EPA-650/2-75-035
April 1975
Environmental Protection Technology Series
£
I ^«^ §
\
-------
EPA-650/2-75-035
APPLICATION OF AUTOMATIC
DATA PROCESSING TECHNOLOGY
TO LABORATORY PROBLEMS
by
Larry Babb and Fredric Jaye
Aerotherm Acurex Corporation
485 Clyde Avenue
Mountain View, California 94042
Contract No. 68-02-1436
ROAP No. 21BCC54
Program Element No. 1AB014
EPA Project Officer: W . B . Kuykendal
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D. C. 20460
April 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are:
1. ENVIRONMENTAL HEALTH EFFECTS RESEARCH
2 . ENVIRONMENTAL PROTECTION TECHNOLOGY
3. ECOLOGICAL RESEARCH
4. ENVIRONMENTAL MONITORING
5. SOCIOECONOMIC ENVIRONMENTAL STUDIES
6. SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to
develop and demonstrate instrumentation, equipment and methodology
to repair or prevent environmental degradation from point and non-
point sources of pollution. This work provides the new or improved
technology required for the control and treatment of pollution sources
to meet environmental quality standards.
This document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 22161.
Publication No. EPA-650/2-75-035
11
-------
ABSTRACT
This report describes the design, acquisition, program-
ming and installation of a multiprogramming data acquisition
system. Two multichannel experiments and a high speed laser
doppler velocimeter were interfaced. The laser velocimeter
is interfaced by a bi-directional ASCII communication link.
The mini-computer operating system supports several foreground
data acquisition programs with concurrent background data
processing.
iii
-------
PREFACE
This report covers work accomplished for the U.S.
Environmental Protection Agency, Control Systems Laboratory,
Research Triangle Park, North Carolina, by the Aerotherm Divi-
sion of Acurex Corporation under Contract 68-02-1436. The
work was performed between June 10, 1974 and August 31, 1975.
Mr. William B. Kuykendal was the Environmental Protection
Agency project officer for the program.
We wish to express our appreciation to Frank Briden,
Nelson Butts, and Bill Lowans, of the Control Systems Labora-
tory, for their active and enthusiastic assistance during the
installation phase. We would also like to thank Mr. John
Barry, of Modular Computer Systems, San Jose, for his invalu-
able assistance in a number of problem areas.
iv
-------
TABLE OF CONTENTS
Section Page
1 CONCLUSIONS AND RECOMMENDATIONS 1
1.1 Conclusions 1
1.2 Recommendations 1
2 INTRODUCTION 3
3 FEASIBILITY STUDY AND DESIGN 7
3.1 Feasibility Study 7
3.1.1 Interaction with Existing
Systems 8
3.1.2 Proposed System Usage 8
3.1.3 Data Source and Data Bases 8
3.2 System Constraints 9
3.2.1 Contractual Requirements 9
3.2.2 Environmental, Physical,
and Electrical Constraints 9
3.2.3 Expansion Capability 9
3.2.4 Schedule 10
3.3 A-D Converter Selection 10
3.4 Computer Selection 10
3.4.1 Computer Requirements 10
3.4.2 Equipment Selection 12
4 IMPLEMENTATION 23
4.1 Data Acquisition Programs 24
4.2 Data Reduction Program 26
4.3 Laser Velocimeter Data Acquisition 29
5 INSTALLATION 34
6 EXPANSION CAPABILITY OF THE SYSTEM 37
6.1 Analog-Digital Interfaces 37
6.2 Data Storage 38
6.3 Data Printout 39
6.4 Memory Expansion 39
APPENDIX A 42
TECHNICAL REPORT DATA 47
-------
LIST OF ILLUSTRATIONS
Figure Page
1 System hardware configuration. 20
2 Typical instruction manual page. 27
3 General multichannel experiment. 28
4 Velocimeter interface. 30
5 PDM-70 front panel. 33
6 System outline. 36
VI
-------
LIST OF TABLES
Table Page
1 System comparisons. 13
2 System comparisons expansion capability. 15
3 Hardware cost comparisons. 16
4 Final decision table. 19
5 MOD COMP II/MAX III system hardware. 21
6 Analog to digital system characteristics. 22
7 LDV command and data messages (ASCII
characters). 31
vii
-------
SECTION 1
CONCLUSIONS AND RECOMMENDATIONS
1.1 CONCLUSIONS
A complete data acquisition system has been designed,
integrated and installed within the Control Systems Labora-
tory high bay area. The system design is based on a third
generation mini-computer utilizing a modern multiprogramming
operating system, cartridge disc storage media, high speed
floating point arithmetic hardware, and wide range analog-
digital conversion equipment. The initial application pro-
grams were based on discussions with EPA engineers concerning
their future test programs and later modifications by the EPA
project officer. While the inital start-up encountered the
usual problems, such as instrument nosie, incorrect conversion
coefficients, revised thermocouple types, that seem to hamper
any major interfacing task, the system is meeting the needs of
the experimental programs. As the use of the system increases,
there will undoubtedly be changes and modifications to the
program and options prompted by both internal and external
forces.
The basic system is a flexible and capable tool and has
the capacity for future growth and expansion to meet new and
changing requirements.
1.2 RECOMMENDATIONS
Aerotherm makes the following recommendations for fu*
ture expansion and improvement of the original data system as
currently installed.
-------
The addition of floppy disc as a data storage de-
vice should be seriously considered. The system
currently has no archival data storage device.
The reduction of data in the background would be
considerably enhanced by addition of a faster
printing device than the 10-character/second tele-
printers now installed.
The flexability of the system would be materially
enhanced by 8,000 or 16,000 words of additional
memory.
Several new system software modifications have re-
cently become or will shortly become available
from Modular Computer Systems. Using the new asyn-
chronous communications handlers, the data acquisi-
tion programs could be combined into one program
per experiment at a net savings in memory and a de-
crease in complexity. It may be profitable to re-
vise the data acquisition programs completely at
some future time. A new revision of the operating
system has been released with several new capabili-
ties which may be useful later. This change could
be extremely useful if combined with an increase in
memory on the system.
-------
SECTION 2
INTRODUCTION
The Control Systems Laboratory (CSL) at the EPA Nation-
al Environmental Research Center (NERC), is the primary center
within EPA for stationary source air pollution control techno-
nology development. In support of this program, CSL operates a
number of pilot scale research faciliites in the G wing of the
EPA Research Triangle Park Center, North Carolina. These fa-
cilities include several combustion research furnaces, two wet
scrubbers, a medium scale wind tunnel and a hot baghouse test
facility.
The full utilization of these facilities is limited by
manpower restrictions. A need became evident for an automatic
data acquisition system designed to eliminate the clerical
tasks of data logging, conversion to engineering units, calcu-
lation of air/fuel ratios and other parameters of interest,
and correction of data to standard conditions.
