Insider`s Guide STM32

Insider`s Guide STM32
The
Insider’s Guide
To The
STM32
ARM®Based Microcontroller
An Engineer’s Introduction To The STM32 Series
www.hitex.com
Published by Hitex (UK) Ltd.
ISBN: 0-9549988 8
First Published February 2008
Hitex (UK) Ltd.
Sir William Lyons Road
University Of Warwick Science Park
Coventry, CV4 7EZ
United Kingdom
Credits
Author:
Illustrator:
Trevor Martin
Sarah Latchford
Editors:
Cover:
Michael Beach, Alison Wenlock
Wolfgang Fuller
Acknowledgements
The author would like to thank Matt Saunders and David Lamb of ST Microelectronics for their assistance in
preparing this book.
© Hitex (UK) Ltd., 21/04/2008
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in
any form or by any means, electronic, mechanical or photocopying, recording or otherwise without the prior
written permission of the Publisher.
Contents
Contents
1.
1.1
1.2
1.2.1
1.2.2
1.2.3
1.2.4
1.2.5
Introduction
4
So What Is Cortex?..................................................................................... 4
A Look At The STM32 ................................................................................ 5
Sophistication ............................................................................................. 5
Safety ......................................................................................................... 6
Security....................................................................................................... 6
Software Development ............................................................................... 6
STM32 Performance Line And Access Line ............................................... 7
2.
2.1
2.2
2.3
2.3.1
2.3.2
2.3.3
2.3.4
2.3.5
2.3.6
2.3.7
2.4
2.4.1
2.4.2
2.4.3
2.4.4
2.4.5
2.5
2.5.1
2.5.2
Cortex Overview
9
ARM Architectural Revision ........................................................................ 9
Cortex Processor And Cortex CPU........................................................... 10
Cortex CPU .............................................................................................. 10
Pipeline ..................................................................................................... 10
Programmer’s Model ................................................................................ 10
CPU Operating Modes.............................................................................. 13
Thumb-2 Instruction Set ........................................................................... 13
Memory Map ............................................................................................. 15
Unaligned Memory Accesses ................................................................... 16
Bit Banding ............................................................................................... 16
Cortex Processor ...................................................................................... 18
Busses ...................................................................................................... 18
Bus Matrix................................................................................................. 18
System Timer ........................................................................................... 19
Interrupt Handling ..................................................................................... 19
Nested Vector Interrupt Controller ............................................................ 20
Power Modes ............................................................................................ 26
Entering Low Power Mode........................................................................ 26
CoreSight Debug Support......................................................................... 26
3.
3.1
3.2
3.3
3.4
3.4.1
3.4.2
3.4.3
3.4.4
3.4.5
3.4.6
Getting It Working
29
Package Types and Footprints ................................................................. 29
Power Supply ........................................................................................... 29
Reset Circuit ............................................................................................. 30
Oscillators ................................................................................................. 31
High Speed External Oscillator ................................................................. 31
Low Speed External Oscillator.................................................................. 31
Clock Output ............................................................................................. 31
Boot Pins And Field Programming............................................................ 31
Boot Modes .............................................................................................. 32
Debug Port ............................................................................................... 32
4.
4.1
4.2
4.2.1
4.2.2
4.2.3
STM32 System Architecture
35
Memory Layout ......................................................................................... 36
Maximising Performance .......................................................................... 37
Phase Locked Loop .................................................................................. 38
FLASH Buffer ........................................................................................... 39
Direct Memory Access .............................................................................. 39
5.
5.1
Peripherals
45
General Purpose Peripherals ................................................................... 45
© Hitex (UK) Ltd.
Page 1
Contents
5.1.1
5.1.2
5.1.3
5.1.4
5.1.5
5.1.6
5.2
5.2.1
5.2.2
5.2.3
5.3
5.3.1
5.3.2
General Purpose IO .................................................................................. 45
External Interrupts .................................................................................... 47
ADC .......................................................................................................... 48
General Purpose and Advanced Timers ................................................... 54
RTC And Backup Registers ...................................................................... 60
Backup Registers And Tamper Pin ........................................................... 61
Connectivity .............................................................................................. 62
SPI ............................................................................................................ 62
I2C ............................................................................................................ 63
USART ..................................................................................................... 64
Can And USB Controller ........................................................................... 65
CAN Controller ......................................................................................... 65
USB .......................................................................................................... 67
6.
6.1
6.1.1
6.2
6.2.1
6.2.2
6.3
6.4
6.5
Low Power Operation
69
RUN Mode ................................................................................................ 69
Prefetch Buffer And Half-Cycle Mode ....................................................... 69
Low Power Modes .................................................................................... 70
SLEEP ...................................................................................................... 70
STOP Mode .............................................................................................. 71
Standby .................................................................................................... 72
Backup Region Power Consumption ........................................................ 72
Debug Support ......................................................................................... 72
7.
7.1
7.2
7.3
7.4
7.4.1
7.4.2
7.5
7.5.1
7.5.2
7.5.3
Safety Features
74
Reset Control............................................................................................ 74
Power Voltage Detect ............................................................................... 74
Clock Security System .............................................................................. 75
Watchdogs................................................................................................ 76
Windowed Watchdog ................................................................................ 76
Independent Watchdog............................................................................. 77
Peripheral Features .................................................................................. 78
GPIO Port Locking .................................................................................... 78
Analog Watchdog ..................................................................................... 78
Break Input ............................................................................................... 78
8.
8.1
8.2
8.3
8.3.1
8.3.2
8.3.3
The FLASH Module
80
Internal FLASH Security And Programming ............................................. 80
Erase And Write Operations ..................................................................... 80
Option Bytes ............................................................................................. 80
Write Protection ........................................................................................ 81
Read Protection ........................................................................................ 81
Configuration Byte .................................................................................... 81
9.
9.1.1
9.1.2
9.1.3
Development Tools
83
Evaluation Tools ....................................................................................... 83
Libraries And Protocol Stacks................................................................... 84
RTOS........................................................................................................ 84
10.
End Note
86
11.
Bibliography
88
© Hitex (UK) Ltd.
Page 2
Chapter 1: Introduction
© Hitex (UK) Ltd.
Page 3
Chapter 1: Introduction
1. Introduction
Over the last six or seven years one of the major trends in microcontroller design is the adoption of the ARM7 and
ARM9 as the CPU for general purpose microcontrollers. Today there are some 240 ARM-based microcontrollers
available from a wide range of manufacturers. Now ST Microelectronics have launched the STM32, their first
microcontroller based on the new ARM Cortex-M3 microcontroller core. This device sets new standards of
performance and cost, as well as being capable of low power operation and hard real-time control.
1.1 So What Is Cortex?
The ARM Cortex family is a new generation of processor that provides a standard architecture for a wide range of
technological demands. Unlike the other ARM CPUs, the Cortex family is a complete processor core that provides
a standard CPU and system architecture. The Cortex family comes in three main profiles: the A profile for high
end applications, R for real time and M for cost-sensitive and microcontroller applications. The STM32 is based
on the Cortex-M3 profile, which is specifically designed for high system performance combined with low power
consumption. It has a low enough cost to challenge traditional 8 and 16-bit microcontrollers.
While the ARM7 and ARM9 CPUs have been successfully integrated into standard microcontrollers, they do show
their SoC heritage. This is particularly noticeable in the area of exception and interrupt handling, because each
specific manufacturer has designed their own solution. The Cortex-M3 provides a standardised microcontroller
core which goes beyond the CPU to provide the entire heart of a microcontroller (including the interrupt system,
SysTick timer, debug system and memory map). The 4Gbyte address space of the Cortex-M3 is split into welldefined regions for code, SRAM, peripherals and system peripherals. Unlike the ARM7, the Cortex-M3 is a
Harvard architecture and so has multiple busses that allow it to perform operations in parallel, boosting its overall
performance. Unlike earlier ARM architectures, the Cortex family allows unaligned data accesses. This ensures
the most efficient use of the internal SRAM. The Cortex family also supports setting and clearing of bits within two
1Mbyte regions of memory by a method called bit banding. This allows efficient access to peripheral registers and
flags located in SRAM memory without the need for a full Boolean processor.
The heart of the STM32 is the Cortex-M3 processor.
The Cortex M3 processor is a standardised
microcontroller including 32 bit CPU, bus structure,
nested interrupt unit, debug system and standard
memory layout.
© Hitex (UK) Ltd.
Page 4
Chapter 1: Introduction
One of the key components of the Cortex-M3 core is the Nested Vector Interrupt Controller (NVIC). The NVIC
provides a standard interrupt structure for all Cortex based microcontrollers and exceptional interrupt handling.
The NVIC provides dedicated interrupt vectors for up to 240 peripheral sources where each interrupt source can
be individually prioritised. The NVIC has been designed for extremely fast interrupt handling. The time taken from
receiving an interrupt to reaching the first line of code in your service routine is just twelve cycles. This is achieved
in part by automatic stack handling which is done by microcode within the CPU. In the case of back to back
interrupts, the NVIC uses a “tail chaining” method that allows successive interrupts to be served with only a six
cycle latency. During the interrupt stacking phase, a high priority interrupt can pre-empt a low priority interrupt
without incurring any additional CPU cycles. The interrupt structure is also tightly coupled to the low power
modes within the Cortex-M3 core. It is possible to configure the CPU to automatically enter a low power on exit
from an interrupt. The core then stays asleep until another exception is raised.
Although the Cortex-M3 is designed as a low cost core, it is still a 32-bit CPU and as such has support for two
operating modes: Thread mode and Handler mode, which can be configured with their own stacks. This allows
more sophisticated software design and support for real-time operating systems. The Cortex core also includes a
24-bit auto reload timer that is intended to provide a periodic interrupt for an RTOS kernel. While the ARM7 and
ARM9 CPUs have two instruction sets (the ARM 32-bit and Thumb 16-bit instruction sets) the Cortex family is
designed to support the ARM Thumb-2 instruction set. This blends both 16 and 32-bit instructions, to deliver the
performance of the ARM 32-bit instruction set with the code density of the Thumb 16-bit instruction set. The
Thumb-2 instruction set is a rich instruction set that is designed as a target for C/C++ compilers. This means that
a Cortex application can be entirely coded in C.
1.2 A Look At The STM32
ST already have four ARM7 and ARM9 based microcontroller families, but the STM32 is a significant step up the
price/performance curve. With volume pricing at just over one Euro, the STM32 is a serious challenge to existing
8-bit microcontrollers.. . The STM32 was initially released with fourteen different variants.These are split into two
groups: the Performance line which operates up to CPU clock speeds of 72MHz and the Access line which runs
up to 36MHz. Both sets of variants are pin and software compatible and offer FLASH ROM sizes up to 128K and
20K SRAM. Since the initial release the STM32 road map has been extended to include devices with larger
RAM and FLASH memories and more complex peripherals.
The STM32 family has two distinct branches. The Performance line which runs to 72MHz and has
the full set of peripherals and the Access line which runs to 36MHz and has a reduced set of
peripherals.
1.2.1
Sophistication
At first glance the peripheral set looks like a typical small microcontroller, featuring peripherals such as Dual ADC,
general purpose timers, I2C,SPI,CAN,USB and a real-time clock. However, each of these peripherals is very
© Hitex (UK) Ltd.
Page 5
Chapter 1: Introduction
feature-rich. For example the 12-bit ADC has an integral temperature sensor and multiple conversion modes and
devices with dual ADC can slave both ADCs together in a further nine conversion modes. Similarly, each of the
four timers has four capture compare units and each timer block may be combined with the others to build
sophisticated timer arrays. An advanced timer has additional support for motor control, with 6 complimentary
PWM outputs with programmable dead time and a break input line that will force the PWM signal to a pre
programmed safe state. The SPI peripheral has a hardware CRC generator for 8 and 16 words to support
interfacing to SD and MMC cards.
Surprisingly for a small microcontroller, the STM32 also includes a seven channel DMA unit. Each channel can
be used to transfer data to and from any peripheral register on memory location as 8/16 or 32-bit words. Each of
the peripherals can be a DMA flow controller sending or demanding data as required. An internal bus arbiter and
bus matrix minimise the arbitration between the CPU data accesses and the DMA channels. This means that the
DMA unit is flexible, easy to use and really automates data flow within the microcontroller.
In an effort to square the circle the STM32 is a low power as well as high performance microcontroller. It can run
from a 2V supply and at 72MHz with everything switched on it consumes just 36mA. In combination with the
Cortex low power modes the STM32 has a standby power consumption of just 2µA. An internal 8MHz RC
oscillator allows the chip to quickly come out of low power modes while the external oscillator is still starting up.
This fast entry and exiting from low power modes further reduces overall power consumption.
1.2.2
Safety
As well as demanding more processing power and more sophisticated peripherals, many modern applications
have to operate in safety-critical environments. With this in mind, the STM32 has a number of hardware features
that help support high integrity applications. These include a low power voltage detector, a clock security system
and two separate watchdogs. The first watchdog is a windowed watchdog. This watchdog must be refreshed in a
defined time frame. If you hit it too soon, or too late, the watchdog will trigger. The second watchdog is an
independent watchdog which has its own external oscillator separate from the main system clock. A further clock
security system can detect failure of the main external oscillator and fail safely back onto an internal 8MHz RC
oscillator.
1.2.3
Security
One of the other unfortunate requirements of modern design is the need for code security to prevent software
piracy. Here the STM32 FLASH can be locked for FLASH READ accesses via the debug port. When READ
protection is enabled, the FLASH memory is also WRITE protected to prevent untrusted code from being
inserted on the interrupt vector table. Further WRITE protection can be enabled over the remainder of the FLASH
memory. The STM32 also has a real-time clock and a small area of battery backed SRAM. This region has an
anti-tamper input that can trigger an interrupt on a state change. In addition an anti-tamper event will
automatically clear the contents of the battery backed SRAM.
1.2.4
Software Development
If you are already using an ARM-based microcontroller, the good news is that the chances are that your
development tools already support the Thumb-2 instruction set and the Cortex family. The worst case is a
software upgrade to get the necessary support. ST also provide a peripheral driver library, a USB developer
library as an ANSI C library and source code that is compatible with earlier libraries published for their STR7 and
STR9 microcontrollers. Ports of these libraries are already available for popular compiler tools. Similarly, many
open source and commercial RTOS and middleware (TCP/IP, file system etc) are available for the Cortex family.
The Cortex-M3 also comes with a whole new debug system called CoreSight. Access to the CoreSight system is
through the Debug Access Port which supports either a standard JTAG connection or a serial wire (2 Pin)
interface. As well as providing debug run control, the CoreSight system on the STM32 provides a data watchpoint
and an instrumentation trace. The instrumentation trace can send selected application information up to the
debug tool. This can provide extended debug information and can also be used during software testing.
© Hitex (UK) Ltd.
Page 6
Chapter 1: Introduction
1.2.5
STM32 Performance Line And Access Line
The STM32 family has two distinct branches: the Performance line and Access line. The Performance line has
the full set of peripherals and runs to the maximum 72MHz. The Access line has a reduced set of peripherals and
runs to a maximum 32MHz. Importantly the package types and pins layouts are the same between both the
Access and Performance line variants. This allows different versions of the STM32 to be interchanged without
having to re-spin the PCB.
© Hitex (UK) Ltd.
Page 7
Chapter 2: Cortex Overview
© Hitex (UK) Ltd.
Page 8
Chapter 2: Cortex Overview
2. Cortex Overview
As we saw in the introduction, the Cortex processor is the next generation embedded core from ARM. It is
something of a departure from the earlier ARM CPUs in that it is a complete processor core, consisting of the
Cortex CPU and a surrounding set of system peripherals, providing the heart of an embedded system. As a
result of the wide variety of embedded systems, the Cortex processor is available in a number of application
profiles. These are denoted by the letter following the Cortex name. The three profiles are as follows:
Cortex-A Series, applications processors for complex OS and user applications.
Supports the ARM, Thumb and Thumb-2 instruction sets.
Cortex-R Series, real-time systems profile.
Supports the ARM, Thumb, and Thumb-2 instruction sets.
Cortex-M Series, microcontroller profile optimized for cost-sensitive applications.
Supports Thumb-2 instruction set only.
The number at the end of the Cortex name refers to the relative performance level, with 1 the lowest and 8 the
highest. Currently performance level 3 is the highest performance level available in the microcontroller profile. The
STM32 is based on the Cortex-M3 processor.
2.1 ARM Architectural Revision
ARM also somewhat confusingly denote each of their processors with an architectural revision. (This is written
ARMV6, ARMV7 etc.) The Cortex M3 has the architectural revision ARMV7 M.
The Cortex-M3 processor is based on the ARMV7
architecture and is capable of executing the
Thumb-2 instruction set.
Thus the documentation for the Cortex-M3 consists of the Cortex-M3 Technical Reference Manual and the
ARMV7 M Architectural Reference Manual. Both of these documents can be downloaded from the ARM website
at www.arm.com
© Hitex (UK) Ltd.
Page 9
Chapter 2: Cortex Overview
2.2 Cortex Processor And Cortex CPU
Throughout the remainder of this book, the terms Cortex processor and Cortex CPU will be used to distinguish
between the complete Cortex embedded core and the internal RISC CPU. In the next section we will look at the
key features of the Cortex CPU followed by the system peripherals in the Cortex processor.
2.3 Cortex CPU
At the heart of the Cortex processor is a 32-bit RISC CPU. This CPU has a simplified version of the ARM7/9
programmer’s model, but a richer instruction set with good integer maths support, better bit manipulation and
‘harder’ real-time performance.
2.3.1
Pipeline
The Cortex CPU can execute most instructions in a single cycle. Like the ARM7 and ARM9 CPUs this is achieved
with a three stage pipeline.
