(SANITIZED)UNCLASSIFIED HIGH FREQUENCY RADIO STUDY, MAY 1955 - JULY 1956(SANITIZED)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-01043R002500180001-0
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 220
• Document Creation Date: December 23, 2016
• Document Release Date: September 9, 2013
• Sequence Number: 1
• Publication Date: August 6, 1958
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
R
STAT
Next 3 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
FINAL REPORT
HIGH FREQUENCY RADIO STUDY
1 May 1955 thru 31 July 1956
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part- Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Section A
TABLE OF CONTENTS
Section
TITLE PAGE
TABLE OF CONTENTS
SUMMARY
ABSTRACT
INTRODUCTION
RADIO AND TERMINAL FACILITIES
CHARACTERISTICS OF THE AN/FGC-29 AND THE
AFSAY D-806 CHANNELING SYSTEMS
PERFORMANCE OF THE AN/FGC-29 CHANNELING SYSTEM
AND ITS RELATION TO THE AFSAY D-806
RELATION OF SHORT PATTERN ERRORS AND TELETYPEWRITER
CHARACTER ERRORS THROUGH AN AN/FGC-29 CHANNEL
MULTIPATH JITTER AND ITS RELATION TO THE AN/FGC-29
AND AFSAY D-806 CHANNELING SYSTEMS
APPENDIX
ANALYSIS OF REDUCED CARRIER RECORDINGS
DETAILS OF SIGNAL-TO-NOISE MEASUREMENTS
FREQUENCY STABILITY MEASUREMENTS ON THE
WESTERN ELECTRIC D-1524.99 FREQUENCY
STANDARD
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 CIA-RDP81-01043R002500180001-0
t;
Section B
SUMMARY
An evaluation of the operation of the AN/FGC-29 channeling
system was performed over the 5000 mile radio circuit between
Kahuku, T.H. and Riverhead, N.Y. One of the purposes of these
tests was to determine how well the AN/FGC-29 would operate using
keying speeds appreciably faster than that for which it was de-
signed. It was found that with keying speeds up to 104.166-- bits
per second there was no appreciable degradation in performance.
Another purpose of these tests was to determine the relative
improvement in operation to be gained by substituting the AN/FGC-29
for the AFSAY D-806 channeling system. In terms of percent errors
the AN/FGC-29 averaged 1/5 the error rate when the MAY D-806 was
producing 1% errors. Normally the AFSAY D-806 handles 250 bits per
second per channel (4 ms bits). To provide the same information
Nita in approximately the same bandwidth the 16 channels of the
AN/FGC-29 would have to be keyed at a rate of only 100 bits per
second (10 ms bits). It was to be expected that a given amount of
multipath delay would have relatively more effect on signal elements
of L. milliseconds in length than on signal elements of 10 milli-
seconds.
Other effects of radio circuit propagation on the AN/FGC-29
channel were determined and are briefly listed below.
1. An 11 db signal-to-noise ratio measured in the channel*
is required in order to average 1% errors using diversity
*The method used to measure signal-to-noise ratio is described in
Appendix BB, page BB1.
Page Bl
Cnnifi7ar1 nntIV Approved for Release ,50-Yr 2013/09/09 CIA-RDP81-01043R002500180001-0
'
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
reception and keying speeds of 71.42 and 100 bits per
second's'
2. Error rates in one channel due to interference from one
keyed adjacent channel will average less than 1% for
adjacent channel levels not more than 38 db greater
than the level in the channel.,
3. When the error rate with a single receiver is 1%, the
application of diversity reception will reduce the
error rate by a factor of about 6.
The operation of the AN/FGC-29 using 100 word-per-minuto
teletypewriter was related to the bit errors produced in a
synchronous 48-element electronically generated pattern. It was
found that in comparing a 71.L-bit per second keying rate with the
100-wpm teletypewriter character errors, an average of 7 times more
percent teletypewriter errors occurred when the 71.4 bit-per-second
keykng was producing 0.5 to 0.05% errors.
Additional ionospheric propagation information is reported in
the appendix of the report. This information is a reduction of the
signal intensity recording into hourly medians for the month for
the Kahuku, T.H. - Riverhead, N.Y. radio circuit covering the
period of November 1953 to May 1956 and Orleans, France - Riverhead,
N.Y. circuit covering the period of August 1954 to November 1954.
Page B2
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release
?
50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
Section C
ABSTRACT
For about six months covering the period of December 1955
through May 1956 two channels of an AN/FGC-29 terminal were tested
over the S000-mile Kahuku, T.H. - Riverhead, N.Y. radio circuit;
The chief objects of these tests were to determine the relative
performance of the AN/FGC-29 Channels at keying speeds appreciably
faster than their design rating, and to determine their relative
improvement when compared to the AFSAY D-806 channeling system.
These evaluations were performed on both daytime and nighttime
frequencies so as to have comparisons with relatively small and
large multipath distortions. In order to facilitate testing, an
electronic short pattern generator was used to provide synchronous
keying for the channels. The pattern consisted of 48 elements in
various mark and space combinations. Synchronism was supplied from
a highly stable source of frequency for both the Kahuku and
Riverhead terminals.
Error rates taken simultaneously and expressed in percent
were used to measure the performance of the AN/FGC-29 channel at
keying speeds of 71.42?, 100, 104.166 and 125 bits per second
It was found that the 29 channel will handle 104.166 bits per
second with no appreciable degradation. This is an information rate
of 1666 bits per second for the entire system of 16 channels.
Presuming that an average of 1% errors represents the lower
limit of usable information, an AN/FGC-29 channel using diversity
reception requires a signal-to-noise (S/N) ratio of about 11 db
Page Cl
Declassified in Part- Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-0104f1Prin9cnn1sznrml (-1
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
measured in a channel* for keying speeds of 71.42 and 100 bits per
second. Since it was found that there is little difference between
the relative performance of the 100 and 104.166 bit rates, it may
be inferred that the 11 db S/N ratio would also apply to the
104.166 bit rate. Appreciable degradation of the 125 bits-per-
second rate however, was observed, a S/N ratio of at least 13 db
being required for 1% errors. Using a local laboratory single
sideband generator and "white noise" injected into the receiver
input, 1% errors resulted for 71.42 and 100 bit-per-second keying
rates with S/N ratios of about 7 db. This value is smaller than
that on the radio circuit because no fading is involved. It was
also determined using the local test setup that when "noise" was
due to an adjacent channel keying, an average of 1% errors were
produced for adjacent channel levels not more than 38 db greater
than the level in the channel.
Diversity and single receiver reception were compared over the
radio circuit for the AN/FGC-29 channeling system. On the average,
diversity operation reduced the error rates from about 25 to 1 when
single receiver reception was producing 0.3% errors to about 6 to 1
for 1.0% errors.
The AFSAY D-806 system normally handles keying rates of
250 bits per second per channel (4 ms bits). In order for the
AN/FGC-29 to transmit the same information rate in the same band-
width each channel would have to be keyed at only 100 bits per
second (10 ms bits). As the length of the signal element is
increased, a given amount of multipath delay variation will produce
*The method used to measure S/N ratio is described in Appendix BB,
page BB1.
npriassifipri in Part - Sanitized Copy Approved for Release
Page 02
50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
a smaller relative variation in the length of the detected signal
element and consequently less difficulty should be experienced from
this cause.
When compared to the AFSAY D-806 system the AN/FGC-29 produced
a lower error rate, the improvement averaging about 5 to 1 when the
AFSAY D-806 was producing 1% errors. Similarly when the AN/FOC-29
produced 1% errors the AFSAY D-806 averaged 66 to 100% more errors.
It was also found that the AN/FGC-29 diversity combining system was
equal to or slightly better in performance than the AFSAY D-806
diversity system, the ratio between the two being 1.5 or less for
AN/FGC-29 error rates greater than 0.03%.
Recordings of the keying element transition on facsimile
equipment were made, the deviation of the transitions (or recorded
dots) from a straight line being a measure of the jitter. In order
to reduce the effects of noise, only those recordings whose signal-
to-noise ratios were equal to or greater than 20 db were scaled.