The Aerotherm Division of Acurex Corporation was con-
tracted to develop a laboratory data acquisition system for
CSL. This program covered the design, implementation and in-
stallation of an automated turn-key data acquisition system
which gathers analog and digital data directly from the pro-
cess instrumentation installed on selected test facilities in
the G wing of the EPA Research Triangle Park Center. The work
was accomplished during the period from June 10, 1974 through
March 15, 1975.
-------
The contract scope of work was divided into three ma-
jor phases:
• Phase I — A survey of system requirements culmi-
nating in a feasibility study and design report to
the Management Information and Data Systems Divi-
sion of EPA.
• Phase II — The acquisition of the system hardware.
This phase included programming and testing at
Acurex Corporation, Mountain View, California.
• Phase III — The installation of the equipment at
CSL's pilot plant experimental facilities in Re-
search Center at Research Triangle Park, North
Carolina. This phase included operator training
and program documentation and reporting.
The initial implementation of this system includes da-
ta acquisition from one of the CSL combustion units, the par-
ticulate aerodynamic test facility and a laser doppler velo-
cimeter. However, the system design is expandable to meet
the requirements of the entire experimental area.
The first experimental facility serviced by the data
acquisition system is the multisegment furnace used by the
Combustion Research Section for NOX control studies. This
facility has 40 analog channels which are connected to process
instrumentation measuring ©2, NO, NOX» CO, C02, unburned hy-
drocarbons, air and fuel flow rates and a number of tempera-
tures at various locations in the system. The majority of the
instrumentation is set up to produce 0-1 volt signals while
the thermocouples are read directly at the low millivolt levels.
A typical test involves some 16-24 hours of initial warm
up of the instrumentation and the furnace itself. A given set
of conditions is set in the furnace and the system allowed to
equilibrate for 30 to 40 minutes. The data acquisition system
-------
can display an engineering unit conversion for any data chan-
nel and compute excess air to assist the operator in making
these adjustments. The test operator may select an "equili-
bration" mode whereby the system will monitor the furnace for
equilibration stability during the waiting period.
After equilibrium is reached, the operator can select
the log data mode and acquire data on all channels for a pe-
riod of time. Typically, this is some 20 to 30 minutes and
may involve some 50 to 100 data sets depending on the data
rate selected. After the data is required, a pause is se-
lected, the furnace resets for a new set of conditions, and
the cycle repeats. Typically, some five or six sets of these
tests are completed each day. A complete test sequence may
involve 30 or 40 tests over a three-week period.
The testing done in the particulate aerodynamic test
facility follows a similar pattern in that the major portion
of the testing is accomplished under approximately steady
state conditions.
The laser velocimeter and its associated electronic
produce two sets of eight BCD coded digits representing posi-
tion/ x and y velocity and particle size. These data sets
occur in bursts depending on the particulate loading of the
air stream being measured. Typically, some 40 or 50 data sets
are taken at a given location and set of conditions. These
data are converted to velocity and statistically analyzed. By
analysis of x and y velocity vectors at a number of points, a
two-dimensional flow map is built up. Normal analysis of a
group of data sets will be to analyze for mean and normal
values and truncate outliers. Analysis of individual values,
including outliers, can give fine scale turbulence measurements.
This report describes the work accomplished on the three
contract phases previously listed. Section 2 presents the work
done on Phase I, the requirement survey and system design.
-------
Section 4 presents the system implementation, Phase II. The
system installation, Phase III is summarized in Section 5.
Section 6 considers expansion capability of the system.
-------
3.1.2 Proposed System Usage
The initial implementation of the data acquisition
system covers one of the combustion research furnaces, the
particulate aerodynamic test facility and the laser doppler
velocimeter. The system has sufficient expansion capability
to serve the three remaining combustion facilities, the two
wet scrubber units and the high temperature fabric filter
test unit. The use of a multiprogramming operating system
makes it possible to independently serve many projects with
the principal current constraint being the memory require-
ments of the application programs.
3.1.3 Data Source and Data Bases
As currently implemented, the data acquisition system
utilizes a 72-channel multiplexed analog to digital connector
as its primary data input device. This unit operates at 100
samples/second and auto-ranges between ±5 millivolts and
±10.24 volts full scale. Conversion accuracy is 0.05 percent
or 11 bits plus sign. As installed, 47 analog channels are
allocated to the segmented furnace facility and 24 channels
are assigned to the particulate test facility. The multi-
plexer is capable of expansion to 512 channels.
Acquired data is currently written on a series of disc
partitions. These partitions will hold 256 data sets of both
raw and converted data. The furnace experiment is allocated
12 of these partitions and the particulate test facility is
allocated 6 partitions.
There are no current provisions in the system for archi-
val or hierarchial storage of data. This task may be more
properly assigned to the Univac 1110 system via a data link.
-------
SECTION 3
FEASIBILITY STUDY AND DESIGN
3.1 FEASIBILITY STUDY
In late August 1974, Aerotherm submitted a feasibility
study and design report (Reference 1) to EPA's Management In-
formation and Data System Division in accordance with EPA
Policy Order 2800.1. This feasibility study and design report
covered the aspects of interaction with other EPA (Electronic
Data Processing) systems, proposed system usage, data sources
and data bases, scheduling, environmental and design con-
straints, and overall system requirements. Some of the major
items covered in the feasibility study are discussed below.
3.1.1 Interaction with Existing Systems
As a first level real-time data acquisition system,
this system does not impact any existing EPA EDP systems.
The system is designed to be self sufficient in its primary
data acquisition and primary analysis tasks. As CSL's data
analysis requirements expand to a more extensive usage of the
on-line data acquisition system, the unit is designed to allow
remote job entry into the EPA Univac 1110 via a Univac 1004
terminal emulation. This option will allow ready access to
the 1110 for data reduction and storage consistent with its
capabilities while retaining the real time response and flexi-
bility of the data acquisition system.
-------
3.2 SYSTEM CONSTRAINTS
The overall system design was influenced by four major
constraints, which are identified below.
3.2.1 Contractual Requirements
1. Approximately 70 analog input channels
2. An interrupt driven, multilevel priority multipro-
gramming operating system featuring foreground and
background modes of operation
3. Program generation software that operates in the
background mode
4. A FORTRAN compiler, assembler and necessary utility
programs
5. An operating system that is capable of generating
and storing random files
3.2.2 Environmental, Physical, and Electrical Constraints
The physical environment in the Wing G High Bay Area
is a typical enclosed pilot plant operation where the noise,
dust, temperature and humidity levels can be somewhat ele-
vated, particularly during the summer months. The electrical
distribution networks of 440 volt 3-phase power along with
frequent motor or pump switching, creates an electronically
noisy environment for sampling.
