|
Most common problems
In System Programming(ISP) of the mega103/603
AVR emulator can not drive an external load
A/D converter is reading the value from the last channel
When I load a file into AVR STUDIO, everything chrashes
Use of the low byte's of the pointers to access the
available memory
Development tools for megaAVR
Is there any starter kits available for AVR ?
Access the AVR internal EEPROM data memory within the
IAR C-compiler
Is the Real Time Clock(RTC) in the ATmega controller
Year 2000 compliant
Minimum Power Down current consumption
Asynchronously reset of I/O ports
NC(No connect) pins
How do I define a string constant in the AVR assembler
?
My assembler does not produce an Intel Hex file, what
is wrong ?
Assembler editor in file size limitation
Upgrade of AVR Simulator for Windows 3.11
PC does not recognize the IAR Dongle
How can I reach the entire program memory with RJMP
and RCALL
Use of the Timer/Counters used to generate the UART
Baud Rate
Single bit set or clear to the I/O with the SBI and
CBI instructions
Why are there three addresses for each I/O Port ?
Why does SBI and CBI only operate on I/O registers $00
-$1F ?
Why do the immediate instructions not work on registers
R0-R15 ?
Why is there no EORI instruction ?
Use of SBR instruction
Check and clear flags
Clearing a register and the carry flag
How do I implement a constant table in the AT90S1200?
How to add an immediate constants
How can Flash and Mask ROM ever cost the same ?
Is the external clock signal multiplied ?
How many clock cycles make up one AVR instruction cycle
?
Why is the AT90S8515 called a 8K MCU ?
Where can I find datasheets ?
What added value can the AVR offer over my current MCU
?
What does "AVR" mean ?
Question: What are the most common problems when starting a new
project with the AVR ?
Answer
The most common mistake is to forget to set up the stack-pointer of
the
devices without hardware stack. On an AT90S8515 this is done like
this:
LDI
R16,low(RAMEND)
OUT
spl,temp
LDI
R16,high(RAMEND)
OUT
sph,temp
It is also common to forget to set up the ports when using them
as output. On the AVR, the ports are set as outputs by writing a '1'
to the Data Direction Register(DDR). For instance to set PORTB to
output high:
LDI R16,0xFF
;Load FF(hex) into the register file
OUT DDRB,R16
;set PORTB as output
OUT PORTB,R16
;set PORTB high
When using the AVR assembler a common mistake is to point to the wrong
location in the program memory while using the Load Program
Memory(LPM) instruction. The program memory is organized as 16-bit
words, while the LPM instruction is reading 8-bit bytes.
The LPM instruction can read either the upper or the lower byte of the
16-bit word. Because of this, it is necessary to multiply the 16-bit
program memory address by two to point to the byte you want to
load.
This is decribed in app.note 102, "Block Copy routines"
Question: What pins are used for In System
Programming(ISP) of the mega103/603 controller ?
Answer
The In System Program interface for mega103/603 controller should be
connected to the following pins on mega103/603:
SCK - SCK
(PB1,pin 11)
MISO - PDO
(PE1,pin 3)
MOSI - PDI
(PE0,pin 2)
RESET - RESET
(RESET,pin 20)
The PEN(Program Enable) pin can be used as an alternative to RESET.
The two pins have the same functionality, (they activate the internal
iRESET signal) but PEN is read only during Power On (VCC approx. 1.2
volts). If PEN is not used, it is recommended to tie this pin to VCC.
RESET is then used as on all other AVR parts to enable serial
programming.
Question: I can not get my AVR emulator to
drive an external load correctly, when I use a
voltmeter to measure the pins on the emulator POD it seems like the
pin can not drive the load.
What is wrong ?
Answer
This is a problem with version 1.21 of the ICEPRO and version 1.10
of the megaICE.
New versions of the emulators are available as software upgrades from
www.atmel.com->products->AVR 8 bit RISC->software.
To upgrade the ICEPRO to version 1.22, download the file UPGRD122.ZIP
To upgrade the megaICE to version 1.11, download the file M111UPGR.ZIP
These files are compressed files containing a new program file for the
emulator and a PC program that automatically updates the emulator and
performs an error check of the code.
