Digital PDFs

DIGITAL-8-11-S

December 1965

59 pages

Original

2.4MB

Document:	digital-8-11-s-d
Order Number:	DIGITAL-8-11-S
Revision:
Pages:	59
Original Filename:	https://svn.so-much-stuff.com/svn/trunk/pdp8/src/dec/digital-8-11-s/digital-8-11-s-d.pdf

OCR Text

43:05

J<F_0_O

FZU_21_DOM

20....(YLOEEOU

.Ofl.<z><—2

wIan

WPPUWDIU<WW<2

U.Q....>....m|..._lm

UUvim

0.9.4.)...

ONES—A

NOCZUKNZA.

.u.»OQ.~>Z:<_—ZQ

oomuom>4.oz

7>><Z>mny

.<_>ZC>_.

g>mm>01cmm44m

PREFACE

The programs discussed in this manual,

though written on the Programmed

Data Processor—8 computer, can also be used without change on Digital's
Programmed Data Processor-5. This compatability between the libraries of
the two computers results in four maior advantages:
l

The PD P-8 comes to the user complete with an extensive

selection of system programs and routines making the full
data processing capability of the new computer immediately

available to each user, eliminating many of the common
initial programming delays.
2.

The PDP—8 programming system takes advantage of the

many man-years of field testing by PDP-5 users.

Each computer can take immediate advantage of the

continuing program developments for the other.
4.

Programs written by users of the PD P—5 and submitted to

the users' library (DECUS Digital Equipment Corporation
Users' Society) are immediately available to PDP—8 users.

CONTENTS

INTRODUCTION. .'

SUMMARY OF THE DZTAK SYSTEM

Interfacing Transducers
Interface Types

Input Signal Buffer

4
4

......................

..........

.............................................

System Inputs

........................................

System Outputs

...............................

............................................

................................................

Variables

........

Arithmetic Operators

Format Statements
GOTO Statements

.........................

Gray Binary Conversion

Clock

.........

...............................

....................................

THE DZTAK PROGRAMMING LANGUAGE

.....

.................................

I I

...............................................

DXTAK Statements

.......

......................................

Format Statements

....................................

Conditional Program Section
ERROR MESSAGES

.................

Variable Names

.....................

Multiplexed Analog—to-Digital Conversion Input

....................................

..........................................

Serial-Parallel

Time

............................

Digital-Parallel Input Signal Buffer

Programming

...........................................

‘

....................

........

...........................................

..
'

.........................................

............................................

Compilation
Run Time

CONTENTS (continued)

Appendix

Page

FLOW CHARTS

PROGRAM EXAMPLES

..................................

.............

........................................

vi.

A23

CHAPTER

INTRODUCHON

The speed of operation and powerful operation code structure of the
make possible a unique programming system for use in data
This system, called DKTAK (for Ea Equisition),

Programmed Data Processor-8

acquisition situations.

permits a complex, program-controlled,

data acquisition system to be adapted to a particular experimental environment through the
use

of a highly sophisticated and precise pseudo code.

The necessity of extensive machine

programming to meet a particular data acquisition requirement is thereby eliminated.
In addition to its data acquisition applications,

DKTAK furnishes the experimenter with a means

of calibrating transducers and is a powerful aid in troubleshooting a complex data-gathering

system.

Programs to accomplish these objectives may be written and readily revised in the

DXTAK language

The D—ATAK system, when used to record experimental data on paper tape,
that may be used as input to a FORTRAN program

produces a tape

Thus the program for the processing of

experimental data may be coded in a widely recognized compiler language if the user so
desires.

A typical

D'ATAK equipment configuration is illustrated in Figure l.

DKTAK does not, of course, require all of the equipment in Figure i.

If, for example,

given system did not include a plotter, the PLOT command would never be used.

Finally, the DXTAK system may be readily adapted to special—purpose peripheral devices by
specification of input/output transfer (IOT) commands appropriate to a given device. Thus a

Type 350 Plotter might be replaced by a more elaborate plotting mechanism if desired.
The DZTAK

language presents for the first time

unified, systematic approach to the data

acquisition, calibration, experiment modification, and troubleshooting requirements of the
scientist engaged in experimental analysis.

The recent use of digital computers in industrial

EEEIS‘ELTRAIN

—v

PARALLEL
DIGITAL
INPUTS

SERML T0
PARALLEL

CONVERTER

SWITCH

MULT'PLEXER

A-To-o
CONVERTER

POP-3

ASR-33

_—_.

CLOCK

HIGH SPEED
TAPE PUNCH
75A

DECTAPE

Figure l

PnggER

CONTROL
552
DUAL OECTAPE
TRANSPORT
555

PARALLEL
STATIC OUTPUTS

Typical Data Acquisition System

process control has shown that it is a valuable tool for acquiring

information, both digital and

analog, and making logical decisions concerning the acquisition of data while concurrently

storing, processing, 'and displaying that information. Certainly

method of data manipulation

similar to industrial control techniques would be advantageous to a scientist conducting an

experiment. DKTAK is a programming system designed precisely for these purposes; this manual

explains how to use the D-ATAK system and language.

CHAPTER

SUMMARY

THE

DRTAK

SYSTEM

INTERFACING TRANSDUCERS
Three basic types of interfaces are used with the PDP-7 and PDP—8 computers in order to rye-3
ceive data from environment sensors.

These interfaces allow data to enter the computer under

control of a simple real-time symbolic compiler that gives the investigator the Following flex-

ibility when he samples the environment:
1

A variety of preprogrammed interface devices that easily connect the

computer to instrumentation.
2.

Simple symbolic assignment of identifying names to physical input var-

iables such as pressure, temperature, etc.

Absolute control in sampling the experimental environment with respect

to time

Computer compatibility with respect to rapid respOnse sensors using up

176 separate sensing devices.

Capability to make logical decisions concerning the acquisition of data

while sampling the environment.

Storage of data for future computations as well' as immediate output for

checking quality while obtaining data.
Transducers are sometimes considered unique devices that do not lend themselves to being connected directly to a computer; however, by the use of basic standard interface

types,‘ most

instruments can be connected directly to a computer with little additional equipment.
us

examine the basic types of interfaces that would be

to various

transducers

necessary to interconnect

'Let

computer

INTERFACE TYPES

Digital-Parallel Input Signal Buffer
This buffer permits the direct parallel insertion of a digital number into the computer, using a
method by which the number immediately becomes a computer word.

Examples of devices

feeding data in by this method are shaft—position encoders, switch registers, and other allied
devices.

Figure 2 shows the general method for digital-parallel input.

———.

——+—->
—-—-—.

—-—~>

..__._.

“r.

TRANSDUCER-4

L-

COMPUTER

——>

—---9

-—_>

._._.

————>

“M'

US$822“
Figure 2

GATING

Digital-Parallel Signal Buffer

This method can accommodate signals or pulses from a minimum range of O to -l0 mv up to a
maximum of 20 to —15v.-

The system can include one buffer for each of several dozen variables.

serial-Parallel Input Signal Buffer
This method would. be used, for example, in telemetry applications where the transmitter is

remotely located and able to transmit a number of input words one bit at a time along
a

single conductor cable or by radio (see Figure 3).

TRANSDUCER

SERIAL

DEC TYPE
xam

PULSE TRAIN

WNGllllllllllll

llllllllllll

TTlllllllTlT
COMPUTER

Figure 3

Serial-Parallel Signal Buffer

This buffer can convert a serial pulse train to a l2-bit word in 6 psec,

or one

bit every 500 nsec.

lt accommodates ranges of input levels from a minimum of 0 to —l0 mv, up to a maximum of 20
to

The flexibility provided in this buffer allows the user to format the data before it is

—15v.

finally assembled as computer words; that is, he can insert octal constants in the specific serial
It provides the programmer with the following additional instructions in

word of his choice.

the computer:

Skip if data flag

Skip if start flag

Clear data flag and start flag.