3.2.3 Expansion Capability
The system must be modularly expandable by adding in-
cremental amounts of core memory, random access discs, inter-
facing additional numbers and types of instruments, increasing
types and numbers of terminals and modifying, and adding or
changing application programs.
-------
3.2.4 Schedule
The system must be installed and operational not more
than nine (9) months after the signing of the contract.
3.3 A-D CONVERTER SELECTION
The first design parameter addressed the selection of
an analog to digital converter. Primary consideration in
this area were the ability to reject transients and noise,
protection from over voltage damage, wide range (5 mv to 10
volts) capability and auto-ranging capability. The minimum
sampling speed desired was 50 to 100 channels/second.
An intergrating type of A-D converter was chosen for
its immunity to high frequency and switching type nose since
it integrates the signal over several noise periods. The
choice of a relay over a solid state multiplexer was made
principally for the relay system's relative ability to reject
overvoltages which might burn out the usual solid state type
multiplexer. The selection of mercury wetted relays minimizes
contact bounce and switch contact nose associated with dry
reed relay multiplexers and contributes to an extended opera-
tional lifetime at the cost of limiting the unit to a 100
sample/second sampling rate.
An alternative type of analog-digital converters is the
"flying capacitor" type which uses relays to isolate a small
capacitor and switch the capacitor from voltage source to the
converter input.
3.4 COMPUTER SELECTION
3.4.1 Computer Requirements
The choice of an operating system was somewhat limited
by the scope of work requirement that the system "shall effi-
ciently execute operational programs in an interrupt driven,
10
-------
multilevel priority foreground and provide a background mode
for mathematical analysis and program generation. The oper-
ating system must support a FORTRAN compiler."
After a careful review of the overall system require-
ments, the following general system specifications were
proposed:
1. A 16-bit mini-computer with hardware integer and
floating point arithmetic, 24,000 to 32,000 words
of core memory expandable to at least 64,000 words,
and equipped with direct memory access from dual
1 million-word capacity (or greater) cartridge
discs, one of which must be removable. The CPU
must be equipped with hardware memory protect and
have a real-time clock of a least 100 milliseconds
resolution. Memory parity checking is a desirable
option.
2. The operating system shall be an interrupt driven,
multiprogramming system capable of execution of at
least 10-disc-resident independent programs in a
multilevel priority structure. Since experiments
may be run completely asynchronously, it is re-
quired that the foreground data acquisition pro-
grams servicing each experiment be truly indepen-
dent and not overlays of a single foreground root
task. As the system must also contain a background
of varying size for analysis and software prepara-
tion, it is desireable that the system be capable
of dynamic core allocation from operator assignable
core pools. These pools should be changeable under
run time conditions. The background core pool
should be checkpointable under foreground priority
demand.
11
-------
3. The FORTRAN run-time system should be a memory
system resident element and must provide intertask
control capability. The FORTRAN system should meet
the requirements of ISA Standard S61.1, "Industrial
Computer System FORTRAN Procedures for Executive
Functions and Process Input-Output."
4. The operating system must provide a random access
file capability and such file must be capable of
read/write protection on a global or task only
basis.
5. The operating system and hardware must provide
hardware memory protection for the executing pro-
gram, both for other programs and the resident por-
tion of the operating system. Outside boundary
memory access shall be on a task by task priviledge
basis which shall be operator programmable or
selectable.
6. The system must be readily expandable for closed
loop real-time process control using standard
modules.
3.4.2 Equipment Selection
After surveying the broad field of mini-computer ven-
dors, the more restricted field of multiprogramming operating
systems and the still more restricted field of process control
and data acquisition oriented systems, we selected three sys-
tems for final consideration. They were the Digital Equipment
Company's PDP-11/40-RSX-11D system, DEC'S PDP-11/40-RSX-11M
system, and Modular Computer System's MODCOMP II/220-MAX III
system (see Table 1).
All of these systems offer these common features: in-
terrupt driven multilevel priority, multiprogramming operating
systems, wide range relay muliplexed A-D converters, FORTRAN
12
-------
TABLE 1. SYSTEM COMPARISONS
Parameter
Equipment
Operating System
Multiprogramming (Tasks)
Memory Protection
File Protection
Random Access Files
Parity Memory
Foreground/Background
Cataloged Procedures
FORTRAN IV with ISA
Extensions
FORTRAN - Task Control
Extensions
Local Software Office
Local Maintenance Office
Local Users
Software Supplied
Operating System
FORTRAN IV Compiler
Macro Assembler
Link/Edit
Source Editor
Source Librarian
Binary Librarian
Copy Utility
File Utility
Diagnostics
Digital Equipment Co.
POP 11/40
RSX-ll/M or RSX-ll/D
Yes
256
Yes
6, R, W, Task Only
Yes
No
Multilevel Priority
No
Yes
Yes
Yes
Yes
Yes
RSX 11/M or RSX-ll/D
Yes
Yes
(Macro-11)
Yes
(Task Builder)
Yes
(Edit-11)
Yes
Yes
(LIB-11)
Yes
(PIP)
Yes
(F1les-ll)
Yes
(All Hardware)
Modular Computer Systems
MODCOMP 11/220
MAX III
Yes
256
Yes
G, R, W, Task Only
Yes
Yes
F, M, B
Yes
Yes
Yes
No
Yes
Yes
MAX III
Yes
Yes
Macro
Yes
(Link-Edit)
Yes
(Source Update)
Yes
(Source Maint. Control)
Yes
(Library Update)
Yes
Yes
(DAMP)
Yes
(All Hardware)
13
-------
IV compiler with ISA S61.1 real-time extensions, modular ex-
pansion including process control equipment, resident reen-
trant FORTRAN run time system, hardware floating point and
integer arithmetic units, core memory, real-time clocks, task
and memory protection on a task by task privilege basis, file
protection and support random access files.
Both vendors maintain maintenance service offices in
the Research Triangle Park area and DEC supports a full scale
sales, software and maintenance office in the Durham area.
After extensive investigation, we concluded that any of these
three systems would fully support the requirements of CSL's
pilot facilities in the initial implementation and all con-
tain sufficient modularity and growth potential to be fully
adquate over an extended life span (Table 2). Table 3 shows
the hardware cost comparisons among the three systems. The
first system considered is based on DEC's PDP-11/40 and the
RSX-11D operating system. RSX-11D offers a full scale multi-
programming system which is engineered for very fast response
times and highly dynamic task swapping. It accomplished this
by having a large portion of the system memory resident and
making maximum use of defined single program partitions. This
leads to a highly responsive system at the cost of additional
memory requirements. It also offers an extensive file manage-
ment system. RSX-11D is a mature operating system with sever-
al hundred installations over an 18-month period. User re-
ports were very satisfactory as to the hardware, software and
company support.
