PROPOSAL TO DESIGN AND DEVELOP D-5 INFRARED SYSTEM PROPOSAL NO. C126-CP65 PART 1
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
05752557
Release Decision:
RIFPUB
Original Classification:
U
Document Page Count:
52
Document Creation Date:
March 9, 2023
Document Release Date:
February 10, 2021
Sequence Number:
Case Number:
F-2011-01575
Publication Date:
July 29, 1965
File:
Attachment | Size |
---|---|
PROPOSAL TO DESIGN AND DE[15858481].pdf | 4.05 MB |
Body:
Approved for Release: 2020/12/28 C05752557
TEXAS INSTRUMENTS INCORPORATED
Apparatus Division
P. 0. Box 6015
Dallas, Texas 75222
PROPOSALMM/MINMOMMIO
SD-5 INFRARED SYSTEM
Proposal No: C126-CP65
Part 1
This document contains se pages of CONFIDENTIAL information
Copy No.
� 13
copies
WARNING: This document contains information affecting the national
defense of the United States within the meaning of Espionage Laws,
Title 18, U. S. C., Sections 793 and 794. The transmission or the
revelation of its contents in any manner to an unauthorized person is
prohibited by law.
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL
TEXAS INSTRUMENTS INCORPORATED
Apparatus Division
P.O. Box 6015
Dallas, Texas 75222
July 29, 1965
PROPOSAL TO DESIGN AND DEVELOP
D-5 INFRARED SYSTEM
ProAgal .0 1 ab- CP65
Part .1 -
exa.s Iiistrurnents hain t e%pasttWo years; 'Substa'htially advanced all
-
bases of infrared sensor system; tedlindid-' The AN/AAS 18 -rip ra
,- , . , .13 g m
advanced state-of-the-art V/H CaPability:td..226_radianS-iperthecOna:in addition
to other advanCes.. The serie of systems developed a truly "hands off"
electronic system, 0.5 milliradian angular resolution (with good NET) and
developed a temperature control system to combat a hostile environment. The
AN/AAS-10 infrared system proved the techniques necessary for multichannel
recording. Only through a combination of these advanced techniques is a D-5
Infrared Reconnaissance System feasible.
The D-5 Infrared Reconnaissance System reflects design constraints
which have been determined through approximately 75 man-years of infrared
system design accumulated at Texas Instruments in the past two years. Typical
of the constraints are the maximum scanner speed of 6000 rpm, video band-
width of 650 kc, and a scan angle of 140�. Having applied a great deal of
experience to the preliminary design of this system, we are confident of our
ability to deliver the proposed system - maximum V/H of 1. Z radians per
second, angular resolution of 1.0 milliradian, and NET of 0.5�C - within six
months after receipt of order.
It is important to realize the versatility of the D-5 IRRS, which is
modular in concept. With minor changes, the same D-5 system could have
maximum V/H of 0.33 radian per second, 0.5 milliradian angular resolution
and NET of approximately 1.0�C. The basic system will be supplied in one or
the other configuration and the change-over kit will be auxiliary equipment.
. V/H max .
NET
D-5 "A"
1.2 sec-1
1.0 mr
O. 5�C
D-5 "B"
� 0.33 sec-1
0.5 mr
1.0�C
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. Cl 26-CP6 5
With the exception of Table 2-1, this proposal specifically describes the
D-5 "A" configuration since it represents the more difficult design situation.
The numbers will change for the "B" configuration but the techniques apply
equally to either configuration discussed.
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
SECTION J.I
SYSTEM DESCRIPTION
A. "PERFORMANCE AND PHYSICAL CHARACTERISTICS
Determination of the maximum velocity-to-height (V/H) ratio for con-
tiguous mapping requires examination of the parametric relationships between
the scan mirror rotational speed (f lines per second), the number of detector
elements (n), and the angular resolution of each detector element (Se). The
equality expressing contiguity of scan lines is:
In addition, the relationship between, optical area (A), focal length (F
bandwidth (Af),, detectivity of the detector (D*), atmospheric and optical
transmissions (-7- , and T ),� and angular resolution (AO) Must be recognized in
order to provide the required noise equivalent temperature (NET). This re-
lationship is -expressed as --
NET = ki(df
Bandwidth is related to rpm
= 27r(rpm)/6 0 (AO .