Read data into the accumulator.

Multiplexed Analog-to—Digital Conversion Input
This method is particularly advantageous in data acquisition when many devices such as therm—

istors, pressure sensors, and strain gages are working together.

One example of where this

method would prove advantageous is a thermistor chain in which each thermistor, pressure
sensor,

conductivity sensor could be individually sampled by the computer. Figure 4 indicates

this relationship.

‘

SENSORS

THRU

"2:112:50?“ —->

ANALOG TO
DIGITAL
CONVERTER

i—u

COMPUTER

Figure 4

Multiplexed Analog-to—Digital Conversion
5

Switching from one input to the next in the multiplexer is accomplished in 2 psec, and up to
64 separate
uses

inputs can be sampled directly by the computer. The analog—to-digital converter

the range of O to -—lOv.

PROGRAMMING
The task of writing a special-purpose program for each installation could be a formidable prob—
lem involving a great deal of time and money.

In order to alleviate this problem,

Digital

Equipment Corporation has written DKTAK, an algebraic compiler that allows the experimenter
to

format his data' acquisition problem in a simple language similar to algebra.

By using the

DKTAK language, he is given a- great deal of flexibility concerning the interface hardware
that he has available.

This program allows him flexibility in choosing the frequency and con-

ditions under which he samples the experimental environment .as well as making possible the

adding or changing of sensors without the major task of reprogramming in machine coding.
This program is available for the
time

PDP-8, a compact 12-bit computer with a l.5-psec cycle

In summary, this program allows the investigator to analyze his sample using the following

inputs.

Up to 96 independent data variables using digital-parallel input.
Up to 64 independent data variables using a multiplexed A—to-D converter.
Up to 25 independent data variables via serial buffer input.
Each independent variable can be sampled at a rate of up to lOO times per second.

System lnputs
Data can come to the computer from the digital-parallel
and the multiplexed analog-digital converter.
a

input buffer, the serial input buffer,

Each Of these devices is

assigned

symbolic name that tells the computer which device is transmitting information.

The symbols are:DGIN:

Digital-parallel signal buffer; input can be converted from Gray
code to binary.

,BUFR:

Serial, buffer-input.

ADCV:

Multiplexed analog4digital converter:

System Outputs
Data output can be distributed to a number

iof‘specificdlly named devices to allow. immediate

presentation as well as permanent storage. The following output symbols and their associated
devices are available:
I

On-line teleprinter.

TYPE

Variables can be typed in decimal Or octal.

Decimal is specified by 1
name

-QPNCH:

High-speed paper tape p’Unehx.; FVariablescanibe*pbnched~:invdeci—“
mal or octal.

Decimal is specified by 1 immediately following.

the_variab|e name
DCTP:

DECtape in binary with; identifying words 5;

X—Y plotter.

J-

The plotter pen is moved to a new position each

time an output is -spec'ifi'ed'.»-'

DIG l

Digital's compact DECtape" Variables are "recorded magnetically
ron

PLOT:

immediately following the variable

“

Up to four digital outputs are available through parallel buffers“...
‘

to other devices such as

DIG 4:

relays, buffers, sense lines, and range

sWitChin'g devices .>
Variables

Input variables to the computer are aSSigned alphanUmeric symbols by' the investigator. They
can

be one to four characters long, and must begin with a lettert-“Some exampleszofthese

Would be:
T 123,

SURF, DEEP, AIR, X, Y, TEMP, H20, H202

In addition, variables that are inputted through the multiplexer have specified channels; that

is, T 123 (i) would be input through channel 1 of the multiplexer.

Time
Four- types of time can be used by the computer:

basic, program, variable, and reference.

Basic Time
Basic time represents the basic interval of .01 sec in which a clock interrupts the program.

Program Time
This time represents the basic rate at which the investigator desires to interrogate the sensors.
It is some multiple of the basic time and is under program control.

‘

Its symbology is simply ex-

pressed as fol lows:
QUNT: 50

The investigator has specified that the program time will be (50-)
x

(.01)

0.5. sec.

0.5 sec; that is, each sensor will be sampled every
If he desires the fastest rate possible, he may express

the. following:
QUNT: 1

In which case each variable will be sampled every 0.0] sec.

Variable Time
In order to allow more flexibility in timing, digital inputs can be sampled at a slower rate than

the program time specifies.
DGIN:Ll

For example, the expression:

(4, L2 (4

specifies that the digital input variables Li and L2 should be sampled
the program time or every (4) x (0.5)

2 sec.

once

in four

cycles of

Reference Time
It is often desirable to know the reference time in order to associate data with time.

Within

the program is a 3-word variable, CLOK, which counts the number of seconds, minutes,_ and

hours that have elapsed since start-up time; it can be used as an output variable to reference
data with time.

Arithmetic Operations

Addition and Subtraction
Variables can have constants added to or subtracted from them as they are sampled, or the

variables can be added to or subtracted from each other.

All arithmetic operations are done in 2's complement arithmetic, with the operands being con-

fixed—point numbers. The following examples mean that the variable T2 will

sidered signed,

have constants or variables added or subtracted before output:
Add a constant to the variable.

T2 + 137

Subtract a constant.

Add a second variable.

T2 + T3

Arithmetic Comparison
Variables can be compared against constants, compared against other variables, or compared

against themselves with respect to sample time. The basic comparison instructions are:
lFEQ X, Y

if X is equal to

IFLS

if X is less than

X, Y

IFGR X, Y

if X is greater than Y

An example of the comparison of a variable against a constant would be:
IFEQ X,

1000;

meaning that is X is equalt010008, execute the operation following the semicolon; otherWise,
go to the next

line.

Every time a variable is recorded and outputted, its value is preserved and is given the name
of the original variable.

Thus, X and @ X represent the current value of the variable X, and

its value when last used as output,

respectively.

The following example of the comparison of a variable and its predecessor:
lFLS X, @ X;
means

that if X is less than it was when last recorded and output, execute the operation fol-

lowing the semicolon; otherwise, go to the next line.

Gray Binary Conversion
Gray binary code can be converted to simple binary under program control if the input method
is digital

(DGIN). This provides the investigator with a rapid means of conversion in order to

intercompare the usual shaftFencoded Gray binary numbers, if the shaft encoder does not convert from

Gray Code to simple binary prior to buffering.

The conversion is accomplished by inserting 1 immediately before the time multiple.

DGIN

Li M

This means that the Gray binary variable Ll is sampled every fourth time through the program
and is converted to a simple binary number before comparison or storing.

Format Statements

Format statements are numbered from l-l5 (decimal).

They containthe namesof variables and

their 'output forms (octal or decimal for the Teletype or punch).
with a device name in every output. statement.
FORM:

Format numbers appear along

Thus, the statement:

l,X,Y‘l,Z

indicates that the variables X, Y, and Z are to be outputted to the Teletype, with X typed in

octal, Y typed in decimal, and Z typed in decimal.

GOTO Statement

Program control can be unconditionally transferred through the use of the GOTO statement.
l0

CHAPTER

DATAK

THE

PROGRAMMING

LANGUAGE

Every DKTAK program consists of four parts:
.

Assignment of input devices.

Definition of output formats.

Unconditional outputs, i.e., outputs controlled by the clock.

The

conditional program section. This is illustrated in the short DZTAK

program be low:

QUNT:

62‘

DGIN:

SWIT (2

INPUT ASSIGNMENT

OUTPUT DEFINITIONS

FORM:1,CLOK,SWIT,X,Y,Z
FORM:7,CLOK,SWIT
<7, PNCH,2>

UNCONDITIONAL OUTPUTS

[5:

CONDITIONAL
PROGRAM
SECTION

Y=I+(x=SWIT-Io):77
Z=¥—(X+]7);OUTP(1,TYPE);OUTP(1,DCTP)
IFEQ@SWIT,SWIT;GOTO 6
GOTO 5

1
END STATEMENT

END

The maximum hardware configuration for which DZTAK is designed is illustrated in Figure 5.