Question: I am using an AVR device with A/D
converter, the device works excellent, and I
love the speed of the AVR. But I have one problem: When I change input
channel on the A/D
converter it seems like the A/D converter is reading the value from
the last channel. If I do a
second read I get the right value. Why is this happening ?
Answer
Two bits in the A/D Control and Status Register(ADCSR) is used
to detect the end of an A/D conversion:
The A/D Start Conversion (ADSC) bit is used to test when a new
conversion can start, but the result of the previous conversion is
not
yet ready in the A/D data register. When the A/D data register is read
the result of the last conversion is present.
To get the result of the latest conversion, test the A/D Interrupt
Flag (ADIF) or enable global interrupt and execute the reading of the
A/D converter in the interrupt routine.
Example code:
ldi R16,1
; Select channel
out
ADMUX,R16
sbi
ADCSR,adif
; Reset interrupt flag
sbi
ADCSR,adsc
; Start A/D conversion
wait: sbis ADCSR,adif
; Wait until the ADIF is set
rjmp
wait
Question: When I load a file into AVR STUDIO,
everything chrashes. It worked the first time
I used the file. What is wrong ?
Answer
The ".AVD" file contains information about the window settings for the
current object file.
Sometimes, the '.AVD' created by AVR STUDIO can get corrupted.
Depending on which part of the file that is corrupted, different
symptoms will appear. When you run into 'strange' problems with AVR
STUDIO, first try to delete the '.avd' file located in the folder
where your source is placed.
Question: In AVR devices like the AT90S2313,
where total Data Memory is less than 256
locations and there is no access to external memory, can you ignore
the upper byte of the X,
Y, and Z pointers and only use the low byte's of the pointers to access
the available memory ?
Answer
In devices where total Data Memory (registers, I/O and SRAM) is less
than 256 bytes, the CPU will ignore the high byte of the pointer when
accessing data memory using LD/ST instructions. This allows the high
byte of each pointer (R27, R29 and r31) to be used for general purpose
storage. The post-increment and pre-decrement will not affect the high
byte. (like LD -Z, Rd).
Note that the high byte of the Z pointer is used when accessing flash
program memory and ADIW/SBIW instructions WILL affect the high byte.
Question: I want to use the megaAVR microcontroller,
what development tools are available
for this part ?
Answer
The megaAVR family of microcontrollers have a complete suite of
development tools:
Evaluation/Programming board: The ATMEL STK300 Starter Kit, developed
specifically for the ATMega series. Programming is accomplished via
the In System Programming(ISP) interface.
Emulator: The ATMEL MegaICE is available now for real-time emulation.
C compiler: The third-party tools company IAR Systems is providing a
C-compiler for our AVR product line, including the ATMega parts. You
can check their products at http://www.iar.com
ATMEL also offers most of the front-end software required to work with
our AVR devices (assembler, simulator) as freely-downloadable software
on our Web site: http://www.atmel.com.
All these products are available through an authorized ATMEL
distributor/sales representative
Question: Is there any starter kits available
for AVR ?
Answer
The STK200 supports all AVR parts with 8, 20, 28 and 40 pins.
The STK300 supports all megaAVR parts.
Both starter kits includes pushbuttons, LED's and RS232 interface. The
I/O ports are accessible through header connectors on the edge of the
board. The starter kits includes an In System Programming(ISP) cable
that can be used to program the AVR controller in the user
application.
The STK300 also includes the AVR application builder which generates
the initialization code for setting up the peripherals.
Question: How can I access the AVR internal
EEPROM data memory within the IAR
C-compiler ?
Answer
The internal EEPROM on the AVR is part of the I/O memory of the AVR
controller. It can not be accessed as normal variables. Special macros
to read and write the EEPROM is defined in the standard library
routine "ina90.h"
The read and write macro has the following prototypes:
_EEGET(VAR,ADR) /* To read the value of EERPOM address ADR */
_EEPUT(ADR,VAL) /* To write VAL to EERPOM address ADR *
The IAR C-compiler does not include the option to define constants to
be downloaded in the EEPROM. The easiest way to do this is to use the
AVR assembler and write code for the EEPROM constants:
.ESEG
.org 0x50
datatable: .db
$01,$02,$03,$04,$05
dtableend: .db
$00
When this is assembled, it generates a .EEP file. This file can be
dowloaded into the controller or the emulator. The .org statement
gives the address of the data. The variables can be accessed with
EEGET() and EEPUT() in the C-compiler.