The proposed compromise between the
metric equations and the system complexity, size, weight, and power considera-
tions is listed in Table 2-1. With these parametric definitions, the NET will be
approximately 0.50�C for the 1-milliradian mercury-doped germanium (Hg:Ge)
detector.
requirements defined by the para-
B. FUNCTIONAL DESCRIPTION
The proposed system is illustrated in Figure 2-1. Major subsystems de-
sign characteristics and functions are described in the following paragraphs.
1. Optical System
The transverse scanning motion of the all-reflective, modified
Cassegrainian optical system is provided by rotating a rectangular scan mirror
about an axis parallel to the aircraft heading. Each side of the four-sided scan
mirror is approximately 8 by 2.5 inches. The mirror will be driven by a motor
at 6000 rpm to generate 400 scans per second.
This rotating element scans a total field of view of 140 degrees,
centered about the aircraft ground track.
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
7
Approved for Release: 2020/12/28 CO5752557,
TA
TO4CNIP
4. SNOCKMOUNTS
.F1L,A
AKCUP
cn7
//- GIMBAL
C00120. �
LA
(scs s,.z)
26.00
REGB,V212 -2CCORDIIR
111.12C.TRONICS BOB '
Figure 2-1. Proposed D-5 Infrared System
(Sheet 101 3)
rIVIINI2M11\100
S9cID-9ZTD (DNI resodozd
Approved for Release: 2020/12/28 005752557
ROTATING MIRROR
Approved for Release: 2020/12/28 005752567'
FILM SUPPLY
71/
z
ACCESS COVER
SECTiOKI f3'13
RoTATED cv4
(sse S14 2)
7/./W7////7/./ 7./
Figure 2-1.�, Proposed .0-5'Infre.rect 4yetern.
:(Sheet Z of 3). ,
FILM TAKE UP
ACCESS COVER
FIOA DRIVE
'IVIINaaIIMOD
S9c10-9ZID '0N Fesoclo/d
Approved for Release: 2020/12/28 005752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. CI 26-CP65
SECT tow A-A
(seE
Figure 2-1. Proposed D-5 In.frared System
(Sheet 3 of 3)
CONFIDNTIAL-
Approved for Release: 2020/12/28 C05752557
4B
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C 126 -CP65
Angular resolution (A6)
Total scan angle
V/H range
Auxiliary data
Table 2-1. Performance and Physical Characteristics
of D-5 IR Reconnaissance System
Configuration "A"
1 milliradian
140 degrees (�70 degrees
from nadir)
0.02 to 1.2 seconds
.5�C (Hg:Ge detector
300 watts, 28 volts dc;
2200 volt-amperes,
115 volts, 3 phase,
400 cps
Approximately 150 pounds Same
Panoramic recording on a Same
5-inch photographic film
strip, 4-inch format,
250-foot supply
Configuration "B"
0. 5
140 degrees (t70)
- �
0.02 to 0.33 second'
1.0�C (Hg:C.te detector
arra
Time Data
Same
Flat relay mirrors reflect the radiation to the primary focusing
element, an ellipsoidal mirror. This mirror, with a secondary spherical
mirror, produces a 25-inch effective focal length optical system with an un-
obscured collecting aperture of approximately 140 square centimeters and
has an optical resolution capability of 0.3 milliradian. The proposed optical
system is illustrated in Figure 2-2.
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. Cl 26-CP65
Detectors and Cryogenic Refrigerator
spectral region of interest for the proposed system is 8 to 14
crims. Mercury-doped germanium (Ge:Hg) is the optimum operational
detector in this spectral region.
The Ge-:Hg detector is- a single-crystal, extrinsic, photoconductive
device. It is a bulk absorber of 2 to 14 micron energy. A metal plate with a
small aperture in it normally defines the sensitive area of the detector. Another
aperture is placed a short distance in front of the detector to shield the sensitive
area from radiation outside the useful field of view.
The sensitive area of the detector is then placed near the focal point
of the optics system such that it subtends an angle
AO -
where AX' is the length of one side of the square sensitive area and F is the
focal length of the scanning optical system.