On the average less than 0.1% AN/FGC-29 errors resulted for
100 bit-per-second keying when 1% jitter values in a one-minute
sample were less than L. milliseconds. It was further found that
for daytime operation of 14, 15, 16, 19, and 20 mc and nighttime
operation of 7, 8, and 16 mc 99% of all the 1% jitter values in a
one-minute sample gave less than about 4.0 milliseconds of jitter.
This indicates that multipath jitter was not a problem for the
AN/FGC-29 during these tests. An additional relationship established
was that when the F/MUF 3000 KM (where F is the operating frequency
and MUF is the measured maximum usable frequency for a 3000 km path)
is greater than 0.5, all the 1% jitter values (for one minute
Page C3
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
go,
samples) were less than L. milliseconds.
Comparison of the mounts of jitter produced through the
AFSAY D-806 and A1/FGC-29 indicated no substantial difference when
the two systems were operated at the same bit rate and using the
same low pass filter cut-off frequency on the detected signal.
Since it is expected that the AN/FGC-29 terminal will be used
on many of the important military world-wide ratio circuits to pro-
vide communications by teletypewriter, a large amount of teletype-
writer performance data may be accumulated. It was, therefore, of
interest to relate teletypewriter character error rates and short
pattern error rates. In these comparisons the teletypewriter was
operated at 100 words per minute and the pattern generator at
71.4 bits per second (14 ms bits). It was found that when the error
rates of the short pattern tests ran between 0.05% and 0.5%, the
character error rates for start/stop teletypewriter were about
7 times greater.
A brief series of comparisons were also made between the
character error rates of a 2 channel time division teletypewriter
signal operating at a keying speed of about 86 bits per second in
the aggregate and the error rates of the pattern generator signal
operating at 100 bits per second. These tests resulted in corre-
sponding ratios of from 5 to 17.
In addition to the above data taken on the channeling equip-
ment simultaneously, recordings of signal intensities were also
taken. Hourly median values for the month were determined and
classification of the distribution of the day-to-day variations
for each hour were also made.
Page C.
Declassified in Part- Sanitized Copy Approved for Release @50-Yr 2013/09/09: CIA-RDP81-01n4fliRnn9nn1sznrirm
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
It was found that on the average, winter signal-intensities
were stronger than in the spring and summer during daytime opera-
tion. More than 3/4 of the day-to-day variations for each hour
followed a log-normal distribution with the standard deviation
varying between 1.5 and 27 db. A correlation of daytime signal
intensities with the smoothed Zurich sunspot numbers indicated that
there was an inverse relationship between the two.
Included in these results are signal intensity recordings not
previously analyzed from contract DA49-170-sc-1131 for the
Kahuku, T.H. - Riverhead, N.Y. circuit November 1953 - November 1954
and Orleans, France - Riverhead, N.Y. August 1954 - November 1954.
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part- Sanitized Copy Approved for Release ? 50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
It was found that on the average, winter signal-intensities
were stronger than in the spring and summer during daytime opera-
tion. More than 3/4 of the day-to-day variations for each hour
followed a log-normal distribution with the standard deviation
varying between 1.5 and 27 db. A correlation of daytime signal
intensities with the smoothed Zurich sunspot numbers indicated that
there was an inverse relationship between the two.
Included in these results are signal intensity recordings not
previously analyzed from contract DA49-17040-1131 for tile
Kahuku, T.H. - Riverhead, N.Y. circuit November 1953 - November 1954
and Orleans, France - Riverhead, N.Y. August 1954 - November 1954.
Declassified in Part -
Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043Ron7snn1Rnnn1_n
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
4
Section D
INTRODUCTION
The chief object of this study was to determine the
suitability of the AN/FGC-29 transmission system for handling the
information stream from such equipment as the AFSAY D-806. This
called for operating the AN/FGC-29 equipment at keying speeds
considerably higher than the keying rate of 100 word per minute
teletypewriter signals, the rate for which this equipment was
designed. Error rates at keying speeds of 71.4, 100, 104.166 and
125 bits-per-second were compared and evaluated over a 5000 mile
radio circuit from Kahuku, T.H. to Riverhead., N.Y. As a further
aid in performance evaluation, simultaneous .comparisons were made
over this radio circuit using an AN/FGC,,29 channel operating at
100 bits-per-second and an. AMY D-806 channel. operating at
250 bits-per-second.
Another object was to determine the relation between
signal-to-noise ratio and percent errors of an AN/FGC-29 channel
operating at the above keying rates and 41so the percent errors
when the channel carried 100 wpm teletypewriter (TTY) and when it
carried 60 wpm diplex TTY.
Still another object was to test and compare the AN/FGC-29
diversity system with the diversity system used in the previous
radio circuit tests of the AFSAY D-806.
Jitter recordings were made and their scaled values related
to short pattern errors, AFSAY D-806 jitter, and the ratio of
operating frequency to maximum usable frequency (F/MUF).
Page D1
_
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Continuous signal strength recordings were made and their
analysis included as supplementary data.
Declassified in Part - Sanitized Copy Approved for Release 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release
?
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Section E
RADIO AND TERMINAL FACILITIES
Transmittin Facilities at Kahuku T.H.
An LD-T2 Western Electric single Adeband transmitter was used
to send test signals from the RCA transmitting station at Kahuku,
T.H. This transmitter provided for operation in the range of 4 to
23 mc, and was capable of accepting two independent audio frequency
bands of from 100 to 6000 cps. These bands would appear in the
radio frequency output signal as upper and lower sidebands of a
pilot carrier. The transmitter also provided automatic frequency
changing by push button control, permitting selection of any one
of 10 preselected frequencies. The time required to accomplish the
frequency change was about 15 seconds.
This transmitter was rated to have a peak-envelope power (PEP)
output of 4 kw. During most of the testing only half of the rated
PEP was used or an average of 1 kw output. Since only two tones
(on one sideband) and a suppressed pilot carrier were transmitted
at any one time there was about 0.5 kw average available for each
tone. The pilot carrier was normally operated at an average level
of 0.1 kw output.
A rhombic antenna was used to direct the signal toward
Riverhead. The dimensions Of this antenna were as follows.
Azimuth 53?271
Side length
Average height above ground 107.61
1/2 Side Angle 66?
Slope of plane of antenna 5? rising in forward direction
Design frequency 11 MC/s
Page El
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Ap
roved for Release
?
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
?":1.
'p
i;y'?
A number of frequencies were made available on a loan basis
by the Army, Air Force, Navy, and RCA Communications, for use with
this transmitter. These frequencies are listed below.
Ennatau Service Limit Hours in GMT
20,288 Air Force
18,870 RCA Communications
16,189 Air Force
15,665 Navy
14,615 Air Force 1330-2230
14,550 Air Force
14,450 Air Force
12,127.5 Navy
11,515 Navy 2330-0210
11,015 RCA Communications
10,738 Air Force
10,115 Air Force
8,180 Army
7,922.5 Air Force
7,710 Army 2330-0430
7,520 RCA Communications
6,790 Army
Use of the above frequencies was subject to prior coordination with
the military and civilian services listed. In practice it was
found that due to interference from other radio stations most of
the frequencies were unsatisfactory. By selection of the sideband
with the least interference and using a 4 kc instead of a 6 ko
receiver IF bandwidth, a satisfactorily interference-free radio
circuit could usually be found.
The terminal equipment used to provide audio test signals to
modulate the transmitter was located in a shielded room. Included
with the terminal equipment was a high precision 100 kc crystal
oscillator for generating an accurate time base reference, pattern
generating equipment, and four frequency shift oscillators.
Appendix C describes 100 kc oscillator stability measurements. Two
Page E2
Yik
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043Roo25oniRnnni_n
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
p
of the F.S. oscillators were for Channels 1 and 3 of the AN/FGC-29
and the remaining two were for Channels 3 and 4 of the AFSAY D-806.
