home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Mega Top 1
/
os2_top1.zip
/
os2_top1
/
DOS_GG
/
COMCAT
/
VOL33.ARJ
/
VOL3357
< prev
next >
Wrap
Text File
|
1992-01-07
|
4KB
|
77 lines
Volume 3357 - Released may 1988 - Abstract (C) Programmatheek
P r o g r a m m a T h e e k volume 3357
gelijk aan EE Library #7
Titel: Feedback Control Engineering (PHASLOCK)
Number Name Size Date Comments
CATALOG 357 846 9-13-87 This list
357.01 CLSLOOP DAT 239 1-01-80 Demo data file
357.02 IFTLOOP BAS 14414 9-15-86 Feedback control analysis using
Fourier transform
357.03 IFTLOOP DOC 384 1-04-87 Short doc
357.04 IFTLOOP EXE 110510 9-15-86 Compiled version (uses 8087)
357.05 OPNLOOP DAT 133 9-18-86 Demo data file
357.06 PHASLOK BAS 21228 9-16-86 Feedback control analysis using root
locus
357.07 PHASLOK EXE 108238 9-16-86 Compiled (uses 8087)
357.08 PLL DOC 3456 1-04-87 Documentation (abstract + details)
357.09 PLOT BAS 24313 1-01-80 General screen plotting (CGA, Hercu-
les, Olivetti)
357.10 SMPLOOP BAS 15109 1-01-80 Feedback control an. using sample data
theory
357.11 STEPS DOC 4736 9-13-87 Instruction on demo
357.12 VOL3357 ABS 2816 9-13-87 Disk abstract
13 File(s) 50176 bytes free on volume VOL3357
THE PROGRAMS ON THIS DISK ARE THE WORK OF JORGAN ANDERSON
OF ALPHA INDUSTRIES, INC. HE DONATED THESE PROGRAMS IN RESPONSE TO
A REQUEST BY THE E.E.P.D. LIBRARY FOR SOFTWARE TO DESIGN AND ANALYZE
PHASE LOCKED LOOPS.
MR. ANDERSON HAS POINTED OUT THAT THE PROGRAMS HAVE NO ERROR TRAPPING
AND VERY LIMITED DOCUMENTATION. HE IS WILLING TO GIVE GUIDANCE TO ANYONE
TAKING ON THE TASK OF CORRECTING THE ABOVE.
THESE PROGRAMS ARE GENERAL PURPOSE FEEDBACK CONTROL PROGRAMS. THEY
TAKE AS INPUT THE POLES AND ZEROES OF THE TRANSFER FUNCTION AND THE LOOP
GAIN. THE PROGRAM CALCULATES AND PLOTS THE OPEN LOOP RESPONSE, CALCULATES
AND PLOTS THE CLOSED LOOP RESPONSE. IT ALSO CALCULATES AND PLOTS THE TIME
RESPONSE.
THE FIRST PROGRAM, "PHASELOK", USES ROOT LOCUS METHODS. IT BREAKS
DOWN WITH CO-LOCATED ZEROES AND POLES. THE NEXT PROGRAM, "IFTLOOP", USES THE
INVERSE FOURIER TRANSFORM TO AVOID THIS PROBLEM. THE LAST PROGRAM, SMPLOOP,
USES A SAMPLED DATA APPROACH.
COMPILED VERSIONS OF ALL THREE FILES HAVE BEEN SUPPLIED BY MR.
ANDERSON. THEY HAVE BEEN COMPILED USING "MICROWAYS 87 BASIC".
NOTE:*******A MATH CO-PROCESSOR IS REQUIRED*********.
IF THIS IS NOT AVAILABLE THE PROGRAMS COULD BE COMPILED WITH ANOTHER COMPI-
LER. THIS WILL RESULT IN THEM RUNNING MUCH SLOWER NOT ONLY DUE TO THE LACK OF
THE CO-PROCESSOR BUT ALSO BECAUSE OF THE LOSS OF OTHER ADVANTAGES OF 87BASIC.
CAUTION: THE USER MUST BE CONVINCED THAT THE PROGRAM IS GIVING THE
PROPER RESULTS. THESE TECHNIQUES DO HAVE CONVERGANCE PROBLEMS. FOR THE LOOP I
ANALYSIZED I FOUND THE DEFAULTS IN SPMLOOP GAVE ME THE CORRECT SOLUTION.
The Phaselock programs are really general purpose feedback
control programs. I have included a data file of a -9db/Octave
loop filter, normalized to unity gain at 1 rad/sec.
The programs IFTLOOP, and SMPLOOP, are similar in operation,
but very different in the transient response algorithm. The
first thing you will notice is that the time response takes
longer to compute. This is because they compute 1024 open loop
gain and phase points, subsequently 1024 closed loop gain and
phase points, and then performs a 1024 point inverse Fourier
transform.
In using these programs you should always run the impulse
response first, making sure it has settled completely to zero at
the end of the time base chosen. This ensures the correct
calculation of the step response.
SMPLOOP includes the sinc x characteristics of a sampling
phase detector, and also includes the excess loop pole generated
by non-ideal sampling. HP has an excellent application note on
sampling phase detectors.