The number of detectors to be used in this application was determined
by two factors. First, according to the equations, system rps is inversely,
proportional to the number of detectors, U. Therefore, the centrifugal,
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP6 5
acceleration of the rotating optics is proportional to n2. To minimize dynamic
misalignment and distortion of the optical components, a three-element array
is best for the D-5 system.
The detectors in the array can be exactly contiguous (or tangent)
when the array is canted perpendicular to the linear axis of the detector array.
By locating the detector heatsink between the rows of Ge:Hg detectors, good
isolation of optical and electrical crosstalk is maintained. Also, there is
additional room for surface contacts and lead wires. Detector arrays in this
configuration have been built and flight proven by Texas Instruments.
A North American Phillips clos
proven on the AN/AAS-18 and used on th
,
Electronics
le cooler similar to that
ystems is proposed.
The electronics will be identical to D-4 lectronids except
, ,
for those minor changes required- by the additional bandwidth requirements of
the D-5 system. ',A- simplified block diagram showing the signal flow is shown
in Figure 2-3. The function of the : preamplifier is to provide the initial gain
and good noise characteristiCS:'';,The'aUiomaiic,leveling circuit prOvicie-S:addi-
tional gain and also compensates for "background changes" in the incoming=
video. The video compensation circuit corrects for atmospheric and scanner.c-`
induced "humps" in the video. The automatic gain control circuit compensates
for infrared background changes and also for "apparent background changes"
- such as a change caused by temperature variations of the scanner or detector -
itself.- The video compression circuit is a nonlinear amplifier used to extend
the dynamic range of the infrared system. The glow modulator driver provides
the light output for exposing the film.
a. System Parameters
The bandwidth for a particular system depends on various
parameters, among them resolution, scan rate, and field of view. In general:
BW - 0
N
Where BW = Needed electronic bandwidth (cps)
N = Effective scan speed in revolutions per second
= Resolution in radians
This equation supposes that the electronics is the limiting link in setting reso-
lution. In general, this is not true because resolution depends on the optical
resolution, the detector field stop, the electronics bandwidth, and the recording
spot size. Basically, the effective resolution on the film is the mathematical
convolution of the impulse responses of all the parts of the electro-optical
system. Here the electronics has an impulse response associated with its
bandwidth.
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
7
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. Cl 26 -CP65 �
PREAMPLIFIER
AUTOMATIC
VIDEO
LEVELING
-ON
AUTOMATIC
VIDEO
COMPENSATION
AUTOMATIC
GAIN
CONTROL
V
GLOW
MODULATOR
DRIVER
TO MECHANICAL
-"RECORDING SYSTEM
Figure 2-3. Simplified Signal Flow, Block Diagram
�
THEVENIN NORTON
Figure 2-4. Simplified Equivalent Circuits for IR Detector
This says that the electronics bandwidth, no matter how great,
will always have an effect on system resolution. It is only a question of when
it becomes appreciable. A value slightly larger than that given in the equation
is found to be adequate.
b. Preamplifier
The purpose of the preamplifier is to obtain the best noise
characteristics from the infrared system. It does this in two ways. First, by
making maximum use of the detector parameters; second, by providing sufficient
gain to make shielding and noise characteristics of the following stages less
severe.
A simplified equivalent circuit for an infrared detector is
given in Figure 2-4. If we assume that the equivalent current or voltage
generator faithfully reproduces the information scanned (or that its frequency
response is ideal), we still note that the frequency response at the output is
limited by the R and C of the equivalent circuit. This problem can be overcome
in several ways. One is loading by a resistor RI. While the original 3-db
frequency was 1 , it is now 1/2 7rRIC, where R' = RR1/ R + RI.
irRC
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
.Another approach is to run:t e signal into a.high impedance amplifier and to
shape the response by "peaking"-later_ori.., While either approach is satisfactory
and has certain noise advantages.the'peaking approach is highly dependent on
detector parameters (C and R) and can result in 'a system with peculiar
frequency response characteritics.
All this has neglected the noise output of the detector itself.