There were also available a 100 word-per-minute transmitter distri-
butor, and through interconnecting line facilities with the Honolulu
Central Office of RCA Communications, a transmitting diplex distri-
butor.
Personnel of RCA Communications maintained and operated the
LD-T2 transmitter, and an RCA Laboratories engineer operated the
terminal equipment.
Receiving_Facilities at Riverheadt NO Y.
The radio signals from Kahuku, T.H. were received at Riverhead,
N.Y. with uwo rhombic antennas. The physical dimensions of these
antennas were as follows:
Height above ground (center)
Height above ground (western end)
Length of side
1/2 of side angle
Azimuth
Damping resistor
Slope of plane of
Length of 200 ohm
line to antenna
antenna
4-wire transmission
coupler (Bldg. 6) 17sot 559.5'
j283-R #284-18R
55' 1061
55' 165'
328' 3601
65.10? 71,3?
284? 284?
850 ohm 850 ohm
00 9.88?
There was a loss of 1 db per 1000 feet at seven megacycles on
the 4-wire transmission line and this loss varied as the square
root of the frequency,
The antennas were fed over 4-wire lines to two antenna multi-
couplers each having a gain of about 10 db. The multicoupler
outputs were fed through two 1000-foot lengths of 75 ohm coaxial
cable and terminated in two receiver inputs. Western Electric LD-Rl
Page E3
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-0104nRnn9cnni szn MI 1
ueclassified in Part - Sanitized Copy Approved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
f
receivers were used during the early part of the radio testing
and these were later replaced by an AN/FRR441 receiving system.
The AN/FRR-411 consists of two sets of Collins R-390/URR receivers
with Hoffman CV457 single sideband converters.
When desired, "White Noise" was inserted between the re-
ceivers and the multicouplers. A photograph of these noise
generators and the AN/FRR-41 receiving system is shown in Figure El,
and a schematic diagram of the noise generator is shown in
Figure E2.
During the Kahuku to Riverhead tests continuous signal strength
recordings were made. One of the LD-R1 receivers was fed from one
of the antenna multicoupler outputs over an independent transmission
line. The multicoupler used was the one receiving its signal from
the sloping rhombic antenna.
A lead was brought out from the automatic gain control portion
of the receiver, isolated with a 2.2 meg. resistor, and fed to a
capacitor of 20 mfds. The de potential across the capacitor was
amplified and applied to an Engelhard recorder. The R-C time
constant of 44 seconds was about right to reduce the recorder trace
scatter due to fading to a reasonably well defined line.
The recorder was periodically calibrated by substituting a
standard signal generator output for the antenna feed to the
multicoupler. To speed up the process, the 14 second time constant
was removed during calibration.
1
CO CO ? CO yip iippo cppi
Instruction Book for Radio Receiving Sets AN/FRR-40 and AN/FPR441,
Order No. 26565-PH, 1 September 1954
Page E4
1 .?.
Declassified in Part- Sanitized Copy Approved for Release 50-Yr 2013/09/0q ? rin_Druno.f
s.
d
4
?
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
At the receiver site were located a pair of distribution
amplifiers, each fed from the audio output of a receiver and pro-
vided with three independent outputs. One pair of outputs was used
to feed a monitor amplifier speaker system. and a calibrated amplifier
for signal-to-noise measurements. The second and third pairs of
outputs were used to feed AN/FGC-29 and. AFSAY D-806 signal detection
equipment.
It was felt that the best and quickest way to evaluate system
performance comparisons was to operate the two systems in question
simultaneously over the same radio circuit at as near the same radio
frequencies as possible and with identical or comparable transmitter
powers. With this in mind, two adjacent AN/FGC-29 transmit channels
were set up at Kahuku and the two corresponding AN/FGC-29 receive
channels were set up at Riverhead. These were channels 1 and 3.
Two adjacent AFSAY D-806 transmit and receive channels 3 and 4,
were similarly set up. The center frequencies and shifts of the
AN/FGC-29 and. AFSAY D-806 channels appear below.
System Channel Center Freguency Shift
AFSAY D-806 3 1390 cps ?100 cps
AFSAY D-806 1. 1755 cps ?100 cps
AN/FGC-29 1 1785 cps ?42.5 cps
AN/FGC-29 3 1955 cps 42.5 cps
(During the tests, channel L. was never used.) In this way the
radio frequency wave propagation for the two paths being tested
was essentially equivalent since the radio frequency separation
of the channels was but a few hundred cycles.
Test signals to modulate the channels were provided by two
identical short pattern generators. Each of these produced a re-
Page E5
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01n4f1Pnn9cnni sznnt-11
Declassified in Part - Sanitized Copy A ?proved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
????????..
petitive pattern 48 bits in length. The pattern consisted of 1, 3
and 9 unit length elements in various combinations. A photograph
of the pattern is presented in Figure E5A. The length of the bit
(or the bit-rate) was determined by the frequency applied which in
turn was derived from the 100 kc frequency standard. For instance,
when 100 cycles was applied to a short pattern generator, the bit
length of the pattern was 10 milliseconds and the bit rate was
100 bits per second.
The following table shows the bit-lengths used and their
corresponding bit-rates.
Bit Lerlgth Bit Rate (and control frequency - cps)
14.00 ms 71.428-- bits/sec. (approximates 100 wpm TTY)
10.00 ms 100.0
9.60 ms 104.166--
8.00 ms 125.0
At Riverhead the above control frequencies were similarly
developed to drive another two short pattern generators identical
to those at Kahuku. However, at Riverhead there was also available
two continuously variable phase changers arranged so that the phase
of the applied frequencies, and hence the phase of the 48-013ment
short patterns, could be adjusted to coincide in phase to the
patterns arriving over the radio circuit from Kahuku.
At Riverhead the locally generated and properly phased short
patterns together with their exciting frequencies were fed to two
coincidence units. Into these units were also fed the corres-
ponding Kahuku signals after having been demodulated by the
terminal equipment (either one channel of the AN/FGC-29 and one
channel of the AFSAY D-806 or two channels of the AN/FGC-29).
Page E6
Declassified in Part - Sanitized Copy A ?proved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R0025001soon1_n
Declassified in Part - Sanitized Copy A. proved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
In the coincidence unit, a 20 microsecond pulse was used to sample
the center of each bit of the short pattern from Kahuku to
determine whether the bit was "high" or "low". Each sample was
compared to the locally generated pattern. As long as the samples
agreed with the locally generated pattern the coincidence unit pro-
duced no output. Figure E3B shows a portion of the 48-element
pattern with the 20-microsecond sampling pulses, centered within
the elements. If, however, a bit from Kahuku became mutilated due
to some cause such as static or fading etc., and the sample did not
agree with the corresponding bit of the locally generated pattern,
the coincidence unit involved would produce an error output pulse.
Each of the two coincidence unit outputs fed an RCA Time Interval
Counter. By means of a single manually operated switch these two
counters were started simultaneously, and by means of this same
switch they were automatically stopped one minute later. However,
during periods when the errors were few, the count was prolonged to
5, 100 and occasionally 20 minutes. In all cases, the data were
reduced to the form of errors (or fractions of errors) per minute
and this figure divided by the total number of bits received in a
minute to give the error rate, which was expressed in percent.
Where % errors of 14 ms bits were compared with % character
errors of 100 wpm TTY and 60 wpm diplex TTY, the TTY % errors were
determined by dividing the number of TTY operation errors
(including line feed, carriage return etc.) by the total number of
operations during an interval of usually 5 minutes. To facilitate
counting errors in the TTY copy the transmitter tape was punched
Page E7
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R0025001801)01_n
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
up as follows:
First the letter "0" was punched, then "RYRY---" until
34 characters were punched, then the letter "I" was
punched, followed by carriage return and line feed. This
was repeated for 9 lines. The 10th line was punched in
the same fashion except that the line finished with "W"
instead of "I", followed by carriage return, line feed
and five spaces. The five spaces were required to glue
together the tape so as to make it a continuous loop.