DATAK STATEMENTS

Clock
All sampling of the input is controlled by the clock, except for the serial-parallel cOnverter
which sends a signal to the PDP—*8 when it has twelve bits in its buffer.

Output to the devices

may be unconditionally controlled by the clock or may be initiated when certain prescribed

conditions are met.

When a variable or quantity is output, its value at output time is stored

and hence may be compared with incoming values to detect changes.
is .Ol sec.
ll

The basic clOck cycle

PULSE TRAIN

up TO
300 BITS

{5%

fri o
I
I)

SERIAL
'2 B'Ts

fgLSEAEKEfE'i

63.3 cps
PUNCH

“P333“

—

‘2 9'T5

CONVERTER

100 cps
CLOCK
ANALOG -To-

DIGITAL
CONVERTER

ON- LINE
TELEPRINTER

l
64-CHANNEL
MULTIPLEXER

TTT

64 ANALOG INPUT

STATIC OUTPUT

12-Bl
STATIC OUTPUT

12- BIT
STATIC OUTPUT

‘

2—BIT
STATIC OUTPUT

'
Figure 5

Maximum Hardware Configuration

The sampling period for The analog-digiTal converter is some mUlTiple Of This as specified by
The sTaTemenT:

QUNT:

XXXX

where XXXX is a given decimal inTeger.

QUNT:
means

Leading zeros need noT be specified.

For example:

ThaT all specified Channels of The analog-'digiTal converTer are To be sampled every

0.05 sec and The values sTored.

Variable Names
"

Variable names may be up To four characTers long, musT sTarT wiTh a leTTer, and may conTain

only leTTers and numbers:

T723

SYMB

HEAD

SWIT

Each variable is assigned two unique addresses.

the most recent input value.
was

last output.

One is used by the input routines and contains

The second address contains the value of the variable when it

The output variable may be addressed in the conditional program section by

preceding the variable name with "@."

This may be used for comparison purposes, etc.

Variable names are assigned to input devices in the following manner:
1.

Analog-to—Digital Converter
ADCV:
Followed by variable names and channel
ADCV:

numbers; thus:

T723 (6), HEAD (4)

This means that channels 6 and 4 of the analog converter—multiplexer will
be sampled and the values stored in registers T723 and HEAD,

respectively.

These will be sampled at some multiple of .Ol sec, where the multiple, is"

specified by a QUNT:

statement.

QUNT:

ADCV:

T723(6), HEAD(4)

Channels 6 and 4 of the converter-multiplexer will be sampled every .05 sec.

Serial- Parallel Converter
BUFR:

The serial-parallel converter sends a start pulse to the compUter, followed by

followed
in the order in which the 712—bit Words are received. llf

data pulses every time twelve bits have been assembled.

by variable names

the computer receiVed.

BUFR:

If the computer receives more data pulses thanthere
If more variables

are

variables, additional data from the converter is ignOred.

are

specified than there are data pulses, the remaining variables are not used.

Each start pulse causes the list to be reset.‘
BUFR:

SYMB,TH|S

After the start pulse from the buffer, succeeding data pulses cause the 12—bit
If a third data pulse is received

words to go into registers named SYMB, THIS.

before the start pulse, it is ignored.

Digital Inputs
The program is capable of handling up to eight l2-bit digital inputs.
are

These

sampled at periods that are multiples of the period specified by the QUNT:

statement.

These inputs may be converted from 12-bit Gray code to l2—bit

binary if specified.
_

QUNT:

DGIN:

SWlT(6, CODE12

This means that the first digital input (usually the switch register) is to be

sampled and stored at a period that is six times the period specified by
QUNT: 2 or every .12 sec.
'

The second digital

input is sampled every (2)

(.02) sec or .04 sec and is to be converted from Gray code to binary before
it is stored.

This is useful when sampling shaft encoders, etc.

'4. Digital Outputs
Output is specified with a format number and a device name.
ments contain the format number and

list of variables.

Format state—

Output form de-

pends on the output device.
relay buffers, etc. The first

DlGl

These are static output devices; i.e.,

3:33

variable in the specified format statement is output to the

DIG4

appropriate device asla l2-bit binary number.

Teleprinter
The entire list of variables in the specified format statement is

TYPE
I

typed in octal (or decimal) on the 33 ASR.
ceded by two

decimalidigits which

Each output is pre-

are the format

number, followed

by the values of the variables separated by spaces, and finally
a

carriage return-line feed. Decimal output is specified in the

format statement.

High-Speed Paper Tape Punch
PNCH

Output format is identical to the teleprinter format.

Since the

farmat number starts every line, this tape maybe processed with
the PDP-8 FORTRAN SyStem.
D EC tape

DCTP

The datum is

recorded

DECtape using block lengths of 20l8.

The first word of each'block indicates whether 'or not it is the

last block.

The second word contains the number of 12-bit words

of data on the black.
a

Each command to output to DECtape causes

12-bit identifier word to be placed on the tape followed by 12-

bit

binary words. Decimal specifications in format statements-are

overridden.

When the DECtape nears the end of the tape,

message

is typed and a fresh tape may be mounted on the second transport.

,X-Y Plotter
PLOT

The first tWO variables in the specified format statements represent,

respectively, the X and Y coordinates.

The plotter pen is moved

’

from its previous position to this specified
set to
OI'GCI

position-.11, It is initially

location (0.0)-—the lower left—hand corner of the plotting

Format Statements
Format statements are of the following form:

FORM:

6,SW|T1, HEAD

This is format statement number 6 which consists of the variables SWlT and HEAD.
to

the teleprinter or punch,

SWIT is Converted to decimal when it. is outputted.

There is a fixed system variable called CLOK.

tively, the hours, minutes, and seconds.

If output is

This is a 3-register variable containing, respec—

The seconds are initially set to 0, and thevhours

and minutes are set from the switch register when the execution of the program is initiated.

When output is to the teleprinter or to the'punch-and CLOK is in the specified formatstatement,
it is outputted as three 2-d-i’git decimal numbers.

When it is output to the DECtape, it is

written as three 12-bit words. CLOK may not be output to the plotter or to any of the four

digital outputs. However, the following is permissible:
‘

FORM: 3

1-2, X,Y,'CLOK

with format statement 12 being specified in output statements to the punch and to the plotter.

Output may be initiated unconditionally (i.e., by the clock) or Conditionally.

Unconditional

output statements contain the format statement number, the device name, and the output period

(a multiple of the clock period established by the QUNT: statement).
The unconditional output statements are delimited by angle brackets < .>.

Example:

‘

A program to:

l -.

Sample the contents of the switch register every .5 sec;

:Punch this vaer in octal every .5 sec; and

3'.

Type the value in decimal, along with the time, every 2 sec;

could be programmed as follows:

QUNT:

DGIN:
FORM:

swna
7,CLOK,SWth

FORM:

6,SW|T

<7, TYPE, 4
6, PNCH,]>
'

END

'The first statement, QUNT: 50, established the time interval as

The second statement, DGIN:

whichIs to be

SWlT(l

identifies the first

sampled every (l) (.5 sec)

.5 sec.

(50)

digital input

(.Ol sec):
as

5 sec.

the variable SWlT

Decimal output is specified for SWIT

The next two statements identify the output variables.
in format statement number 7.

The unconditional output statements are interpreted as follows:

7,TYPE,4

output to the teleprinter under control of format statement number 7 every (4) (.5 sec)

2 sec.

6, PNCH,l output to the punch under control of format statement number
6 every

(1) (.5 sec)

.5 sec.

Format numbers are decimal and may be from l—l5.

‘

Conditional Program Section
The conditional program section consists of algebraic and control statements that may be used
to

initiate output, test for tolerance levels,

alter variables.

