e021044.pdf

MICRO CONTROLLER

Microcontroller

Basics Course

part 1: the TASM assembler

By B. Kainka

This course is for everyone who always wanted to how microcontrollers

work and how to use them, but was always afraid to ask. It is intended

to explain the fundamentals, starting from scratch. The Elektor Electron-

ics 89S8252 Flash Microcontroller Board (presented in last month’s issue)

is used as the hardware platform.

Nowadays, we all take working with

computers for granted, and not only

that, we often work with quite pow-

erful equipment. The heart of a com-

puter is its processor, such as a Pen-

tium III. Relatively speaking, a micro-

controller is both much less and

much more than the processor of a

typical PC. It is less because it

processes smaller programs, uses

less memory and is usually much

slower. However, it is also more

because it already has many ele-

ments on a single chip that are

spread out over the complete moth-

erboard of a PC, namely working

memory, timers, interfaces and port

connections. What makes microcon-

trollers attractive is that is in the lim-

iting case, a complex problem can be

solved using only a single IC. Using

programming alone, anyone can pro-

duce a special IC that does exactly

what he or she wants — and at a rel-

atively low cost.

A microcontroller is thus some-

Figure 1. The 89S8252 Flash Board,

which is used in this course, is a general-

purpose microcontroller system.

Elektor Electronics

1/2002

MICRO CONTROLLER

thing like a logical circuit with many

possible inputs and outputs. What

this circuit does is determined by a

program. Perhaps you want to build a

digital counter, or would you rather

have a stopwatch? Do you want to

create a special logical gate, or per-

haps a universal clock generator? Do

you need to decode a complicated

digital signal or control a digital cir-

cuit? In all of these cases, a micro-

controller can help you. There are

many examples of problems whose

solutions previously required an

enormous board full of ICs and now

can be solved quite elegantly by a

single IC, namely a microcontroller.

Consequently, some knowledge of

programming is worth having. There

are many different approaches that

can be taken to achieve this goal.

The hardware basis for this

course is the 89S8252 Flash Micro-

controller Board described in last

month’s issue of Elektor Electronics

(see Figure 1 ). As already

announced, for programming soft-

ware we will use the following three

programming languages: assembler,

Basic and C. Our first experiments

will be carried out in assembler. Why

should we use assembler in particu-

lar? Isn’t it rather difficult, perhaps

too difficult for beginners? The

answer is no, since the initial exam-

ples will be very small and easy to

understand. The advantage or using

assembler is that it allows us to

work very close to the hardware, so

we can see exactly what is happen-

ing. High-level languages (such as

BASIC), by contrast, hide much of

what actually takes place.

In our first experiment, all we

Figure 3. The ﬁrst sample program in TASMedit.exe.

want to do is to switch the levels on

one of the microcontroller’s output

ports. After all, operating a switch is

the ﬁrst step in automation. Also, the

results can be observed using a volt-

meter connected to lead P1.4 via

connector K4 (see Figure 2 ).

In order to change the level on

this lead, we will use a small assem-

bler program. Put brieﬂy, assembler

is a notation used to write instruc-

tions for a processor or microcon-

troller. Every microcontroller has an

instruction set, which ultimately con-

sists of numerical values and associ-

ated functions. The following series

of six numbers represents a small,

complete program for an 89S8252

microcontroller — in fact, it is what is

called a machine-language program:

troller reads the numbers from the memory,

one after the other, and it then knows what it

has to do. In normal language, we can

express this as follows:

74 : So, I’m supposed to transfer a numerical

value to the accumulator

(that’s my memory) — but which one?

0F : Here it is: OF —– good, I’ve made a note

of it.

F5 : OK, now I have to write the value to a

which one?

90 :I see, the register for Port 1 is located at

address 90. There you are.

80 : And now I have to make a short jump —

but to where?

FE :Two bytes back from the location that

would have been the next one. OK, I’m

jumping!

80 : The same jump again — OK, I guess I’ll

just have to keep on running around in a

circle.

116,15,245,144,128,254

It is generally customary to write

computer programs using hexadeci-

mal numbers instead of decimal

numbers, since the former are easier

to read. In hexadecimal notation, the

above program looks like this:

This is how the microcontroller ‘thinks’ and

acts, since clever engineers have trained it to

behave this way. Ultimately, a microcontroller

is nothing more than a very complex circuit

made up of logic gates. This circuit responds

to the states of its input lines, which in this

case are the data lines connecting the pro-

gram memory to the central processing unit.