At Riverhead the R-bar and the Y.--bar were removed from the TTY
machines. When receiving the above tape the TTY copy under perfect
conditions appeared as a blank strip of paper with a column of
along the left margin and a column of "I"s down the center,
with every 10th I replaced by the letter W. Errors appeared as
random letters between the columns of Ots and Its. This made the
errors easy to spot. If a carriage return was in error, it was
indicated by the appearance of an I (or W) on the right hand
margin of the paper. With this arrangement, almost all types of
errors could be detected and identified. A transposition of R
and Y was of course lost. It was felt that this method of
detecting TTY operation errors was as accurate as visually
counting errors in "Quick brown fox----" or solid "RY" copy,
because in both these arrangements errors are occasionally
overlooked.
The system referred to as "Diplex" is a synchronous
commercial system in which two 60 wpm TTY are multiplexed,
resulting in a bit-length of 11.66 ms or a bit-rate of 78.6 per
Page E8
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
second2. For further details of the frequency dividers, phase
changers, pattern generators, coincidence units etc., see 1st, 2nd
and 3rd quarterly reports for this project.
In all tests involving two channels of the AN/FGC-29, the
channels were interchanged each half hour. That is, the test
material on channel 3 would appear on channel 1 on alternate half
hours. This was done to equalize the effects, if any, of
QRM (interference). In comparisons involving the AFSAY D-806
and AN/FGC-29 channeling systems, power levels were adjusted
according to the number of channels each system normally transmits.
In the case of the AFSAY D-806 system the power was apportioned on
the basis of 7 channels whereas the AN/FGC-29 was apportioned
power based upon 16 channels. For the approximately I kw
average sideband power transmitted the AFSAY D-806 channel con-
tained about 695 watts and the AN/FGC-29 channel contained about
305 watts. Where two channels of the AN/FGC-29 system were
compared, each contained approximately 0.5 kw average power.
The last five minutes of each half hour transmission was
utilized to measure "noise". This was done by having all modulation
removed from the transmitter during this period. The level of the
noise in one of the AN/FGC-29 channels was then measured. When the
modulation was returned to the transmitter the signal level in this
same channel was measured. The ratio in db of these two levels was
called the signal-to-noise ratio.
For further details of S/N measurements see Appendix BB.
2See EM-63-62 "Recent Developments in Time Division Multiplex"
Page E9
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
Declassified in Part- Sanitized Copy Approved for Release @50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
4 ?
Two types of facsimile equipments were used to record multipath
jitter and hit distribution. One was a relatively high speed RCA
experimental machine and the other was a standard U.S. Army
TT-1E/TXC-1 facsimile machine.
The RCA machine provided fax recordings on a strip of paper
8-1/2" wide fed continuously from a large roll at the rate of
10-7/16" per minute. The effect of a recording stylus moving
across the paper was obtained by drawing the paper across a
rotating drum 8-1/2" long, and 2-1/2" in diameter containing a
one-revolution helix of 0.037 inch platinum iridium wire fastened
to its cylindrical surface. This helix pressed the paper against
a "striker bar" parallel to the axis of the drum. Each revolution
of the drum caused the point of pressure on the paper between the
helical wire and the striker bar to traverse the 8-1/2" dimension
of the paper. By a wet chemical process the paper was treated so
that if a current passed through it as it passed between the helix
and the striker bar it changed color. If a steady voltage were
applied between the striker bar and the helix, straight lines
parallel to the short dimension of the paper would appear as the
drum rotates. If, instead of a steady voltage, short pulses were
applied of the appropriate frequency commensurate with the speed of
rotation of the drum, then a series of dots would appear on the
paper forming one or more straight lines parallel to its long
dimension.
For these tests the speed of the rotating drum was adjusted
so that frequency transitions of 50-cycle and later 25-cycle
reversals produced at least two lines lengthwise on the paper.
Page El0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
?
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Variations in the linearity of these lines were due to variations
in the timing of the mark-to-space or space-to-mark transitions
and were caused by "jitter". Random spots between the straight
lines were due to noise crashes.
The Army machine TT-1E/TXC-1 was an electro-mechanical fax
transceiver of the revolving drum type for the transmi3sion and
reception of page copy. The drum normally rotated at a rate of one
revolution per second as established by a tuning fork and motor-
driven gear reduction arrangement. The equipment normally scanned
a single 12" x 18" page of copy in 20 minutes.
In order to have this machine record two cycles of the short
pattern when the short pattern had several repetition rates, a high
precision variable frequency audio oscillator (General Radio Type
1107-A) was substituted for the fork frequency standard.
Recordings on this machine of the short pattern gave data that
could be analyzed for jitter, and, when the output of the coincidence
unit was recorded, gave an indication of the distribution of errors
among the short pattern elements.
During a large part of the teat one or more induced delay
curves were taken daily and sent to the NSA for analysis by their
personnel.
An induced delay curve was obtained by noting the % hits per
minute obtained when the sampling pulse was alternately displaced
by the phase changer various amounts either side of the short
pattern bit center. If the transitions from mark to space or
space to mark of the short pattern were instantaneous and the
sampling pulse of infinitely short duration and there was no jitter
Page Ell
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
there would be an instantaneous transition from zero % hits per
minute to 50% hits per minute as the sampling pulse passed from the
correct bit to either the preceding or following bit. If the short
pattern consisted of 10 ms bits the calibrated phase changer would
indicate that the sample pulse could be moved ?5 ms from element
center before errors would be produced. Since the sampling pulse,
had finite duration (about 20 ?s), the mark-space transitions were
not instantaneous, and the frequency dividers produced some jitter,
there would be some reduction of this ?5 ms figure in the back-to-
back condition. The addition of transmit and receive roofing
filters still further reduced the available bit length. The
following table shows the % reduction of the bit length due to the
pattern generating and sampling system aad the roofing filters when
operating back-to-back i.e., without the radio circuit link.
% Reduction Due to
Rate % Reduction Due to Generating System
Bits Sec. pit Length Generating System and Filters
71.4 14 ms
100. 10 ms
250. 4 ms
AMP
0.5
7.1
0.7
14.5
0.75
10.0
Figs. E4, E5, and E6 show sample induced delay curves taken
over the Kahuku-Riverhead radio circuit using 4, 10, and 14 ms bits
respectively. Figs. E4 and E5 were taken simultaneously using a
nighttime frequency (6790 kc) and illustrate a relatively large
amount of jitter. These are typical of approximately 600 induced
delay curves supplied to the NSA for the months of March and April
1956.
Page E12
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Fnotorrarh of ; heceivinn ,:sten
and MIs Ge-i era tl!Ir nme:1
91263
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
1.0???????????
??????????????????????????????????????????????????????
3 -t
LL Is
kto
1 'SI
L:1
?
1.1 .11
140o.
fr'!PI
?41
e
4-1
? 0,1
111
0,7 044
At/VIII
09018 4
4
?ory
VO "
;
?1"
? C
-r
t's
y
t t
lo >)
t
41 114 it
/ oSz fr - Teici
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
0-1-00081-009ZOnl?1701-0-1-8dCl-V10 60/60/?1,0Z JA-OS ? eSeeiei -104 pamdd Ado pazillues - ;Jed ui pawssepaa
(7/
c/ .4 11 r)(t
Y.+ 0 -Z-
1 113 111 11111111111111
1? 1 Imo i nil 011451=11311.5.6==
11111111
II 1111 RI 11111 11111111 MI 11111111111111111/1
I
I?
21A 2 i
I
I I I i 11111/11 I 1111/111111111 1111111111111111111111
I 11111 111 11111111111111111111111111111111111I11111111
1111111 1 11111111111111111 /1/111.1111111/0.1111111111111111/1
"lir 1111111111111111111111111111111401/111111 I
ill 11 iliguillimplui 1 I 111: iimplumpoill
II
1 a laallillilliiiialliT i 1 11111111111111111161111
- illir-1-411.-=.....212
I Li PiteMNIVALLE i tarageriPPII FININLI
. r2"2.2"? MCM:=2:::::::= Willign
11144 i===::1111011111111
"12522""::"
51111=11211111011111111111 S as= wiess=saissa=llassal
1=14-11/ Wiliiiiiiii111111 Al Iiiiiihibliiiiffli1411111
I
5 Allggsgagag aggmlaiglo Ellimmg IggvAggifigiNgicg, A _
111111 MO Miaill111111111111111i1118141111 1111 1
II
mill minim orisilloglimilimplogragillii Min
I
111114111111miumillismalimilgillaiimingEmillumill
milimmailimaimmligilliggilwpggelliggimitelling
ElsaingradgmlEmielealgEminimating. EgiNEN411Mil
ill iiiiiiiiiiiillillitiiiiiiirimilii=iiiiiiiiigulli ilk,
.1i
- sr moor mom in up lim op rmormlimmu il
1 1 111111 trilillilillidai ilhailhalliili
I gm
114:14 i, ma niminimpli: limatipmeill gum
wilm.
mu
in?.... .....1 rum. mums .... mom
Ninthlaillamaill. a 111111111111111?31 .....
gm!!
1 Abiall sir
U al ma SX2111 lill iiiiiiiiiiill Iiiiiiiiiiiet Ill.
a X gs 22....
1111
a 1 11111=_S-E= - _=====.==. u=... E:ES=
Nais Wiraigmfiiiimaf gaggRIEgagsmaamm5=wagneal=gara guremagE il
120.3E3EE ==EE=-- ---"Ti65-1==="% : p
BB -==ESS=SEE=--== - =.
S-==-lailiii 8
aszl= =-Er.,_saea==a-A-g=am=N=g-seg=as==-ggsEEgi:E==.-4,aassa,== g.g,-.4asE,74
...J.,.? =5117.=E=Ltam;=z?...4=12==
--Ala
Na2gP-5/=MaqgijrIE-EiMngENE?P4g-AWE=rg2I===95.
%.1
-4.5.P4A 00--A=7 :-:,1 ..41A4gggm=g-ea444;4=-..-1.44.4 5.:;24qTin. ;-s====
.. . ..... ....,.. -A..E...= - --.. L....=., . -----
mmilvan :._ 1..Nd--=...
====== 1== z=g,E======-c--r-i-ix.--==-==.=5====a=3 Eg==E=
----=----=---=-=agmam as MB IIIIMOMMI
EPEE EOWVWVOW.=..P.....=
iggwaglglftg3ggaggeSMS.'="ASIS5g.ggggSg?ggiSIIMEM!
gM'=*
.''." IlligNIN1Tigigg========....==ggE4:4gPg4ggIgg.HeA iliP:g
= =,...
=... g gagt?gallgINENIMIR-7P=,== =.= mIA4MAgE=.51.1m-sa.ti E===
WM
gT-1?
111 4mW0u-Mm-m4p gaWlgoMlmiimwr
Ow wmM lT slmma1im1 o.osrIos tiR Mgoig g 6Mm g-g0iml pr0sislm.woEmmo=in-Si mE.im.JmaSmmPu.ai.FsawAmmIoi1i1 :gui:1mmtmawalwmia!luml! iia E.m44m1--.W- 0m; ng4lo:ieom;0 m7mpIaOIM
1111111111illlgiiiiiiiiiiiliiiiiiiiiiiiiiiiIl4i
i11 ii i"1".11 l i
linl o m.iaa.msm aap..
ara amosamwmiiawormmliamaammm o
w
2:1IFSEr:::::......41.0:::ELi:-i.ir:222"1221alibi: sptal4
? 'Mrii EE1IOI gilia01 tEE-4m:
BEBE= BO SEISM i EISS=4,2SPE=E5415,P=SEg;=S*ftSESS:
g0 1_111Paglirli==irganigiiigligli ===-01.04-;"="41i
laiNE-1111tabilliP I=TAIIMEE1=- I747.:77:17271:d
i
111 - -.:.= . 10!0 11 4 gi m=gg itHiggggilliOgi iMgffigginigi
wtv443.71g,14g1,7q4atTim si5:77-14klovia5
x a BEEeS=EgEmt Em==21.71=R.v!..#?ERE311:2"44 tE'Ea3=
glii==g=E=gsga- ==;=Eig,E=AvagAU3ginEEET4
lilli - 111101111111U1 1 gitti? vrg_ iNglogg=tatilikm4ggEgq
hili hill INORMII 5101AIMPR. n5Rt"1- iBtLI4.0gt':-; TT; 3554411
flgiNTAINI4111 liehtEM T,gaRAgiNtligyirWEttil
?'/
WV61.2 p c 4 C -I 1 ?v.
The Kahuku 18 MC median levels, around 2-3 PM, EST, for the
winter of 195344, and spring of 1954, respectively, were 47 and
29 db >1 ?v. The medians were lower for the corresponding seasons
of 1955-6.
yight,
For the Kahuku 6 to 7 MC night frequencies, the median levels
around 2 AM, EST, in the spring of 1954 were around 18 and 31 db
>1 ?v. In the winter of 1955-6, the 2 AM medians were around 48
and 51 db >1 ?v.
9,Emlation with Sunsota.besi...2
During the all-daylight path period, using the day frequency,
and based upon a limited number of months there was a strong trend
that the receiver input decreased with an increase of the smoothed
sunspot number.
Types of Frevenc Distributions
Most of the day-to-day variations of the carrier levels
followed a log-normal distribution. The occurrences of log-normal
distributions were 83.1% and 71% for the Kahuku and Orleans
circuits respectively.
Page AA14
Declassified in Part- Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 : CIA-RDP81-01043R0n7snn1Rnnni_n
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
,
Standard Deviation (6')
The following were the extreme ranges of the hourly values of
the log-normal distributions:
Kahuku - Riverhead 1.5 - 27 db
Orleans Riverhead 3.2 - 22.5 db
The following were the monthly range of average monthly
for all hourspand average 6 for all months, for Orleans:
Orleans: Aug. -Nov.,1(64
monthly range (all hours)
Log-normiq
8.8 - 14 db
average for all months 10.3 db
For the Kahuku circuit the monthly range of average monthly
p for all hours, are 2.3 - 18.2 db for the log-normal distri-
bution. The average Or for all months, for different parts of the
day are:
Log-normal
Morning All- Evening All-
Transition Daylight Transition Darkness
6.0 db 6.1 db 7.6 db
7.7 db
For the Kahuku to Riverhead circuit, much greater day-to-day
variability was observed during the evening transition and
all-darkness periods than during the morning transition and the
all-daylight periods.
Bibliography
1. CRPL-F Series, plot of sunspot numbers
2. Signal Corps RPU No. 9, August 1950
3. RCA Engineering Report F-43 100, S. Goldman
Page AA15
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R0025001soon1_n
Declassified in Part - Sanitized Copy Approved for Release
`,.71;:cr)
E" 4
0 (.7.-1
C
0
IHH
1-4
44'4
%% Cr,'
V
HI
H
Itt;;?1
if.1 (I)
4
?
OR*
50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
?(N 1,'S PLOT OV HOURLY AVERAGE SO pr CARRIER S IGO 1.).
HON YANUKUIS S SB TRANSMISS IONS HOE TVU AT R IVkBH ni)
go'
'(
3
ii:P91,14131,11( 1951-1
902
70
60
III: 111111111 I I
1111111111 . III PP
. .
m
:Rs 111 PIO' I 111711111111111111
.
11111111"11
I
I
m.....5 ..111.151.1210 or 11111110111
is
Itimilimmiligirdrumahliktribillitillinallpr- inq
....
www. . m ow . wrill I E A .
w
pwwwwwwwwm w wwwwww wwww wpw parill
wwwlaw 5:1111w1wpwalpw sawwWww
_ a
. al
wpm
I _ um
. Mem?m"mann 21121121151.M2 =MAU SEM
mmounimmmimmilmrimmlim. mom
omm.. Immo ..w. immommommommilmil mmommom mismilm
mnsw.was. nwm. smawmincomm.
m. nowwwsmommom. nom. m 1 nil . 111111 :11 i
mama. .......mmummm m m nu ww
...m.mm.............. .11.1......m mommimm am..
mrsisommanumml..smicermil iiiii.maimi
2 1112221111125111111:1152211:151f111 121:1112.
? ? .... . um m....... ....v. sai . . ....... ?
I
1 . ........................A.v...... ....m.. se
..... mom .....m.mmsamwm....
ow ...miwn won WWWWWwftMWWWWWw 111 1
MWWWW
1 1=2:2116:111:111111111:111=0:11111 Alp
Immissammomm miummilms mwelmatimmimmommws.4. ? is
I
........ mummonsusamionagial 1.1 Ws
I 151111111111111111airparr
WI
WWWWWWWWWWWWWWWWWWm .t. wwwt m i 1
::: 61.51.1
.................. Vir. MMW -.A
ill :::1:!!!!211
............. ....,..n.
.... ............... ? wilells
WWWWWWWWwWwwwWwwwwwW WWWWWWWWW6A W WwWWW
WwWwWWwwwWWWwwwwwww malwWWwieww WOWWWWwWwW
WWWWWWWWwWWWWWWW WW WWWWWIM 11121111221
WWWWWWW M W MW WWSWWW
11 WWWWwWWWWWWWWWWW WM WIMONEWWWwW
WWWMWWWWWWWW1WwWWWWWW 111=WimMOMM
WWWwWWWWWWwW wiWWWWWWW WWWWwWwl
WNW* UMW WM WwwWWWWWWWm.AWNWWW
WW W Mime .. POI 121'11
1111112211
. .........111: m . ..11.
a lam
. .... .2. MmINME m own 1
101.1 111........... ........m...
Mie mm n=sommarom ? wm
gm lla
O
am um. n
now.. miwiwn
inweinnm. w mom. nm...11... ono =ow pow imm.
rill mw wilmwriman: :nom
""1121112 A= II 22 2111.11..m.m mogml.
IL 11 111 11:11E
? . .... 12.1.2 NM 11141212221
.1121 12 .122.221112.
HirlErliffl IMLIBIUMIIPISIPIerIP
... ... 1
21.1111.201 61 114111661111112 ' MA .2
... ..... .
11122221121111 SMI.111111.111111112111 m 11
a a ii a
????????
004.
hoar
20
012
4
AM
r,
10 11 12
EST
PM
Declassified in Part- Sanitized Copy Approved for Release ?50-Yr 2013/09/09 : CIA-RDP81-01043R00
Pi ,tire AAI.
 Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
rx1
C11?"???
cr.4
E-4
16
0
8 8
C'Jr4
? rt4
Irtp H0
rz E4
41 1:D
t.)
E-4
074 M
C4
ri 4:4
9
E-4
4 E-4
H
Ci)
a M
E-4
0 0
H E-4
c/) A
VE131
.ce.
C:41
'P. A
X
0
4.
,to
MASS PLOT OF HOURLY AVERAGE SUPPRESSED CARRIER
PROM KAHUMPS SSR TRANSMISSIONS In:CP.:IVO AT
14,1150 Ito
W 11 12
80
1W4

sMW.m.aCWE..aCWo..
60
30
20
10

-1012
.mmoaMm
1 .ftsmiImNEI1I..MmMio1M1 .ua1WM.l.lMmm11m=.r.IOIuaI W.omm1 m..Wm
W5wWmu
.em
I.0an
III
M1IE 6l11
..
. .
.rm.mm.. .wIDsl..opMa.uspMmM.a1si..mIOeN? ?mMA ?a
ftm.m1aMu..m1nMmMm.m.wIwWWImu..OomW
le1
an.1Mm
I
FME1111
WWWPWWW
1:111EMM1 Em1 1.11WW
lallalliala
111111
::::::::::L_
WWE ihm
mi:1Ir4lma
.o
ms1
mmaao mIl
MrMMMNWOM ira=l1
MOMMIWIRM WIMMIIIIIMM
1111111011MWSIMM
.......... mil 1..1111111
........
-Emma
........... 1....
.... .11.1 .....
mum .... 1...:.
: 1 nun
. . ...........==!!!!!!!!!!11..
i
mma:11.1 . ...............
Insummig 1 comprimpcm-
.. .. . ...
.... ... .............
. ...... ...
mommem===11 mmummummminsammarmammgm
immummarmwm nommumw
mosammummomm mommemommaRaimm
ACCOMMX: . ..
. ..
smargi 11.......
....... ..
III :MIER
I= Imp*
I
it Egral al=
lid i III
imp 1
III III
__
1, ?
SIGNAL .T.di.v1.152,
EIVERHEI'M
11:1 11 12
,??????
7-71-
milmsammomin
uninammumemn
wwwwwwwwwwn
winammormamn
mommummminn
11111111111
rannirrim
I
111111111
1 IMMO
sm. .....
WW1
. ail
........
........
AM
10 11 12
EST
PM
10 1
12
gure AA15
o A. roved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
MASS PLOT OF HOURLY AVERAGE SUPPRESSED CARRIER LEVELS
FROM KAHUP, SSB TRANSMISSIONS
RECEIVED AT RIVERHEAD
4';????
$-4
rt".i ft)
a)4.)
0 4.)
rr.4.
0
00
r--1
? 0
0 C-4
('J
TS 4)
0 0
00
(1) `'e)
H-1
1-i
t> 0 A
0 C.;
a).'-)
;-4 5
ro
cd
0 cti
E-4
b0
ctl 0
rd

Declassified in Part - Sanitized Copy Approved for Release
?
80
! ? '
; ?
98 99 99-5
uti 95
sou, Alk
- - -
-
IVO
-
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
99.8 99_9
0
-2 0
4C-)
10
9.9?
Declassified in Part - Sanitized Copy Approved for Release
?
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
99.9 99.8 99.5 99
98
95
90
80
70 60 50
30 20 10
11=111.....1111111."MelailinraratinaMi=0 IMAM. SO= WIWI ?am:m=0m
01?111.1111 *EOM
II-1-????????111".."1"1"."-iss-wwwigusirjeuiraS.R'nlissurarragrAili
99.99
neimmara_41_Men7=ri meausammIsiams.
150111011111011011 0110?11
agnomINMI ogli sideM
Miams.gc,.
NEM
' ?
- I
Emma MILMMagars
-I' am w-fin "mb
_
Effamwe'
0.5 0.2 0.1 0.05
0.01
1110110111!;10
e 0011111
!MINIBOOM
MOM
- - 1
Jo
,ag
MENIMMINM.
ISIMENIS!
)1M9M
MENSAWair
-
_
? -
N\
tA,L
'CU
IMMO
; 02 0.5
10
20
LLL 'I
30 40 50 60 70
>
Declassified in Part - Sanitized Copy Approved for Release
?
80
! ? '
; ?
98 99 99-5
uti 95
sou, Alk
- - -
-
IVO
-
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
99.8 99_9
0
-2 0
4C-)
10
9.9?
Declassified in Part- Sanitized Copy Approved for Release @50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Appendix BB
DETAILS OF SIGNAL-TO-NOISE MEASUREMENTS
In the transmission of information over radio circuits, one
of the most important factors leading to the reliability of the
system is the ratio of the received signal energy and the energy
of noise and static present when the signal is removed. This
describes a ratio of "signal plus noise-to-noise" (S + N/N)?
which is commonly referred to as "signal-to-noise" (S/N). For
large values', the difference between these ratios is small.