There are four types of statements allowable in this section of the program:
1.

Algebraic Expressions
:

Variable

Variable Operator Variable

For example:
:

Y=A+B.

These statements may be nested to any reasonable degree:
:

Y=A+(B=C+D)

There are four basic operators:
A+B

2's Complement Addition

Arithmetic Operators

A+B

2's Complement Subtraction

Arithmetic Operators

A.'B
A<-B

Inclusive OR
Clear bit. For every bit in
B that is a l, clear the cor-

Boolean Operators

responding bit in A. AAB

Boolean Operators

Truth tables for the logical operations are as follows:

A
0

E
'0

£13.

A___"3

Examples:
'

5:8

5:3

.513

_A~.B_

6000
0077

7000

5000

7000

001 7

Cl 16

0060

7000
0077

0060

Expressions within parentheses are evaluated first.

0000

The Boolean operators

have a higher priority than the arithmetic operators:

Y=A+B?—C

is interpreted as if it had been written:

Y=A+(B«C)

The equal

sign (=)

means

replace the single variable on the left with the value

of the expression on the right.

X=B+_(Y=C-D)

is perfectly valid.
same

level.

Thus:

Two equal signs may not be used in one expression at the

Thus:
:
‘

A=B=0

although:

is not valid,

A=(B=0)

is valid.

‘

An arithmetic statement is terminated by either a carriage-return or a semicolon (;).

The single quote (') acts as a line continuer as it does in all pro-

gram sections

Y=A+ B;

Z=Y+ D‘—A

is a valid line and is evaluated from left to right.

Unconditional Transfer

Any statement may begin with a statement number delimited by a colon (:.)
There is one transfer instruction:
:

which says:

GOTO XX

transfer control to statement number XX.

Example:
6:
:

Y=A+(B=C-D)
GOTOé

The GOTO statement must be the last statement on a line, although it need
not be

the first.

Conditional

Y=A+(B=C-D) ; GOTO 7

Expressions

In addition to the unconditional transfer instruction, there must be condi-

tional expressions.

Conditional expressions are of the following general

form:
'

FUNCTION EXPRESSION,

EXPRESSION;

If the conditions are met, the instruction following the semicolon is executed;
if they are not met, execute on the next line.

The expressions may be a

single variable or a statement within parentheses.

The conditional functions

are:

lFEQ

X,Y;

If equal.

lf X=Y, the statement is true.

arguments are considered unsigned.

The

IFLS

X,Y;

If less than.

IF X is less than Y in absolute

value, the statement is true.

The arguments

considered to be signed integers.

are

Thus:

7770 <2H

IFGR

X,Y;

iswgreater than Yin absolute value, the statement is true. The argu—
If greater than.

considered to be signed

ments are

IFBE

X,Y;

If X

if bit equal.

integers.

IF there are any corresponding

1's in X or Y, the statement is true.

IFBE l77,l; is true.
For example, if:
6:

IFEQ A,B; GOTO 6
GOTO 5
V

This waits in a loop until A and B are unequal.
6:
:

IFEQ A,B; GOTO 5
GOTO 6

To reverse it:

This waits in a loop until A and B are equal.
4.

Output Statements
Output statements are expressed in the conditional section as follows:
OUTP (format number, device name)

Whenever a variable is outpUt, its value at output time is recorded in a loca—
tion that has,

as a

name, the original variable name preceded by

"@.

The conditional program section is delimited by square brackets [ ]
The last statement in a program is END.

CHAPTER

ERROR

MESSAGES

COMPILATION

During compilation, there may be two diagnostic messages:
LANGUAGE ERROR

There was some error in the source language.

Since

the source language is typed out as it is processed,
the programmer can see where the error occurred.
5 # NOT FOUND

A statement number was referenced by a GOTO
statement in

the conditional program section, but

has not been defined anywhere in the program.

Both of these errors cause the compiler to halt.

It must be restarted at

2008.

RUN TIME
At run time, there may be several diagnostic messages:
EX INTR

Extraneous interrupt.

Some device that is not used

by DATAK has caused an interrupt.
INTR OVR

Interrupt overflow. DATAK was interrupted at a
rate

FORM OVR

higher than its processing 'rate.

Format overflow.

The output devices are incapable

of handling the specified output rates.

The format

list is about 30 positions long.

STACK ERROR

Indicates an error in the arithmetic statements that
was

The above errors are catastrophic; that is,

not detected

by the compiler.

recovery from these errors is possible, and the

object system halts.
2i

DT ERROR

The DECtape error flag was raised For something

other than an end—zone condition.

DZTAK will

clear the flag and attempt to proceed.
TIMING ERROR

The. DECtape buffer was filled before the second
buffer could be written.

DKTAK will continue

although some information will be lost.
READY TO SWITCH

DECtape block number 2400 is being written.
The write routines will switch from DECtape
unit 1 to DECtape unit 2 at block number 2700.

The next time the switch will be from DECtape
unit 2 to DECtape unit 1, etc.

APPENDIX

FLOW

CHARTS

EXPLANATION OF FLOW CHARTS

DATAK consists of two basic elements: the compiler and the operating system.
of the compiler is to read the source

The function

language from the ASR-33 Paper Tape Reader or Keyboard.

The compilerproduces the structured lists and other elements that control the operating system,
and compiles the
at run time.

halts.

logical and arithmetic statements into an interpretive code that is executed

When the compiler reads the word END, it initializes the operating system and

The switch register is set to the current time of clay and when the CONTINUE

Key is

depressed, the CLOK registers are set to this value and the operating system is started.
The operating system consists of three major sections:
1

The arithmetic

interpreter, which executes the interpretive code generated

by the compiler from the arithmetic and logical statements.
2.

The output controlling routines, which call routines for specific devices

and initiate output through use of the interrupt system.

The interrupt system, which handles all of the actual output operations,

counts elapsed

time, and initiates sampling and unconditional output.

The arithmetic interpreter may call the output controlling routines (conditional outputs), as

may-the clock service routines (unconditional outputs).

Either type’of routine may be inter-

rupted except for the DECtape flag servicing routines, which are accorded the highest priority.

Page A3 of the flow charts shows the overall flow of events at run time.

After each interpre-

tive instruction is executed, the status of the output waiting list is tested.

If there is an out—

put word waiting to be processed, it is transmitted to the appropriate device handling routine.

Page A4 shows the overall flow of the compiler.
Pages A5 and A6 show the list structures used by the input/output routines.
Al

Page A7 shows the interrupt servicing technique.

”the. interrupt

was

not caused

tape; the C(AC), C(L) and the return address are pushed down onto a list.

causing the interrupt has been determined,

it is cleared and’the

Page A8 shows the clock service routine.

It counts to 1

registers.

It also counts to 1 time quantum

by the DEC-

When the flag

interru‘lpt is reenabled.

before incrementing the CLOK

sec

(specified by QUNTz) before testing to see if any

input or output is to be initiated.

It may call the routines ADCSER (analog-digital converter

service routine), DISERV (digital

input service routine) or 95191 (unconditional output con-

troller).
Page A9 illustrates the serial-parallel buffer service routine.
Page A9 shows the teleprinter and punch service routines.

Both of these routines have buf—

fers which are cleared before new data is read in.

Pages A10 and A11 show the routines called by the clock service routine to initiate sampling
or

output.

The digital

input routines contain a list of eight IOT's that the user may define to

correspond to his particular equipment configuration.
Page A11 shows the output-control routines. FORMAT is used to place an output word on the
waiting list and to test for overflow. TFORM tests to see whether or not the waiting list is
empty.

If it is not empty, the top word on the list is saved and all others are moved up one

position on the list.

FOP1 is then called to operate on this format-output word.

empty, TFORM exits.
to test the waiting

FOP1 calls the appropriate device routine and then returns to TFORM

list status.