The second system considered is the Modular Computer
System's MODCOMP II/MAX III system. This combination offers
a full scale multiprogramming system which was designed as
an IBM 1800 replacement for process control work. Since it
uses a more dynamic core allocation procedure than RSX-11D
(from a pool instead of fixed partitions), it is somewhat
slower (60-70 millisec to load a task instead of 20-35
14
-------
TABLE 2. SYSTEM COMPARISONS EXPANSION CAPABILITY
CURRENT/MAXIMUM
in
MEMORY INSTALLED/MAXIMUM
DISC UNITS INSTALLED/MAXIMUM
A/D CONVERTER INSTALLED/MAXIMUM
CHANNELS
EXPANSION FRAME
MPX CARDS
D/A CONVERTERS
LINE PRINTER
TERMINAL INTERFACES
(20 MA, RS232C)
TERMINALS (TTY)
DIGITAL I/O
MAGNETIC TAPE INSTALLED/MAXIMUM
DEC 11/40
32K/64K
2/8 SINGLE
72/320
32
8 CH
UDC11 SYSTEM
MODULES
60 LPM
YES
ASR 33
UDC 11 FRAME
VARIOUS MODULES
0/8
MODCOMP II
32K/64K
2/4 DUAL
72/512
128
8 CH
IOIS MODULES
50-150 LPM
YES
ASR 33
IOIS FRAME
VARIOUS MODULES
0/4
APPROXIMATE
6,500
5100
2,050
415
795
5,600
500
1,600
2,300
250
1st 11,000
2-8 8,000
PRICE ($)
6,500
lO.OOO(DUAL)
500
500
500
7,000
1 ,250/2
1,800
150
-------
POP 11/40-RSX 11D
TABLE 3. HARDWARE COST COMPARISONS
MODCCMP II/MAX III
Basic System
DEC 1142D
2 Teletypes
AFC11 A/D &
72 Channels
Floating Point
& EIS
Card Reader
LDV Interface
$48,150
4,100
11,990
2,700
4,860
1,830
$73,630
Basic System
11/220
2 Teletypes
Console Term.
A/D & 72 Channels
LDV Interface
Floating Point
Card Reader
Ill
$32,000
3,750
1,800
9,500
2,600
4,000
4,000
$57,650
PDF 11/40-RSX
Basic System
11-40/BH
2 Discs
16 K Memory
Bootstrap
2 Teletypes
AFC 11 & 72 Ch.
Floating Point
& EIS
Card Reader
LDV Interface
RSX 11M License
TIM
$16,100
16,100
4,900
540
4,100
11,990
2,700
4,860
1,830
3.000
$65,920
Basic Unit Includes:
48 K Memory
Dual Disc
Bootstrap
Necessary System
Features
Console Terminal
Basic Unit Includes:
32 K Memory
Dual Disc
Bootstrap
Necessary System
Features
Basic System Includes:
15 K Memory
Console TTY
-------
with RSX-11D). MAX Ill's file management requires that disc
partitions be preassigned and is somewhat less extensive than
RSX-11D. MAX III supports a full scale background batch pro-
cessor which is capable of operating under procedures for
ease of operation. Each task receives individual hardware
memory protection during execution. There are several levels
of file protection — read only, write only, read-write by
owner, read only by others, global access, system only access
and several other combinations.
The MAX III/MODCOMP II system, is a mature stable sys-
tem which has several hundred installations over a 3-year
period. User reports were very satisfactory regarding hard-
ware, software and vendor support.
The third system considered, was the Digital Equipment
Company's PDP-11/40-RSX-11M system. This system combines the
proven PDP-11/40 hardware in combination with a new operating
system called RSX-11M. This new system was announced for
first release in November 1974. Since at the time of the de-
sign study period (August 1974) RSX-llM had not yet been re-
leased, we did not have any user background on the system.
Background operation for program preparation and mathematical
analysis will not be available until May 1975, when Version 2
is released.
Delivery schedule and economics were critical factors
in system selection. At the time of the system design, DEC
was quoting 120- to 130-day delivery dates placing potential
equipment delivery in late December. Since the contract required
installation of the complete turn-key system by February 10, 1975,
this schedule did not allow much time for program debugging and
changes. The MODCOMP hardware was quoted for 45-day delivery.
The second consideration was economic. Table 3 is a
cost comparison of the three candidate systems. A PDP-11/RSX-
11D system would consume so much of the program budget as to
17
-------
make any application programming minimal. The choice of a
PDP-11/RSX-11M system offered a better economic balance to
the program. However, we were reluctant to rely on a first
release of a new operating system, especially when program-
ming changes would be necessary in the following spring when
version 2 was released.
Aerotherm recommended in the design report that the
Modular Computer System's MODCOMP II/MAX III system offered
substantive advantages to EPA in this program and should be
purchased. The final decision table is shown in Table 4.
In early September, the EPA project officer approved the pur-
chase of the recommended system (Tables 5 and 6).
The computer system was delivered to Aerotherm on
October 8, 1974. Figure 1 shows the complete system hardware
configuration.
18
-------
TABLE 4. FINAL DECISION TABLE
DIGITAL EQUIPMENT
COMPANY
vo
MODULAR COMPUTER
SYSTEMS
ADVANTAGES
1. LOCAL SOFTWARE & MAINTAINANCE
OFFICE
2. EXTENSIVE FILE SYSTEM
3. MANY LOCAL USERS (80-100)
DUKE, UNC, EPA
4. MORE DYNAMIC SYSTEM
5. LARGE COMPANY - FUTURE
IMPROVEMENTS MORE RAPID
1. COMPUTER & SOFTWARE DESIGNED
FOR PROCESS APPLICATIONS
2. LOWER COST ALLOWS ADDITIONAL
FEATURES
3. DELIVERY DATE - OCTOBER ALLOWS
MORE PROGRAM TESTING
4. MATURE OPERATING SYSTEM -
VERSION E, 3 YRS OPERATING
EXPERIENCE
5. FOREGROUND/MIDDLEGROUND/BACKGROUND
6. PARITY MEMORY
7. CANNED PROCEDURES
DISADVANTAGES
1. MINIMUM SYSTEM FOR BUDGET
2. DELIVERY TO AEROTHERM
DEC-JAN
3. RSX-ll/M VERSION 1
RELEASE NOV 18
4. BACKGROUND BATCH SUPPORT
IN VERSION 2 (APRIL-MAY)
5. FORTRAN COMPILER OUTPUTS
THREADED CODE (DIFFICULT
ERROR TRACING)
6. NON PARITY MEMORY
LOCAL SUPPORT IS MAINTAIN-
ANCE ONLY. SOFTWARE
-(ATLANTA)
PREASSIGNED FILES -
MINIMUM SYSTEM
ONLY 2 LOCAL USERS
(BURLINGTON MILLS & UNC)
SMALLER COMPANY - MAJOR
ADDITIONS WILL OCCUR MORE
SLOWLY
-------
M
O
(+u>tjiarf\
' y yic-
S.S5
\
•x. .