Like the ARM7 and ARM9 CPUs
the Cortex-M3 has a three stage
pipeline. However, the Cortex-M3
also has branch prediction to
minimise the number of pipeline
flushes.
Whilst one instruction is being executed, the next is being decoded and a third is being fetched from memory.
This works very well for linear code, but when a branch is encountered the pipeline must be flushed and refilled
before code can continue to execute. In the ARM7 and ARM9 CPUs branches are very expensive in terms of
code performance. In the Cortex CPU the three stage pipeline is enhanced with branch prediction. This means
that when a conditional branch instruction is reached, a speculative fetch is performed, so that both destinations
of the conditional instruction are available for execution without incurring a performance hit. The worst case is an
indirect branch where a speculative fetch cannot be made and the only course of action is to flush the pipeline.
While the pipeline is key to the overall performance of the Cortex CPU, no special considerations need to be
made in the application code.
2.3.2
Programmer’s Model
The Cortex CPU is a RISC processor which has a load and store architecture. In order to perform data processing
instructions, the operands must be loaded into a central register file, the data operation must be performed on
these registers and the results then saved back to the memory store.
The Cortex-M3 is a load and store architecture. All data has to be moved into a central register file before a
data processing instruction can act on it.
© Hitex (UK) Ltd.
Page 10
Chapter 2: Cortex Overview
Consequently all the program activity focuses around the CPU register file. This register file consists of sixteen
32-bit wide registers. Registers R0-R12 are simple registers that can be used to hold program variables. The
Registers R13-R15 have special functions within the Cortex CPU. Register R13 is used as the stack pointer. This
register is banked, which allows the Cortex CPU to have two operating modes each with their own separate stack
space. This is typically used by an RTOS which can run its ‘system’ code in a protected mode. In the Cortex CPU
the two stacks are called the main stack and the process stack. The next register R14 is called the link register.
This register is used to store the return address when a call is made to a procedure. This allows the Cortex CPU
to make a fast entry and exit to a procedure. If your code calls several levels of subroutines, the compiler will
automatically store R14 on the stack. The final register R15 is the program counter; since this is part of the central
register file it can be read and manipulated like any other register.
The Cortex-M3 has a CPU register file of 16 32-bit wide registers. Like the
earlier ARM7/9 CPUs R13 is the stack pointer. R14 is the link register and
R15 is the PC. R13 is a banked register to allow the Cortex-M3 to operate
with two stacks: a process stack and a main stack.
2.3.2.1 XPSR
In addition to the register file there is a separate register called the Program Status Register. This is not part of
the main register file and is only accessible through two dedicated instructions. The xPSR contains a number of
fields that influence the execution of the Cortex CPU.
The Program Status Register contains status fields for instruction execution. This register is
aliased into the Application, Execution and Interrupt Status Registers
The xPSR register can also be accessed through three special alias names that allow access to sub-ranges of
bits within the xPSR. The top five bits are the condition code flags and are aliased as the Application Program
Status Register. The first four condition code flags N,Z,C,V ( Negative, Zero, Carry and Overflow) will be set and
cleared depending on the result of a data processing instruction. The Q bit is used by the DPS saturated maths
instructions to indicate that a variable has reached its maximum or minimum value. Like the ARM 32-bit
instruction set, certain Thumb-2 instructions are only executed if the instruction condition code matches the state
of the Application Program Status Register flags. If the instruction condition codes do not match, the instruction
passes through the pipeline as a NOP. This ensures that instructions flow smoothly through the pipeline and
minimises pipeline flushes. In the Cortex CPU, this technique is extended with the Execution Program Status
© Hitex (UK) Ltd.
Page 11
Chapter 2: Cortex Overview
Register. This is an alias of bits 26 – 8 of the xPSR. This contains three fields: the “If then” field the “interrupt
continuable instruction” and the Thumb instruction field. The Thumb-2 instruction set has an efficient method of
executing small ‘if then’ blocks of instructions. When a conditional test is true, it can set a value in the IT field that
tells the CPU to execute up to four following instructions. If the conditional test fails, these instructions will pass
through the pipeline as a NOP. Thus a typical line of C would be coded as follows:
If (r0 ==0)
CMP r0,#0
ITTEE EQ
Then r0 = *r1 +2;
LDR r0,[r1]
ADDr0,#2
compare r0 to 0
if true execute the next two instructions
load contents of memory location into r0
add 2
While most Thumb-2 instructions execute in a single cycle, some (such as load and store instructions) take
multiple cycles. So that the Cortex CPU can have a deterministic interrupt response time, these instructions must
be interruptible. When an instruction is terminated early, the interrupt continuable instruction field stores the
number of the next register to be operated on in the load or store multiple instruction. Thus once the interrupt has
been serviced, the load/store multiple instruction can resume execution. The final Thumb field is inherited from
the earlier ARM CPUs. This field indicates if the ARM or Thumb instruction set is currently being executed by the
CPU. In the Cortex-M3 this bit is always set to one. Finally, the interrupt status field contains information on any
interrupt request that was pre-empted.
© Hitex (UK) Ltd.
Page 12
Chapter 2: Cortex Overview
2.3.3
CPU Operating Modes
While the Cortex processor is designed to be a low gate count, fast and easy to use microcontroller core, it has
been designed to support the use of a real-time operating system. The Cortex processor has two operating
modes: Thread mode and Handler mode. The CPU will run in Thread mode while it is executing in non-interrupt
background mode and will switch to the Handler mode when it is executing exceptions. In addition, the Cortex
CPU can execute code in a privileged or non-privileged mode. In privileged mode, the CPU has access to the full
instruction set. In unprivileged mode certain instructions are disabled (such as the MRS and MSR instructions
which allow access to the xPSR and its aliases). Additionally, access to most registers in the Cortex processor
system control space is also disabled. Stack usage can also be configured. The main stack (R13) can be used by
both Thread and Handler mode. Alternatively, Handler mode can be configured to use the process stack (R13
banked register).
The Cortex-M3 can be used in a ‘flat’ simple mode. It
is also designed to support real-time operating
systems. It has Handler and Thread modes that can
be configured to use the main and process stacks
and have privileged access to the Cortex system
control registers.
Out of reset the Cortex processor will run in a ‘flat’ configuration. Both Thread and Handler modes execute in
privileged mode, so there are no restrictions on access to any processor resources. Both the Thread and Handler
modes use the main stack. In order to start execution, the Cortex processor simply needs the reset vector and the
start address of the stack to be configured before you can start to execute your application C code. However, if
you are using an RTOS or are developing a safety-critical application, the chip can be used in a mode advanced
configuration where Handler mode (exceptions and the RTOS) runs in privileged mode and uses the main stack
while application code runs in Thread mode with unprivileged access and uses the process stack. This way the
system code and the application code are partitioned and errors in the application code will not cause the RTOS
to crash.
2.3.4
Thumb-2 Instruction Set
The ARM7 and ARM9 CPUs can execute two instruction sets: the ARM 32-bit instruction set and the Thumb 16bit instruction set. This allows a developer to optimise his program by selecting the instruction set used for
© Hitex (UK) Ltd.
Page 13
Chapter 2: Cortex Overview
different procedures: 32-bit instructions for speed and 16-bit instructions for code compression. The Cortex CPU
is designed to execute the Thumb-2 instruction set which is a blend of 16 and 32 bit instructions. The thumb-2
instruction set gives a 26% code density improvement over the ARM 32-bit instruction set and a 25%
improvement in performance over the Thumb 16-bit instruction set. The Thumb2 instruction set has some
improved multiply instructions which can execute in a single cycle and a hardware divide that takes between 2 – 7
cycles.
The Cortex processor benchmarks give
a performance level of 1.2 DMIPS/MHz,
which is 1.2 Clock cycles per instruction.
The Thumb-2 instruction set also has: improved branching instructions including test and compare, if/then
conditional execution blocks and for data manipulation byte ordering and byte and half word extraction
instructions. While still a RISC processor, the Cortex CPU also has a rich instruction set that is specifically
designed as a good target for a C compiler. A typical Cortex-M3 program will be written entirely in ANSI C, with
minimal non-ANSI keywords and only the exception vector table written in Assembler.
© Hitex (UK) Ltd.
Page 14
Chapter 2: Cortex Overview
2.3.5
Memory Map
The Cortex-M3 processor is a standardised microcontroller core and as such has a well-defined memory map.
Despite the multiple internal busses this memory map is a linear 4 Gbyte address space.
The Cortex-M3 defines a fixed 4
Gb memory map that specifies
regions
for
code
SRAM
peripherals, external memory
and devices and the Cortex
system registers. This memory
map is common to all Cortexbased devices.
The first 1Gbyte of memory is split evenly between a code region and a SRAM region. The code space is
optimised to be executed from the I-Code bus. Similarly, the SRAM is reached with the D-code bus. Although
code can be loaded and executed from the SRAM, the instructions would be fetched using the system bus, which
incurs an extra wait state. It is likely that code would run slower from SRAM than from on-chip FLASH memory
located in the code region. The next 0.5 Gbyte of memory is the on-chip peripheral region. All user peripherals
provided by the microcontroller vendor will be located in this region. The first 1 Mbyte of both the SRAM and
Peripheral regions is bit-addressable using a technique called bit banding. Since all the SRAM and all the user
peripherals on the STM32 are located in these regions all the memory locations of the STM32 can be
manipulated in a word-wide or bitwise fashion. The next 2 Gbyte address space is allocated to external memorymapped SRAM and peripherals. The final 0.5 Gbyte is allocated to the internal Cortex processor peripherals and
a region for future vendor specific enhancements to the Cortex processor. All of the Cortex processor registers
are at fixed locations for all Cortex-based microcontrollers. This allows code to be more easily ported between
different STM32 variants and indeed other vendors’ Cortex-based microcontrollers. One processor to learn, one
set of tools to invest in and large amounts of reusable code across a wide range of microcontrollers.
© Hitex (UK) Ltd.
Page 15
Chapter 2: Cortex Overview
2.3.6
Unaligned Memory Accesses
The ARM7 and ARM9 instruction sets are capable of accessing byte, half word and word signed and unsigned
variables. This allows the CPU to naturally support integer variables without the need for the sort of software
library support typically required in 8 and 16-bit microcontrollers. However, the earlier ARM CPUs do suffer from a
disadvantage in that they can only do word or half-word aligned accesses. This restricts the compiler linker in its
ability to pack data into the SRAM and some valuable SRAM will be wasted. (This can be as much as 25%
depending on the mix of variables used.)
The Cortex-M3 can make unaligned memory accesses, which ensures that the SRAM is efficiently
used.
The Cortex CPU has addressing modes for word, half-word and byte, but is able to make unaligned memory
accesses. This gives the compiler linker complete freedom to order the program data in memory. The additional
bit banding support on the Cortex CPU allows program flags to be packed into a word or half-word variable rather
than using a byte for each flag.
2.3.7
Bit Banding
The earlier ARM7 and ARM9 CPUs were only able to perform bit manipulations on SRAM and peripheral memory
locations by using AND and OR operations. This requires a READ MODIFY WRITE operation which is expensive
in terms of the number of cycles taken to set and clear individual bits and the overall code space required for each
bit manipulation.
The bit banding technique allows atomic bit manipulation while keeping the Cortex-M3 CPU to a
minimal gate count.
© Hitex (UK) Ltd.
Page 16
Chapter 2: Cortex Overview
To overcome this limitation it would be possible to introduce a dedicated bit set and clear instructions, or a full
Boolean processor, but this would increase the size and complexity of the Cortex CPU. Instead, a technique
called bit banding allows direct bit manipulation on sections of the peripheral and SRAM memory spaces, without
the need for any special instructions. The bit addressable regions of the Cortex memory map are composed of the
bit band region (which is up to 1Mbyte of real memory or peripheral registers) and the bit band Alias region which
takes up to 32Mbyte of the memory map. Bit banding works by mapping each bit in the bit band region to a word
address in the Alias region. So by setting and clearing the aliased word address we can set and clear bits in the
real memory.
Bit Banding is supported over the first 1Mb of the SRAM and Peripheral
regions. This covers all the resources of the STM32.
This allows us to perform individual bit manipulation without the need for special instructions and keeps the
overall size of the Cortex core as small as possible. In practice, we need to calculate the address of the bit band
alias word for a given memory location in the peripheral or SRAM space. The formula to calculate the alias
address is as follows:
Address in the bit band alias region
Where bit word offset
= Bit band alias base address + bit word offset
= Byte offset from bit band base X 0x20 + bit number x 4
This is much easier than it may look at first glance. For a practical example, the GPIO output data register is
written to in order to set and clear individual IO lines. The physical address of the Port B output register is
0x40010C0C. In this example we want to be able to set and clear bit eight of this word using the above formula.
Word address
Peripheral bit band base
Peripheral bit band Alias base
Byte offset from bit band base
Bit word offset
Bit Alias address
= 0x40010C0C
= 0x40000000
= 0x42000000
= 0x40010c0c – 0x40000000 = 10c0c
= (0x10c0c x 0x20) +(8x4) = 0x2181A0
= 0x42000000 + 0x2181A0 = 0x422181A0
We can now create a pointer to this address using the following line of C:
#define PortBbit8
(*((volatile unsigned long *) 0x422181A0 ))
This pointer can then be used to set and clear the IO port bit:
PB8 = 1;
//led on
Which generates the following assembly instructions:
MOVS
LDR
STR
r0,#0x01
r1,[pc,#104]
r0,[r1,#0x00]
© Hitex (UK) Ltd.
Page 17
Chapter 2: Cortex Overview
Switching the LED off:
PB8 = 0;
//led off
Generates the following assembly instructions:
MOVS
LDR
STR
r0,#0x00
r1,[pc,#88]
r0,[r1,#0x00]
Both the set and clear operations take three 16-bit instructions and on the STM32 running at 72 MHz these
instructions are executed in 80nsec. Any word in the peripheral and SRAM bit band regions can also be directly
addressed word-wide so we could perform the same set and clear using the more traditional AND and OR
approach:
GPIOB->ODR |= 0x00000100;
LDR
r0,[pc,#68]
ADDS
r0,r0,#0x08
LDR
r0,[r0,#0x00]
ORR
r0,r0,#0x100
LDR
r1,[pc,#64]
STR
r0,[r1,#0xC0C]
//LED on
GPIOB->ODR &=!0x00000100;
LDR
r0,[pc,#40]
ADDS
r0,r0,#0x08
LDR
r0,[r0,#0x00]
MOVS
r0,#0x00
LDR
r1,[pc,#40]
STR
r0,[r1,#0xC0C]
//LED off
Now each set and clear operation takes a mixture of 16 and 32-bit operations, which take a minimum of 14 bytes
for each operation and at the same clock frequency take a minimum of 180 nSec. If you consider the impact of bit
banding on a typical embedded application that sets and clears lots of bits in the peripheral registers and uses
semaphores and flags in the SRAM, you are very clearly going to make significant savings in both code size and
execution time and it is all handled in the STM32 header file for you.
2.4 Cortex Processor
2.4.1
Busses
The Cortex-M3 processor has a Harvard architecture with separate code and data busses. These are called the
Icode bus and the Dcode bus. Both of these busses can access code and data in the range 0x00000000 –
0x1FFFFFFF. An additional system bus is used to access the Cortex system control space in the range
0x20000000-0xDFFFFFFF and 0xE0100000-0xFFFFFFFF. The Cortex on-chip debug system has an additional
bus structure called the Private Peripheral Bus.
2.4.2
Bus Matrix
The system and data busses are connected to the external microcontroller via a set of high speed busses
arranged as a bus matrix. This allows a number of parallel paths between the Cortex busses and other external
bus masters such as DMA to the on-chip resources such as SRAM and peripherals. If two bus masters (i.e. the
Cortex CPU and a DMA channel) try to access the same peripheral, an internal arbiter will resolve the conflict and
grant bus access to the highest priority peripheral. However, in the STM32 the DMA units are designed to work in
concert with the Cortex CPU, as we will see when we examine the operation of the DMA unit.
© Hitex (UK) Ltd.
Page 18
Chapter 2: Cortex Overview
2.4.3
System Timer
The Cortex core also includes a 24-bit down counter, with auto reload and end of count interrupt. This is intended
to provide a standard timer for all Cortex-based microcontrollers. The SysTick timer is intended to be used to
provide a system tick for an RTOS, or to generate a periodic interrupt for scheduled tasks. The SysTick Control
and status register in the Cortex-M3 System control space unit allows you to select the SysTick clock source. By
setting the CLKSOURCE bit the SysTick timer will run at the CPU frequency. When cleared the timer will run at
1/8 CPU frequency.
The SysTick Timer is a 24-bit auto-reload timer located within
the Cortex-M3 processor. It is intended to provide a timer tick
for a Real Time Operating System.
The SysTick timer has three registers. The current value and reload value should be initialised with the count
period. The control and status register contains an ENABLE bit to start the timer running and a TICKINT bit to
enable its interrupt line. In the next section we will look at the Cortex interrupt structure and use the SysTick timer
to generate a first exception on the STM32.
2.4.4
Interrupt Handling
One of the key improvements of the Cortex core over the earlier ARM CPUs is its interrupt structure and
exception handling. The ARM7 and ARM9 CPUs had two interrupt lines: the fast interrupt and the general
purpose interrupt line. These two interrupt lines had to support all of the interrupt sources within a given
manufacturer’s microcontroller. How this was done varied according to the implementation, so while the
techniques used were broadly the same, the implementation differed between manufacturers. The ARM7 and
ARM9 interrupt structure suffers from two further problems. Firstly it is not deterministic; the time taken to
terminate or abort an instruction under execution when the interrupt occurs is variable. This may not be a problem
for many applications, but it is a big issue in real-time control. Secondly, the ARM7 and ARM9 interrupt structure
does not naturally support nested interrupts; further software is required: either Assembler macros or an RTOS.