Page A12 contains the digital-output routines.
one

If the list is

——

each for DIGl,

The user may define up to four lOT instructions,

DIGZ, DIG3, and D104, to correspond to his equipment configuration.

Pages A12 through A15 contain routines used for teleprinter and punch output.

Pages A15 through A20 contain the DECtape output routines. DECtape uses two buffers in core
memory and one may, be filled while the second is being written.

The write routine

determines whether or not the tape is nearing the end; if it is, the message "READY TO SWITCH"
is typed, and

3008 blocks later write-out is continued

another DECtape unit.

The DECtape

search routine is informed of the setting of the DT flag by the interrupt service routine lOSERV.

Page A21 contains the plotter service routines. The plotter attempts to draw a diagonal line
from its previous position to its new position.

Page A22 contains the arithmetic interpreter.
tions.

It calls routines to carry out indicated opera-

This is the main "background" program for DZTAK.

If there are no arithmetic state-

D-ATAK executes a loop in the interpreter.

ments in the source prOQTam,

INTPGO
EVERY
INSTRUCTION

TFORM

“$031338
0

ARITHMETIC
INTERPRETER

LIST EMPTY

LIST

(ARITHMETIC AND
LOGICAL STATEMENTS)

FOR I
CALL OUTPUT
CONTROLLING ROUTINE

EXECUTE INTERPRETIVE

INSTRUCTIONS; CALL

EXECUTION ROUTINES:
DIGITI

(STATIC)

DIGITZ

I FE

XFER

LOAD
STORE

DIGIT3

IFG

DIGIT4

IFBE

PLOTF

(START MOTION)

PUFOR
TTFOR

BUFFERED
OUTPUT

—+

OUT

SL3.
5TB

SUBT

DECTPE

ADD
DECR

FORMAT
PLACE FORMAT WORD

ON WAITING LIST

UNCONDITIONAL
OUTPUT

ADCSER
SAM PLE
DIGITAL
INPUTS

CLOCK SERVICE
I IN TE RR U PT EN Aa LE D )

smngsAriaLge
STORE

TELETYPE SERVICE
ROUTI N E

PROGRAM INTERRUPT
SERVICE ROUTINE

PUNCH SERVICE
ROUTINE

DECTAPE HAS
HIGHEST PRIORITY

SERIAL BUFFER
SERVICE ROUTINE

INTERRUPT IS ENABLED
AFTER EACH PROGRAM
INTERRUPT

Figure A]

DECTAPE SERVICE

ROUTINE

'_
'

lQS—E—R—v
f

Overall Flow (Run Time)

DISMISS
INTERRUPT

PROGRAM
INTERRUPT

START
INITIALIZE
ALL LISTS

25w

(INPUT SYMBOL)
GETSYM
DISPATCH

COMPILE
BUFR:

UCGOI
<

UNCONDITION
OUTPUT
COMPILER

COM PILE
DGIN :

235m

ADCV
COMPILE
V ADCV:

ACOMP
c
'

ARITHMETIc
COMPILER

(END)

COMPILE
FORM:

STOPCM

59m;

INITIALIzE
OPERATING
SYSTEM
COMPILE

OUNT

QUNT:

(CARRIAGE -RETURN)

HALT

READ

TIME

FROM SR
START
OPERATING

l Tom
Figure A2

Overall Flow (Compile Time)

CONTROL LIST FOR ADC

ATABLE.

CHANNEL #
ADDRESS
......

1281°
POSITIONS

TERMINATOR
gggggg

CONTROL LIST FOR SERIAL/ PARALLEL CONVERTER

BUFTAB,

C(SCONTR)

ADDRESS

NUMBER OF INPUTS -I

ADDRESS
25 IO

POSITIONS

ooooooo

CONTROL LIST FOR DIGITAL INPUTS

DTABLEP

cooE=o BINARY INPUT

TYPE CODE
CURRENTCOUNT

ADDRESS

32.0
POS|T|°N3

‘

GREY INPUT

7777 TERMINATOR

RESET COUNT
C

.........

77 7 T

Figure A3

Input Device Control Lists

FORMAT WORD

[Oh 1213l41516-l7181911011i]
-——————§ NUMBER OF 12-BIT WORDS

*
——,+ DEVICE

NAME

TYPE CODE

FORMTD,

ooo

DIGITAL 1

OOI~

DIGITAL 2

o ——->

OIO

DIGITAL 3

DIGITAL 4-

Ioo

PLOTTER

'0'

PUNCH

‘110

TELEPRINTER

III

DECTAPE

TYPE CODE
ADDRESS

......

.....

7 7 7 7

TYPE CODE

ADDRESS
............

7 7 7 7
ETC

UNCONDITIONAL OUTPUT LISTS

UTABLE,

TYPE CODE

CURRENT COUNT

0 -CLDK
1 —DEcINIAL

FORMAT WORD

2 -OCTAL

3-MESSAGE OUTPUT

RESET COUNT

9°“)
POSITIONS
LONG

7 7 7 T—TERMINATOR

..........

FORMAT WORD

RESET COUNT

C(UCOUNT

Figure A4

NUMBER OF UNCONDITIONAL
OUTPUT STATEMENTS

Output Control Lists

PROGRAM INTERRUPT

HARDWARE
FUNCT|ON

C(PC)90

SERVICE
DECTAPE ER

SAVE

C(AC),C(L)

E RFLAG

SERVICE
DECTAPE FLAG

SAVE

YES

DEC TAPE
F LAG

C(AC),C(L)

DTFLAG
N0

POP

ammo C(AC)
FROM LIST

ENAB
I NTERRUPT

ERRORYES

ORFLow

(INT OVRNHALT)

DISABLE
INTERRUPT

@Es
NO

YES

SERVICE
BUFFER
BUFFER

YES

SERVICE
TELEPRINTER
TELEPR

@Es
NO

YES

SERVICE
PLOTTER
PLTSER

YES
KEYBOARD ?
'

STOP THE
PROGRAM

YES

EXTRANEOUS
INTERRUPT

(HALT)

INTERR

(Ex INTR)

Figure A5

Interrupt Service Routines

CALLED FROM IOSERV

CLKCLR
IN
STOPPING
STATUS?

CONTINUE

333qu

“0

M
+1—>
SECONDS

CLOCK'
CLEAR FLAG
ENABLE
INTERRUP‘I’
SECONDS
=60?

I
(WAIT FOR
I SECOND)

YES

+1->

§ECNT
o=> SECONDS
+19 MINUTES

YES»
YES
"°

o=> MINUTES
+19 HOURS

+1—> QCNT

(l QUANTUM
OF TIME ?)

:49 ?
YES

(RESET
QUA NTU M)

QCNTe—
QUANT

ANY ADC
INPUTS?

SERVICE
ADC INPUTS

ANY DIGITAL

SERVICE
DIGITAL

INPUTS

ADCSER

—-

ceou
ANY
UNCONDITIONAL
OUTPUTS
NO

IRETURNTO POP I

Figure A6

Clock Service Routine

SERIAL / PARALLEL BUFFER

CALLED FROM IOSERV

BUFFER

CLEAR
FLAG

F’

INITIAUZE
usr

SET

(3%?ng

CLEAR
FLAG

ENABLE
INTERRUPT

+1—>
BTEMPZ

'55

-1=>
BTEMPZ

GET NEXT
ADDRESS

READ
BUFFER

STORE
IN THIS

ADDRESS
7

EXIT TO POP

Figure A7

TELEPR

Buffer Service Routine

CALLED FROM

“E32596

YES

CLEAR FLAG

GET NEXT
CHARACTER

CGHR‘QET’EE

TER M INATOR ?