/""
f .
H
t^,
i j
/)/& C&A/K£KT£&
I I
& I _T~ X
» *~ 1 <£X/^
-------
Item
1
Table 5. MOD COMP II/MAX III SYSTEM HARDWARE
Model Qty Description
11/220
Computer System Consisting of:
. MODCOMP II 16 Bit CPU
. 32,768 Words of Core Memory
800 nsec Cycle Time
. Memory Expansion to 65,536 Words
. Direct Memory Processor
. Executive Features
. System Protect
. Multiply/Divide
. Hardware Fill
. Power Fail Safe/Auto Start
. Memory Parity
. 4903 Peripheral Controller Interface
. 3751 Controller for TTY & Paper Tape
Reader
. 4128 Moving Head Disc
. Two 0001 Cabinets
Hardware Floating Point
ASR-33 Console Teletypewriter
*
Programmable Power On/Off Switch for TTY
Peripheral Controller Interface
Asynchronous Communications Interface
Two 110 Baud Channels
481066 1 Asynchronous Communications Interface
4800 Baud Channel
2
3
4
5
6
3512
4233
3747
4903
4810AA
1
1
1
1
1
8 4223 2 Remote ASR-33 Teletypewriter
9 4411 1 Card Reader - 300 CPM
10 1406 1 Wide Range Analog Input Subsystem
11 1421 9 Wide Range Analog Multiplexer Switch
7.4 Hz Filter + 10.24V Input
21
-------
TABLE 6. ANALOG TO DIGITAL SYSTEM CHARACTERISTICS
INPUT - ± 5 MV TO ± 10.24V FULL SCALE
RESOLUTION - 1/2048 OR 0.05%
CHANNELS - UP TO 512
SPEED - 100 SAMPLES/SECOND
TYPE - DUAL SLOPE INTEGRATION
NOISE IMMUNITY - DC - INFINITE
60 Hz-86DB
MINIMUM
ro
to
SPECIAL FEATURES
AUTOMATIC RANGE SELECTION OR PROGRAM CONTROLLED
INVALID DATA FLAG
AUTOMATIC DISCONNECT ON OVERVOLTAGE
MERCURY WETTED 3 WIRE RELAYS - LOW NOISE
- LOW VOLTAGE OFFSET
- 10 BILLION CYCLE LIFE
TRANSFORMER ISOLATION - NO DIRECT ELECTRICAL CONNECTION TO
COMPUTER
-------
SECTION 4
IMPLEMENTATION
Implementation of a multiprogramming data acquisition
system is at its simplest a resource allocation problem.
There are certain hardware resources which must be "dedicated"
to a particular task. For example, analog-digital converter
channels must be physically rewired to change from one pro-
ject, location, instrument to another. Of the 72 channels
available on the system, 48 are assigned to the furnace exper-
iment and 24 to the aerodynamic particulate test facility
(wind tunnel). Similarly, the teletype output ports repre-
sented by the dual channel 4810 interface, are allocated (by
physical wiring) to the same two experiments. A third 4810
interface channel is allocated to the laser velocimeter inter-
face, but this has three potential locations where it can be
plugged into the system.
The remaining resources of the machine are under the
"command" of the operating system and thus may be reassigned
dynamically to meet the immediate needs of the system. For
example, the two cartridge disc units are split into a number
of "partitions" which are used by specific programs. The
first disc (called the system disc) is primarily used as a
repository for programs, libraries, program sources, and also
contains several large reusable scratch areas for program gen-
eration and data analysis. The second disc is divided into
20 identical partitions for data storage by the data acquisi-
tion programs. These disc partitions are dynamically assigned
as required by different programs.
23
-------
The principal limiting resource on the system at this
time is the amount of memory available. The system has cur-
rently installed 32,000 words of a possible 64,000 words of
memory. the operating system requires approximately 13,000
words of this space which leaves about 19,000 words for dyna-
mic assignment to programs as required. Within this 19,000
words of memory, we must fit various combinations of furnace
data acquisition, wind tunnel data acquisition, and program
generation.
The program generation programs are a part of the oper-
ating system and are furnished by Modular Computer Systems.
The FORTRAN compiler is furnished in two forms; one requires
10,000 words of memory to execute and is heavily overlayed
and rather slow. The extended compiler is quite fast (I/O
limited) but requires 14,700 words of memory to run. The out-
put of both compilers is identical. All program generation
software except the extended compiler will execute in 10,500
words of memory or less.
4.1 DATA ACQUISITION PROGRAMS
In order to conserve memory, the data acquisition pro-
grams are written in the form of a root task plus four over-
lays which represent mutually exclusive functions.
With the exception of I/O unit numbers, and the A/D
channels to be sampled, the data acquisition programs for the
furnace and wind tunnel experiments are identical. The data
acquisition task for each experiment is handled by two pro-
grams called "COMMAND" and "RUN-TIME." COMMAND acts primarily
as a communications interface to the test technician allowing
him to enter commands and comments. The program interprets
the command and passes various flags to the "RUN-TIME" program
via a segment of global common which is accessible to both
programs. The RUN-TIME program operating asynchronously under
24
-------
its own timing loops actually performs most of the work such
as data taking, converting and logging to the disc.
The division of labor between the programs was required
by the inability of the "asynchronous communications I/O
handler" within the operating system to correctly handle the
full duplex hardware installed on the system. The manufac-
turer states that a new "asynch" handler will be available
later in 1975 which will correctly use the full duplex hard-
ware. When this is available, we recommend that the programs
be revised and combined since this will simplify and shorten
the programs and use less memory.
The "RUN-TIME" program is broken into a root area and
four exclusive overlays. The root area contains the code to
acquire and convert the data, write it on the disc and keep
its timing loops. The various overlays (which are loaded dur-
ing specific times) are "INITIALIZE," "DISPLAY," "COMMENT,"
and "EQUILIBRATE."