So-long-as the equivalent-noise of,.the preamplifier is much below the detector
noise, the preamplifier does not appreciably detract from the system perfor-
mance. This was found to be the case with the RS-7 scanner, and the noise
figure of the preamplifier, as defined by
S/N in
f = 20 log10 S/N out
was found to be les than 1 db. Therefore, �the system was found to be nearly
"detector noise limited."
c. Automatic Leveling Circuit
In general, ground information of interest is the relative
radiance of objects, not their absolute radiance as in a normal photograph.
In ordinary photography this background change is taken care of by changing the
exposure. To accomplish exposure control in the infrared mapping system, the
system must be effectively "A-C coupled" from preamplifier to film.
However, another problem arises. The video information
appearing at the preamplifier output appears as shown in Figure 2-5. The
scanner looks at the ground only during the active portion of the scan. Part of
the other time it looks at the casting or at the detector itself. If this waveform
were fed directly into a light-modulating amplifier, then the average light out-
put would have to be constant, because it is an ac coupled video system. Any
CONFIDENTIAL._
Approved for Release: 2020/12/28 C05752557
9
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
GATED SERIES
TRANSISTOR
VIDEO
I N PUT
VIDEO
OUTPUT
Figure 2-6. Simplified Leveling Circuit
I. INPUT ; OUTPUT"
Figure 2-7. Leveling Circuit Waveforms
changes in the waveform, representing changes in casting temperature, ground
�teriaperatu- re, etc., would shift the exposure during the acttial video time. What
�.
is needed is eleveling circuit" which senses incoming radiation only during the
active video time and disregards all information outside the field of view. The
circuit shown in Figure 2-6 performs this function and was used during the test
of this video processing scheme. The series transistor is turned "on" only
during the active scan period and is off the remainder of the time. A simple
analysis shows that the output point is tied directly to the input waveform during
the active scan and left free to return to the voltage on C during the unwanted
portion of the scan. But RC forms an integrator and C is charged to the average
level of the video during the active portion. Therefore, the output waveform is
returned to its average level over the field view. The circuit performs much
like an automatic exposure control on a camera.
Similar in function to several circuits used with other infrared
systems, this circuit has the advantage of being much simpler. It does not need
adjustment, is much more stable, and can accept a much greater range of sig-
nals than other circuits. A typical output waveform is shown in Figure 2-7.
d. Automatic Video Compensation
"Humps" which occasionally appear in the video are believed
to be caused by the air through which the infrared radiation must pass. Energy
received from the edges of the scah has to pass through a longer distance of
cold air than energy arriving at the center of the scan. As expected, an
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
CORRECTION
WAVEFORM
GENERATOR
GAIN
CONTROL
VIDEO IN 0
SUMMING
CIRCUIT
SYNCHRONOUS
" HUMP"
DETECTOR
OVIDEO OUT
increase in altitude is found to aggravate the condition. Because of the limited
dyiiamic range of any recording, system, and'because-this"hiimp:does not repre-
sent true infrared information: but rather a-manufactured side-effect, a means -
s a sirriplifie-d block diagram Of tie scheme used
to construct th D-3, D-4 utoniatic.videooiripeniation'circuit. "A correction
waveform similar in s ape to,the,expected hump characteristics is generated.
A controlled amount of this correction waveform is Summed into the video
until-a "hump detector" is at zero. Thig-hu-iii-15 detector-is-a gated-samplirig
circuit which analyzes the average value of the video at the center of the scan
and compares it with the average value at the edges. The response of this
system is made as long as it is possible to make a system using small electro-
lytic capacitors and still maintain temperature stability. Its time constant is
on the order ,of about 4 seconds.
e. Automatic Gain Control
Just like an automatic leveling control is needed to take care
of slow background changes, an automatic gain control is needed to correct for
slow changes in the peak-to-peak amplitude of the signal. These changes can
be due to many factors, among them atmospheric attenuation, day-to-day changes,
changes in the detector caused by changes in the operating temperature of the
cryogenic cooler, or simply by the type of terrain being mapped. The automatic
gain control senses the signal component of the video information in a range
of frequencies and adjusts the gain correspondingly. It has the capability of
correcting for approximately a 30-db range of input signals. Its response is
fairly long, being approximately 4 seconds. A shorter response time than this
was found to have an adverse effect on small, sharp targets.
Automatic video compensation precedes the automatic gain
control. Failure to do this would result in gain controlling a possible hump,
which is not really video information but system generated-effect.
CONFIDENTIAL' 11
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557 n
CONFIDENTIAL. Proposal No. C126-CP65
f. Video Compression
This circuit is basically a nonlinear amplifier and limiter.
Shaping of its input-output characteristics is done with diodes breaking at pre-
set points.
Video compression has two main advantages. First, it pre-
vents the following amplifiers from saturating on large signals by providing a
form of gradual, controlled "saturation." Also, it keeps the film operating in
its proper dynamic range. Secondly, the gain can be increased considerably,
bringing out less pronounced targets without the usual adverse effects caused
by too much gain. It tends to make the video system less critical to gain or
signal content changes in the video.
e recording system consists. of a set,_ofmodulators recor-
ding light sources ,-d-riveii by the vide-C;Chain:;f:t1:1;e.'d-4eOlOr-a:?''1:a�)?. A set of ,
modulators is composed oftl_g_e� 'modulators,- each of which represents
one channel of video information. i:iIder, tile moment, operation of the
-
modulator set. The modulated light flux from the three sources passes through
_ ._
the exposure control filter and a: set-of boiripensatioii-slits on a rotating drum to
flood the entrance of a stationary fiber optic assembly.
This light is transmitted through the fiber optic to the exit
end which is mounted at the center of rotation of the compensation drum. The
fiber bundle is so constructed that the light emerging is confined in a 90-degree
cone. The exit end of the tapered fiber optic is imaged on the photographic film
by a microscope objective which is mounted on the rotating drum. There are
four of these objectives mounted at 90-degree intervals around the periphery of
the drum. These four microscope objectives correspond to the four sides of tie
scan motor shift so that the recording objectives rotate in precise unison with
the scan mirror.
The film is formed into a circular arc by a platten and, as the
drum rotates, the three individual images expose three parallel lines across the
full recording width of the film. The exit end of the fiber optics is so designed
that each of the respective images is contiguous with the adjacent image.
There are several pecularities of the multichannel recorders
that are not found in single-channel devices. One of these is transferring infor-
mation from three separate light sources through the processing optics and
arranging these sources on the film so that they appear contiguous. This is
accomplished with a three -channel, tapered, fiber-optic assembly that rearranges
the sources such that their images correspond to the detector array configuration.
CONFIDENTIAL,; 12
Approved for Release: 2020/12/28 C05752557
SCAN DRIVE
MOTOR
FILM DRIVE
MOTOR AND
GEARBOX
22778
Approved for Release: 2020/12/28 605752557�
COLLIMATING LENSE
BANDING.AND EXPOSURE CONTROL FILTER
COMPENSATING
MASK
FRONT DRUM PLATE
STATIONARY
SHAFT
:-
3 CHANNELTAPERED
FIBER OPTICS g
MICROSCOPE
OBJECTIVE
TO, FILM TAKE�UP SPOOL .
DRIVE ROLLER
TO FILM DRIVE SERVO AMPLIFIER
Figure 2-9. Recorder Opticw
Approved for Release: 2020/12/28 605752557
EXPOSURE CONTROL GEARBOX AND
DRIVE MOTOR
SYNCHRONIZING PULSE TIPS
(-De. TO
SYNC
CIRCUITRY
FILM
TrvIINaaUNOD
S9d0-9210 "oM yesodo�Id
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
FILM VELOCITY
Another problem inherent in three-channel recording systems
is that of banding. Banding is primarily an exposure control problem. Figure
2-10 illustrates the basis of the problem. Two successive scans by a three-
channel recorder that is ptilling film at a rate below maximum V/H are illustrated.
In this condition, the exposed emulsion from the first scan does not move suf-
ficiently far to be missed completely by the second scan. In this case, the last
channel of the first scan is over-lapped by the first channel of the second scan,
thereby creating a band that has twice the exposure of the surrounding area.
This overexposure will be repeated periodically. Essentially the problem is
one of channel-to-channel, rather than line-to-line, overscanning.
FAR effect is another serious multichannel problem in wide-
angle scanners. This problem concerns the ability of the recorder to reproduce
information at larger scan angles and will be discussed in detail shortly.