CALLED FROM

L95_ER_V

was

INTERRUPT

SEFTL28NE

SETFngNE

PUNCH

JMP POP

Figure A8

Teletype Service Routine

Figure A9

Punch Service Routine

lCALLED

DISERV

BY CLOCK

SAVE

TYPE

CODE

ADCSER

Eta-5‘0

O=>
COUNTER

FROM

SELECT

CURRENT
COUNT

l
REQUEST
CONVERSION
-

YES

DELAY
3O usoc

GET
STORAGE
ADDRESS

GET STORAGE
ADDRESS

RESET
CURRENT
COUNT

READ

CONVERTER

EXECUTE
IOI *
C (COUNTER)

GREY

STORE
VALUE

CONVERT
GREY ->
BINARY

YES

GET NEXT
CHANNEL

NUMBER

STORE
VALUE

RNIINATOR

GET NEXT
TYPE
CODE

‘

UPDATE
LIST
POINTER

RETURN To C351

+‘ 9
COUNTER

$¢¥E

No
TE RMINATOR '2

CODE
YEs
RETURN TO CLOCK

Figure A10

Analog—Digifal‘S‘ervice Routine

Figure All

A10

Digital Input ServiC'e Routine

CALLED BY

CLOCK SERVICE
ROUTINE

SET
COUNT

COM

+I->

ELEMENT 1

”0
YES

EM.

ELEMENT 2T0
WAITING LIST

FORMAT

CALLED BY
ANY DEVICE FORMAT HANDIERS

FORMAT WORD
ONTO LIST

ELEMENT 3 -)
ELEMENT I

UPDATE
POINTER BY +3

RETURN TO CLOCK SERVICE

Figure A12

PRINT
FORM ova“
HALT

RETURN

Unconditional Outputs
TFORM
TEST FORMAT
WAITING LIST

CALLED BY INTPGO

l
DISABLE
INTERRUPT

FOPI
LOOK-UP
DEVICE
HANDLER

@ms

RETURN TO
INTPGO

TOP OF UST

LOOK-UP
FORMAT

I-> FORM

LIST

I
SLI OE-UP
UST ENABLE
I NTERRUPT

PLOT F

PUFOR

TTFOR

mmTI

mmTz

PROCESS
C (FORM)
FOPI

mst

mmT4

DECTPE

Figure AI 3
L‘

T
RETURN

TFORM

Figure AM

TEST STATUS OF
WAITING UST

Call Output Device

Format Routines

Format Output Routines

iCALLED

BY FOPI

FETCH
ADDRESS OF

VARIABLE

I
FETCH
VARIABLE

l
+19

ADDRESS

I
STORE
VARIABLE
AT ADDRESS

(AT VARIABLE)

I
OUTPUT
VARIABLE

I
RETURN TO
FOPI

Figure‘AI5

Digital Output Format Routine

CALLED BY
FOP!

DISABLE
INTERRUPT

PUFOR

C(FORM) ONTO

—H

WAITING LIST
ENABLE
INTERRUPT

—1 =>

FLAG

I
GET FORMAT
# CONVERT
TO OCTAL
ASC II

(OCTAL)

I
0 =>
0 SIEE

PLACE
IN
BUFFER

(OSUBR)

CALL

TELETYPE
FORMAT
ROUTINE

(TGO’I)

SET PUNCH
BUSY FLAG

I
RETURN TO
FOPI

Figure A16

Punch Format Routine
A12

(FORMAT)

CALLED BY
FOP!

TTFOR

C(FORM) ONTO
DISABLE

TELETYPE
BUSY _?

1 N T E RR U PT

__.

RETURN
To

WAITING LIST
_—.
ENABLE
INTERRUPT
,

€921

(FORMAT)

0 =>
0 FLAG

l
GET FORMAT #
CONVERT TO
ASCII
OCTAL
O S) OSIEE

(OCTAL)

PLACE
IN
BUFFER

(OSUBR)

FETCH CODE
WORD

T601

TERMINATOR

GET HRS
CONVERT 10—.
DECIMEL

GET
VAmABLE

VAWABLE

PUT IN
BUFFER

SAVE

“’

2335522

“"

CONVERTTO
DEaMEL
PUT IN
BUFFER

._.

away
IN

GET
ADDRESS

l
.

ADD 5le
BITS

l
'

STORE

(ADDRESS
or TABLE)

l
GO TO
OEXIT

PUT TABLE
IN BUFFER

Figure A17

Teleprin’rer Forma’r Rou’rine

A13

FOR

MlNUTES,

SECONDS

ADEfiESS

+ 1

SAME

—'

BUFFER

H OR TELETYPE
CALLE
FORMAT ROUTINES

OFLAGIO?

DI INTO
TELEPRINTER
BU FFER

BUFFER

DZ INTO
PUNCH
BUFFER

02 INTO
TELEPRINTER
BU F FER

I
03

INTO
TELEPRINTER
BUFFER

03 INTO
PUNCH
BUFFER

I
D4 INTO
TELEPRINTER
BUFFER

I
SPACE INTO
PUNCH
BUFFER

SPACE INTO
TELEPRINTER
BUFFER

I
XIT
SUBROUTI NE

Figure AI 8

OSU BR

OEXIT
CALLED BY TGOI

PLACE cm:
IN PUNCH
BUFFER
-

‘

PLACE CR.I.F IN
TELEPRINTER
BU FF ER

0FLAG=D‘

SET BUFFER
POI NTER
PUNCH FIRST
CHARACTER

SET BUFFER
POINTER
TYPE FIRST
CHARACTER

EXIT TO
PU FOR

SET BUSY
FLAG
0‘) TFLAG

J
EXIT TO
FOPI

Figure A19

A14

OEXIT

CALLED BY
TELEPRINTER AND PUNCH
FORMAT ROUTINES
‘

OCTAL

CONVERT
C(AC) TO
OCTAL ASCII

l
RESULTS T0

01, 02,03,134

l
EXIT
ROUTINE

CALLED BY
TELEPRINTER AND PUNCH
FORMAT ROUTINES

DECMAL

CONVERT
C(AC) TO
DECIMAL ASCII

i
RESULTS TO

01.02.015.04

l
EXIT
ROUTINE

Figure A20

Conversibn Routines

7A15

CALLED BY
FOPI

TO NUMBER or
WORDS IN BUFFER

SWITCH
TO WRITE

2991.4

PLACE
FORMAT
WORD IN
BUFFER

I
He won
I

COUNT

I
GET CODE

TERMINATOR

EXIT
ROUTINE
TO FOP!

PLACE HRS

+2-9WORD
IN

BUFFER

COUNT

GET VARIABLE
AND SAVE IN
TS ADDRESS +1

I
PLACE WORD

Figure A2I

BUFFER

DEC’rape Forma’r Routines

A16

DWRT

l
SWHCH

ALL BUFFER
PGNTERS

l
0‘€> WORD
OF NEW
BUFFER

l+v—>
BLOCK

TYPE

"READY TO
swncn“

swncu
UMT
NUMBERS

+3=>

92925

0‘=>
HRST WORD

—1 =>

FIRST woao

v
TYPE

DECTAPE
ausv ?

"TIMING
ERROR"

59M

PUTRFORMAT

wo ° 0"
WAITING LIST

SET
BUSY FLAG

Figure A22

DECfope Write

A17

5X” To
FOP1
—-——

99
DSTART

WAIT

SEARCH FOR
BLOCK
SAVE

DSERCH

(JMS WAIT)

C(PC)=>
WAIT

LOAD WRITE
FUNCTION

+19
DSTA RT

I
LOAD MAC
REGISTER

a
WAIT FOR
FLAG
WAIT

JMP EXIT
0 =>
FUNCTION.
MOTION

WAIT FOR
FLAG

WAIT

DTFLAG

l
CLEAR FLAG

CLEAR

BUSY FLAG

I
'

JMP

C(WAIT) ==>
C(PC)
(JMP i WAIT)

EXIT

Figure A22

DEC’rope Write (continued)

AI8

DSERCH

lCALLED

BY DWRT

S ET
DIRECTION
TO FORWARD

I
-10 '>
DCNT

DTURN

+1.)
DCNT

ERFLAG

MOVING
FORWARD

SET
DIRECTION
TO REVERSE

SET
DIRECTION
TO FORWARD

'BL°°K=>

-IBLOCK-1)

”BLOCK

=> DBLOCK

I
LOAD UNIT.