"INITIALIZE" is used only once at the start of the pro-
gram to load the instrument calibration, etc., from the con-
figuration file and to assign a disc partition for data stor-
age. "COMMENT" is used whenver the command "COMMENT" or "CO"
is intended. The overlay is loaded and asks the operator to
enter any one line comment which is tagged with time and data
and written on the comment disc file. "DISPLAY" is loaded in
response to the operator's request to type out the current
readings on certain channels. The program will print the mil-
livolt A-D readings and the engineering units for each selected
channel on a one time basis. "EQUILIBRATE" can be used by the
test operator to survey the stability of his experiment. Up
to 10-operator-selected channels will be compared from reading
to reading with operator-input range values to determine wheth-
er the experiment is in equilibrium after adjustments. The
program will signal when equilibrium is reached or when the
unit is not equilibrated.
25
-------
Figure 2 reproduces a page from the operating manual
(Reference 2) which details the various commands available
to the operator. The specific internal details of these pro-
grams are given in Section I and II of the program documenta-
tion (Reference 3) and will not be discussed further here.
The program pairs for each experiment require approximately
6000 words of memory per pair. The overall program relation-
ships are indicated in Figure 3.
4.2 DATA REDUCTION PROGRAM
The data reduction and analysis programs fall into
three categories: utility, generalized and specific. All
data reduction programs operate in the background under the
jurisdiction of the background Job Control processor. Most
of the programs will operate in a background size of 6000
words or less.
The utility programs serve the function of being gener-
al interfaces between the data files and the data reduction
programs. There are two programs (REFORMAT and GET DATA)
which will interogate the raw or engineering unit data files
and extract specific channels, sequences of channels, etc.,
and write them on a scratch disc file in a card image format
tailored to specific data reduction programs. "COMMENT PRINT"
and "DATA PRINT" are two programs which exist mainly to print
the contents of the data file in a understandable manner.
The generalized data reduction program include a linear
regression, a multiple linear regression, general statistics,
a polynomial curve fit, and specific programs for multirange
NO and hydrocarbon monitoring instrumentation. These programs
X
are all general in their usage and will read card image data
from punched cards, scratch disc file, and the operator console,
The programs are fully described in the program documentation
and operators manuals.
26
-------
C. RUN TIME OPERATIONS
Valid commands are:
LOG - starts the data taking cycle at the normal rate
FAST - take data at the FAST rate defined 1n the configuration
NORM - take data at the NORMAL rate defined in the configuration
PAUSE - have everything going but don't take data for awhile
RESUME - start taking data again after a pause
COMMENT - I want to write a comment 1n the records. The program will
answer TYPE. Type in any message .you wish up to 70 characters
long.
DISPLAY - type out the current values for up to 5 data channels
The program will answer "CHANNELS?". Type in the desired
channel numbers (up to 5) in 13 format. Example: 00100200300
9022. Include the zeros as necessary, no spaces please. The
program will print the raw and converted data for each channel
requested.
EQUILIBRATE - will keep an eye on up to 10 chennels and check for steady
state. The program asks "ENTER CHANNEL AND RANGE". You may
Input up to 10 channels 1n the following manner:
XXXYYYYY where XXX Is the channel I and YYYYY is the
acceptable range in engineering units.
(020 5.7)
End the list with a $$ (end of file) marks. The program will
ask "SURE?". If you answer $$ again, it will go into the
equilibrium wait mode. If you answer anything else, it will
start over asking for channels again. When 1t begins checking
for equilibrium, it reads the specified channels at the data
rate selected. The program will print "HR:MIN:SEC EQUILIBRIUM
STATUS « X" every time the equilibrium status changes. If
X - 0, the system is not in equilibrium. If X « 1, the system
1s equilibrated. The equilibrium mode may be terminated at any
time by entering the command LOG or Resume.
TIME - prints out the current time
EOT - End of Test - prints out the current record numbers on the data
files for Information stops taking data. You can now set new
conditions on the equipment and wait for It to equilibrate,
then begin taking data again.
PJJIT - QUIT for the day - prints out record numbers on the data files.
Writes an end of file on the disc files. Shuts down the data
program.
This completes the real time portion of the experiment.
Figure 2. Typical instruction manual page.
27
-------
>. loin
orawcDtt
oo
Figure 3. General multichannel experiment.
-------
There are several specialized data analysis programs
available in the system. Program BMP is a program to perform
specific data analyses on data from the segmented furnace ex-
periment. Emissions are corrected for excess air and a num-
ber of process variables are printed in a tabular format.
4.3 LASER VELOCIMETER DATA ACQUISITION
The laser doppler velocimeter is connected to the data
system via a two-channel asynchronous interface and a Digital
Equipment Company PDM-70 interface unit (Figure 4). The PDM-
70 receives an ASCII character string from the computer and
performs certain electronic functions depending on the com-
mands contained in the message string (Table 7). Its basic
function is to indicate a ready or busy status to the LDV
operator and to receive, convert and transmit to the computer
two 8-digit BCD messages from the LDV's data multiplexer.
The first 8 digits represent a 4-digit position and the next
4 digits the x plane velocity. The second 8 digits are 1
marker digit, 3 size parameter digits (future expansion) and
4 y plane velocity digits. The marker digit is used to iden-
tify the first and second messages returned by the LDV multi-
plexer. The first digit of the position is always 0 or 1
(3-1/2 digit DVM) and the marking digit is set with a thumb
wheel to 2-9 thus providing an easy separation. Data messages
not meeting the x-y sequence criteria are discarded.
There are two LDV data acquisition programs used with
the LDV. The LDV test program uses a nearby teletype to in-
teract directly with the operator to assist him in setting up
the LDV for use. The program will ask how many data sets are
required, obtain the data and print means, standard deviations,
and maximum and minimum readings on the teletype. No data is
saved other than the printed output.
29
-------
Laser
Velocimeter
RDM-70
Converter
u>
o
IV Data
Multiplexer
Data
Data Ready
Asynchronous
Data Links
Dual Channel
Interface
Print Hoi doff
Computer
Figure 4. Velocimeter Interface.
-------
TABLE 7. LDV COMMAND AND DATA MESSAGES
(ASCII Characters)
Computer Command
LDV Reply or Action
LF, DC1, 3, SOH, 1, DCS, NULL
(wait for reply)
LF, DC2, 3, DCS, 7, 0, EOT
LF, DC1, 0, SOH, 3, DCS, NULL
LF, DC1, 0, SOH, 3, DCS, NULL
LF, DC2, 3, DCS, 0, 7, EOT
"EOT" or "D EOT" toggle in by operator
(Turn on busy light, turn off ready
lights)
PPPPXXXXEOT
First data string
X Velocity
Position
MSSSYYYYEOT Second data string
T
Y Velocity
Size
Marker
Turn off busy lights and turn on ready
31
-------
The normal data acquisition program operates without
a remote teletype and will acquire 50 data sets and writes
the data and time date information on a selected disc parti-
tion, the LDV operator must enter (via the PDM-70 console
switches) (Figure 5) either an "EOT" (04) character or a "D"
(44) character to signal the program to take another data
set or to quit taking data and exit. The system signals the
operator of the "busy" or "waiting" status via the front
panel lights.
There are two data reduction programs which operate in
the background. The first program was written by Mark Wagman
at EPA and uses punched cards to enter data and truncation
and limit parameters before calculating means and standard
derivations and printing histograms before and after trunca-
tions beyond selected data limits. The second data analysis
program is an adaptation of the first to read data sets from
the disc file where they were stored. The operator is asked
for range setting information and whether or not to print a
histogram for each data set. These options provide a quick
look form of the data analysis without excessive waiting time
because of the low speed of the teletypewriter used for
printing.
32
-------
Output data lights (LEDS)
U>
OJ
T Buss R Buss
o o o o o
Data
Enter
/ X
0000 OOOO
A A
Clear
O
Power
8421 8421
\ /
Input data entry switches
Figure 5. PDM-70 front panel.
-------
SECTION 5
INSTALLATION
This data acquisition system was shipped to EPA in
mid-January 1975. The installation phase of the program be-
gan on January 25, 1975, and except for the laser velocimeter,
was completed on February 14, 1975. The installation con-
sisted of emplacing the system in a room on the mezzanine of
the high bay area and installing conduits, cabling and ter-
mination panels at several locations in the high bay area
(Figure 6). With the active assistance of the EPA co-project
officer and several of the EPA test technicians, the installa-
tion phase was accomplished with only minor hitches. A room
air conditioner was azquired and installed to prevent over-
heating in the computer rooom since the central air handling
system in the area was inadequate. The only major problem en-
countered during the installation was involved with the laser
velocimeter interface. The LV interface problems were associ-
ated with the computer hardware and software. The first pro-
blem was associated with the 4800 Baud communications channel
which would not return a data interrupt on the receive channel.
The problem was subsequently corrected by a Modular Computer
Systems field service engineer. The second problem was in the
asynchronous communcations data handler, a part of the operat-
ing system software. The LDV interface apparently was capable
of receiving a down-link channel command and replying with the
data on the up-link channel faster than the program could be
ready to receive the data. This problem was solved by separ-
ating the channels on the dual interface and installing the
34
-------
new program on a return trip to EPA. At the time of the
return a number of minor revisions were made to several of
the data acquisition and reduction programs. After a initial
period of operating correctly for several weeks the analog to
digital connection developed an apparently random zero drift
of 1 to 2 millivolts. Repeated service by the manufacturer
has failed to disclose any apparent cause for this behavior
which is still under investigation. The data acquisition
sections of several programs were modified to read channel 48
at the beginning of each AD channel scan. Channel 48 is per-
manently shorted so its reading should represent the instan-
taneous value of the zero offset. This value is then sub-
tracted from the other channel readings to correct the zero
offset pending a conclusive hardware solution.
Another problem which has occurred periodically is that
occasionally a single data point will read a very high un-
usual value. This is probably due to a recurrent nonperiodic
noise spike or pulse somewhere in the center getting into the
A/D connection at the crucial time. The source of the noise
is unknown and is not periodic or frequent enough to track
down at this time. Since all the current data programs in
use compute averages from all the instantaneous data readings,
the occasional periodic bad value is not significant in the
overall data reduction scheme. It would be simple to pass
the data through a simple screening program prior to data
analysis if the problem persists.
35
-------
Figure 6. System outline.
-------
SECTION 6
EXPANSION CAPABILITY OF THE SYSTEM
The Modular Computer System, Model 11/220 computer
system, was selected as the basis of this data acquisition
system in part because of its expansion capability. The
MODCOMP II computer is designed specifically as a process
control computer to be a replacement for the IBM 1130, 1500
and 1800 series of process computers. As such, MODCOMP of-
fers a wide variety of process control peripheral units on a
standard equipment basis. These units are fully supported .
with software. The first area of expansion covers the analog
and digital control interfaces available.
6.1 ANALOG - DIGITAL INTERFACES
The unit as installed has a wide range A-D converter
with 72 multiplexed input channels. This can be expanded to
128 channels by plugging in additional 8-channel multiplexer
cards. Addition of another multiplexer card frame will hold
an additional 128 multiplexer channels. The unit can be ex-
panded to 512 channels with more expander frames and multi-
plexer cards. This system can also contain additional high
and low speed A-D converters, each of which may service 512
channels.
The addition of process control capability requires
that one additional cabinet be added to the system along with
an Input-Output Interface System unit. This unit is essen-
tially a frame with power supplies into which a wide variety
of printed circuit cards are plugged. These cards include
37
-------
relays, power controllers/ digital switches in and out, timers
and digital to analog converters. Expanders for this unit are
readily added. This option is fully supported with software
via the ISA Standard S61.1 extension to FORTRAN. Through this
interface unit and its modules, the system could support al-
most any forseeable process control function.
6.2 DATA STORAGE
The data storage facilities of the system are currently
limited to the partitions available on the second disc unit.
This approach is adequate for the current real time needs of
the system, but are not convenient for longer term storage
of data. There are four options which could increase the sys-
tem capacity and capability in this area. The system can
support the addition of up to two more cartridge disc units.
These may be removable or fixed discs. This approach would
also add additional capability for data manipulation since
these are moderately high speed units. A second option which
is considerably less expensive, is the addition of a "Floppy"
disc controller and unit to the system. The floppy disc of-
fers a convenient method of saving data in external files.
The floppy disc cartridges are about $8 each and may be readily
kept in a file cabinet or desk drawer much like a 45 RPM re-
cord which they physically resemble.
Data can be readily copied from large discs to and from
the floppy quickly and conveniently, for storage data, retrie-
val and access.
A third approach would add a Model 4811 interface, a
modem and a Univac 1004 emulator program to the system. In
this manner, the system could act as a remote job entry termi-
nal to the control EPA Univac 1110 facility. Presumably, the
1110 could be used for data storage, large scale data manipu-
lations, plotting, etc. The 4-6 thousand words of memory re-
quired by the 1004 emulator program is the only major drawback
to this approach.
38
-------
The last data storage approach is the addition of mag-
netic tape to the system. The tape transports are available
in a variety of industry-compatible speeds, density and track
combinations. All are fully supported by system software.
This option is the most expensive of any except the addition-
al high speed disc units and is not recommended unless other
data acquisition equipment on hand currently produces data
records on tape. In this case, tape may be a desirable inter-
change media between systems. The system can support up to
four magnetic tape drivers in any combination on each control-
ler unit.