Question: Is the Real Time Clock(RTC) in the
ATmega controller Year 2000 compliant ?
Answer
The real time clock function of the AVR controller consist of an
asynchronous counter with an oscillator optimized for use with a 32.768kHz
watch crystal. The timer is set up to generate a periodic interrupt
that is
used to update the clock and calendar variables. The RTC can be made
Y2K
compliant by applying code that takes care of the millennium update.
Question: I am unable to get the Power Down
current consumption down to the value in the
databook. What should I do ?
Answer
When an AVR enters Power Down mode, the states of the I/O pins are left
unchanged. If you have some unconnected pins that are tri-stated (input,
no
pull-up), these pins are floating and will most probably oscillate
and cause
extra current draw.
To solve this, activate the pull-up resistors on all unconnected inputs.
This
will cause a defined logical level to be applied to these pins. Turning
on
pull-ups on unused I/O pins is a good practice also in active mode,
as
oscillating I/O pins can cause problems in some cases.
Secondly, you should review what is connected to your output pins.
Any load
on an output pin will draw extra current - also in Power Down mode.
Question: With my current microcontroller I
observe activity on the I/O ports during power up.
This is a problem for my application since the MCU is able to shut
off power to the system by
setting one of the I/O pins high. Even if this pin is pulled low by
a resistor, it sometimes goes active
high during power up. Consequently, when power is turned on, the system
shuts down immediately
and to the user it appears that the system does not work. If I switch
to the AVR, can I expect the
same sort of problem ?
Answer
Most microcontrollers have a synchronous reset, i.e. the MCU needs a
number of valid clocks while RESET is asserted to reset properly. During
power up, there is a chance that VCC rises to a value close to the
operating
voltage before the crystal oscillator starts. In this period, Power-On
reset
is active, but is not effective and the I/O pin values and directions
are
set to a random state. The I/O ports on the AVR are asynchronously
reset and
are guaranteed to be in tri-state during the entire power-up cycle.
Switching to an AVR will fix your problem.
Question: What shall I do with NC(No connect)
pins on the AVR controller ?
Answer
Do not make any connections to the NC(No connect) pins. The NC pins
are
reserved for future requirements.
Question: How do I define a string constant
in the AVR assembler ?
Answer
Strings can be defined both as constants in flash memory or in EEPROM
data memory.
Example: Define a string constant in flash:
.CSEG
fstring: .db "This is a string in flash",0x00
To define a string constant in EEPROM:
.ESEG
eestring: .db "This is a string in EEPROM",0x00
Question: My assembler does not produce an
Intel Hex file, what is wrong ?
Answer
By default, AVRASM generate Motorola S-Record file
format with file extension .ROM.
To change this to Intel Hex in WAVRASM, do the following:
-
Select 'Options' from the menu
-
At the box "Output file extension", write HEX
-
At "Output file format" click on "Intel intellec 8/MDS"
The assembler will now generate an Intel HEX file when assembling.
Question: I am working on a quite big assembly
program. When I add new lines to the
program, characters at the end of the file seem to disappear. What
do I do to prevent this?
Answer
The editor in assembler has a file size limitation of 30K. To overcome
this limitation, the file
must be split into several modules which is joined by using ".include".
Alternatively, use an
external editor and the command-line version of the assembler.
Question: When will the AVR Simulator for Windows
3.11 be upgraded to support new
devices?
Answer
This product is being phased out and is no longer supported by Atmel.
For debugging of
all peripherals and also new parts, use the AVR Studio, which can be
downloaded from
the web. We still keep it on the web for users that do not run Windows95
or NT.
Question: My new Pentium 200 MHz computer does
not recognize the IAR Dongle. It
works fine with my good, old 75. Is this a known problem?
Answer
Yes, on Pentium P133 and faster machines, the hardware lock might have
synchronization
problems. Add the following line to your AUTOEXEC.BAT file:
Set SSI_ACT =10,20,20
Restart your computer and try again. If the hardware lock is still not
working, try
increasing the numbers gradually in steps of 10 or 20 (the maximum
limit is 200). Not that
too high numbers, e.g. 200,200,200 also may cause communications to
fail.
Question: The AT90S8515 has got 8kB of program
memory. As the RJMP and RCALL
commands can take me 2K locations backwards or forwards, how can I
reach the entire
program memory? According to the Instruction Set Summary, this part
has no JMP or
CALL.
Answer
The program memory is organized as 4K x 16, so there are only 4k program
memory
locations. In the assembler, select "Options >> Wrap Relative Jumps".
This will enable you
to jump over the boundary of the program memory. As an example, if
you do a relative
jump from $FFE back to $00A, the program counter will be incremented
by 12, and wrap
around the boundary of the program memory. This feature is only applicable
to 8K
devices, as 4K devices do not require wrap-around, and 16K devices
need JUMP and
CALL.
Question: When I use the UART, are any of the
Timer/Counters used to generate the
Baud Rate?
Answer
No, the Baud Rate generator has a separate timer that is used by the
UART only. All
Timer/Counters can be used as normal even if the UART is operating.
Question: When I do a single bit set or clear
to the I/O with the SBI and CBI instructions,
will that effect other bits on the same port ?
Answer
No, unlike most microcontrollers available, with the AVR you will always
be 100% safe
when operating the I/O with single bit operations. This also applies
to operations on the
entire port. Refer to the question regarding three addresses for each
I/O Port.
Question: Why are there three addresses for
each I/O Port ?
Answer
To allow you to build 100% safe systems, the AVR supports true read-modify-write
to the
I/O ports. If you want to read the physical value on the I/O Pins,
read the PIN register.
When you want to update outputs, read the PORT latch to ensure that
you write back the
correct data to all outputs on the port. This method always gives the
desired result
independent of the physical level of the pins. This feature saves you
from having a copy of
your port in memory (and use many instructions) to build a safe system.
When you do a
SBI or CBI instruction to set/clear single bits in I/O you must always
use the PORT
address.
Question: Why does SBI and CBI only operate
on I/O registers $00 -$1F?
Question: Why do the immediate instructions not work on registers
R0-R15?
Question: Why is there no EORI instruction?
Answer
All AVR instructions, with a few exceptions are 2 bytes long. This
means that there is only
65,536 (64K) possible combinations to make an instruction set from.
When we specified
the AVR instruction set, some compromises had to be made to utilize
these 64K
combinations as efficiently as possible. Unlike a CISC controller,
where an instruction can
be one, two, three or even more bytes, we could not implement on the
AVR all the
instructions we would like to have. As an example, an immediate instruction
which
addresses all 32 registers, needs 8 bits for the constant and another
5 bits to address a
register file. This instruction would occupy 8K of the total instruction
space. In other
words, if we made eight immediate instructions which worked on the
entire register file, no
more instructions could be implemented. Making the instruction word
17 bits long, was
never been considered as an economical nor convenient solution.
The process of designing the AVR instruction set included a lot of gives
and takes. We
have consulted C compiler experts, who have given us a lot of advise
on how to tune the
instruction set to yield compact C-code. As an example, the compiler
experts advised us
to sacrifice the ADDI for a SBCI (subtract immediate with carry).
For those instructions that are missing, convenient work-arounds exist.
The code
efficiency of the AVR should prove that we have found a good compromise
between
which instructions to implement, and which to omit.
Question: I am using "sbr r30,3" to set bit
3 in register 30. This does not seem to affect bit
3, but bits 0 and 1 become set. What am I doing wrong?