If we wanted to know what goes on inside

a microprocessor in detail, we would have a

lot of work on our hands. However, it is suffi-

cient for us to know the machine instructions

of a processor and be able to use them. This

means that the microcontroller itself remains

a sort of ‘black box’, whose inner functions

P1.7

P1.6

P1.5

P1.4

P1.3

P1.2

P1.1

P1.0

74 0F F5 90 80 FE

We have to write this sequence of

numbers into the microcontroller’s

program memory. We can use a pro-

gram called MicroFlash for this pur-

pose. Ultimately, the program num-

bers end up in the program memory

of the microcontroller. The microcon-

010208 - 2 - 11

Figure 2. The results of the ﬁrst exper-

iment can be checked using a volt-

meter.

1/2002

Elektor Electronics

MICRO CONTROLLER

Is it really necessary for us to

remember that a particular register

for Port 1 is located at position 90h?

After all, it would be nicer if we

could also write this as text. This is

easily done; we simply have to

define a certain bit of text as being

equivalent to a numerical value. The

assembler will then replace this text

with the corresponding numerical

value at every location where the

text is found. To make such a deﬁni-

tion, we use the assembler directive

‘.equ’ (equate). An assembler direc-

tive always begins with a full stop,

which informs the assembler that it

is not an assembly-language instruc-

tion. In the following example, the

word ‘P1’ is assigned the value 90h.

In the actual program, the value 90h

is thus replaced by ‘P1’:

Figure 4. TASM in the DOS window.

; ﬂash1.asm port output

we do not fully know but whose behaviour is

easy to understand. That’s how it is with

modern technology — it has become nearly

impossible to regard everything at all possi-

ble levels of understanding.

So, now we know that a microcontroller

has its own language, which actually con-

sists of nothing but numbers. However, there

is an obvious problem: the language that a

microcontroller can read easily and ﬂuently is

not exactly suitable for people. We are not

made to work with numbers, but rather with

words. Consequently, words (which are easier

to remember) have been devised to represent

the individual machine-language instruc-

tions. The programmer writes these words in

a text ﬁle, and a special program then trans-

lates them into the language of the micro-

controller. This program is called an assem-

bler, and the programming language is also

called assembler (or assembly language).

Assembler is thus a notation that you and I

can use to tell a microcontroller what it

should do, as in:

first line, the numerical value

0Fh (= 15), which is identiﬁed by the

‘#’ symbol, is loaded into the accu-

mulator a. The accumulator is a reg-

ister or memory with a size of eight

bits, so it can hold numerical values

between 0 and 255.

In the second line, the value in

the accumulator is then copied to

address 90h (= 144). At this address

there is a register whose leads are

routed to the exterior of the IC,

namely to the Port 1 pins. The word

‘sjmp’ (short jump) causes a jump in

program execution, in this case to

the location ‘loop’. The word ‘loop’

has been chosen completely arbi-

trarily and simply represents an

address, in this case a position in the

series of instructions. The assembler

treats such words, which are called

labels, as addresses and replaces

them with the appropriate numerical

values. The sjmp instruction can

cause a jump of up to 127 bytes

backwards or 128 bytes forwards. A

single byte is thus sufficient to spec-

ify the jump destination. Here the

jump is calculated relative to the cur-

rent position in the program.

The word ‘main’ at the beginning

of the program is also arbitrarily cho-

sen. The only actual assembler key-

words here are thus ‘mov’ and

‘sjmp’. We humans can easily

remember such words without the

aid of an electronic brain.

However, there is still a problem.

P1 .equ 090H ;Port 1

main mov a,#0Fh

mov P1,a

loop sjmp loop

.end

This listing also shows us something

else: the jump labels and newly

deﬁned words are all located at the

beginning of the line, and all assem-

bler instructions are located some-

what indented. Furthermore, there is

also a comment, which plays

absolutely no part in the translation.

A comment starts with a semicolon

(;).

The speciﬁc notation varies some-

what from one assembler to the

next. Here we are using the share-

ware assembler TASM (Table-Driven

Assembler, a program written by

Thomas N. Anderson). TASM is very

simple and can translate programs

for many different types of microcon-

trollers, as long as it has the appro-

priate instruction table.