During thcse tests the audio output of the receiver was
passed through an AN/FGC-29 receiving channel filter having a
bandwidth of about 115 cps and thence to a calibrated amplifier
with a high speed recording meter. The amplifier input was
adjusted so that during a one-kinute period the maximum reading
resulting from signal plus noise peaks gave a peak reading of
zero db. Since the inertia of the meter movement, while
relatively low, could not permit the pointer to accurately follow
the peak excursions of the signal and particularly the noise, a
number of readings were made. Usually four such readings were
taken during four consecutive one-minute periods. Following the
last reading the modulation was removed from the. sidebanda leaving
the radiated carrier power unchanged. By using carrier operated
automatic gain control in the receiver, the receiver gain did not
change between the signal and noise readings. The receiver audio
output with the sideband energy off was reduced. The amplifier
input pad was readjusted to bring the peak output of the receiver
Page BB1
4f,
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R0025nniRnnni_n
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
back up to zero db on the output meter during a one-minute period.
Four consecutive "noise" readings were usually taken.
When these four readings in db were subtracted from the four
readings obtained with the sideband energy radiated, four "signal
plus noise-to-noise" (signal-to-noise or S/N) ratios in db were
obtained! The recorded value of S/N was found by converting the
db values to ratios, averaging these ratios, and converting the
average ratio to db.
During a number of noise meter. S/N measurements, simultaneous
"peak-to-peak" readings were taken on an oscilloscope. These
comparative measurements were made both on the Kahuku-Riverhead
radio circuit and also in the laboratory. The procedure followed
in determining the "peak,4o-peak" S/N ratio was the same as that
indicated above except that voltage ratios were read directly from
the oscilloscope face instead of the db setting of the attenuator.
Comparative "peak-to-peak" and "noise meter" measurements in the
laboratory were made by injecting locally generated white noise
with 'the signal. Figure B51 is a block diagram of the circuit
which was used.
Figure BB2 is a plot of the peak-to-peak 4/N in db Versus
the noise meter reading in db based upon data taken in the
laboratory. A similar plot is given in Figure BB3 for data taken
on the Kahuku-Riverhead radio circuit. On both plots, the dotted
line represents the calibration cullre that was determined from the
plotted points. It may be noted that the noise meter measurements
made on the back-to-back circuit with white noise were approximately
2 db higher than the simultaneous peak-to-peak noise measurements.
Page, BB2
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
ueciassitied in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-R0P81-01043R002500180001-0
The noise meter measurements made on the Kahuku-Riverhead circuit
were approximately 3 db higher than the simultaneous peak-to-peak
measurements for signal-to-noise ratios in the order of 10 db.
The db difference between the two measuring methods increased with
the signal-to-noise ratio so that the noise meter measurements at
Signal-to-noise ratios in the order of 50 db were about 7 db
higher than the simultaneously-measured peak-to-peak signal-to-
noise ratios.
Page B133
"De _classified in Part - Sanitized Cop A proved for Release 50-Yr 2013/09/nqIA
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
BLOCK DIAGRAM OF NOISE MEASURING ARRANGEMENT
KAHUKU-RIVERHEAD
RECEIVING ANTENNA
AitIMIPOMMIONOWINIPONOMMOMOMPO
NOISE
GENERATOR
SINGLE SIDEBAND
GENERATOR
SIGNAL-TEST
MODULATION
CALIBRATED
OSCILLOSCOPE
I RADIO RECEIVER
R-390
itomemINISIIIMINISOINIMPIONMIONP
CONVERTER
CV-157
AN/FGC-29
CHANNEL FILTER
(Terminated in
600 Ohms)
NOISE METER
FIGURE BBI
Declassified in Part - Sanitized Copy Approved for Release @50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
1 t
k4
itill115111110
hillinitiMPUMMINVIIIREI ff
vs,
1110 akiliiNtragiilirliniiilf 143: I
,L
3
Cilniql oprihmoPolaw
sio sionomplaq
slum lloglloommodor
ova mommainnot
1111 01901111111111111111111111111R 11111111
11
iIIii!.
11101111Wili
10011011111111TMARIUI
NNIENNINIMINDIMIni MILM
1101111,11Fillif ifielP5111111111111111160
INIESIONNETIMMPRIEliiin
MEM MIN; MEM Hi Mt ORM
mopmemom. MEMPRIVINIA MERNIUM I
' ROMMONINENNI 1211104111
MEM 111011tHlitgr malipon IAN& IMF
N14
EER &ORM Mk TIMM
pordosompan pips? mommolifflip
Hmomiwlfdbtilddro
nu:re i ;If'
111F ftlitifill 11 iiiiiii '1 fj,,IIIIIIHRIJI iinifief i ,. N $ 11111 ail 1111111T ilt, 1
t 4 ,
L lip m
2011 'alio it pill mom I inlo I III I1 dim ?
ut i 0 moll a i , ?,,r ?,?ll
t, ..?
a m t ti .mmi 1,1 iilli,41 \iI1,4 lit1:1 .11 1 i I rar.i
.t., ,
. i
li 4 ti It 1 I i t ,ii , ill 111.1. Iti I
1* T
is ril
. ii1111 Bali li -.-1 s 4 4!
..,
1 s I it 1 4,1,1 L;Itl'ill ?,,
i,,...1: eh 1111il
ME '1111 III 1 .1/4.1 1 ..4
ivAil-il.li . ,
t, H,i',, no
., 1 ,
, 1 1.
L/ r4,11 1.,
, 1 , ri I
11
, , 1, 1 . , .11/ ,
14
I /
1 iii r 1,
16' d II.
..7 117
, .
1 , II 1, li,,,tii.il 1
, 1 nt i ,
t, wi 1
r-FiJ 14 1 , I l' i 1
wi 1 s
,
. i , T.,1 I.', 1 it I. T,i
.I.,4.
1,,,,,
4 Ir , t 1 it !t r r 1 . i H t I 1 t I 11 I/
r
I iii I t I I
I I 1 1 ' I r r- 1
diii ii
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
/11
LIII 11111
1\1
111111111 11111
;
11111111111
\L,
Ani"""irillinli tiNti IMPIlimk111
i.
1111,411111111111:1:1;11iittiviiiilil!ikli:11
y
11111111111111111111111111611411111\1141011111411101
illinemn du 1114 ill - "1 ul,"! tipp .:1001 %Liu tAr irsti
1"11 A prilthic ,ILIIIIIItT I
1111 111011006
PaiRmid 11116411115"11111ilit"1" ligliM1811111111 . I
AINIINI111111
" 41111164 111 IM14411111111 orbit,
: lifili : ? ?
1111111111111111111111110 1 L 1 lim 1 1
ow mismos im i am imilition mom h 1
IIIlow Irsonstown r Iwo mossio
1\im
AMIN II 111101110111111111111111111011 11 IMMIIIV NIS
fltqilliLIlit
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
ueclassified in Part - Sanitized Co .y A proved for Release
50-Yr 2013/09/09. CIA-RDP81-01043R002500180001-0
Appendix CC
FRE UENCY STABILITY MEASUREMENTS ON THE WESTERN ELECTRIC D-1 2 99
FREQUENCY STANDARD
Comparison of the Western Electric D 1 2 99 Freency Standard
ABliaLLFE
A breadboard frequency multiplier of the type shown in
Figure CC1 was constructed to supply harmonics of the D-152499
100 ko Frequency Standard. As shown in Figure CO2, the 5 me
harmonic of the Standard was coupled into a receiver tuned to the
5 me WWV standard frequency transmissions. The coupling of the
local harmonic was adjusted to supply a signal at the input to the
receiver at approximately the same level as the incoming WWV
signal, The Frequency Standard was detuned slightly from 100 ko,
and the resulting beat note between the 5 me harmonic of the
Standard and WWV was direct-coupled from the receiver second
detector to a Sanborn chart recorder. In this manner, a recording
of the beat signal versus time was obtained, a sample of which
is shown in Figure CC30
Short strips of the beat signal recording, each between one
and two minutes in length, were taken at 15-minute intervals during
portions of a two-day period. The average beat frequency during
each recorded period was determined from the chart strips and
plotted as a parts per million frequency difference between the
5 me harmonic of the Standard and WWV, (Figure CC5) As indicated
in Figure CCS, the maximum frequency difference between the 5 me
harmonic of the Standard and WWV did not exceed 0.272 parts per
million. The minimum frequency difference was not less than
Page CC1
Declassified in Part - Sanitized Co
50-Yr 2013/09/0g ? r _R noo4 _
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
0.260 parts per million. This indicates an overall frequency
variation between the Frequency Standard and WWV of 0.0l2 parts
per million, and is based on measurements taken during portions of
a two-day period.