These are the problems associated with multichannel recording
systems that are different from those of single-channel systems. Unfortunately,.
all problems suffered by single-channel recorders are shared equally with their
multichannel counterparts. Accuracy in alignment, sufficient light, jitter,
maintainability, and reliability are a few of the common problems.
14
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
b. Recorder Component Operation
(1) Collimating Lens System
The purpose of the lens system is to provide collimated
light to the fiber-optics system. This is necessary because of the extremely
narrow acceptance angle of the tapered-fiber bundle. Also, the parallel light
will cast a sharper shadow through the compensation mask.
(2) Compensating Mask
The purpose of the mask is to allow selective illumination
of the entrance end of the fiber bundle. Starting from zero-scan angle, the mask
illuminates progressively larger areas of the fiber while the illuminated area of
the fiber translates across the fiber, end. This translation rate depends on the
, _
video-channel number and is symmetrical about the center channel.
. �
3) Tapered Fiber-Optics Bund
The fiber-Optics ,bnndle-proyi es a means of transmitting -
the video modulated light from several separated sources and combining it so
that it appears to come from sources 'adjacent to each other. This bundle is an
aligned tapered-fiber bundle. The alignment provides the means Of reimaging
the compensation mask in the object plane of the objectives, and the taper pro-
vides a wide angle of divergence at the exit end of the bundle so that the objec-
tive is constantly illuminated throughout the entire recording angle.
(4) Microscope Objectives
The microscope objectives focus the light energy from
the fiber bundle to the film plane. Since the fiber-bundle end is on the axis
of the rotating optics, the objective is in focus throughout the entire recording
scan.
' (5) Modulators
A cold-cathode glow discharge tube is proposed as the
recording light source. This current-controlled light modulation device ha
been proved in a series of infrared line scan systems (RS-7, RS-9 and th
series) and is currently used as the recording light source in the AN/AAS-10
(XE-1) multichannel system.
(6) Film Drive
The range of the V/H ratio requires a film drive system
with a 60:1 dynamic range. This capability can be provided by incorporating
a slightly modified. AN/AAS-18 film drive subsystem. This system is described
here.
The servosystem is essentially an amiplifier and motor
generator regulator loop that uses the input voltage from the V/H servosystem
relative to a reference voltage to achieve linear speed control. Figure 2-11 is
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
15
riviimaaiamoo
Approved for Release: 2020/12/28 005752557
Figure Z-11. Film,Drive Block Diagram
P/1
S9dD- 9Z ID '0N resodoad
Approved for Release: 2020/12/28 005752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126 -CP65
the block diagram of the system. An ac amplifier is connected to the control
phase of a two-phase servomotor. The motor shaft is connected to a generator
whose output is a 400-cps voltage with an amplitude that is proportional to
speed. The generator output (opposite in phase to the V/H input) is fed to the
amplifier where it is compared to the V/H input to produce an error signal.
This error signal is amplified to drive the servomotor. The V/H input and
generator feedback signal (opposite in phase) are fed to a preamplifier, and the
error is amplified to a level sufficient to drive a phase-sensitive demodulator.
The demodulator output contains a dc component which represents the in-phase
error component and an ac component which represents the quadrature com-
ponent. The ac component is eliminated by a resistor-capacitor filter so that
it will not saturate the amplifier:-. The filtered dc output is reconverted to 400
cps by a chopper, and this signal is amplified in the drivef,circuits._;;I.ThiS out-
put is fed to another demodulator. which converts it to adc signal to control the
magnetic amplifier .' The filtered dc Output is:reCOOerted;i6:.400.cpS by a chopper
and this signal is amplified in; the driver circuits.This�outPiit-is fed to another
demodulator which. Converts it to a dc signal to contiol,the'thagi,ietiC aMplifier.. -
The magnetic amplifier is connected directly. to the-Conti61:phase:Of the film'
drive motor. The voltage gain from the preamplifier,tO'.the.-mOdulatar_ output
is 20; The combined driver7magnetic,amplifier.vOl4gegain-is approximately
500k, giving an overall voltage. gain of aPPrOximatelY.10,,000., The filth drive
motor turns a roller which causes the filth to move through the magaZine. So
that film tension may be maintained in the magazine, the film supply reel is
restrained by a brake and the-take-up reel is driven by the take-up motor. - The
film drive motor is coupled to the take-up reel shaft by a magnetic particle
clutch. The film supply reel and brake are restrained by a magnetic particle
clutch (coupled to the brake). The torques on the two reels are varied with the
amount of film on each reel by mechanical arms that sense the diameter of the
spool and film and control the amplifier input to the magnetic clutch and brake.