At the end of the program there is

a loop, in which a jump to the desti-

nation ‘loop’ takes place, always and

forever. In other programming lan-

guages, such a situation would be

called a fatal endless loop, which is

practically equivalent to a crash. In

such a loop, the processor is in a

state that it cannot exit under its

own power. In general, it should

always be clear what should be

main mov a,#0Fh

mov 090H,a

loop sjmp loop

This is already much more readable. Actually,

here we only need to know two special

words: ‘mov’ and ‘sjmp’. Both of these words

are called mnemonics, which means markers

used in place of the actual machine-language

instructions. The word ‘mov’ (move) means

‘move’, ‘shift’ or ‘load’. Following it comes

first the location where something is to be

loaded and then what is to be loaded. In the

Elektor Electronics

1/2002

MICRO CONTROLLER

done once the current task has been

completed. However, in this case

this loop is very important. There is

only one task, namely changing the

state of the port. If we were to leave

the processor to its own devices, it

would execute commands that just

happen to be in the program mem-

ory and perhaps belong to a com-

pletely different program. Conse-

quently, a limit must be set by

means of an endless loop: this far

and no further! In fact, the processor

is trapped in this loop, with the only

means of escape being a reset. After

that, the same program could be

started again, or we could load a

new program and then run it. Load-

ing a program also takes place in the

reset state; the program memory is

thus not ﬁlled by the processor itself,

but by special functional blocks in

the microcontroller that program the

flash ROM. Each time the board is

switched on, a short reset is auto-

matically executed. Following this,

the microcontroller finds the most

recently loaded program and runs it.

Software

To load a program into the microcontroller on the 89S8252 Flash Board, you will need the

Windows program MicroFlash.exe , which can be found on the Elektor Electronics website

( www.elektor-electronics.co.uk ) on the Free Downloads page, see the list for the Decem-

ber 2001 issue.

For the programming course, the TASM assembler is all you need to get started, but

later on you will need the Rigel READ51 C compiler and the BASIC-52 Basic compiler. The

TA SM assembler is a popular program, which can be obtained together with TASMedit

from the download list for this issue on the Elektor Electronics website. Please register the

software and pay the programmer (T.N. Anderson) his well-earned fee. No payment is

required for the C compiler, which Rigel make available free of charge for private and edu-

cational use. This compiler can be obtained from www.rigelcorp.com . BASIC-52 is a Basic

interpreter created by Intel, which was mask-programmed in the program memory of an

80C52 microcontroller that received the designation 80C52-AH-BASIC. This microcon-

troller was used for nearly two decades by electronic engineers and programmers and

became internationally famous, mainly as a result of articles in Elektor Electronics.. Several

years ago, Intel ceased production of this IC, but they released the programming language

as open source for general use. The language has been further developed and also adapted

for use with other microcontrollers. Probably the most advanced version, V1.3, was pre-

sented in the February 2001 issue of Elektor Electronics and is available from Readers Ser-

vices on diskette (order number 000121-11 ).

The diskette for this course (Readers Services order number 010208-11 ) contains the

TA SM assembler, TASMedit and the ﬁrst sample programs, along with BASIC-52, Micr-

Flash.exe and a small Basic terminal emulator program with its own sample programs.

Number formats

The fact that we use a decimal number system is probably due to the fact that we happen

to have ten ﬁngers. The ‘natural’ number system for a computer is the binary system. The

hexadecimal system represents a compromise, in which the range of numerals runs from 0

to 15, with the understanding that the numerals above 9 are represented by the letters A,

B, C, D E and F.

Using

the TASM assembler

Now it’s time to get down to busi-

ness! What we want to do is to write

this first program, translate it and

send it to the microcontroller. For

this, we need some software. We

will use the well-known shareware

assembler TASM, which is located

on the working diskette for the

course in the form of a zip ﬁle; it can

also be downloaded from the Elektor

Electronics website. The file

TASM.ZIP must be unpacked into a

working directory on the hard disk

that will contain the program

TASMedit and the sample programs.

The special feature of TASM is

that is can be used for different

types of microcontrollers. For each

type there is a table of available

machine instructions, which must be

identified when the program is

started. This is done using a com-

mand line; in this case we use

Decimal

Hexadecimal

Binary

00h

00000000b

01h

00000001b

02h

00000010b

03h

00000011b

...

0Ah

00001010b

0Bh

00001011b

0Ch

00001100b

0Dh

00001101b

0Eh

00001110b

0Fh

00001111b

10h

00010000b

11h

00010001b

...

253

FDh

11111101b

254

FEh

11111110b

255

FFh

11111111b

In assembler programs, it is generally possible to choose which notation you want to use. If

you are describing how an 8-bit port is being driven, binary notation is particularly clear.

For example, the rightmost bit represents pin P1.0 and the leftmost bit represents pin P1.7.

Here eight leads require eight bits, or one byte.