Answer
"SBR" can be used to set more than one bit in a register (like "CBR"
can be used to clear
more than one bit in a register). The second argument in the instruction
does not specify the
bit number, but the mask to be ORed with the register. In your case,
$03 is ORed to R30,
and consequently the two lowermost bits are set. To set bit 3 only,
try one of the following:
sbr r30,$08
sbr r30,0b00001000
sbr r30,(1<<3)
All these lines will yield the same result. Just pick the one you find
most convenient.
Question: In my multi-tasking system, I need
to define flags that can be checked and
cleared without being interrupted. How can I do that?
Answer
Use one register per flag. To set a flag, use:
ldi flag,01
Alternatively, if you want to use a low register (r0-r15) for flag,
clear it initially after reset
and use
inc flag
each time you set the flag.
When the flag is checked and cleared, do this:
lsr flag
;shift flag into carry (flag is cleared)
brne flag_was_set ;branch if flag was set
As long as all your interrupt routines restore the status register before
exiting, this method is
completely safe.
Question: I need to clear a register and also
the carry flag. Can this be done with one
instruction?
Answer
Yes, for example to clear r10 and the carry flag, do the following:
sub r10,r10
Question: How do I implement a constant table
in the AT90S1200?
Answer
Since the S1200 does not have the LPM (Load Program Memory) instruction
the most
efficient way is to place small tables in the EEPROM. This will give
you a very efficient use
of the memory and you can access your table directly with the EEPROM
address register
and the EEPROM data register.
.ESEG ;define
EEPROM segment
.db my_var00=$45
.db my_var01=$4c
.db my_var02=$5f
.CSEG ;define
general code segment
If your table does not fit in the EEPROM, you can place the rest of
it in Flash like this:
.def tp =r16 ;table pointer
.def output=r17 ;output from table
table:
ldi output,$45
cpi tp,01
breq end
ldi output,$4c
cpi tp,02
breq end
ldi output,$5f
cpi tp,03
breq end
end:
ret
To access element number 2 in this table, do the following:
ldi tp,2
rcall table
Upon return from the table subroutine, "output" will contain "$4c".
Question: How do I add an immediate constants
to a register when there is no ADDI
instruction?
Answer
You simply subtract with the negative value. The following code example
adds 5 to register
20:
subi r20,-5
Note however that with this solution, the carry flag can not be used
to detect an overflow.
If you already have the constant in another register, you can use a
general ADD. If the
constant is 1, you can use INC to increment the register.
16-bit addition is done the same way. The following code example adds
$0b3c to r17:r16:
subi r16,low(-$0b3c) ;subtract low byte
sbci r17,high(-$0b3c) ;subtract high byte
Question: How can Flash and Mask ROM ever cost
the same ?
Answer
This applies when the MCU has got on-chip EEPROM. The EEPROM contains
a charge
pump and some sense amplifiers. These resources can be used by the
Flash as well.
Hence, when there is an EEPROM on the chip, Atmel can give you Flash
program
memory without increasing the price.
Question: To achieve single cycle execution,
is the external clock signal multiplied ?
Answer
No, the clock is used as is. No clock division or multiplication is
performed.
Question: How many clock cycles make up one
AVR instruction cycle ?
Answer
Only one. All it takes to e.g. add two numbers in the register file,
is one positive and one
negative transition on the XTAL1 pin.
Question: Why is the AT90S8515 called a 8K MCU
?
Answer
Most CISC controllers have instructions that vary in size. Some instructions
consist of a
single byte, whereas others can be up to four bytes long. Most AVR
instructions are two
bytes long (a few are also four bytes), so the CPU does not have to
fetch one byte at a
time and determine when to start execution. In our code size benchmarks,
we always
compare the number of bytes. Hence, a C program that fills 8K in an
80C51, will certainly
fit into an 8K AVR.
Question: Where can I find datasheets ?
Answer
They can be downloaded from www.atmel.com.
Alternatively, get a datatbook from your
local distributor or Atmel office. You can also send an email to literature@atmel.com.
Question: What added value can the AVR offer
over my current MCU?
Answer
More compact code, especially for C-programming
4 - 10 times higher speed
Program Flash memory, in-system downloadable
On-chip EEPROM, in-system downloadable
Lower Power Consumption
Question: What does "AVR" mean?
Answer
Nothing, it's just a name.
|