MOTION.
FUNCTION

WAIT
FOR FLAG

21m

LOAD MAC

O—G

WAIT
FOR FLAG

WAIT

Figure A23

DEC’rOpe Search

A19

?
THIS BLOCK
+0 BLOCK

N0
3

o ?

YES

THIS BLOCK
BLOCK
-

MOVING IN
CORRECT DIRECTION ?

DTURN

TURN AROUND

EXIT SEARCH

ROUTINE

(TO DWRT)

CONTINUE IN
THIS DIRECTION

ERFLAG

CLEAR
FLAG

STOP TAPE

I
TYPE

"0T ERROR"

Figure A23

CLEAR BUSY
FLAG
JMP EXIT

DECfape Search (continued)

A20

CALLED BY
FOPI

PLACE
C(FORM)
ONTO WAITING
———>
LIST
ENABLE
INTERRUPT

DISABLE
INTERRUPT

RLOTTER
BUSY 2

RETURN
T0
FOPI

SET BUSY
FLAG

—1

PWAIT

pCOUNT

FETCH x
STOREIN
AT ADDRESS
X/4 => TX

FETCH Y
STORE IN
AT ADDRESS
Y/4 => TY

C(cp)

C(ppgwimp>
PWAIT)

l'
_‘

(JMS RWAII)

1 €>
PCOUNT

CLEAR
FLAG

EEI§§§

STORE

YES

MFLAG

Exn T0
FOP1

—4 +> Rx
REN LEFT

+4 €> Rx
PEN RIGHT

SET DONE
FLAG

WAR

O=>

WAN

+1->

(PWAW)

MFLAG

(RwAn)

PCOUNT

”0

€> RY
DRUM DOWN

EflT T0
POP

Exn

—1

+1€>PY

YES

$’

WAR

WAW

(PWAIT)

EmT

T0
FOP!

___

YES

Figure A24

Plotter Format Routines

A2]

INTPGO
FETCH NEXT
INSTRUCHON

TEST STACK
UMWS

WITHIN

PMNT

"STACK

LIMITS

ERROR"

HALT

YES

TEST
WAHTNG IJST
STATUS

TFQRM
ADDRESS OPERATION
NO

BITS

0-14 9

BITS 0-1LOOKUP

3' LE“
_

‘

STACK

OPERATION

BITS 2-11:
RELATIVE
ADDRESS

YES

i
BITS 7—11:
LOOKUP

2 = STORE

MAKE ABSOLUTE;
STORE IN PSEUDO
PROGRAM
COUNTER

I= LOAD

IFE
m
In;
IFBE
OUT

BITS 2—11=
RELATIVE

1
2
=
3
=4
=
5
=

<sufg =‘6/
sum- :10
ADD = II
DECR =12

RETURN

+1 —> PSEUDO

PROGRAM
COUNTER

Figure A25

BITS 2—11:
RELATIVE

ADDRESS;

MAKE ABSOLUTE

FETCH AND
PUSH ONTO
STACK

FETCH TOP
OF STACK;
STORE IT

;

Arifhmefic lnIerpre’rer Interrupted by
Clock and l/O Devices

APPENDIX

PROGRAM

PROBLEM I:

EXAMPLES

A SIMPLE PROGRAMMING PROBLEM

An in situ pressure, temperature, and salinity sensing instrument is lowered into the ocean.
Data is transmitted along a single conductor cable and is brought into the computer using a

serial buffer input.

We want to sample the ocean in the Following manner:

From the surface to 100 meters, record at each meter the pressure, tem—

perature, and salinity.
2.

From 100 meters to I000 meters, record the pressure, temperature, and

salinity whenever the absolute change of temperature is greater than .05°C
or

the absolute change of salinity is greater than .02 "/00.

Also record the

pressure, temperature, and salinity every I00 meters from 100 meters to

I000 meters.

Let us assume that the oceanographic sensors have the following precision; that is,

equal to'the following:
I unit of pressure

I meter

I unit of temperature
I unit of salinity

.0I

.OI°C

°/°°

A program to accomplish this sampling is written as follows:

FORM:

PRES, TEMP, COND
I, PRES, TEMP, COND

[

OUTP (1,

IFGR PRES,

BUFR

DCTP)
144; GOTO 2
IFGR (PRES -@PRES), 1; OUTP (1, DCTP)
,

GOTO 1

A23

unity is

(PRES -@PRES), 144; ‘IFLS (TEMP -@TEMP), 5;'
(COND -@COND), 1; GOTO 2
OUTP (1, DCTP); GOTO 2
|FLS

IFLS

END

This program says, in effect:
BUFR:

PRES, TEMP, COND

Three variables named PRES,

FORM:

TEMP, and COND are to be sampled using the serial buffer.

1, PRES, TEMP, COND

Three variables named PRES, TEMP, and COND are to be outputted together.

:OUTP (1,

DCTP)

This says output is to be recorded on magnetic tape.
l:lFGR PRES, T44; GOTO 2

This states that if the absolute change of pressure is greater than 100

(1448), the control of the

sampling will be transferred to statement number 2; otherwise, it will go to the next line.
:lFGR(PRES -@PRES), l; OUTP (l, DCTP)
This line states that if the absolute change of the pressure between two successive readings is

greater than 1, output onto magnetic tape according to Format l; that is, OUTP (l, DCTP)
which means store data on DECtape using Format l; otherwise, go to the next line.
GOTO i

This says to go to statement number 1 and test the environment again.
2:

IFLS
IFLS

(PRES -@PRES), 144; IFLS (TEMP -@TEMP), 5;'
(COND —@COND), 1; GOTO 2

This states that if the'absolute change of pressure is less than 100 meters 'or the absolute change
of temperature is less than

.05°C, or if the absolute change in salinity is less than .02 °°§

then go to statement number 2 which begins the tests over again.

A24

Otherwise go to the next line.

:OUTP (1,

DCTP); GOTO 2

This says to output data onto magnetic tape and transfer control to statement number 2.

The

sampling and testing procedure begins again.
END

This last instruction is self explanatory.
As shown in the above description, the computer has been programmed to make logical decisions specified by
as a means

can

the investigator in sampling the marine environment.

of storing data.

It also has been used

In the above instance, data has been stored on magnetic tape and

be used in other programs to determine variables such as Sigma T, anomaly ofspecific

volume, and sound velocity. Table Al shows a portion of the calculated output from stored
data on magnetic tape transport number 1 that can be run immediately after the sample prOgram.

INPUT SOURCE?

TABLE Al

REDUCED DATA

SALIN.

SIGMA-T

OBSERVED VALUES
DEPTH

TEMP.

DELTA-A

SOUND—VEL
‘

0000

8.35

34.17

+26.590

0001

8.28

34.19

+26.616

+145.53
+143.08

+l483.4

+140.45
+137.20

'+1482.9
+1482.4
+1482.0

+l483.2

0002

8.20

34.21

+36.644

0003

8.13

34.24

+26 .678

0004

8.05

34.26

+26.706

0005

7.98

34.28

+26.732

+134.59
+132.19

0006

7.91

34.31

+26.766

+128.98

0007

7.83

34.33

+26.793

+126.38

+1481.7

0008

7.76

+26.819

7.69

0010

7.61

34.39

,+26.873

+123.94
+121.45
+118.89

+1481.4

0009

34.35
34.37

PROBLEM 2:

+26.845

+1482.7
'

+1482.2

+1481.2

+1480.9

A MORE SOPHISTICATED PROGRAM

For a better demonstration of the flexibility of this programming technique, consider the follow-

ing program

An investigator desires to use a thermistor, pressure, and conductivity chain towed from an

oceanographic vessel.