One of the key criteria of the Cortex core is to overcome these limitations and provide a standard interrupt
structure which is both extremely fast and deterministic.
© Hitex (UK) Ltd.
Page 19
Chapter 2: Cortex Overview
2.4.5
Nested Vector Interrupt Controller
The Nested Vector Interrupt Controller is a standard unit within the Cortex core. This means that all Cortex-based
microcontrollers will have the same interrupt structure, regardless of manufacturer. Thus application code and
operating systems can be easily ported from one microcontroller to another and the programmer does not need to
learn a whole new set of registers. The NVIC is also designed to have a very low interrupt latency. This is both a
feature of the NVIC itself and of the Thumb-2 instruction set which allows multi-cycle instructions such as load
and store multiple to be interruptible. This interrupt latency is also deterministic, with several advanced interrupt
handling features that support real-time applications. As its name implies, the NVIC is designed to support nested
interrupts and on the STM32 there are 16 levels of priority. The NVIC interrupt structure is designed to be
programmed entirely in ‘C’ and does not need any Assembler macros or special non-ANSI directives.
The STM32 processor includes a Nested
Vector Interrupt Controller which can
support a maximum of 240 external
peripherals.
Although the NVIC is a standard unit within the Cortex core, in order to keep the gate count to a minimum the
number of interrupt lines going into the NVIC is configurable when the microcontroller is designed. The NVIC has
one non-maskable interrupt and up to a further 240 external interrupt lines which can be connected to the user
peripherals. There are an additional 15 interrupt sources within the Cortex core, which are used to handle internal
exceptions within the Cortex core itself. The STM32 NVIC has been synthesised with a maximum of 43 maskable
interrupt lines.
2.4.5.1 NVIC Operation Exception Entry And Exit
When an interrupt is raised by a peripheral, the NVIC will start the Cortex CPU serving the interrupt. As the Cortex
CPU enters its interrupt mode, it will push a set of registers onto the stack. Importantly this is done in microcode,
so there is no instruction overhead in the application code. While the stack frame is being saved, the starting
address of the interrupt service routine is fetched on the instruction bus. Thus the time taken from the interrupt
being raised to reaching the first instruction in the interrupt routine is just 12 cycles.
The NVIC will respond to an interrupt with a latency of
just six cycles. This includes a microcoded routine to
automatically push a set of registers onto the stack.
© Hitex (UK) Ltd.
Page 20
Chapter 2: Cortex Overview
The stack frame consists of the Program Status Register, the program counter and the link register. This saves
the context that the Cortex CPU was running in. In addition registers R0 – R3 are also saved. In the ARM binary
interface standard these registers are used for parameter passing, so saving these gives us a set of CPU
registers that can be used by the ISR. Finally R12 is also saved; this register is the intracall scratch register. This
register is used by any code that runs when function calls are made. For example, if you have enabled stack
checking in the compiler, the additional code generated will use R12 if it need a CPU register. When the interrupt
ends the process is reversed, the stack frame is restored automatically by microcode and in parallel the return
address is fetched, so that the background code can resume execution in 12 cycles.
2.4.5.2 Advanced Interrupt Handling Modes
As well as being able to handle a single interrupt very quickly, the NVIC is designed to efficiently handle multiple
interrupts in a very real time application. The NVIC has several clever methods of handling multiple interrupt
sources the minimum delay between interrupts and to ensure that the highest priority interrupt is served first.
2.4.5.2.1 Interrupt Pre-emption
The NVIC is also designed to allow high priority interrupts to pre-empt a currently running low priority interrupt. In
this case the running interrupt is halted and a new stack frame is saved in the standard 12 cycles after which the
high priority interrupt runs. When the high priority interrupt is finished, the stack is automatically POPed and the
low priority interrupt can resume execution.
2.4.5.2.2 Tail Chaining
If a high priority interrupt is running and a low priority interrupt is raised, the Cortex NVIC uses a method called tail
chaining to ensure that there is a minimum delay between servicing interrupts.
Multiple interrupts will be ‘tail chained’ so
there is a minimum delay from the end of
one interrupt to the start of the next.
If two interrupts are raised, the highest priority interrupt will be served first and will begin execution in the standard
12 cycles. However, at the end of the interrupt routine the Cortex CPU does not return to the background code.
The stack frame is not restored, only the entry address of the next highest priority ISR is fetched. This takes just
six cycles and then the next interrupt service routine can begin execution. At the end of the pending interrupts the
stack is restored and the return address is fetched, so the background code can begin execution in the standard
12 cycles. If the low priority interrupt arrives while the running interrupt is exiting, the POP will be abandoned and
the stack pointer will be wound back to its original value. There is an additional 6 cycle delay while the new ISR
address is fetched. This gives a latency of between 7 – 18 cycles before the new interrupt service routine can
begin execution.
A low priority interrupt which occurs as
the stack is being unwound from a
currently
executed
interrupt
will
automatically be tail chained with a
delay of 7 – 18 cycles
© Hitex (UK) Ltd.
Page 21
Chapter 2: Cortex Overview
2.4.5.2.3 Late Arrival
In a real-time system there may often be a condition where we have started to serve a low priority interrupt,
only for a high priority interrupt to be raised. If this condition occurs during the initial PUSH the NVIC will switch to
serve the higher priority interrupt. The stacking continues and there will be a minimum of 6 cycles from the point
at which the high priority interrupt is raised, while the new ISR address is fetched.
A late arriving high priority interrupt will
pre-empt a low priority interrupt without
incurring an additional stacking overhead.
Once the high priority interrupt has finished execution, the original low priority interrupt will be tail chained and
begin execution six cycles later.
2.4.5.3 NVIC Configuration And Use
To use the NVIC we need to do three things. First configure the vector table for the interrupt sources we want to
use. Next configure the NVIC registers to enable and set the priorities of the NVIC interrupts and lastly we must
configure the peripheral and enable its interrupt support.
2.4.5.3.1 Exception Vector Table
The Cortex vector table starts at the bottom of the address range. However rather than start at zero the vector
table starts at address 0x00000004 the first four bytes are used to store the starting address of the stack pointer.
The Cortex exception
table contains the start
address or an ISR which
is
loaded
into
the
Program counter as the
CPU
enters
the
exception.
Each of the interrupt vector entries is four bytes wide and holds the start address of each service routine
associated with the interrupt. The first 15 entries are for exceptions that occur within the Cortex core. These
include the reset vector, non-maskable interrupt, fault and error management, debug exceptions and also the
SysTick timer interrupt. The Thumb-2 instruction set also includes system service call instruction which when
executed will generate an exception. The user peripheral interrupts start from entry 16 and will be linked to
peripherals as defined by the manufacturer. In software, the vector table is usually maintained in the startup by
locating the service routine addresses at the base of memory.
© Hitex (UK) Ltd.
Page 22
Chapter 2: Cortex Overview
__Vectors
AREA
EXPORT
RESET, DATA, READONLY
__Vectors
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
DCD
__initial_sp
Reset_Handler
NMI_Handler
HardFault_Handler
MemManage_Handler
BusFault_Handler
UsageFault_Handler
0
0
0
0
SVC_Handler
DebugMon_Handler
0
PendSV_Handler
SysTick_Handler
; Top of Stack
; Reset Handler
; NMI Handler
; Hard Fault Handler
; MPU Fault Handler
; Bus Fault Handler
; Usage Fault Handler
; Reserved
; Reserved
; Reserved
; Reserved
; SVCall Handler
; Debug Monitor Handler
; Reserved
; PendSV Handler
; SysTick Handler
In the case of the SysTick timer we can create a service routine by declaring a ‘C’ routine with the matching
symbolic name:
void SysTick_Handler (void)
{
….
}
Now with the vector table configured and the ISR prototype defined, we can configure the NVIC to handle the
SysTick timer interrupt. Generally we need to do two things: set the priority of the interrupt and then enable the
interrupt source. The NVIC registers are located in the system control space.
The NVIC registers are located in the Cortex-M3 System
control space and may only be accessed when the CPU is
running in privileged mode.
The Cortex internal exceptions are configured using the system control and system priority registers, while the
user peripherals are configured using the IRQ registers. The SysTick interrupt is an internal Cortex exception and
is handled in the system registers. Some of the internal exceptions are permanently enabled; these include the
reset and NMI interrupts, but also the SysTick timer, so there is no explicit action required to enable the SysTick
interrupt within the NVIC. To configure the SysTick interrupt we need to set the timer going and enable the
interrupt within the peripheral itself:
SysTickCurrent = 0x9000;
SysTickReload = 0x9000;
SysTickControl = 0x07;
//Start value for the sys Tick counter
//Reload value
//Start and enable interrupt
The priority of each of the internal Cortex exceptions can be set in the system priority registers. The Reset, NMI
and hard fault exceptions are fixed to ensure that the core will always fallback to a known exception. Each of the
© Hitex (UK) Ltd.
Page 23
Chapter 2: Cortex Overview
other exceptions has an eight bit field located in the three system priority registers. The STM32 only implements
16 levels of priority so only four bits of this field are active. However it is important to note that the priority is set by
the four most significant bits.
Each of the USER peripherals is controlled by the IRQ register blocks. Each user peripheral has an Interrupt
Enable bit. These bits are located across two 32-bit IRQ Set Enable registers. There are matching IRQ Clear
Enable registers that are used to disable an interrupt source. The NVIC also includes pending and active registers
that allow you to determine the current condition of an interrupt source.
Each interrupt source has an enable bit in the NVIC
and in the peripheral. In the STM32 there are
sixteen levels of priority.
There are sixteen priority registers. Each priority register is divided into four eight bit priority fields, each field
being assigned to an individual interrupt vector. The STM32 only uses half of this field to implement 16 levels of
priority. However, you should note that the active priority bits are in the upper nibble of each priority field. By
default the priority field defines 16 levels of priority with level zero the highest and 15 the lowest. It is also possible
to format the priority field into priority groups and subgroups. This does not provide extra levels of priority, but
helps management of priority levels when you have a large number of interrupts by programming the PRIGROUP
field in the Application Interrupt and Reset Control Register.
The PRIGROUP field
splits
the
priority
levels into groups and
subgroups. This is
useful for software
abstraction
when
dealing with a large
number of interrupts.
© Hitex (UK) Ltd.
Page 24
Chapter 2: Cortex Overview
The three bit PRIGROUP field allows you to split the 4-bit priority fields into groups and subgroups. For example,
PRIGROUP value 3 creates two groups, each with four levels of priority. In your application code you can now
define a high priority group of interrupts and a low priority group. Within each group you can specify subgroup
levels of low, medium, high and very high. As mentioned above this does not provide you anything “extra” but
provides a more abstracted view of the interrupt structure which is useful to the programmer when managing a
large number of interrupts. Configuring a peripheral interrupt is very similar to configuring an internal Cortex
exception. In the case of the ADC interrupt we must first set the interrupt vector and provide the ISR routine:
DCD
ADC_IRQHandler
;
void ADC_Handler void
{
}
Then the ADC must be initialised and the interrupt must be enabled within the peripheral and the NVIC:
ADC1->CR2
ADC1->SQR1
ADC1->SQR2
ADC1->SQR3
ADC1->CR2
= ADC_CR2;
= sequence1;
= sequence2;
= sequence3;
|= ADC_CR2;
//Switch on the ADC and continuous conversion
//Select number of channels in sequence conversion
//and select channels to convert
ADC1->CR1
= ADC_CR1; //Start regular channel group, enable ADC interrupt
GPIOB->CRH
= 0x33333333; //Set LED pins to output
//Rewrite on bit
NVIC->Enable[0] = 0x00040000; //Enable ADC interrupt
NVIC->Enable[1] = 0x00000000;
© Hitex (UK) Ltd.
Page 25
Chapter 2: Cortex Overview
2.5 Power Modes
We will have a look at the full power management options within the STM32 later. In this section we will look at
the power management modes within the Cortex core. The Cortex CPU has a sleep mode that places the Core
into its low power mode and halts execution of instructions within the Cortex CPU. A small part of the NVIC is kept
awake, so that interrupts generated from the STM32 peripherals can wake the Cortex core up.
2.5.1
Entering Low Power Mode
The Cortex core can be placed into its sleep mode by execution of either a Wait For Interrupt (WFI) or Wait For
Event (WFE) instruction. In the case of the WFI instruction, the Cortex core will resume execution and serve the
pending interrupt. Once the ISR routine has completed, there are two possibilities. Firstly, the Cortex CPU can
return from the ISR and continue execution of the background code as normal. By setting the SLEEPON EXIT bit
in the System Control Register, the Cortex core will automatically enter the sleep mode once the ISR has
completed. This allows a low power application to be entirely interrupt-driven, so that the Cortex core will wake
up, run the appropriate code and then re-enter the sleep mode with minimal code being used for power
management.
The WFE interrupt allows the Cortex core to resume execution from the point it was placed into the sleep mode. It
does not jump to a service routine. A wake-up event is simply a peripheral interrupt line that is not enabled as an
interrupt within the NVIC. This allows a peripheral to signal the Cortex core to wake up and continue processing
without the need for an interrupt service routine. The WFI and WFE instructions are not reachable from the C
language, but compilers for the Thumb-2 instruction set provide intrinsic macros that can be used inline with
standard C commands:
__WFI
__WFE
In addition to the SLEEPNOW and SLEEPONEXIT low power modes the Cortex core can issue a SLEEPDEEP
signal to the rest of the microcontroller system.
The System control register configures the Cortex processor sleep modes. The STM32 has
additional low power modes that use the DeepSleep signal which is exported from the
Cortex processor.
This allows additional functions such as the PLL and user peripherals to be halted, so that the STM32 can enter
its lowest power modes.
2.5.2
CoreSight Debug Support
All of the ARM CPUs have their own on-chip debug system. The ARM7 and ARM9 CPUs have as a minimum a
JTAG port which allows a standard debug tool to connect to the CPU and download code into the internal RAM or
FLASH memory. The JTAG port also supports basic run control (single step and setting breakpoints etc) as well
as being able to view the contents of memory locations. The ARM7 and ARM9 CPUs can also provide a real-time
trace through an additional debug peripheral called the embedded trace macro cell (ETM). While this system
works fine, it does have some limitations. The JTAG debug peripheral can only provide debug information to the
development tools when the ARM CPU is halted, so there is no possibility of real-time updates. Also, the number
of hardware breakpoints is limited to two, though the ARM7 and ARM9 instructions sets include a breakpoint
© Hitex (UK) Ltd.
Page 26
Chapter 2: Cortex Overview
instruction which can be patched into the code by the development tool (typically called soft breakpoints.)
Similarly for the real-time trace to work, the ETM must be fitted by the manufacturer at additional cost.
Consequently this is not always supported. With the new Cortex core a whole new debug system called
CoreSight has been introduced.
The Cortex CoreSight debug system uses a JTAG or serial wire
interface. CoreSight provides run control and trace functions. It
has the additional advantage that it can be kept running while
the STM32 is in a low power mode. This is a big step on from
standard JTAG debugging.
The full core sight debug system has a debug access port which allows connection to the microcontroller by a
JTAG tool. The debug tool can connect using the standard 5 pin JTAG interface or a serial 2 wire interface.
In addition to the JTAG debug features, the full CoreSight debug system contains a Data Watch trace and an
embedded trace macro cell. For software testing there is instrumentation trace and FLASH patch block. The
STM32 implements the CoreSight debug system with the omission of the embedded trace macro cell.
In practice, the CoreSight debug structure on the STM32 provides an enhanced real-time version of the standard
JTAG debug features. The STM32 CoreSight debug system provides 8 hardware breakpoints which can be nonintrusively set and cleared while the Cortex CPU is running. In addition the Data Watch trace allows you to view
the contents of memory locations non intrusively while the Cortex CPU is running. The CoreSight debug system
can stay active when the Cortex core enters a low power or sleep mode. This makes a world of difference when
debugging a low power application. Additionally the STM32 timers can be halted when the CPU is halted by the
CoreSight system. This allows you to single-step your code and keep the timers in sync. with the instructions
executing on the Cortex CPU. The CoreSight debug infrastructure significantly improves the real-time debug
capabilities of the STM32 over earlier ARM7 and ARM9 CPUs whilst still using the same low cost hardware.
© Hitex (UK) Ltd.
Page 27
Chapter 3: Getting It Working
© Hitex (UK) Ltd.
Page 28
Chapter 3: Getting It Working
3. Getting It Working
A minimal STM32 design really can be very minimal. To make the STM32 work you really just need to add a
power supply. The microcontroller contains its own internal RC oscillators and an internal reset circuit. This
section will look at the main hardware considerations you will need to address to build a practical design.
3.1 Package Types and Footprints
The STM32 Access line and Performance line variants are designed with matching package types, to allow an
easy hardware upgrade without any need to redesign the PCB. All the STM32 microcontrollers are available in
LQFP packages ranging from 48 pins up to 144 pins.
3.2 Power Supply
The STM32 requires a single power supply which must be in the range 2.0V to 3.6V. An internal regulator is used
to generate a 1.8V supply for the Cortex core. The STM32 has two other optional power supplies. The real time
clock and a small number of registers are located on a separate power domain, which can be battery-backed to
preserve data when the rest of the STM32 is placed in a deep power down state. If the design is not using battery
back up, then VBAT must be connected to VDD.
The STM32 runs from a single 2.0V-3.6V supply.
There is an additional backup power domain and a
separate supply for the ADC converter (144 pin
package only).