6.3 DATA PRINTOUT
The remaining major expansion capability in this sys-
tem is concerned with those options which act to increase the
throughput or output of the system. One of the major bottle-
necks in the system is the lack of a moderate or high speed
device for generating hard copy data output. Currently the
system uses 10-character/second teleprinters as output de-
vices. A background data analysis typically takes 5 to 10
seconds to get all the data from the disc, 2 to 3 seconds to
analyze it and 5 to 10 minutes to print the results. The ad-
dition of a small medium speed printer in the 100- to 150-
character/second region would increase system efficiency
radically.
6.4 MEMORY EXPANSION
The other throughput expansion option is the addition
of memory from the 32,000 words now installed to the maximum
64,000 words. This expansion may be done in 8,000 or 16,000
word modules. Since most of the data acquisition programs
occupy about 6,000 words of memory (as do many of the data
reduction programs), it is generally possible to run any three
39
-------
of the four major programs concurrently, such as Data Reduc-
tion/Background, Furnace Data Acquisition, Wind Tunnel Data
Acquisition, and Laser Velocimeter Data Acquisition.
It is presently possible to run only one major data
acquisition task while doing program generation in the back-
ground. The addition of even 8 or 16,000 words of additional
memory would eliminate this potential bottleneck.
40
-------
REFERENCES
1. "Feasibility Study and Design of an Automated Data Acqui-
sition System," Aerotherm Interim Report 74-110, August
30, 1974.
2. Users Manual: "Data Acquisition System," Aerotherm UM-75-
59, January 1975, revisions, March 1975.
3. "Program Documentation, Data Acquisition System," Aerotherm
UM-75-61, March 1975.
4. MAX II/III Reference Manaul, "Modular Computer Systems,"
210-600303-001 and -002, May 1973.
41
-------
APPENDIX A
GLOSSARY OF TERMS
ASCII
1. American standard code for Information Interchange
2. A code used to transmit symbolic characters
3. A mode of transmission to and from an I/O device
Asynchronous
1. Not in step with
2. A term which specifies that an event has no time
relationship with another event.
3. Not synchronized with time
Background
1. The MAXII task operating at software priority level
255, used for batch processing (nonreal-time)
operations
2. A special area of the MAXIII core pool which is set
aside and used to execute the background task.
Batch Processing
1. Processing of programs sequentially
2. A method of program execution where control passes
from one overlay program to another as specified
by a control statement processor (called Job Control)
which is loaded itself as an overly between other
programs.
42
-------
Cataloging
The process of storing object programs or other data
in a random access disc partition. A directory is maintained
in a central place that allows the cataloged item to be found
quickly elsewhere on the disc.
Direct Access
A method of locating a particular record on a device
without passing through other records to find it; sometimes
called random addressing.
Directory
An ordered data array in core or on disc which contains
pointers and keys to the location of other data items on disc
or in cores.
Driver
Executive
A section of program code directly connected to an
interrupt.
In MAXIII, a driver is usually the section of code
between an interrupt entry point and an I/O device
handler.
1. A collection of programs and data bases whose sole
purpose is to control other programs and provide
services for them. Sometimes called a monitor
2. The resident core elements of an operating system
43
-------
APPENDIX A (Continued)
Foreground
1. The highest priority group of multiprogram levels
in MAXIII
2. A collection of privileged tasks
3. A region of memory in which foreground tasks and
overlay programs execute
Full Duplex
A method of interconnecting a computer system and a
peripheral device so that both the input and output data
paths are unique and can be operated asynchronously and
concurrently
Global
1. Capable of being shared by more than one program
2. Global common — a public core area which may be
concurrently accessed by several tasks
Half Duplex
A method of communicating with an I/O device in which
input and output operations must share a common data path and
must be programmed alternately
Job Control
The name of the root overlay program of the background
test. This overlay processes Job Control statements or direc-
tives which set up I/O devices and files and then transfers
control to and loads other background programs on top of itself.
It is reloaded automatically any time a background program
exits or is aborted.
44
-------
APPENDIX A (Continued)
Multiprogramming
The concurrent execution of two or more programs on a
single computer by means of switching between them as certain
events occur, or as particular programs need to wait for an
external event.
Object
A representation of a program which is in a data format
suitable for processing by a loader or editor. In addition
to being a specification of machine instructions, data arrays,
addresses, it also has information which permits the program
to be relocated into any area of memory.
Partition
1. Core Partition — a contiguous area of memory de-
fined for a specific purpose
2. Disc Partition — a defined contiguous area of disc
medium (integral number of tracks) which is given
a unique name and made to take on the characteri-
tics of a unique logical device
Physical Device
The device hardware integrated with its physical re-
cording or transmission medium; as opposed to a logical device
which is the entity to which a programmer can assign his files
Procedure
A precise step by step method for solving a problem
or producing a result
A named symbolic skeleton or prototype that basical-
ly defines a step by step Job Control sequence, with
variables are supplied by the parameters of the cal-
ling statement ($DO) which specifies the name of the
procedure ._
43
-------
APPENDIX A (Concluded)
Real-Time
1. Pertaining to program elements which are executed
during (or close to) the time a physical process
is actually being performed
2. The processing of information in a sufficiently
rapid manner so that the results are available
quickly enough to allow another program element
or human operator to influence the process being
investigated
46
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing)
•REPORT NO.
EPA- 6 50/2-75-035
2.
3. RECIPIENT'S ACCESSIOC+NO.
J" TITLE AND SUBTITLE
Application of Automatic Data Processing Technology
to Laboratory Problems
I. REPORT DATE
April 1975
6. PERFORMING ORGANIZATION CODE
AUTHORtS)
Larry Babb and Fredric Jaye
8. PERFORMING ORGANIZATION REPORT NO.
, PERFORMING ORGANIZATION NAME AND ADDRESS
^.erotherm/Acurex Corporation
485 Clyde Avenue
Mountain View, CA 94042
10. PROGRAM ELEMENT NO.
1AB014; ROAP 21BCC-054
11. CONTRACT/GRANT NO.
68-02-1436
^SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
ERC-RTP, Control Systems Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 6/74-3/75
14. SPONSORING AGENCY CODE
[g SUPPLEMENTARY NOTES
-^ABSTRACT
The report describes the design, acquisition, programming, and installation of a
multiprogramming data acquisition system. Two multichannel experiments and a
high-speed laser doppler velocimeter were interfaced. The laser velocimeter is
interfaced by a bi-directional ASC n communication link. The mini-computer
operating system supports several foreground data acquisition programs with
concurrent background data processing.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
pata Processing
leasers
Speed Indicators
Combustion
Research Projects
Air Pollution Control
Stationary Sources
Laboratory Problems
Process Instrumentation
13B
9B
20E
14B
21B
^^DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (Tl
Unclassified
eport)
53
20. SECURITY CLASS (THISpage)
Unclassified
22. PRICE
Form 2220-1 (»-73)
47
------- |