The actual beat frequency between WWV and the Standard 5 mc
harmonic was approximately 1-1/3 cps, which is a frequency offset
of 0.266 parts per million from WWV. It was found by experiment
that a beat signal of between one and two cps between the two
mc signals gave the best results for measuring purposes. A
beat signal less than one cps was sometimes difficult to observe
when examining the recorder charts because of the normal fading
of the incoming WWV signal. Also, a great enough frequency offset
had to be used to insure that the Standard harmonic never went
through zero beat with respect to the WWV signal during the
measuring periods since an ambiguous condition would have existed
if this had occurred. On the other hand, it was desirable to set
the Frequency Standard as close as practicable to its normal
operating frequency.
During the signal-tests, the W. E. Frequency Standard at the
Kahuku end of the circuit was adjusted by listening to the beat
note between the 5 mc harmonic of the Standard and WWVH with a
radio receiver. During periods when the incoming WW-VH signal was
stable, it was possible to detect beat notes with periods of seven
seconds or less. During most of the signal-test program, the 5 mc
harmonic of the Kahuku Frequency Standard was held to within
1/3 cps of WWVH.. This is within 0.067 parts per million of the
standard frequency. The WWVH signal as received at Kahuku was
Page CC2
Declassified in Part - Sanitized Copy Approved for Release 50-Yr 2013/09/09 ^ - --
Declassified in Part - Sanitized Co .y Approved for Release
50-Yr 2013/09/09: CIA-RDp81-01043R002500180001-0
most stable during midday or midnight and, whenever possible,
frequency measurements were made at this time. In the region
of sunrise and sunset hours, it was almost impossible to obtain
accurate frequency measurements because of severe disturbance in
the received WWVH signal.
Fre uenc Variation Measurements Between Two Western Electric
D-1S2Ls.99 FrequencyStandards
A second frequency multiplier of the type shown in Figure CC1
was constructed and the circuit described in Figure 006 was used
to obtain chart recordings of the frequency variations between the
two standards. The 100 me harmonic from each standard was fed
into a Hallicrafter 3-27 radio receiver. The output frequency of
one standard was kept at 100 kc, but the other standard was
detuned slightly so that the beat signal between the two 100 me
harmonies produced an audio frequency in the vicinity of 40 cps.
(It had been determined by experiment that the measuring arrange-
ment of Figure C06 operated best when the beat signal between the
100 me harmonics of the standards was not less than 30 cps nor
greater than 50 cps.
The beat signal present at the audio output of the receiver
was fed through a low-pass filter to reduce the effects of
.unwanted audio voltages, and then into an RCA Type 306A Audio
Frequency Meter. The meter was of the type that produced a de
output voltage proportional to frequency. The output of the
audio frequency meter was direct-coupled through a low-pass
filter to a Sanborn chart recorden, Since the Sanborn recorder
is capable of responding to frequencies greater than 40 cps,
Page CC3
Declassified in Part - Sanitized Co
y Approved for Release js_)-Yr 2013/09/0g ? riA_01-100.,
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
the filter network was required to reduce the beat signal ripple
components. A high impedance battery bias supply provided a
bucking voltage to facilitate centering the trace on the chart
recorder.
Figure 004 is a copy of a portion of the recorder chart
strip and the trace that was obtained. The calibration was
accomplished by feeding audio signals of known frequency into
the frequency meter. The chart revealed the presence of a
periodic frequency variation between the standards, the variation
having a period of about 3-1/3 minutes and a peak-to-peak swing
of 1.2 cps which is a frequency variation oi 0.012 parts per
million. During a seven-hour period, the maximum variation in
frequency between the two standards was less than 0.02 parts
per million, It should be stated that during this test each
standard was operated from its awn regulated de power supply;
both standards were in the same room and were operated from the
same ae power source,
No attempt was made to determine the cause of the periodic
frequency variation between the two standards, Each Western
Electric D-152499 Frequency Standard was subsequently compared
against a third standard of a different type, and the same
periodic frequency variation was present in both chart recordings.
This appears to indicate that the periodic variation was present
in both Western Electric standards. It has been suggested that
the periodic frequency variation between the two standards might
be a result of "hunting" in operation of the crystal oven
heater circuit.
Page CC4
- -
Declassified in Part- Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIATRDP81-01043R002500180001-0
CA.
F r e
UtVIC' MLitt lier
-J
61i5/56 H
CCA
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part- Sanitized Copy Approved for Release @50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Bloc* D Circuif Use,d to Compot-e
\NE Ft-e9tAeylct Stov\Acork v1V WW V 0 VI
Five hilejck cj
1411.????? moYame so.
RCA AR-88
RectitiveY ivoel
WWV 0\1 5/c
Otitrui from receivo-
secoml defecio
SqvAoryt dio
Rcc'e ssehlblti
Mold 152
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Sqmple RCcbr
QV\ 11) c
(I
71
Fi9,CC3
cAort sirs;rd o Ben-I Bc4.??vc.ev s me, ww v
1-vver.wylcti stowlotiS Mc, Harwovtic,
&last II)
PM
60 10
2o An
I s SO A
1111/\ .11\
III/ 87
11011 )111
I f? ;
30
I I
?
Vik) Ph. Ijir
\ 1441f?
motics
ovte secolicl
...111??????
01111111101411+1101ItIt 0 1 1 .1 I1-1- ? ? -1-4-1-1---1 ? 0--4 .1. *--W-1 1 1 IA -14. 1 1 A -1
_
-- 87 begt cjcleS 642 sec:1)1401S
11???????????????
Fi,CC4
Sqmple Re corck? Clolort Shitrc BeQt h-eeit4encl
betwtevl 100 Mc, Hatmorqc,s Di Two W.E, 0-152 49'3
Ft.e(iti enc S+cmcilarcis
6 hAtimate
?????
Ttio VIA (4 iI E
w oc) P? - iiII 01A. ,
$
?????46....06
winti(S
41.1.? - ? ? ? ..
..0.1.14.101.4.10111.........??????????? .10.11.4.0.11.114.?????????????????.....1
-UcS 4 t, ?') (
el 1. er(t(e
(eocti
r e C
N
CC3 cOld CC4.
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09 : CIA-RDP81-01043R002500180001-0_
0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
?
ay
-
1
II
Wk
?
1601
-Fr,
7
6
- c).
Raab
11111111111111111111111111L
1111.111
wt
QX
Iwe
IL
at iivAL
t?
43.1
?
:5
1 ?
?
?
Ii
,A A
" 111MILVINIELMILIMINEEM
1111111111111111111111
1111111111111111111
"
MEW .111111M111 IMMO=
1111111111111111
111 1111111=1111111111111111
1111111=1111111111
MEM re WM 11 111111
e
Oft
c3
c a
00
-Fe
IIIIIIII
1111h0111111111111111111EMIEL911
11111111111111111M111117ASAVINE MI ENE
VAILIMEMPAlla EMI II
gm MEM_ _ 111111111111
? Ram am
irn
Emu Ea immummumm.
oa:00 09:0. 1 2:00 13:00 !4:00 15:0o 16.13o
A tA 9L1 St 11) 195 5 liEsTstetlet ,Stct act ra Time
irvi 4 VA plc
0 7: 00
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0
Block Drawic c Circuit Used Mei) su-rt
e UeVICIt cl i a ? i? In S
Fej-weevt Two Wes-rt..-c
15 2 4 '3
1--trencti St a yt. GI -
Electric,
Declassified in Part - Sanitized Copy Approved for Release
50-Yr 2013/09/09: CIA-RDP81-01043R002500180001-0