When TASM translates a program, you can specify the format in which the results are to be

stored. In the binary format, only the bytes that represent the individual machine-language

instructions are written. A text editor cannot make any sense of such a ﬁle.

The Intel hex format uses text lines containing hexadecimal numbers. In addition to the

actual code, there is a start address and a checksum for each line. This format can also be

viewed as text.

TASM -51 -b ﬂash1.asm

ﬂash1.bin

to specify our particular example

program, the instruction table

TASM51.tab and binary output for-

1/2002

Elektor Electronics

MICRO CONTROLLER

mat. Naturally, not everyone likes to work

with command lines like this. Consequently,

they won’t be used at all in our course.

In general, we want to work only with

Windows, but TASM is still a pure DOS pro-

gram. For this reason, a Windows interface for

the program, called TASMedit.exe, has been

written. It includes its own editor and allows

the user to immediately see the result of the

translation, including possible error mes-

sages. The flash download tool for the Elek-

tor Electronics Flash Board is also integrated

into this program. No effort has been spared to

make things as easy as possible for course

participants!

The program has two text windows (see

Figure 3 ). On the left there is the assembler

source text editor. It can be used to manually

enter a program or load a program from the

hard disk. The TASM button starts the

assembler in the background. The window on

the right displays the assembler’s list file

along with any error messages that may be

present. Download progress when the pro-

gram is being sent to the microcontroller

board is also shown in this window.

Clicking on the TASM button first gener-

ates a file called Work.asm, which contains

the current content of the Editor window.

This working text is then translated. TASM is

called from the Windows interface using the

command line

Figure 5. Setting the property ‘close when done’.

about its work, but after the third

time or so it will only annoy pam-

pered Windows users. Consequently,

from now on we would like to have

the window be automatically closed.

This is no problem, since Windows

provides a solution. First click on the

TASM.EXE ﬁle with the right mouse

button and open the Properties

menu. Under Properties / Program

you will ﬁnd the setting ‘Close when

done’. Enable this setting (see Fig-

ure 5 ). Windows then generates a

link in the form of a file called

TASM.PIF. From now on, the DOS

window will automatically close

after TASM has ﬁnished its job.

Once a program has been suc-

cessfully translated, the RUN button

can be used to transfer it to the

Flash Board system and start it. For

this, you have to select a PC COM

port and connect it to the board’s

programming connector (K2). If you

have also looked after the most

important prerequisite (applying the

supply voltage to the Flash Board),

you can then start to test the pro-

gram. In the case of our first exam-

ple, all you have to do is to observe

the states of the Port 1 outputs in

order to see whether the result is

successful. Using a high-impedance

meter, you should see almost exactly

5V on port leads P1.0 through P1.3

and nearly 0 V on P1.4 through P1.7

(to be precise, around 30 mV ﬂowing

into ground).

In the ground state without any

program running, or following a

processor reset, all of the port leads

take on the High state, with a volt-

age of 5 V on each pin. This can eas-

ily be checked using an oscilloscope

or multimeter. The newly loaded and

started program changes the states

of four lines. P1.4 through P1.7 should

now be Low, which means that they

have a voltage of around 0 V, while

P1.0 through P1.3 remain High. The

program has transferred the value 15

(= 0Fh) to Port 1. This bit pattern can

be seen on the port pins.

TASM -51 -b -work.asm work.bin

This means that the binary format is always

used here. The result of the translation is

stored in a ﬁle called Work.bin. This is the ﬁle

that is read by the download module when

the RUN button is actuated. The assembler

also generates a ﬁle called Work.list contain-

ing the list ﬁle, which holds the translation in

a readable form. The fact that the same file

names are always used for the translation is

an advantage for experimental work with the

assembler, since the source text only has to

be saved after the latest attempt has been

successful. This means that we do not have

a whole collection of garbage data on the

hard disk from all the unsuccessful attempts,

but instead only the intentionally saved

source text and the work files for the most

recent attempt. If you forget to save the lat-

est version of the source text, or if the PC

crashes while you are working, you can

always use the Work.asm file to recover the

fruits of your hard work.

When TASM is automatically started from

the Windows interface, it appears in a DOS

window (see Figure 4 ). This window must be

closed before you can proceed. At ﬁrst, it may

be very enlightening to see how TASM goes

(010208-2)

This concludes our brief introduction

to working with the assembler. In the

next instalment of the course, we will

discuss small sample programs that

are primarily intended to help inves-

tigate the port properties of the

microcontroller. We will look at

inputs, outputs and achievable

speeds.

Elektor Electronics

1/2002

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: