltage can be
measured by the internal A/D converter. This allows the measurement of voltage-output
sensors (for example, Hall effect magnetic field sensors), resistive sensors (such as lightdependent resistors, thermistors, etc.). Current-output sensors can also be connected if an
external current-to-voltage conversion resistor is connected in parallel with the sensor. The
5 V supply can serve as a supply for active sensors or can be used as the input voltage of a
voltage divider formed by a resistor of known value and a resistive sensor. See Chapter 9 for
more information about connecting sensors to the microcontroller.
121
Development kit
Figure 15.2. The extension board schematic.
122
Development kit
Figure 15.3. Extension board top side layout.
123
Acknowledgements
15 Acknowledgements
The work has been supported by the European Union and co-funded by the European Social
Fund, project number: TÁMOP-4.1.2.A/1-11/1.
We thank Silicon Laboratories and their local distributor HT-Eurep Ltd. (Hungary) for
providing the development kits to support education. The technical documents, application
notes, knowledge base and user forum of Silicon Laboratories provided very valuable help in
our work.
We are grateful to the reviewers Dr. György Györök and Dr. Péter Makra who read the
manuscript carefully; they corrected several errors and recommended changes also.
124
References
16 References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
Chew Moi Tin, Gourab Sen Gupta: Embedded Programming with FieldProgrammable Mixed-Signal µControllers,
Silicon Laboratories MCUniversity Course Material
http://www.silabs.com/products/mcu/Pages/MCUniversity.aspx
80c51 family architecture, NXP (Philips Semiconductor)
http://www.lpcware.com/content/nxpfile/80c51-family-architecture
80c51 family programmers guide and instruction set, NXP (Philips Semiconductor)
http://www.lpcware.com/content/nxpfile/80c51-family-programmers-guide-andinstruction-set
80c51 family hardware description, NXP (Philips Semiconductor)
http://www.lpcware.com/content/nxpfile/80c51-family-hardware-description
Keil 51 Assembler,
http://www.keil.com/c51/a51kit.asp
SDCC (Small Device C Compiler) User Manual,
http://sdcc.sourceforge.net/
I. Scott MacKenzie, The 8051 Microcontroller (3rd Edition), Prentice Hall, 1998
C8051F410 datasheet, Silicon Laboratories, 2008
C8051F41x-DK User Guide, Silicon Laboratories, 2006
Silicon Laboratories Application Notes,
http://www.silabs.com/products/mcu/Pages/ApplicationNotes.aspx
Silicon Laboratories Knowledge Base,
http://www.silabs.com/support/knowledgebase/pages/default.aspx
Silicon Laboratories User Forum,
http://www.silabs.com/support/forums/Pages/default.aspx
W. Jung, Op Amp Applications Handbook, Newnes. 2006.,
http://www.analog.com/library/analogDialogue/archives/3905/op_amp_applications_handbook.html
W. Kester, The Data Conversion Handbook, Newnes, 2005.,
http://www.analog.com/library/analogDialogue/archives/3906/data_conversion_handbook.html
W. Kester, Mixed-Signal and DSP Design Techniques, Newnes, 2003.,
http://www.analog.com/en/content/mixed_signal_dsp_design_book/fca.html
H. Zumbahlen , Linear Circuit Design Handbook, Newnes, 2008.,
http://www.analog.com/library/analogDialogue/archives/4309/linear_circuit_design_handbook.html
C. Kitchin and L. Counts, A Designer's Guide to Instrumentation Amplifiers 3rd
edition, Analog Devices, Inc. 2006.,
http://www.analog.com/en/power-management/powermonitors/ad8557/products/CU_dh_designers_guide_to_instrumentation_amps/fca
.html
HD44780 datasheet, see for example http://lcdlinux.sourceforge.net/pdfdocs/hd44780.pdf
80C51FA/FB PCA Cookbook, Intel application note AP-415.
Determining Clock Accuracy Requirements for UART Communications,
http://pdfserv.maximintegrated.com/en/an/AN2141.pdf
http://www.ftdichip.com/
125
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
advertisement
© 2021