He will sample at the same time a telemetering buoy that transmits

A25

data from these current meters.

In addition, he desires to obtain Loran lines of position and

sample the ship's speed and ship's heading.
l

These can be summarized as follows:

Log the time on magnetic tape every 50 meters of distance traveled. Ship's

speed is l0 knots; it will cover 50 meters in approximately 9.70 sec.
2.

Sample each thermistor, conductivity, and pressure sensor in the chain

every half second.

This defines the program time as QUNT: 62.

Sample the Loran, ship's speed, and ship's heading every 2 sec, thus,

Ll(4, L2(4, SP(4, HEAD(4.
4.

Conditional Output

Since the near—surface values of temperature and

conductivity will fluctuate the most, it might be most desirable to set thres—
holds so that relatively large changes of temperature and salinity will be

However, deeper values will not change as significantly, so small

stored.

incremental changes have more meaning and thus should be outputted and

Arbitrary values have been chosen and are shown in Table A3.

stored.

Determination of Octal Constants to be Used in Testing

In order to test the

variable it must be determined to what its unit value corresponds.

This is

found by dividing the range of the thermistor, pressure transducer,

other

device by the precision of measurement; thus if the range of the thermistor is

20°C and the precision of measurement is 1 part in 2000, then each unit

equals .0l °C

TAB LE A2

SUMMARY

Variable

Thermistor
Pressure

Precision

Range
20°C

l00 meters

UNITY Corresponds to

1:2000

.o1°c

lz2000

.05 meters
.01

Conductivity

20 °/0°

1:2000

Vane

360°

1:120

3°

Compass

360°

1:120

3°

l centimeter per second

Rotor

A26

°°°

General Output Requirement

Every variable should be recorded on magnetic tape it the specified conditions are met.

Plot T0 versus SO if it is recorded.

Type Current Meter Data in Decimal if it is recorded.
Punch TO,

P0, and SO if they are recorded.

By outlining the problem, the investigator'will have thresholds established for recording changes
in the variables listed in Tables A3,

do the work

A4, and A5.

Figure A26 shows the equipment needed to

The/program listing to sample these variables and record those which exceed the established
thresholds is given below.

ADCV:

.TO(0), T10), T2(2), T3(3), T4(4). T5(5), P0(6),'
PT(7), P2(10), P3(H), P4(T2), P5(13), 50(14), 51(15),'
52(16), 53(17), 54(20), 55(21)

BUFR

V1, C], RI, v2, C2, R2, V3, C3, R3

DGlN:
QUNT;

FORM;
FORM:
FORM:

L1 (4,

L2(4, SP(4, HEADM

I, T0, T1, T2, T3, T4, T5, P0, P1, P2, P3,'
P4, P5, $0, $1, $2, $3, $4, $5
2, T0, SO
3,v1 ,Cl ,RT ,V2,C2,R2,V3,'
C3
R3
4, T0, P0, 50
5, L1, L2, SP, HEAD, CLOK
<5, DCTP, 22>
OUTP(T, DCTP); OUTP(2, PLDT); OUTP(3, TYPE); OUT P(4, PNCH)
lFEQ (V1+C2), 3; GOTO T
IFEQ (V2+C2), 2; GOTO t
IFEQ (V3+C3), 1; GOTO 1
IFLS RT, I2; IFLS R2, I2; IFLS R3, 12; GOTO 2
,

FORM:

l
3

—l

OUTP (3,

TYPE)
IFLS (T0 -@TO), l2; IFLS (P0 —@PO), I2; IFLS (SO -@SO), 5;'
IFLS (T1 -@Tl), 7; IFLS (P1 -@P1), l2; IFLS (ST —@ST), 5;'
IFLS (T2 -@T2), 5; IFLS (P2 -@P2), 2; IFLS ($2 -@'52), 4;'
IFLS (T3 -@T3), 3; IFLS (P3 —@ P3), 6; IFLS (S3 —@ S3), 3;'
IFLS (T4 -@ T4), 2; IFLS (P4 —@ P4), 4; IFLS (S4 -@ S4), 2;'
IFLS‘(T5 -@T5), 2; IFLS (P5 -@ P5), 2; IFLS ($5 -@ 55), T;'
GOTO 2

OUTP(T, DCTP); OUTP(2, PLDT); OUTP(4, PNCH); GOTO3
END

A27

TABLE A3

Depth
in
Meters

MultiThermistor
Variable
plexer
Name
Channel #

MULTIPLEXER A-D CONVERTER ASSIGNMENT (ADCV)
(Data Originating in Thermistor Chain)

Record if

Pressure

Multi-

Conductivity MultiVariable
plexer

Record if

Change of

Variable
Name

Channel #

Change of

Absolute

plexer

Record if
'

Absolute

Name

Channel #

Change of

(0)

.l°C

(6)

.5 meter

(I4)

.05 °°°

(l)

..07°C

(7)

.5 meter

(I5)

.05 °°°

(2)

.05°C

(IO)

.5 meter

(I6)

.04 °°°

(3)

.03°C

(I I)

.3 meter

(I7)

.03 °°°

(4)

.02°C

(I2)

.2 meter

(20)

.02 °°°

(5)

.Ol°C

(I 3)

.1 meter

(2i)

.OI

°°°

83V
DIGITAL INPUT BUFFER
ASSIGNMENT (DGIN)

SERIAL DATA INPUT BUFFER ASSIGNMENT (BUFR)
(Data Originating in Moored Current Meter String)

TABLE A4

TABLE A5

(Data Originating in Ship's Instrumentation)
Vane

Rotor

Compass

Variable Variable
Name
Name

Record/if

Variable
Name

Record if

Variable
Name

Time

Gray

Multiple

Code ?

Current Meter I

Vl + CI

9°

Rl > IO

Loran Line

Current Meter 2

V2 + C2

6°

R2 > ID

Loran Line

Current Meter 3

V3 + C3

R3 > ID

Ship's Speed

Ship's Head

HEAD

yes

SHIP'S KOINSTRUMENTATION

MOORED CURRENT
METER STRING

SERIAL/ PARALLEL

PARALLEL BUFFERS

BUFFER

(BUFR)

(DGIN)

——:

PARALLEL

1388
ADCV

TELETYPE
33- ASR

PDP-8 COMPUTER

PAPER TAPE
PUNCH
75B

*
a

100 CPS
CLOCK
555 / 552

(ADCV)

63V
139

DECTAPE

(CLOK)

MULTIPLEXER

CONTROL

X-Y

PLOTTER
Y

THERMTSTOR CHAIN

PARALLEL BUFFERS

OUTPUT CONTROL SIGNALS

Figure A26

Equipment Configuration for Problem 2

Dl-ITAK ADDENDA
STANDARD DZTAK CLOCK IOT'S
IOT I

Skip 4 Clock flag

IOT 2

Clear Clock flag and Connect Flag to interrupt

IOT 4

Disconnect Clock flag from interrrupt

INSTRUCTION LIST
I.

Skip if Flag 0

Clear Flag and enable Clock

Disable Clock
IOS

If:

INI

' AM

“I

R202

‘L

abcx

p
K

R40t

T
T

80 <>—-———~

”BE

ENABLE

R202

BH
u

v
‘=

ex
M8
mrsq

3‘8

BM
an
a?
an
as

Lat
SPARE-8U

SPARE-3V

Standard DKTAK Clock

Refer to FLIP CHIP Catalog for further information about module logical design.

Eflfifllafl

Los
Washington, D. C.
Parsippany, N. J.
Palo Alto
Ann Arbor
Angeles
Chicago
Denver
Huntsville
Orlando
Pittsburgh
Carleton Place, Ont. Reading, England Paris,
France
Munich, Germany Sydney, Australia
-

5444

PRINTED IN U.S.A.

20-6/65