The second optional power supply is used to power the ADC. If the ADC is used, the main VDD power supply
range is limited to 2.4V to 3.6V. On the 100 pin package, the ADC has additional voltage reference pins VREF+ and
VREF-. The VREF- pin must be connected to VDDA and VREF+ can vary from 2.4 to VDDA. On all other packages the
voltage reference is internally connected to the ADC voltage supply pins. Each of the power supplies requires
stabilisation capacitors as shown below.
With an internal reset and an internal voltage regulator, the STM32 only needs seven external capacitors.
© Hitex (UK) Ltd.
Page 29
Chapter 3: Getting It Working
3.3 Reset Circuit
The STM32 contains an internal reset circuit that holds the device in reset as long as VDD is below 2.0V with a
hysteresis of 40mV.
The internal power on reset and power down reset
ensure the processor only runs with a stable
power supply. No external reset circuit is required.
3.3.1.1.1 Basic Hardware Schematic
Strictly speaking, an external reset circuit is not a necessary part of an STM32 design. However, during
development the nRST pin can be connected to a simple reset switch. nRST is also routed to the JTAG debug
port, so that a development tool can force a reset on the microcontroller. The STM32 has a number of internal
reset sources that can detect faulty operating conditions and we will have a look at these in the safety section
later on.
© Hitex (UK) Ltd.
Page 30
Chapter 3: Getting It Working
3.4 Oscillators
The STM32 has internal RC oscillators which are capable of supplying a clock to the internal PLL. This will allow
the microcontroller to run at its maximum 72 MHz clock frequency. The internal oscillators are not as accurate or
stable as an external crystal; consequently for most designs you will need at least one clock source.
3.4.1
High Speed External Oscillator
The main external clock source is used to derive the Cortex processor and the STM32 peripheral clocks. This
clock source is called the High Speed External (HSE Osc) Oscillator and can be a crystal/ceramic resonator or a
user provided clock source. If a user clock is selected, it can be a square, sine or triangular waveform, but it must
have a duty cycle of 50% and a maximum frequency of 25 MHz.
The External Oscillator can be run from a crystal or external clock source.
If an external crystal/ceramic resonator is used, it must be in the range 4 MHz – 16 MHz. Since the internal PLL
multiplies the HSE Osc frequency up by integer values, the external clock should be a factor of 72 MHz so that
you can easily derive the full operating frequency.
3.4.2
Low Speed External Oscillator
The STM32 can have a second external oscillator called the Low Speed External (LSE Osc) Oscillator. This is
used to clock source to the real-time clock and the windowed watchdog.
Like the HSE Osc, the LSE Osc can be an external crystal or a user-provided clock which can again have a
square, sine or triangular waveform, as long as the duty cycle is 50%. In both cases the LSE Osc should have a
frequency of 32.768 KHz, as this will provide an accurate working frequency for the real-time clock. The internal
low speed oscillator can be used to supply the real-time clock, but it is not highly accurate so if you plan to use the
RTC to any extent in your design you should fit the LSE oscillator.
3.4.3
Clock Output
One of the GPIO pins can be configured to be a Microcontroller Clock out pin. In this mode the MCO pin can
output one of four internal clock sources. We will look at this in more detail when we examine the internal clock
tree configuration.
3.4.4
Boot Pins And Field Programming
The can be started in one of three different boot modes. These modes are selected by two external boot pins
BOOT0 and BOOT1. By changing the boot mode the microcontroller will alias different areas of the memory map
to the bottom of memory. This allows us to execute code from user FLASH, internal SRAM or system memory. If
system memory is selected, the STM32 will start to execute a factory- programmed bootloader which allows the
user flash to be reprogrammed.
© Hitex (UK) Ltd.
Page 31
Chapter 3: Getting It Working
3.4.5
Boot Modes
For normal operation BOOT0 must be tied to ground. If you wish to use the other modes you must provide
jumpers to allow different settings on the two boot pins.
The external boot pins are used to select
which region of memory aliased to the
first 2k of memory. This can be user
Flash, the internal bootloader or the first
2k of SRAM
This is typically necessary if you want to do a field upgrade using the internal bootloader. If you do plan to use the
bootloader, USART1 is the default serial interface used to download code from a PC, so you will need to add an
RS232 driver chip.
3.4.6
Debug Port
Finally, we need to add the hardware debug port to allow a debugger to connect to the STM32. The Cortex
CoreSight debug system supports two connection standards: the five pin JTAG port and the 2 pin Cortex serial
wire port. Both of these configurations sacrifice GPIO pins for use by the debugger. After reset, the Cortex CPU
places these pins in their alternate function setting so that the debug port is available. If you wish to use them you
must program the alternate function registers to convert them back to GPIO pins. The five pin JTAG interface is
brought out to a 20 pin IDC type connector which has a standard pinning for all JTAG tools. The serial wire
interface uses Port A 13 for the serial data and Port A 14 for the serial clock.
© Hitex (UK) Ltd.
Page 32
Chapter 4: System Architecture
© Hitex (UK) Ltd.
Page 33
Chapter 4: System Architecture
© Hitex (UK) Ltd.
Page 34
Chapter 4: System Architecture
4. STM32 System Architecture
The STM32 is composed of the Cortex core which is connected to the FLASH memory by the dedicated
Instruction bus. The Cortex Data and System busses are connected to a matrix of ARM Advanced High Speed
Busses (AHB). The internal SRAM is connected directly to the AHB bus matrix as is the DMA unit. The
peripherals are located on two ARM Advanced Peripheral Busses (APB). Each of the APB busses is bridged onto
the AHB bus matrix. The AHB bus matrix is clocked at the same speed as the Cortex core. However, the AHB
busses have separate prescalers and may be clocked at slower speeds to conserve power. It is important to note
that APB2 can run at the full 72MHz while APB1 is limited to 36MHz. Both the Cortex and the DMA unit can be
bus masters. Because of the inherent parallelism of the bus matrix, they will only arbitrate if they are both
attempting to access the SRAM, APB1 or APB2 at the same time. However, as we will see in the DMA section,
the bus arbiter will guarantee 2/3 access time for the DMA and 1/3 for the Cortex CPU.
The internal bus structure provides a dedicated bus for program instructions and a bus matrix
which provides several data paths for the Cortex and DMA units to access the on-chip
resources.
© Hitex (UK) Ltd.
Page 35
Chapter 4: System Architecture
4.1 Memory Layout
Although the STM32 has numerous internal busses, to the programmer it offers a linear 4Gbyte address space.
As the STM32 is a Cortex-based microcontroller, the memory map must conform to the standard Cortex layout.
Therefore the program memory starts from 0x00000000. The on-chip SRAM starts from 0x20000000 and all the
internal SRAM is located in the initial bit band region. The user peripherals are memory mapped starting from
0x40000000 and are also located in the peripheral bit band region. Finally all the Cortex registers are at their
standard locations starting from 0xE0000000.
The STM32 memory map follows the Cortex
standard. The first 2K of memory is mapped from
FLASH, System Memory or SRAM depending on
the condition of the boot pins.
The FLASH memory region is composed of three sections. First is the User FLASH memory starting at
0x0000000. Next is the System Memory also called the big information block. This is 4K of FLASH memory that is
factory programmed with a bootloader. The final section from 0x1FFFF800 is called the little information block and
contains a group of option bytes that allow you to configure some system settings for the STM32. The bootloader
is designed to download code over USART1 and program it into the User FLASH memory. To place the STM32
in bootloader mode the external BOOT0 and BOOT1 pins must be held low and high respectively. With the boot
pins held in this pattern, the system memory block will appear at 0x00000000. After a reset the STM32 will begin
to execute the bootloader rather than the application code held in the User FLASH. A bootloader application for
the PC is available for download from the ST website. This program will communicate with the bootloader code
and can be used to erase and reprogram the User FLASH memory. The PC download software is also delivered
as a DLL which allows you to write custom bootloader software for field upgrading or production programming.
The bootpins also allow the internal SRAM to be mirrored at 0x00000000 in place of the User FLASH. This allows
programs under development to be downloaded and executed from the internal SRAM. This speeds the download
process and saves continual burning of the FLASH memory.
© Hitex (UK) Ltd.
Page 36
Chapter 4: System Architecture
4.2 Maximising Performance
In addition to the two external oscillators the STM32 has two internal RC oscillators. After reset the initial clock
source for the Cortex core is the High Speed Internal Oscillator which runs at a nominal 8 MHz. The second
internal oscillator is a Low Speed Oscillator running at a nominal 32.768 KHz. This oscillator is intended for the
real time clock and watchdogs.
The STM32 has a sophisticated clock tree with two external and two internal oscillators, plus a
Phase Locked Loop. The High Speed External Oscillator may also be monitored by a clock
security system.
The Cortex processor can be clocked by either the Internal or External High Speed Oscillators or from an internal
Phase Locked Loop. The Phase Locked Loop can be driven from either the Internal or External High Speed
Oscillator. So it is possible to run the STM32 at 72 MHz without an external oscillator. The downside is that the
internal oscillator is not an accurate and stable 8 MHz clock source. In order to use the serial communications
peripherals or do any accurate timing functions the external oscillator should be used. Regardless of whichever
oscillator is selected, the Phase Locked Loop must be used to derive the full72MHz clock frequency for the Cortex
core. All of the oscillator PLL and bus configuration registers are located in the Reset and Clock Control (RCC)
group.
The Reset And Clock Control unit controls the
clock tree bus bridges and backup domain.
© Hitex (UK) Ltd.
Page 37
Chapter 4: System Architecture
4.2.1
Phase Locked Loop
After reset the STM32 will derive its CPU clock from the HSI oscillator. At this point the external oscillator is
switched off. The first step in running the STM32 up to full speed is to switch on the HSE oscillator and wait for it
to stabilise
After reset the STM32 runs from the Internal High
Speed Oscillator. The External Oscillator must be
switched on.
RCC->CR |= 0x10000; //HSE on
// Wait until HSE stable
while(!(RCC->CR &0x00020000))
{
;
}
The external oscillator can be switched on in the RCC_Control register. A ready bit indicates when the external
oscillator is stable. Once the external oscillator is stable, it can be selected as the input of the PLL. The output
frequency of the PLL is defined by selecting an integer multiply value which is stored in the
RCC_PLL_configuration register. In the case of an 8 MHz oscillator, the PLL must multiply the input frequency by
9 to generate the maximum 72 MHz clock frequency. Once the PLL multiplier has been selected, the PLL can be
enabled in the control register. Once it is stable, the PLL ready bit will be set and the PLL output can be selected
as the Cortex CPU clock source.
Once the HSE oscillator is on it can be used to
supply the PLL. Once the PLL is stable it can
become the system clock
//HSE clock,PLLx9
RCC->CFGR = 0x001D0000; //Enable PLL
RCC->CR |= 0x01000000;
while(!(RCC->CR & 0x02000000))
{
;
}
//Set the remaining control fields
RCC->CR |= 0x00000001;
//Set the remaining configuration
//fields
RCC->CFGR |= 0x005D0402;
© Hitex (UK) Ltd.
Page 38
Chapter 4: System Architecture
4.2.1.1.1 Bus Configuration
Once the PLL has been selected as the system clock source the Cortex CPU will be running at 72 MHz. To get
the remainder of the chip running at its optimum speed you will need to configure the AHB and APB busses
through their bridge registers
//Enable clocks to the AHB,APB1 and APB2 busses
AHBENR
= 0x00000014;
RCC->APB2ENR = 0x00005E7D;
RCC->APB1ENR = 0x1AE64807;
RCC-
//Release peripheral reset line on APB1 and APB2 buses
RCC->APB2RSTR = 0x00000000
RCC->APB1RSTR = 0x00000000;
After rest many of the peripherals have their clock disabled and are held in reset. Before using a
peripheral enable its clock and release it from reset.
4.2.2
FLASH Buffer
When we looked at the system architecture of the STM32 we saw that the Cortex-M3 core is connected to the
internal FLASH by a dedicated I-Bus. This bus is running at the same frequency as the CPU, so with the PLL
enabled the core will be trying to run at the full 72 MHz. Since the Cortex CPU is essentially a single cycle
machine, it will be trying to access the FLASH every 1.3ns. When the STM32 starts up it is running from the
internal oscillator at 8 MHz, so the access time of the internal FLASH is not an issue. However, once the PLL is
enabled and becomes the clock source, the FLASH access time is simply too long (35 ns.) to allow the Cortex
CPU to run at maximum performance. In order to allow the Cortex CPU to run at 72 MHz with zero waitstates, the
FLASH memory has a prefetch buffer which is made up of two 64-bit buffers. Both of these buffers can do a 64bit-wide read of the FLASH memory and then pass the individual 16 or 32-bit instructions to the Cortex CPU for
execution. This technique works well with the conditional execution features of the Thumb-2 instruction set and
the branch prediction of the Cortex pipeline. During normal operation the programmer does not need to take any
special precautions because of the FLASH buffer. However you must make sure that it is enabled before
switching to the PLL as the main clock source. The FLASH buffer is controlled by the FLASH access control
register. As well as enabling the prefetch buffer, you must tune the number of waitstates required for the FLASH
prefetch buffer to read the 8 bytes of instructions from the FLASH memory. The latency settings are as follows:
0< SYSCLK <24MHz
24< SYSCLK <48MHz
48<SYSCLK <72MHz
0 waitstate
1 waitstate
2 waitstate
These waitstates are between the prefetch buffer and the FLASH memory and do not impact on the Cortex CPU.
As the CPU is executing instructions held in the first half of the buffer, the second half is loading so that the CPU
can seamlessly continue execution at its optimum rate.
4.2.3
Direct Memory Access
While the Cortex CPU can be used to move data between peripherals and the internal SRAM, many of these
operations can be automated with the internal DMA unit. The STM32 DMA unit has seven independently
configurable channels that can perform autonomous transfers from memory to memory, peripheral to memory,
memory to peripheral and peripheral to peripheral. The memory to memory transfers will be performed as fast as
the DMA channel can move the data. In the case of the peripheral transfers, the DMA unit is placed under the
control of a selected peripheral and the data is transferred on demand to or from the controlling peripheral. As
well as transferring blocks of data, each DMA unit can continually transfer data to a circular buffer. Since most
communications peripherals do not contain any FIFO buffers, the DMA units are used to stream data to and from
© Hitex (UK) Ltd.
Page 39
Chapter 4: System Architecture
buffers in SRAM. The DMA unit has been specifically designed for the STM32 and as such it is optimised for the
short but frequent data transfers that you typically find in microcontroller applications.
Each DMA memory to memory transfer is made up of four phases each 1 cycle long except for the
bus access phase that takes five cycles for each word transferred
Each DMA transfer is made up of four phases: a sample and arbitration phase, an address computation phase,
bus access phase and a final acknowledgement phase. Each phase takes a single cycle, with the exception of the
bus access phase. The bus access phase (which is where the actual data transfer takes place), takes three
cycles for each word transferred. The DMA unit and the Cortex CPU are designed to work together in an
interleaved fashion so the DMA will not block the CPU and vice versa. We will see how this works in a moment,
but it is only necessary to prioritise DMA transfers between different DMA channels. Each DMA channel is
assigned one of four priority levels by the application software. During the arbitration phase the highest priority
level will be granted the bus. If two DMA units have pending transfers and both have the same priority level, the
unit with the lowest channel number will be granted the bus.
The DMA unit is designed for fast but short data transfers of the kind found in typical small embedded systems.
The DMA unit only occupies the bus during the bus access phase.
The DMA unit can perform both the arbitration and address computation phase while another DMA Channel is in
its bus access phase. As soon as the active channel finishes its data transfer on the internal bus, the next DMA
channel transfer is ready to begin immediately while the original transfer finishes off its transfer by performing its
acknowledgement phase. So the DMA channels not only transfer data faster than the CPU, but are also tightly
interleaved and only occupy the bus for actual data transfer.
© Hitex (UK) Ltd.
Page 40
Chapter 4: System Architecture
During the bus access phase three cycles per transfer are free for the CPU. In a memory to memory transfer
this guarantees the Cortex-M3 60% of the bus even when the DMA is running continuously (remember this is
for data transfer only). The Cortex has a separate I-code bus to fetch instructions.
In the case of a memory to memory transfer each DMA channel only occupies the data bus during its bus access
phase and takes five cycles to transfer each word of data. That is, one cycle to read the data and one cycle to
write, interleaved with idle cycles where the bus is left free for the Cortex CPU. This means that the DMA units will
consume a maximum of 40% for the data bus bandwidth even during continuous maximum data transfer. In the
case of peripheral to peripheral and peripheral to memory transfers the situation is a little more complicated.
Transfers over the AHB bus take two cycles at the AHB clock frequency and transfers involving the APB take two
cycles at the APB bus frequency and a further 2 cycles at the AHB clock frequency. Each DMA transfer consists
of two bus transfer periods and a free cycle. So for example, a transfer from the SPI peripheral to the SRAM will
consist of a transfer from the SPI, plus a transfer to the SRAM plus one free cycle thus;
SPI to SRAM DMA transfer = SPI transfer (APB) + SRAM transfer (AHB) + free cycle(AHB)
= ( 2 APB cycles + 2 AHB cycles) + 2 AHB cycles + 1 AHB cycle
= 2APB Cycles + 5 AHB cycles
Remember that this only refers to data transfers as all of the Cortex program instructions are fetched on the
separate I-Bus.
The DMA unit has seven channels. Each channel is
fully orthogonal
The next good news about the DMA unit is that is very easy to use. The first thing you must remember to do is to
switch in its clock and release it from reset. This is done in the AHB clock enable register within the reset and
clock control unit.
RCC->AHBENR |= 0x00000001;
// enable the DMA clock
Once the DMA unit is powered up, each DMA channel is controlled by four registers. Two registers hold the
source and destination addresses for the peripheral register and the memory location. The size of the transfer is
held in the “number of data” register and the configuration register defines the overall characteristics of the DMA
transfer.
© Hitex (UK) Ltd.
Page 41
Chapter 4: System Architecture
Each DMA channel is controlled by four registers
and has three interrupt sources, finished, halffinished and error.
Each DMA channel can be assigned a priority of “very high”, “high”, “medium” and “low”. The transfer word size
can also be defined separately for the memory and the peripheral. For example, we can flow a 32-bit word into
the DMA channel (3 cycles) from memory and then flow four 8-bit words to the UART data register (a total of 35
cycles rather than 64 cycles if everything was moved as 8-bit quantities.) It is also possible to increment the
memory and peripheral addresses, so for example you can transfer data repeatedly from the ADC results register
but increment the memory address, so that the results are written to an array in memory for processing. The
Transfer Direction Bit allows us to specify if the flow of data is memory to peripheral or peripheral to memory. For
a memory to memory transfer we must set bit 14 to enable a “fast as you can” data transfer between two SRAM
buffers. Though you can use the DMA channels in a polled mode, each DMA channel has three interrupt sources:
transfer finished, half-finished and transfer error. Finally, once the DMA transfer is fully configured, the Channel
Enable Bit must be set and the transfer will begin. A memory to memory transfer can be performed with the
following code:
This code shows a simple DMA memory to
memory transfer and uses an internal timer to
count the number of cycles taken.
DMA_Channel1->CCR
DMA_Channel1->CPAR
DMA_Channel1->CMAR
DMA_Channel1->CNDTR
=
=
=
=
0x00007AC0;
(unsigned int)src_arry;
(unsigned int)arry_dest;
0x000A;
TIM2->CR1
= 0x00000001;
DMA_Channel1->CCR
|= 0x00000001;
while(!(DMA->ISR & 0x0000001))
{
;
}
TIM2->CR1 = 0;
TIM2->CNT = 0;
TIM2->CR1 = 1;
for(index = 0;index <0xA;index++)
{
arry_dest[index] = arry_src[index];
}
TIM2->CR1 = 0;
//configure for mem2mem transfer
//set source and destination
//set size of transfer
//start a timer
//start the DMA transfer
//wait till the transfer ends
//halt the Timer
//Clear the count
//restart timer
//repeat the operation using the CPU
//halt the timer
}
The above code transfers ten words of data between two arrays in SRAM first with the DMA and then using the
Cortex CPU. In both cases a timer is started at the beginning of the transfer and when the transfer ends. In this
example the DMA unit takes 220 cycles and the CPU takes 536.
© Hitex (UK) Ltd.
Page 42
Chapter 4: System Architecture
Each peripheral with DMA support is assigned to a specific channel. When enabled, the peripheral
becomes the flow controller for the DMA transfer. This allows it to sink or source data as it requires
without taking any CPU cycles.
While memory to memory transfers can be useful for initialising regions of memory and performing block
transfers, most of the time the DMA channels will be used to move data between memory and the various user
peripherals. In this case, each of the DMA channels is mapped to a selection of peripherals. First we must
initialise the peripheral and enable its DMA support, then we must configure the matched DMA channel to transfer
data on request of the supported peripheral. We will have a more detailed look at the ADC later, but in a simple
conversion mode it can perform continuous 10-bit conversions to a single results register. Without the DMA the
Cortex CPU would be continually responding to ADC conversion complete interrupts and these would start to
steal useful amounts of CPU runtime. However, by using the DMA the ADC can request a DMA transfer at the
end of each conversion. The DMA will transfer the ADC results data to an incrementing address in the SRAM.
The ADC data can then be processed once a suitable sample of data has been transferred.
ADC dma pics
The STM32 peripherals do not
contain any internal buffering. By
using the DMA in its circular mode,
any size of memory can be used as a
peripheral buffer. This can be
combined with the DMA’s halffinished and finished interrupts to
provide a circular double buffer.
To make this process more efficient, we can enable the circular buffer support so that the ADC data will
continuously write to our buffer. Then, by using the half complete and transfer complete interrupts, we can create
a double buffer. So when the first half of the buffer is full, an interrupt will be generated and we can process this
data while the DMA continues to fill the second half. Once the second half is full, we can process this data while
the DMA starts to refill the buffer from the top. All the other peripherals with DMA support are handled in a similar
way. It should be noted that the communication peripherals have separate transmit and receive DMA channels.
For example, the SPI can simultaneously flow data in both directions.
© Hitex (UK) Ltd.
Page 43
Chapter 5: Peripherals
© Hitex (UK) Ltd.
Page 44
Chapter 5: Peripherals
5. Peripherals
This chapter will present an introduction to the user peripherals on the existing STM32 variants. For convenience,
these are split into two groups: general purpose microcontroller peripherals and communications peripherals. All
of the peripherals on the STM32 have a high level of sophistication and are tightly integrated with the DMA unit.
Each peripheral has some form of extended hardware functionality, which can be useful in minimising the amount
of CPU time required to drive a given peripheral. In other words, there are lots of clever features that can help you
to automate the hardware, reducing the CPU effort in driving the peripherals.
5.1 General Purpose Peripherals
The general purpose peripherals on the STM32 consist of: general purpose IO, external interrupt controller,
analogue to digital converters, general purpose and advanced timer units, and real-time clock with backup
registers and tamper pin.
5.1.1
General Purpose IO
The STM32 is well served with general purpose IO pins, having up to 80 bidirectional IO pins. The IO pins are
arranged as five ports each having 16 IO lines.
Each digital pin may be configured as GPIO or as
an alternate function. Each pin can simultaneously
be configured as one of 16 external interrupt lines.
These ports are names A-E and are all 5v tolerant. Many of the external pins may be switched from general
purpose IO to serve as the Input/Output of a user peripheral, for example a USART or I2C peripheral. Additionally
there is an external interrupt unit which allows 16 external interrupt lines to be mapped onto any combination of
GPIO lines.
Each GPIO port can configure individual pins as
input or output with different driver configurations.
It has registers to write word-wide or for atomic bit
manipulation. Once a configuration is defined it
can be locked.
Each GPIO port has two 32-bit wide configuration registers; these two registers combine to give a 64-bit wide
configuration register. Within these 64 bits each pin has a four bit field that allows its characteristics to be defined.
The four bit configuration field is made up of a two bit wide mode field and a two bit wide configuration field. The
© Hitex (UK) Ltd.
Page 45
Chapter 5: Peripherals
mode field allows the user to define the pin as an input or an output, while the configuration field defines the drive
characteristics:
As well as being able to define a port pin as an input or output, its drive characteristics may also be selected. In
the case of an input, an internal resistor can be connected as a pull up or pull down resistor. For an output, each
port pin may be configured with a push pull or an open drain driver. Each output pin can also be configured with a
maximum output speed of 2MHz,10MHz or 50MHz.
Once the port configuration has been set, these parameters can be protected by writing to the configuration lock
register. In this register each pin has a lock bit which when set will prevent any writes to the matching
configuration and mode fields. When all of the required lock bits have been set, the lock can be activated by
writing a sequence of 1,0,1 to bit 16 in the lock register, followed by two reads of the same bit which will return 0,1
if the lock has been successfully activated. The input and output data registers allow port wide access to the IO
pins. Atomic bit manipulation is supported by either using the Cortex bit banding technique on the input and
output data registers, or through two dedicated bit manipulation registers. The bit set/reset register is a 32-bit wide
register. The upper 16 bits are mapped to each port pin. Writing a logic 1 to these locations will reset the matching
port pin. Similarly, writing a logic 1 to any of the lower 16 bits will set the matching port pin. The second bit
manipulation register is a bit reset register. This is a 16-bit wide register where writing logic 1 in the lower 16 bits
will reset the matching port pin. The combination of port registers, bit banding and atomic bit manipulation
registers allows you very fine control of all the STM32 port pins and can be used very efficiently for IO intensive
applications.
5.1.1.1 Alternate Functions
The alternate function registers allow you to remap the port pins from GPIO to alternate peripheral functions. To
allow flexibility in hardware design, a given peripheral function can be mapped to one of several pins.
A late-arriving high priority interrupt will pre-empt
a low priority interrupt without incurring an
additional stacking overhead.
The STM32 alternate functions are controlled in the remap and debug IO register. Each of the digital user
peripherals (USART,CAN, timers, I2C and SPI) has a 1 or two bit field which allows mapping to several different
pin combinations. Once the alternate function pins have been selected, the GPIO configuration registers must be
used to switch from IO to alternate function. The remap register also controls the configuration of the JTAG debug
© Hitex (UK) Ltd.
Page 46
Chapter 5: Peripherals
pins. After reset, the JTAG port is enabled with the data trace disabled. The JTAG can be switched to serial wire
(two pin debug) or disabled, the unused pins in each case can be used as GPIO.
5.1.1.2 Event Out
The Cortex processor can generate an event out pulse that is intended to wake up a separate microcontroller
from a low power mode. Typically, the event out pulse would be connected to the wake up pin of a second
STM32. The event out pulse is generated by executing the SEV Thumb-2 instruction. In the STM32 the event
control register is used to route the event out pulse to a selected GPI pin. The event control register contains
fields to select the port and pin within the port. Once the port pin is selected, the event out enable bit is set to
finish the configuration.
5.1.2
External Interrupts
The external interrupt unit has 19 interrupt lines that are connected to interrupt vectors via the NVIC. Sixteen of
these interrupt lines are connected to GPIO pins and can generate an interrupt on a rising or falling edge, or both.
The remaining three exit lines are connected to the RTC alarm interrupt, the USB wake up and the Power voltage
detect unit. The NVIC provides individual interrupt vectors for EXTI lines 0-4, the RTC alarm, Power voltage
detect and the USB wake up. The remaining EXTI lines are connected in groups of lines 5-9 and lines 10 – 15 to
two additional interrupt vectors. The EXTI is important for power control on the STM32. As it is not a clocked
peripheral, it can be used to wake up the microcontroller from its STOP mode where both the main oscillators are
halted. The EXTI can generate both interrupts to exit Wait for interrupt mode and events to exit Wait for event
mode.
The STM32 has 16 external
interrupt lines which may be
connected to each of the port
pins.
The 16 EXTI lines dedicated to the GPIO pins can be mapped to any combination of port pins. This is done
through four configuration registers. In these registers each EXTI line is mapped to a four bit field. This field
allows each EXTI line to be mapped onto any of the five IO ports, so for example EXTI line zero can be mapped
onto pin 0 of port A, B, C, D or E. This scheme allows any external pin to be mapped to an interrupt line. The
EXTI can also be used in conjunction with an alternate function that has been remapped to an external pin.
//Map the external interrupts to port pins
AFIO->EXTICR[0] = 0x00000000;
//Enable External interrupt sources
EXTI->IMR
= 0x00000001;
//Enable wake up events
EXTI->EMR
= 0x00000000;
//Select falling edge trigger sources
EXTI->FTSR
= 0x00000001;
//Select Rising edge trigger sources
EXTI->RTSR
= 0x00000000;
//Enable interrupt sources in the NVIC
NVIC->Enable[0] = 0x00000040;
NVIC->Enable[1] = 0x00000000;
Once mapped the external interrupt pins can generate an interrupt on a rising and/or falling edge.
© Hitex (UK) Ltd.
Page 47
Chapter 5: Peripherals
Once the EXTI configuration registers have been set, each external interrupt can be configured to generate an
interrupt or event on rising or falling edges. It is also possible to force an EXTI interrupt by writing to the matching
bit in the software interrupt register.
5.1.3
ADC
The STM32 features up to two independent analogue to digital converters, depending on variant. The ADC has
an independent supply which can be between 2.4V to 3.6V, depending on the package type. The ADC reference
is connected internally to the ADC supply, or brought out to a dedicated pin. The ADC converters offer a 12 bit
resolution with a 1 MHz conversion rate. With up to 18 multiplexed channels, 16 can be available to measure
external signals. Of the remaining two one is connected to an internal temperature sensor and the second is
connected to an internal reference voltage.
The STM32 is a highly featured 12 bit 1MHz sample rate converter with internal band gap and
temperature sensor.
5.1.3.1 Conversion Time And Conversion Groups
When the ADC is configured, it is also possible to individually program the conversion time for each channel.
There are eight discrete conversion times ranging from 1.5 cycles to 239.5 cycles.
Each
ADC
channel
may
be
configured with an individual sample
rate.
Each ADC has two basic conversion modes: regular and injected. In regular mode conversion allows you to
specify a channel or group of channels to be converted on a round robin basis. The regular conversion group can
be configured to have up to 16 channels. Additionally, the order in which the channels are converted can also be
© Hitex (UK) Ltd.
Page 48
Chapter 5: Peripherals
programmed and in one conversion cycle a channel can be converted several times. The regular group
conversion can be started by software, or by a hardware event from a range of timer signals or EXTI line 1. Once
triggered, the regular group can perform continuous conversions. Alternatively it can operate in a discontinuous
mode, whereby a selected number of channels are converted and then conversion halts until the next regular
group trigger occurs.
The regular group conversion sequence can run in a continuous round robin cycle. Or, in discontinuous
mode it can convert a selected number of channels after each trigger event.
Each time a regular group conversion is made, the result is stored in a single results register and an interrupt can
be generated. The 12-bit result is stored in a 16-bit wide register and the converted value can be aligned left or
right.
The 12-bit result can be left or
right aligned in the 16-bit
results register.
ADC1 has a dedicated DMA channel that can be used to transfer each conversion value from the results register
to a buffer in memory. Using this method a regular group conversion cycle can be copied to memory, with a single
DMA interrupt being generated at the end of the group conversion cycle. If you want to be clever, you can make
the memory buffer double the size of the regular group conversion cycle and use the DMA half-finished and
finished interrupts to create a double buffer. This can be combined with the DMA circular buffer mode to place a
lot of ADC results-handling under hardware control.
ADC1 has DMA support
which can automatically
transfer results to a userdefined buffer in SRAM.
The second conversion group is called the injected group. The injected group is a conversion sequence of up to
four channels that can be triggered by a software or hardware event. Once triggered, it will halt the regular
conversion group, perform its sequence of conversions and then allow the regular group to continue. Like the
regular group, any sequence of channels can be configured and a channel can be converted more than once in a
conversion sequence. However, unlike the regular group, each injected conversion has its own results register
and an offset register.
The injected group results
registers are sign-extended
and can be left or right
justified.
© Hitex (UK) Ltd.
Page 49
Chapter 5: Peripherals
The offset register can be programmed with a 16-bit value that will automatically be deducted from the ADC
result. If the result is a negative value, the injected group results register is sign-extended. Like the regular group
the result can be left or right justified.
5.1.3.2 Analogue Watchdog
In addition to the two conversion modes, the ADC has an analogue watchdog. This watchdog can be
programmed with high and low threshold values to detect over or under voltage conditions. Once triggered, the
analogue watchdog can generate an interrupt. The analogue watchdog can be used to monitor a selected regular
and injected channel, or all injected and regular channels. In addition to voltage monitoring, the analogue
watchdog could be used as a zero voltage crossing detector.
The
analogue
watchdog
can
monitor a channel or
all channels for a
user defined high
and low threshold.
5.1.3.3 Basic ADC Configuration
The ADC registers breakdown into six groups, with the status
and control registers defining the operating configuration of the
ADC.
The ADC has register blocks to configure: the individual sample time, regular and injected conversion sequences
along with the injected group offset values and watchdog threshold values. The overall ADC configuration is
through the status and control registers.
© Hitex (UK) Ltd.
Page 50
Chapter 5: Peripherals
The two control registers define the ADC operating mode. A single channel interrupt driven conversion is
shown below.
ADC1->CR2
=
ADC1->SQR1 =
ADC1->SQR2 =
ADC1->SQR3 =
ADC1->CR2 |=
ADC1->CR1
0x005E7003;
0x0000;
0x0000;
0x0001;
0x005E7003;
=0x000100;
//Switch on the ADC and enable continuous conversion
//set sequence length to one
//select conversion on channel zero
//rewrite on bit
//Start conversion
//interrupt
NVIC->Enable[0] = 0x00040000;
NVIC->Enable[1] = 0x00000000;
of
regular
channels,
enable
//enable ADC interrupt
In the ADC interrupt read the result register and copy the conversion result to a bank of port pins.
void ADC_IRQHandler (void)
{
GPIOB->ODR = ADC1->DR<<5; // Copy ADC result to port pins
}
If the interrupt is not used a DMA channel can be used to transfer the ADC result directly to the port pins.
DMA_Channel1->CCR
= 0x00003A28; //Circular mode,
//peripheral and memory increment disabled
//Load destination address into peripheral register,GPIO port data register
DMA_Channel1->CPAR = (unsigned int) 0x4001244C;
//Load source address into memory register
DMA_Channel1->CMAR = (unsigned int) 0x40010C0C;
DMA_Channel1->CNDTR = 0x1;
//Load number of words to transfer
DMA_Channel1->CCR |= 0x00000001;//Enable the DMA transfer
In the ADC the DMA support must be enabled.
ADC1->CR2
© Hitex (UK) Ltd.
|= 0x0100;
Page 51
ADC
Chapter 5: Peripherals
5.1.3.4 Dual Conversion Modes
For a low cost general purpose microcontroller, the ADC on the STM32 is very sophisticated. You should take
time to understand all of its features, as the ADC hardware can be configured to perform some operations that a
more basic ADC would require additional software intervention to achieve. If all this isn’t enough, on the STM32
variants with two ADC converters there are additional dual conversion modes.
The ADC dual conversion modes
synchronise operation of the two
converters providing eight additional
modes.
In the dual conversion modes ADC2 is slaved to ADC1 allowing eight additional conversion modes.
5.1.3.4.1 Injected Simultaneous Mode And Regular Simultaneous Modes
The first two dual conversion modes synchronise conversion of the regular and injected conversion groups on the
two ADC converters. This is very useful if you need to measure two quantities such as voltage and current
simultaneously.
© Hitex (UK) Ltd.
Page 52
Chapter 5: Peripherals
5.1.3.5 Combined Regular/Injected Simultaneous Mode
A further combined mode allows both the regular and injected groups on both ADCs to have their conversion
sequences synchronised.
5.1.3.6 Fast Interleave And Slow Interleave Modes
The fast and slow interleave modes synchronise the conversion of both ADC regular conversion groups, but
unlike the simultaneous mode there is a delay between the start of conversion on ADC 1. In fast interleave mode
this is seven ADC clock cycles after the start of ADC2 conversion. In slow interleave mode the delay is 14 ADC
clock cycles. Both these modes can be used to increase the overall sampling rate by combining the two
converters.
5.1.3.7 Alternate Trigger Mode
In alternate trigger mode a hardware trigger on ADC1 will first trigger an injected group conversion on ADC1. The
next trigger will start a conversion on the injected group of ADC2.
© Hitex (UK) Ltd.
Page 53
Chapter 5: Peripherals
5.1.3.8 Combined Regular Simultaneous And Alternate Trigger Mode
The alternate trigger mode can also be combined with the regular group simultaneous mode. This synchronises
the regular conversions on both ADCs and makes alternate conversions on the two injected groups.
5.1.3.9 Combined Injected Simultaneous And Interleaved Mode
The final conversion mode performs an interleaved conversion on the two ADC regular groups but performs
synchronised simultaneous conversion for the two injected groups.
5.1.4
General Purpose and Advanced Timers
The STM32 has four timer units. Timer 1 is an advanced timer intended for motor control. The remaining timers
are general purpose timer units. All of the timers have a common architecture; the advanced timer simply has
additional hardware features. In this section we will look at the general purpose timers first and then move on to
the advanced timer.
© Hitex (UK) Ltd.
Page 54
Chapter 5: Peripherals
5.1.4.1 General Purpose Timers
All of the timer units are based on a 16-bit counter with a 16-bit prescaler and auto-reload register. The timer
counter can be configured as count up, count down or centred counting (count up then count down). The clock
input to the timer counter can be selected from eight different sources. These include: a dedicated clock
generated from the main system clock, a trigger out clock from one of the other timers or an external clock
through the capture compare pins. The timer trigger inputs and the external clock sources have a gated input to
the timer counter which is controlled by an external trigger pin ETR.
The four timers within the STM32 have a 16-bit counter with 16-bit prescaler and a four
channel capture compare unit. They may be clocked from the system clock, external events
or other timers.
In addition to the basic timer counter, each timer unit has a four channel capture compare unit. This unit can
perform simple capture and compare functions but also has a number of special modes that allow common
operations to be performed in hardware. Each of the timers has both interrupt and DMA support.
5.1.4.1.1 Capture Compare Unit
Each capture compare channel is controlled by a single register. This register has different functions, depending
on the setting of the selection bits. In capture mode it has input filters and a special PWM measurement mode,
plus support for encoder inputs. In compare mode it has standard compare functions and a PWM generation
option plus a one pulse mode.
Each Capture/Compare channel has a single mode register. The
Capture compare selection bits define the operating mode of the Cap
Com channel
© Hitex (UK) Ltd.
Page 55
Chapter 5: Peripherals
5.1.4.1.2 Capture Unit
The basic capture unit has four input channels connected to configurable edge detectors. When a rising or falling
edge is detected, the current timer count is captured into the channel’s 16-bit capture/compare register. When a
capture event occurs, the timer counter can be reset or halted. In addition, an interrupt or DMA transfer can be
triggered.
The four capture units
have input filter and edge
detection. A capture event
can trigger an interrupt or
DMA transfer.
5.1.4.1.3 PWM Input Mode
The capture unit can also be configured to use two capture channels to automatically measure an external PWM
signal; both duty cycle and period can be measured.
In PWM measurement mode two
channels are used to automatically
capture the period and duty cycle of the
PWM waveform.
M3->CR1
TIM3->PSC
TIM3->ARR
TIM3->CCMR1
TIM3->CCER
TIM3->CCMR1
TIM3->CCER
TIM3->SMCR
= 0x00000000;
= 0x000000FF;
= 0x00000FFF;
= 0x00000001;
|=0x00000000;
|=0x00000200;
|=0x00000020;
= 0x00000054;
TIM3->CCER
TIM3->CR1
|=0x00000011;
= 0x00000001;
//default
//set max prescaler
//set max reload count
//Input IC1 mapped to TI1
//IC1 triggers on rising edge
//Input IC2 mapped to TI1
//IC2 triggers on falling edge
//Select TI1FP1 as input,rising edge trigger
//resets the counter
//enable capture channels
//enable the timer
In PWM mode the input signal is routed to two capture channels. At the beginning of a PWM cycle the main
counter is reset by capture channel 2 (using the rising edge of the PWM signal) and it will start counting up. On
the falling edge of the PWM signal capture channel one is triggered, capturing the duty cycle value. On the next
rising edge at the beginning of the next cycle capture 2 is again triggered, resetting the timer and capturing the
PWM period value.
© Hitex (UK) Ltd.
Page 56
Chapter 5: Peripherals
5.1.4.1.4 Encoder Interface
The capture unit on all of the timers are also designed to interface directly to an external encoder. A typical
application of such an encoder is used to detect speed and angular position of a motor.
Each timer may be interfaced to a
linear or rotary encoder to capture
position,
speed
and
direction
information.
In this configuration the capture pins are providing the clock count to the timer counter. The count is then used to
determine position. To obtain speed information a second timer must be used. This will provide a time base so
that we can detect the number of encoder ticks that have occurred in a given period.
5.1.4.1.5 Output Compare
Each of the STM32 timer units also provides four channels of output compare. Using the basic compare mode,
when the timer count matches the 16-bit value stored in the channel capture/compare register, a capture event is
generated. This capture event can be used to: modify the state of the associated capture/compare channel pin,
generate a timer reset, an interrupt or a DMA transfer.
In compare mode, each channel can
be used to generate an interrupt, or
change the state of the cap/com pin,
when contents of the compare
register match the timer count.
© Hitex (UK) Ltd.
Page 57
Chapter 5: Peripherals
5.1.4.1.6 PWM Mode
In addition to the basic compare mode, each timer has a dedicated PWM generation mode. In this mode the
PWM period is defined by the value stored in the timer auto reload register. The duty cycle is defined by the value
stored in the channel capture/compare register. This allows each timer to generate up to four independent PWM
signals. As we will see later the timers can be synchronised together so it is possible to generate up to 16
synchronised PWM signals.
Each timer has a dedicated PWM mode that can generate edge or centre-aligned PWM waveforms.
Each channel can generate either an edge-aligned or centre-aligned PWM signal. In edge-aligned mode, the
falling edge always coincides with the timer reload update event. Changing the capture/compare value simply
modulates the rising edge of the PWM signal. In centre-aligned mode, the timer is configured as a centre counter
(count up then count down). When a match is made with the capture/compare channel, the channel output pin is
inverted.
TIM2->CR1
TIM2->PSC
TIM2->ARR
TIM2->CCMR1
TIM2->CCR1
TIM2->CCER
TIM2->DIER
TIM2->EGR
TIM2->CR1
=
=
=
=
=
=
=
=
=
0x00000000;
0x000000FF;
0x00000FFF;
0x00000068;
0x000000FF;
0x00000101;
0x00000000;
0x00000001;
0x00000001;
//default
//set max prescaler
//set max reload count
//Set PWM mode
//Set PWM start value
//Enable CH1 output
//enable update interrupt
//enable update
//enable timer
5.1.4.1.7 One Pulse Mode
In the basic compare and PWM mode the timer units will generate a continuous output waveform. Each of the
timers also has a one pulse mode option. This is really a special case of the PWM mode, where an external
trigger (an external pin or another timer trigger output) can start the PWM mode running for one cycle. This
generates a single pulse with a programmable start delay and pulse width.
The one pulse mode allows you to define
a single shot pulse with configurable
delay and duration.
5.1.4.2 Advanced Timer
The advanced timer is timer unit 1. This timer contains additional hardware specifically intended for motor control.
Three of the advanced timer channel output pins have complementary outputs. This provides a six channel PWM
unit. As this unit is intended for three phase motor control, each channel has programmable dead time and there
is a global break input line. There is also a Hall sensor interface in addition to the encoder interface.
© Hitex (UK) Ltd.
Page 58
Chapter 5: Peripherals
The advanced timer has the same basic structure
as the general purpose timers. Three of the
compare channels have complementary outputs.
There is an additional break input and Hall
sensor interface.
Each of the three complementary PWM channels has a programmable dead time which places a delay between
the switch-off of a PWM output and the switching on of its complementary channel.
The advanced timer complementary PWM
programmable dead time for motor control.
outputs
have
5.1.4.2.1 Break Function
The advanced timer can place its PWM outputs and their complementary outputs into a predefined configuration
in response to a break input. This input can be from a dedicated external break pin or from the clock security
system which monitors the external high speed oscillator. Once enabled, the break function operates entirely in
hardware and guarantees to place the PWM outputs into a safe state if the STM32 system clock fails, or if there is
a fault on the external hardware.
5.1.4.2.2 Hall Sensor Interface
Each of the timers including the advanced timer is designed to easily interface with a Hall Sensor to allow easy
measurement of angular motor speed. The first three capture pins of each timer can be connected to channel 1
through an XOR gate. As the motor rotates and passes each sensor, a capture event will be generated on
channel one. This captures the current timer count into the channel one capture register and also causes a reset
on the timer. Thus the count value in the capture register can be related back to the motor speed.
© Hitex (UK) Ltd.
Page 59
Chapter 5: Peripherals
5.1.4.3 Timer Synchronisation
Although each of the timer units is a completely independent timer, any or all of the timer units can be
synchronised together. This allows complex timer arrays to be designed in hardware, reducing the amount of
software overhead required to perform a complex time-based function.
Each of the timers has trigger inputs from the other three timers as well as external
inputs from the cap/com pins.
Each of the timer units has a trigger output which is routed as an input to the other three timers. Additionally, a
capture input pin from timer 1 and timer 2 (TI1FP1 and TI2FP2) is routed to the trigger controller of each timer
block. The timers may be synchronised in several different modes. The examples below show a couple of typical
configurations.
All of the timers may be cascaded in
a highly configurable fashion to build
complex timer arrays.
One timer acts as a master with two slaves. The master can provide a clock to the two slave timers, creating one
large timer. Alternatively, it can provide a time delay that is used to trigger or gate the two slaves. Similarly, an
external trigger can be used to gate the activity of each timer.
5.1.5
RTC And Backup Registers
The STM32 contains two power domains: the main STM32 system and peripheral power domain and the backup
domain. Located within the backup domain are ten 16-bit wide registers, the RTC and the independent watchdog.
The backup registers are simply ten memory locations that can be used to hold critical data values during the
STM32 standby mode when the main power domain is switched off. In the low power modes both the RTC and
© Hitex (UK) Ltd.
Page 60
Chapter 5: Peripherals
the independent watchdog can be kept running and may be used to wake up the main STM32 system, or to
perform a chip reset.
The STM32 contains a basic real-time clock. This is a 32- bit counter that is optimised to increment each second if
clocked from a 32.768 KHz clock source. When configuring the clock tree, the RTC oscillator can be selected
from: the low speed internal oscillator, the low speed external oscillator, or the high speed external oscillator via a
fixed divide by 128 prescaler. A further RTC prescaler allows you to get an accurate seconds count. The RTC
counter itself can generate three interrupts: a seconds increment, a counter overflow and an alarm interrupt. The
alarm interrupt occurs when the RTC counter reaches the value stored in a matching alarm register.
The real-time clock may use the
internal or external low speed
oscillator. It provides a seconds
counter with an alarm register.
RTC is located in the backup power domain which is powered by the VBAT voltage supply and the alarm interrupt
is also connected to EXTI line 17. This means that when the main power domain of the STM32 is placed in a low
power mode, the RTC will keep running. Through the EXTI it can generate an interrupt of events on the Cortex
NVIC to wake up the main STM32 power domain. This configuration of the RTC is crucial for low power designs
that need to spend most of their time in stop mode, but need some method of auto wakeup.
5.1.6
Backup Registers And Tamper Pin
The backup power domain also contains ten 16-bit registers which act as battery-backed SRAM. The data held in
these registers can be cleared by writing to the RCC backup control register. An external tamper pin can also be
enabled in the same register. This pin can be configured to be high or low at startup. During normal operation a
change in logic level will trigger a tamper detect event, which will clear the backup registers. A tamper interrupt
can also be enabled, which allows the application software to take defensive action if a tamper condition is
detected.
© Hitex (UK) Ltd.
Page 61
Chapter 5: Peripherals
5.2 Connectivity
As well as having an excellent set of general purpose peripherals, the STM32 has five different types of
communication peripheral. For communication between integrated circuits on the same PCB, the STM32 has SPI
and I2C interfaces. For communication between different modules, there is a CAN bus interface and for
communication to a PC, there is a USB device interface. Finally, there is the ever popular USART.
5.2.1
SPI
For fast communication between integrated circuits, the STM32 provides two SPI peripherals which can provide
full duplex communication at rates up to 18 MHz. It is important to note that one SPI peripheral is located on the
APB2 high speed peripheral bus, which can run at speeds up to 72 MHz. The second is on the low speed APB1,
which can run at speeds up to 37 MHz. Each SPI peripheral has programmable clock polarity and phase and the
data can be transmitted as 8 or 16-bit words, MSB first or LSB first. This allows each SPI peripheral to be
configured as a master or slave, which can communicate with any other SPI device available.
Each SPI peripheral can operate as
master or slave up to 18 MHz. Two
DMA channels are provided for
efficient data transfer
To support high data communication rates, each SPI peripheral has two DMA channels: one for transmitting data
and one to read received data to memory. Using the DMA support allows high speed streaming of bi-directional
data under hardware control. In addition to the standard SPI peripheral features, the STM32 SPI peripheral
contains two hardware CRC units. One CRC unit is used for transmitted data and one for reception. Both units
can generate and check CRC8 and CRC16 codes. This feature is particularly useful if you want to use either SPI
peripheral as an interface to an MMC/SD card.
The SPI peripheral contains a
hardware CRC unit which is designed
to
support
interfacing
with
multimedia and SD memory cards.
© Hitex (UK) Ltd.
Page 62
Chapter 5: Peripherals
5.2.2
I2C
The STM32 can also communicate to other integrated circuits through a dedicated I2C interface. The I2C
interface is capable of operating as a slave or bus master and is also capable of bus arbitration in a multi-master
system. The I2C interface will support standard bus speeds of up to 100 kHz and fast speeds of up to 400 Khz.
The peripheral also supports the I2C seven and ten bit addressing modes. The I2C peripheral is designed to
simply read and write I2C data to and from the bus. Your software must control the I2C ‘engine’ to provide the
protocol necessary for communication with various different bus devices. The I2C peripheral provides two
interrupts to the Cortex processor: one for error containment and the other to control the communication address
and data transmission. In addition, the DMA unit provides two DMA channels which can read and write data to the
I2C transmit buffer. Thus, once the initial address and data transfer negotiation has been done, data can be
streamed to and from the STM32 under hardware control.
The two I2C peripherals have
enhanced support for the
system management bus and
the power management bus.
They
include
hardwarepacked error correction.
All of the above features make the STM32 I2C peripheral a fast and efficient bus interface. However there are
some additional enhanced features that extend the basic I2C functionality. The STM32 I2 C peripheral contains
hardware packet error checking (PEC). When enabled, the PEC will generate an 8-bit CRC error detection byte.
This byte is automatically placed at the end of the transmitted data stream. The PEC will also error check
received data against the PEC error protection byte.
The STM32 I2C peripheral is also designed to support two further communication protocols. These two protocols
are System Management Bus (SMBus) and Power Management Bus (PMB). System Management Bus is a
protocol defined by Intel in 1995 for use within PCs and servers. System Management Bus defines a data link
layer which includes the use of the PEC and an additional networking layer standardising configuration
communication between the PC BIOS and different manufacturers’ devices. When operating in SMBus mode, the
I2C peripheral has additional support for some SMBus features in addition to the PEC. These include support for
the SMN address resolution protocol, host notify protocol and the SMBALERT signal. The Power Management
Bus protocol is a version of System Management Bus designed for use within power conversion systems. PMBus
is intended to allow configuration, programming and real-time monitoring of power systems.
© Hitex (UK) Ltd.
Page 63
Chapter 5: Peripherals
5.2.3
USART
Although serial communication ports have largely disappeared from PCs, they are still widely used in many
embedded applications as a simple serial communication interface. Because of their utility and ease of use they
will be with us for many years to come. The STM32 has up to three USARTs, each with several enhanced
operating modes which support the latest serial communications applications. Each of the three USARTs are
capable of up to 4.5 Mbps communication. Each USART has a fully programmable serial interface with
programmable data size (8 or 9 bits), parity stop bit and baud rate. One USART is located on the APB2 bus which
runs at up to 72 MHz, while the others are located on APB1 which runs at 36MHz.
The
USARTs
are
capable of supporting
asynchronous
communication
with
UARTS and modems as
well as LIN IrDA and
smart cards.
The baud rate generator on each USART is a fractional BAUD rate generator. This is more sophisticated than a
simple clock divider and allows standard BAUD rates to be derived from any bus frequency. Like the other serial
communications peripherals, each USART has two DMA channels which are used to transfer Tx and Rx data to
and from memory. When used as a UART, the USART supports a number of special communication modes. The
USART is capable of single wire half-duplex mode communication, using just the Tx pin. For modem
communication and hardware flow control each USART has additional CTS and RTS control lines.
The USART’s support single wire
half-duplex communication.
Each USART can also be used for local interconnect bus (LIN). This is an automotive standard for
low cost networking to a cluster of microcontrollers. Each USART can be used as a serial infra red
(SIR) encoder/decoder. This conforms to the IrDA standard for infra red communication for bit rate up
to 115200bps, using half-duplex NRZ modulation with low power operation when the USART is
clocked between 1.4MHz and 2.12MHz. Each USART has an additional smart card mode which
conforms to the ISO 7618-3 standard.
The USARTS can support smartcard and IrDA communication.
© Hitex (UK) Ltd.
Page 64
Chapter 5: Peripherals
In addition to high speed UART type operation, the USART can be configured for synchronous communication
which allows a three wire connection to SPI peripherals. When in this mode, the USART acts as an SPI master
and has programmable clock polarity and phase so it can communicate with any SPI slave.
In synchronous operation the USARTS can be used
as additional SPI masters.
5.3 Can And USB Controller
The two remaining communication peripherals on the STM32 are the CAN controller and the USB full speed
device interface. Both of these communication protocols are quite complicated. If you are new to either of these
protocols you should read the CAN tutorial and/or the USB tutorial that comes with this book. Both of the USB and
CAN peripherals require a relatively large amount of SRAM for message-filtering message buffers. The STM32
has a dedicated 512 byte region of SRAM that is shared between the CAN and USB peripherals. This memory
can only be accessed by these peripherals. It is also assigned exclusively to the CAN peripheral or the USB
peripheral. This means that you cannot use the CAN peripheral and the USB peripheral simultaneously, although
it is possible to switch from one to the other in the same application.
5.3.1
CAN Controller
The STM32 CAN controller is a fully-featured CAN node that supports CAB 2.0A and 2.0B active and passive with
data rates up to the maximum 1 Mbit/s. The CAN controller also has extensions to support fully deterministic
communication defined under the time-triggered CAN protocol TTCAN. When enabled, the TTCAN extensions
support automatic message retransmission and will place a message timestamp in the last two data bytes of the
CAN message packet. When enabled, these extensions allow the application software to use the CAN peripheral
for hard real-time control.
The CAN peripheral supports CAN
2.0B with extensions for timetriggered CAN.
The full name of the CAN controller is the bxCAN peripheral, where the bx stands for basic extended. A basic
CAN peripheral is defined as having a single transmit and receive buffer, whereas an extended CAN peripheral
has multiple transmit and receive buffers. The bx CAN peripheral is a hybrid of the two CAN peripheral
architectures. The bxCAN peripheral has three transmit mailboxes and two receive mailboxes. Each of the
receive mailboxes has a FIFO queue three messages deep. This design is a trade-off between having a low
performance CAN module with a small silicon footprint and a high performance module that takes a large amount
of the die area.
© Hitex (UK) Ltd.
Page 65
Chapter 5: Peripherals
The CAN peripheral has
three transmit mailboxes
with
automatic
time
stamping for TTCAN.
The next most important feature of a CAN controller is its receive message filtering. Because CAN is a broadcast
network, every message transmitted is received by every node on the network. In a CAN network of any
reasonable complexity there will be a large number of messages sent over the CAN bus. In such a network the
CPU of a CAN node will spend all its runtime responding to CAN messages. To avoid this problem all CAN
controllers have some form of message filtering that blocks unwanted messages from reaching the receive
buffers. The CAN controller on the STM32 has 14 filter banks which can be used to block all CAN messages
except selected message identifiers or groups of message identifiers.
The 14 message filters have
two configurations that can
be used to filter individual
messages.
Each filter bank consists of two 32-bit registers. Each filter bank can be configured in one of four modes. The
basic method programs each register of the filter bank with a message ID. When a message arrives, it must
match this ID or be rejected. This mode has two configurations. In the first, the filter bank registers are used 3 bits
wide and are able to filter the 11-bit and 29-bit message ID fields as well as the RTR and IDE bits in 16-bit mode.
The same filter banks can be
used to filter groups of
messages.
In the second configuration, the first 32-bit register is written with the message ID and the second register is used
as a message mask. The mask register marks bits in the ID register as ‘care’ or ‘don’t care’. This allows a group
of messages to be received through a single filter bank. When a message is received through the receive filters, a
filter match index is stored with the message in the receive FIFO. This provides the application software with a
shorthand method of determining the message data without having to read and decode the message packet ID.
© Hitex (UK) Ltd.
Page 66
Chapter 5: Peripherals
All CAN controllers have two operating modes: a normal mode for receiving and transmitting message packets
and an initialising mode for setting the communication parameters. The STM32 has a low power sleep mode. In
sleep mode the clock to the bxCAN module is halted, but the mailbox registers may still be accessed. The bxCAN
module will wake up when it detects activity on the CAN bus; it may also be reactivated by the application
software. There are two additional sub modes when it is operating in normal mode. The first is silent mode, where
the CAN controller may receive messages, but cannot transmit and does not generate error frames or message
acknowledge bits. This mode is intended for passive monitoring of a CAN network Secondly, there is a loopback
mode where all transmitted messages are looped back into the receive buffer. This mode is intended for selftesting and is also useful during code development. Both modes may be combined and this is ideal for self-testing
when connected to a running network.
5.3.2
USB
The USB interface on the STM32 is a full speed (12Mb/sec) device interface, which can be controlled by a USB
host such as a PC. The USB peripheral is a complete USB layer 1 and 2 protocol interface, which implements the
USB physical layer interface and the data transfer layer with all packet error checking and retransmission. The
USB device interface also supports the USB suspend and resume operations for low power operation. Developing
a USB-based application does require a good knowledge of the USB specification and its application classes. A
full USB developer’s kit is available from the ST website. This provides a software stack to initialise the USB
interface and has support for commonly used USB classes such as Human Interface Device (HID), mass storage,
audio and legacy communications port. Using this stack, or a similar third party software stack, greatly speeds up
development rather than reinventing the wheel.
The USB peripheral supports
USB 2.0 device
communications.
The USB interface supports up to eight endpoints, which are user configurable as endpoints for control, interrupt,
bulk or isochronous pipes. The endpoint packet buffers are stored in the 512bytes of SRAM which is shared with
the CAN controller. When the device is initialised, the application software divides this SRAM into a series of
buffers.
The SRAM is configured into the endpoint buffers by a buffer description table held at the base of the SRAM.
Here each endpoint is provided with a start address in the SRAM and a count to indicate its size. Each active
control, interrupt and bulk endpoint is allocated an endpoint packet buffer, while isochronous endpoints are
allocated a double buffer. This allows data to be received into one buffer while data in the second buffer is
processed. When the next packet is received, the new data goes into the second buffer while the first is
processed. This double buffer approach supports the streaming of real-time data such as audio.
The 512bytes of SRAM shared with the Can controller is used to store the USB packet data. During initialisation
this region of memory is divided into individual buffers for each of the active endpoints. The endpoints used by
isochronous pipes have a special double buffer so that data can be received into one buffer while an earlier
packet is being read from a second
© Hitex (UK) Ltd.
Page 67
Chapter 6: Low Power Operation
© Hitex (UK) Ltd.
Page 68
Chapter 6: Low Power Operation
6. Low Power Operation
As well as being a high performance microcontroller, the STM32 has several low power modes in addition to its
normal RUN mode. When used judiciously, the SLEEP, STOP and STANDBY low power modes make powering
applications from batteries a practical prospect. The STM32 squares the circle by being a low power
microcontroller with a high performance processor. In the Cortex overview we saw how the Cortex processor can
enter a low power mode in which the CPU and Cortex peripherals are halted and consume minimal power. When
the Cortex Processor enters a low power mode, it can export a SLEEPDEEP signal to the surrounding
microcontroller, signalling it to enter a low power mode. All of the low power modes are entered by the Cortex
CPU executing a WFI or WFE instruction. The low power mode that the STM32 then enters depends on the
setting in the power control registers. In the next section we will have a look at each of the power modes in turn
and look at comparison of their power consumption and wake up times.
6.1 RUN Mode
RUN mode is when the STM32 is executing program instructions and is at its highest level of power consumption.
This section looks at various ways to reduce overall power consumption during program execution. It is important
to remember that all of these features can be used dynamically as the code runs. This means that it is possible to
run code in a low power, low performance configuration and then switch to high power, high performance
configuration in response to an interrupt or program event.
During normal operation the Cortex processor and most of the STM32 can run at 72 MHz. When it is running at
full speed, the STM32 consumes in excess of 30mA. The power consumption of the STM32 can be reduced by
first gating the clocks of any unused peripherals. This stops any unused areas of the chip from consuming power.
The peripheral clocks can be switched on and off dynamically through the Reset clock control module.
Additionally, big power savings can be made by slowing down the system clock. If high speed operation is not a
necessity, the PLL can be switched off and the STM32 can be clocked directly from the HSE oscillator. Further
power reduction can be achieved by switching off the HSE oscillator and using the HSI oscillator. This has the
disadvantage that the HSI oscillator is not as accurate a clock source as the HSE oscillator. Similarly, if the
windowed watchdog and the real-time clock are not being used, the LSI oscillator can be switched off in order to
shave off a bit more power consumption.
6.1.1
Prefetch Buffer And Half-Cycle Mode
If you are running directly from the HSE oscillator at a maximum of 8MHz, you can also disable the FLASH pre
fetch buffer and enable the half-cycle operation. This incurs extra wait states, but reduces the RUN mode power
consumption.
Full
speed
power
consumption is
around 34mA,
but at 8 MHz
(9.6 DMIPS) the
power
consumption is
below 1mA.
© Hitex (UK) Ltd.
Page 69
Chapter 6: Low Power Operation
6.2 Low Power Modes
Careful configuration of the STM32 RUN mode can reduce power consumption to around 8.5mA. In order to get
a true low power application we have to make use of the STM32 low power modes.
6.2.1
SLEEP
The first level of low power operation is the SLEEP mode. By default, when an WFE or WFI instruction is
executed the Cortex processor will halt its internal clocks and stop executing the application code. In SLEEP
mode the remainder of the STM32 will continue to operate. The STM32 will leave SLEEP mode when a peripheral
generates an interrupt. When the STM32 enters SLEEP mode with all peripherals enabled and it is running at
72MHz from the HSE through the PLL, its SLEEP mode power consumption will be around 14.4mA. However, if
the STM32 is prepared for low power operation by: firstly disabling all peripheral clocks (except for the peripheral
used to wake up the Cortex processor) and secondly switching to the HIS oscillator (which can be further divided
down to 1MHz or below) we can get power consumption figures of around 0.5mA.
SLEEP mode power consumption
can be as low as 0.14mA.
In low power applications you should try to enter SLEEP mode as often as possible, in order to consume
minimum power. The next issue is how long it takes for the STM32 to exit its low power mode and resume
processing. The figures below show the wake up time for the Cortex CPU to resume processing using the HIS RC
clock.
© Hitex (UK) Ltd.
Page 70
Chapter 6: Low Power Operation
6.2.2
STOP Mode
The STM32 can be configured to enter the low power STOP Mode by setting the SLEEPDEEP bit in the Cortex
power control register and clearing the Power Down Deep Sleep (PDDS) bit in the STM32 power control register.
When configured for STOP mode, execution of a WFI or WFE instruction will halt the Cortex processor and switch
off the HIS and HSE oscillators. The FLASH, SRAM and peripherals are still powered, so the state of the STM32
is preserved. Like SLEEP mode, STOP mode can be left via an STM32 peripheral generating an interrupt.
However, in STOP mode all the peripheral clocks are halted, with the exception of the External Interrupt
peripheral. The use of the EXTI peripheral allows the STM32 to exit STOP mode when there is a state change on
any GPIO pin. In addition, the EXTI has a line which can both request an interrupt and generate an interrupt from
a real-time clock Alarm event. As the real-time clock has its own dedicated oscillator (either the LSI or LSE
oscillator) it can provide a periodic interrupt to wake up the STM32 from STOP mode.
Once the STM32 has entered STOP mode, its power consumption drops mA in RUN mode to around 24 uA.
Further power savings can be made by placing the internal regulator in a special low power mode when it enters
STOP mode. The low power mode for the voltage regulator is selected by setting the LPDS bit in the STM32
power control register. With this bit set when the STM32 enters STOP mode, its power consumption will drop to
14uA. If the RTC is being used, a further 1.4 uA will be consumed.
The wake up times you can expect in STOP mode are a worst case of 5.5 usec with the voltage regulator fully on
and 7.3 usec with the regulator in its low power mode.
© Hitex (UK) Ltd.
Page 71
Chapter 6: Low Power Operation
6.3 Standby
The STM32 can be configured to enter its standby mode by setting the SLEEPDEEP bit in the Cortex power
control register and setting the Power Down Deep Sleep bit in the STM32. Now, when the WFI or WFE
instructions are executed, the STM32 will drop into its lowest power mode. In Standby mode the STM32 is really
switched off. The internal voltage regulator is switched off and the HSE and HIS oscillators are off. In this mode
the STM32 consumes a mere 2uA.
In Standby mode power consumption
is 2uA with a wakeup time of 50uS.
You can exit Standby mode by using an RTC alarm event in the same way as STOP mode. Additionally, you can
use an external STM32 reset or a reset from the independent watchdog. Standby mode can also be exited by a
rising edge on pin 0 of PortA. This pin must be configured as wakeup pin WKUP, by setting the EWUP bit in the
power control and status register. As the lowest power mode, Standby mode takes the longest to leave and it will
take around 50usec before the Cortex CPU will restart processing instructions. Once in Standby mode, all data in
the SRAM, Cortex and STM32 registers is lost. An exit from Standby mode is effectively the same as a program
reset.
6.4 Backup Region Power Consumption
The backup region containing the battery backed RAM and the RTC will be kept alive during all power down
modes. This power domain will consume around 1.4 uA at 3.3V.
6.5 Debug Support
On traditional microcontroller systems, debugging an application which uses low power modes can be extremely
painful. As soon as the microcontroller enters low power mode it stops responding to the debugger, which then
throws an error or ceases to work. Within the STM32 it is possible to configure the low power modes to keep the
HSI oscillator running in each of the low power modes, providing a dedicated clock path to the CoreSight debug
architecture. This means that you can fully debug low power applications without having to remove entries into
low power mode. This eliminates debug timeout problems. The STM32 enhanced debug features are configured
with the DBG_MCU register.
© Hitex (UK) Ltd.
Page 72
Chapter 7: Safety Features
© Hitex (UK) Ltd.
Page 73
Chapter 7: Safety Features
7. Safety Features
The STM32 has also been designed with a number of inherent features that will detect incorrect operation of the
application code or the STM32 itself. To ensure that there is a reliable power supply, the STM32 has its own
internal reset that will place the chip in reset if the supply voltage is below minimum VDD. Additionally there is a
programmable power voltage detect circuit that can be used to detect power failure early. It will then generate an
interrupt to place the chip into a safe state. The clock tree also includes a clock security system which monitors
the HSE oscillator. If this fails, the CSS will force the STM32 to fail back to the HSI oscillator. Correct program
execution can be monitored by two internal watchdogs. Firstly, a windowed watchdog which must be refreshed at
a specific rate. Secondly, an independent watchdog which is clocked by a separate oscillator from the main
system clock. Additionally, the on-chip FLASH memory has a data retention of 30 years at 85 degrees C. This is a
‘best in class’ data retention for a general-purpose microcontroller. These safety features are not suitable for the
highest levels of safety-critical equipment (at high software integrity levels hardware such as watchdogs must be
separate external devices.) However, the STM32 does allow you to develop rugged self-correcting systems using
the techniques employed by safety-critical applications such as avionics and automotive systems, but using very
low cost hardware. This brings a new level of quality and reliability to even the simplest and most cost-sensitive
applications.
7.1 Reset Control
The STM32 has a number of reset sources other than the external reset line. A reset on the STM32
microcontroller can be forced from: the internal watchdogs, a software reset via the NVIC, the internal Power on
reset/Power down reset and the low power voltage detect circuits. If a reset occurs, a set of flags in the RCC
control and status register can be read to determine the cause of the reset. The state of these flags will persist
until the next power on reset or until a logic one is written to the remove reset bit.
The STM32 may be reset from a number of sources. After reset the
RCC control and status register reports the last reset source.
7.2 Power Voltage Detect
As part of the internal power supply supervisor the STM32 contains a power monitoring unit called the Power
Voltage Detect (PVD). The PVD has a programmable threshold that can be set in the in steps of 0.1V from 2.2V
to 2.9V. This threshold is configured in the power control register.
The power supply is monitored by the
power voltage detect unit, which can
generate an interrupt if the supply dips
below a configured threshold.
© Hitex (UK) Ltd.
Page 74
Chapter 7: Safety Features
The output of the PVD is connected to line 16 of the external interrupt unit. Since the EXTI lines can be
configured to generate an interrupt on a falling or rising edge, or both, the PVD unit can be used to generate an
interrupt for both under and over voltage conditions.
7.3 Clock Security System
In most STM32 applications the main system clock used for the Cortex processor and the STM32 peripherals will
be derived from an external crystal connected to the HSE pins. This clock tree contains a Clock Security System
that monitors the external crystal. If this crystal fails, it will cause the STM32 clock system to fail back to the
internal 8 MHz oscillator.
The Clock Security System generates
an interrupt if the external oscillator
fails and switches to the internal RC
oscillator.
The Clock Security System is enabled by setting the Clock Security Enable bit in the RCC control register.
The Clock Security System is enabled
by setting the CSS enable bit in the
RCC control register.
The CSS has an interrupt line which is connected to the break interrupt of advanced timer 1, which in turn is
connected to the Cortex NVIC non-maskable interrupt line. This ensures that if the main oscillator fails, the PWM
outputs of the advanced timer will immediately be placed in a pre-programmed safe state by hardware control.
This ensures that any hardware driven by the advanced timer PWM outputs will not be allowed to run while not
under control of the Cortex processor. It is particularly important for motor control applications.
© Hitex (UK) Ltd.
Page 75
Chapter 7: Safety Features
7.4 Watchdogs
The STM32 contains two completely separate watchdogs. The independent watchdog is completely separate
from the main STM32 system. It is located within the backup power domain and derives its clock from the internal
Low Speed Oscillator (LSI). The windowed watchdog is part of the main STM32 system and is clocked via the
peripheral bus 1 clock. Both watchdogs must be individually enabled and can be used simultaneously.
The STM32 has two internal
watchdogs one of which has its own
separate oscillator.
7.4.1
Windowed Watchdog
The windowed watchdog is a more advanced version of a traditional on-chip watchdog. Once enabled, the
watchdog will count down and will generate a reset on a transition from 0x40 to 0x3F i.e. when bit T6 is cleared.
An additional count value is stored in the windowed watchdog configuration register. This provides an upper count
value. If the application software refreshes the watchdog count while the actual watchdog counter value is greater
than the configuration value, a reset will also be generated. The windowed watchdog provides a programmable
refresh window which is the only valid time that the watchdog can be written to. This allows you to build extra
confidence that the application software is running within its expected parameters.
The windowed watchdog is a six-bit down counter, which is clocked from PCLK1 via a 12 bit prescaler that divides
PCLK1 down by 4096. The prescaler has a further 4 bits that are user programmable allowing a further divide by
1,2,4 or 8. The prescaler bits are contained in bits 6,7 of the control register.
Hence the timeout period of the windowed watchdog is given by:
Twwdg = Tpclk1x4096x2POW (WDGTB) x (reload value+1)
With Pclk1 running at its maximum 36 MHz, the windowed watchdog minimum timeout period is 910uSec and the
maximum period is 58.25mSec.
© Hitex (UK) Ltd.
Page 76
Chapter 7: Safety Features
Once the windowed watchdog has been configured, it can be enabled by setting the watchdog activation bit in the
control register. Once the windowed watchdog has been enabled by software it cannot be disabled, except by a
reset.
7.4.2
Independent Watchdog
Although the independent watchdog is fabricated on the same silicon as the main STM32 system, it has its own
oscillator separate from the main STM32 clock. The independent watchdog is also located within the VDD voltage
domain, which is kept alive in the STOP and STANDBY modes.
The independent watchdog is a 12-bit count down timer, which will force a reset on the STM32 when it
underflows. It is clocked from the low speed internal oscillator via an 8 bit prescaler. The LSI oscillator has a
nominal frequency of 32.768, but in practice this can vary between 30 KHz to 60 KHz. The independent watchdog
is initialised by first setting the prescaler register, which divides down the LSI oscillator in powers of two between
4 to 256. The minimum timeout period for the independent watchdog is 0.mSec and the maximum period is just
over 26 seconds. The timeout period value is programmed directly into the reload register.
Watchdog debug support
The independent watchdog is a
countdown watchdog with its own
oscillator. It is also located in the
backup domain so that it can remain
active during Stop and Standby
modes.
The option bytes in the FLASH memory small information block can be used to configure the independent
watchdog to start after a reset, or by software command. If under software control, the independent watchdog can
be started by writing 0xCCCC to the key register. The Independent watchdog will count down from an initial value
of 0xFFF. The value 0xAAAA must be written to the key register to refresh the watchdog. This causes the reload
value to be loaded into the down counter register, refreshing the count value.
Traditionally it is very difficult to debug small microcontrollers if the watchdog is enabled. As soon as the CPU is
halted, the watchdog cannot be updated. It will timeout and force a reset, which destroys the debug session.
Normally a watchdog has to be disabled so that it does not upset the debugger. Consequently it is very difficult to
test and prove that watchdog refreshes are occurring at an optimum rate. Within the STM32 MCUDBG register it
is possible to configure both the independent watchdog and the window watchdog to halt when the Cortex-M3
CPU is under control of the CoreSight debug system. This allows you to step through your code with both
watchdogs enabled and they will be incremented in sync. with the number of cycles executed on the CPU.
© Hitex (UK) Ltd.
Page 77
Chapter 7: Safety Features
7.5 Peripheral Features
The user peripherals have also been designed with a number of features that help to ensure the safe operation of
the STM32. These are fully described in the relevant user peripheral section but will be reviewed here:
7.5.1
GPIO Port Locking
When the GPIO ports are initialised, each IO line will be configured as an input or output. Once configured, the
STM32 GPIO port configuration can be locked. This prevents any further accidental changes to the port
configuration. Each port may be locked on a bitwise basis.
7.5.2
Analog Watchdog
Each of the Analog to digital converters has two analog watchdogs. These watchdogs can be set to generate an
interrupt on over-range or under-range voltages.
7.5.3
Break Input
For motor-based applications, the break line within the advanced timer can be used to place the three
complementary PWM outputs into a predefined state, in response to an input on the break pin, or a failure in the
main STM32 oscillator.
© Hitex (UK) Ltd.
Page 78
Chapter 8: The FLASH Module
© Hitex (UK) Ltd.
Page 79
Chapter 8: The FLASH Module
8. The FLASH Module
The on-chip FLASH memory of the STM32 is arranged in three main regions. First, there is the main FLASH
memory designed to hold program instructions. This memory is 64 bits wide to provide efficient memory access
with the prefetch buffer. For flash program and erase operations this memory is divided into 4K pages. This
memory has a WRITE endurance of 10000 cycles and data retention of 30 years at 85 degrees C. Most
microcontroller FLASH memory data retention is rated at 25 degrees C, so the STM32 has an exceptional FLASH
memory. Aside from the main program memory, there are two smaller memory regions: the big information block
and the small information block. The big information block is a further 2k of FLASH memory holding a factory
programmed bootloader, which is designed to download code over USART 1. The small information block
contains six configuration bytes, which are used to define the reset properties of the STM32 and its memory
protection.
8.1 Internal FLASH Security And Programming
The internal FLASH memory can be updated by the internal bootloader, by a JTAG tool, or by in-application
programming through a dedicated set of registers called the FLASH program and erase controller FPEC. The
FPEC is also used to program the option bytes in the small information block.
The FPEC module is used to allow inapplication programming of the
FLASH memory. The FLASH memory
can also be read-protected from
debug tools and write-protected.
8.2 Erase And Write Operations
After reset, the FPEC registers are protected and must be unlocked by writing a special sequence to the key
register. To unlock the FPEC you must write 0x45670123, followed by 0xCDEF89AB. If there is a mistake in this
sequence, the FPEC will stay locked until the next reset. Once the FPEC has been unlocked, it is possible to
erase and WRITE operations on the main FLASH memory. Within the main FLASH memory block it is possible to
perform a mass erase or an erase of a selected 4k page. A mass erase is done by simply setting the mass erase
and start bits in the control register. When the busy bit in the same register is reset, each location in the main
FLASH memory will be reset to 0xFFFF. A page erase is equally easy to perform. First, you must program the
start address of a FLASH page into the Address register, then set the page erase and start bits in the control
register. Again, when the busy bit is clear, the page will be erased. New data can be written to a FLASH memory
cell only after it has been erased. A WRITE operation is performed by setting the program bit in the control
register and then performing a half-word write to the desired location. If the FLASH location is erased and not
write protected, the FPEC will program the new value into the FLASH memory cell.
8.3 Option Bytes
The small information block contains eight user-configurable option bytes. Four of these bytes are used to define
write protection on the main FLASH memory. The fifth is used to set read protection which prevents access to
regions of memory when the chip is in debug mode. A sixth byte is used to configure low power and reset
operation. The final two bytes are simple FLASH memory cells that are available for user-defined options. Before
the Option bytes can be written to, the FPEC must be unlocked as described above. Then the Option bytes must
be unlocked by writing the same two keys to the option key register. The Option bytes have a separate program
© Hitex (UK) Ltd.
Page 80
Chapter 8: The FLASH Module
and erase procedure to the main FLASH memory. The small information block is erased by setting the OPTER bit
in the control register and then the STRT bit. Once the BSY bit is reset, the small information block is erased. To
program an Option byte, set the OPTPG bit in the FLASH control register and perform a half-word write to the
Option byte. Each Option byte is stored as a half-word. The Option byte is stored in the lower byte of the halfword and its complemented value is stored in the upper half-word. You must write a correct value in the lower
half-word and the FPEC will automatically calculate the complemented value.
8.3.1
Write Protection
When it is set, each bit in the write protection option bytes enables protection over a given FLASH page. Write
protection can be disabled by an erase of the small information block.
8.3.2
Read Protection
When the read protection is set, all read accesses to the FLASH memory are disabled when the device enters
debug mode. Access to the SRAM is still possible and code may be downloaded and executed in this region. So it
is possible to disable the read protection by running a program out of SRAM. However, when read protection is
disabled a mass erase of the internal FLASH is also performed, to ensure protection from software piracy. When
read protection is enabled, the FLASH memory is also write protected to prevent a malicious program from being
inserted into the memory region containing the vector table. The STM FLASH memory is protected if the read
protection byte and its complement are set to 0xFF. The memory can be unprotected by writing 0xFA and its
complement as a half-word to the read protection Option byte.
8.3.3
Configuration Byte
The configuration Option byte contains three active bits. Two of these bits govern how the STM32 enters Standby
and Stop modes. Either mode can be configured to generate a reset on entry. This will configure the digital IO
pins as inputs, reducing the overall power consumption of the STM32. The PLL and external oscillator will also be
disabled and the chip will revert to using the internal high speed RC oscillator as the main system clock. The final
bit in the configuration Option byte configures the activation of the independent watchdog. This watchdog has a
hardware watchdog mode where it will start immediately after a processor reset, or software watchdog where it
must be started under software control.
© Hitex (UK) Ltd.
Page 81
Chapter 9: Development Tools
© Hitex (UK) Ltd.
Page 82
Chapter 9: Development Tools
9. Development Tools
The adoption of ARM7 and ARM9 into standard microcontrollers has lead to an explosion in development tools’
support for these CPUs. All of the major compiler developers such as GCC, Greenhills, Keil, IAR and Tasking
provide ARM development tools. With the introduction of the Cortex processor, all of these development tools
have been extended to support the Thumb-2 instruction set. If you are already using an ARM-based
microcontroller, the chances are that it will already generate code for the STM32. The worst case scenario is that
you will have to get an upgrade from your supplier.
If this is your first project with an ARM-based microcontroller, you will be able to select a toolset from your
preferred manufacturer. While it is hard to find a bad toolset these days, two compilers are worth discussing
further. Firstly the “GCC” or “GNU” compiler is an open source tool which can be downloaded and used for ‘free’.
The GCC compiler has been integrated into a number of commercial IDEs and debuggers to make low-cost
development tools and evaluation kits. While the GCC compiler is a reliable and stable compiler, our experience
has been that its code generation is not as efficient as the commercial compilers. There is also generally not a
direct support route if you run into problems, which can slow down development. Of the commercial compilers,
the ARM RealView compiler is the original and most refined C compiler and was developed by ARM for use with
their CPUs. The RealView compiler is available as part of the ARM RealView toolset. This toolset is aimed at
system-on-chip developers and is not really suitable for microcontroller projects. However since January 2006
the RealView compiler has been integrated into the Keil Microcontroller Development Kit (MDK-ARM). As its
name implies, MDK-ARM is a complete tool chain designed exclusively for ARM-based microcontrollers. The
MDK is easy to use (selecting about 4 options configures a whole project) and provides a tightly-integrated tool
chain, which is controlled by one manufacturer.
If you are making a decision between using the GCC compiler and a commercial compiler, you will partly be
driven by the project budget. A one off ‘simple’ project is unlikely to have the budget to justify a commercial
toolset. However, if you plan to standardise on ARM-based microcontrollers, then an ‘expensive’ toolset will
soon pay for itself both in reduced development time and a more compact final image. It is also important to bear
in mind your relative level of experience. If you are a hard-core embedded developer, then you are likely to be
able to develop a whole project with the GCC compiler. If, however, you are less experienced, or do some C
coding, then it is possible to get into a huge mess.
9.1.1
Evaluation Tools
Most compiler vendors will also provide an evaluation kit or starter kit. This is traditionally a hardware board and a
cut down or time-limited version of their toolset. An up-to-date list of evaluation kits is available on the ST
website. One of the best evaluation tools is the Hitex STM32 Performance Stick. Costing about 50 Eur, the
Performance Stick is a complete evaluation tool for the STM32. It is designed as a USB dongle that allows you to
develop and debug an unrestricted amount of code with the GCC or Tasking compilers, via the HiTOP IDE. In
addition to the STM32, the Performance Stick hardware has a second microcontroller in the shape of the
STR750. This microcontroller uses its ADC and timers to measure the STM32’s power consumption and interrupt
latency. This information is sent to a ‘dashboard’ application on the PC. The dashboard allows you to manually
experiment with the different features of the STM32 and get some verification of the data sheet values for power
consumption, wake up time etc.
The Hitex Performance Stick is a very low-cost evaluation tool for the STM32. It provides an unlimited
development environment based on the HiTOP debugger and the GCC compiler. For full product
development, the same IDE and compiler are available for the Hitex Tantino JTAG debugger.
© Hitex (UK) Ltd.
Page 83
Chapter 9: Development Tools
9.1.2
Libraries And Protocol Stacks
To support rapid code development, ST have provided a STM32 firmware library as a free download from their
website. The firmware library provides low- level driver functions for all of the on-chip peripherals. This gives you
some basic building blocks on which to start building your application. The most complex peripheral on the current
STM32 variants is the USB device controller. In order to help you build common USB class devices, ST also
provide a free USB developer’s kit. Like the firmware library, the USB developer’s kit can be downloaded from the
ST website. The USB developer’s kit provides a USB library and demonstration applications for HID, Mass
Storage, Audio and Device Field Upgrade.
With the increasing complexity of
microcontroller peripherals, it is
important to select a tool chain which
is well supported with protocol and
application software stacks.
As new variants of the STM32 are released, they will have more and more complex peripherals (Ethernet MAC,
TFT interface etc). As this complexity increases, it becomes just about impossible to develop all the application
code yourself. So when selecting development tools it is also important to consider the availability of protocol
stacks, such as a TCP/IP stack and other application software, such as a GUI, which may be required on future
projects. Ideally these should be from the same vendor and well integrated into your chosen toolset.
9.1.3
RTOS
If you are moving from an eight or sixteen bit microcontroller, the chances are that you are not currently using an
RTOS. As we have seen, the Cortex-M3 provides you with significantly more processing power than comparablypriced microcontrollers and is designed to support a small footprint RTOS. Thus, if you have not been using an
RTOS it is worth considering when you start work with the STM32. The use of an RTOS gives you the advantage
of more abstract code development, enhanced code re-use, easier project management and enhanced
debugging. The use of an RTOS also provides a structure to your code, which forces you to plan the application
before you dive in and start writing. There are more RTOSes available for ARM and Cortex than for most
embedded CPUs. Many compiler vendors will provide their own and have ports for third party RTOSes, but one
of the most popular open source operating systems is “FreeRTOS”, which is available from www.freertos.org. A
commercial version of FreeRTOS is called “SafeRTOS”, which has been tested to meet the IEC 61508 safety
standard and is also available from the same site.
© Hitex (UK) Ltd.
Page 84
Chapter 10: End Note
© Hitex (UK) Ltd.
Page 85
Chapter 10: End Note
10.
End Note
If you have read through this far, I think you will agree that the Cortex STM32 is a new generation of very low cost
general purpose microcontrollers. Centred around a high performance processor with a deterministic interrupt
system with sophisticated peripherals, the STM32 is suitable for many industrial and consumer applications.
Additionally the low power modes make it suitable for battery-powered and hand held products.
© Hitex (UK) Ltd.
Page 86
Chapter 11: Bibliography
© Hitex (UK) Ltd.
Page 87
Chapter 11: Bibliography
11.
Bibliography
Cortex-M3 Technical reference manual
ARMv7-M architectural reference manual
ARM Architectural reference manual Thumb2 supplement
STM32F103xx User Manual
STM32F10xxx FLASH Programming manual
© Hitex (UK) Ltd.
Page 88
ARM Ltd
ARM Ltd
ARM Ltd
ST Microelectronics
ST Microelectronics
Chapter 11: Bibliography
© Hitex (UK) Ltd.
Page 89
This book is intended as a hands-on guide for anyone intending to use the ST
Microelectronics STM32 family of Cortex-M3 microcontrollers.
Over the last six or seven years one of the major trends in microcontroller design
is the adoption of the ARM7 and ARM9 as the CPU for general purpose
microcontrollers. Today there are some 240 ARM-based microcontrollers available
from a wide range of manufacturers. Now ST Microelectronics have launched the
STM32, their first microcontroller based on the new ARM Cortex-M3
microcontroller core. This device sets new standards of performance and cost, as
well as being capable of low power operation and hard real-time control.
www.hitex.com
D E V E L O P M E N T TO O L S
ISBN 0-9549988-8-X
9 780954 998882
Hitex (UK) Ltd., Sir William Lyons Road, Science Park, Coventry, UK, CV4 7EZ.
Tel +44 (0) 2476 692066
Was this manual useful for you? yes no
Thank you for your participation!

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

Download PDF

advertisement