(7) Auxiliary Data
In a manner similar to that used on th D-3 and D-4,
time data will be printed on one margin of the film.
c. FAR Effect Cothpensation
The fixed-axis resolution (FAR) effect is a problem associated
with any line scan system which scans a flat object plane at a constant angular
rate. This effect arises from the fact that, as the resolution cone scans farther
from the line of flight, the ground area viewed increases. In single-line
scanners, this effect manifests itself as the familiar "bow-tie" degradation.
In the conventional system, this effect can be tolerated be-
cause the limited degradation it produces does not compromise the value of the
imagery. This is not true when a multichannel capability is required, in which
case the FAR effect produces ambiguous and distorted imagery. A solution to
the FAR effect must be provided in the multichannel system if it is to produce
useful imagery.
CONFIDENTIAL 17
Approved for Release: 2020/12/28 C05752557
4,3 .
Approved for Release: 2020/12/28 005752557.
SCAN 5 ENVELOPE
FLIGHT
DIRECTION
RESOLUTION .
ELEMENT
ON GROUND
SCAN 2 ENVELOPE
� �
DIR ECTI�YA OF FILM PULL
SCAN 4 ENVELOPE
TARGET
SCAN LIMIT
SCAN I ENVELOPE
TARGET
Figure 2-12. Printout Illustrating Bow-Tie Distortion
Approved for Release: 2020/12/28 00752557
0
s 9dp- 9z ID �oN resodoad
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
The recording system described here has been designed to
eliminate this effect. Flight tests have proved that it does eliminate it.
FAR effect is a problem associated with a multichannel, wide-
angle line recorder. Its solution is concerned with preserving the resolution
capability of the recording system at scan angles greater than 50 degrees off
the nadir.
Before discussing FAR effect, it may be illuminating to refer
to single-channel scanning-recording relationships to supply some background
to the problem. Figure 2-12 illustrates a single-channel line scanner mapping
a section of terrain where one target is on the flight line and another is off at.
he limit of ground scan., .To the left of this is a schematic representation of :a
recorder imaging the info' imatiori:on film. Note that,:,the',scan pattern�on.,the :
iOr-und is not Condition:::-"exist's--'because,'-.
as the resolution elernent'O'n'the'gi7oundrseans farther from the flight line.- its
Width increases, although thea�gular resolution of :the scanner remains
� nt.,.,The'width"of t e,resoliitiOni'elethent-as a-�funCtionfOf:-S-Can angle can
to be -
his :the-aircraft altitude
6 is the angular resolution of the scanner
6 is the scan angle.
Hence, for a 140-degree full field of view, the width of the ground resolution
element along the direction of flight at the scan limit is three times the width
at the flight line. The recorder does not exhibit bow-tie distortion because the
film is formed in an arc of a circle and the recording optics rotate inside this
arc. When the light source is imaged on the film, the spot image remains
'constant size through the entire scan angle, since the source remains constant
size and the optics remain a constant distance from the film. This results in
a straight, constant-width line across the film for each scan.
As the aircraft moves over the target complex, the scanner
"sees" the target on the flight line on the third illustrated scan. This infor-
mation is transmitted to the recorder, which images a dot on the film at the
proper place. On the same scan, the scanner also sees another target at the
scan limit since it also falls between the same envelope lines.
The recorder puts a dot on the edge of the film to represent
� the target. The aircraft and film both move forward one resolution element
before the next scan. The next scan does not include the target ofi the flight
line but does include the scan limit target. This target still sends a signal to
CONFIDENTIAL 19
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. Cl 26-CP6 5
the detector and the recorder, which puts a second dot adjacent to the first dot.
On the next scan, the target is still in the field of view of the detector and
another dot is imaged by the recorder. If another scan were made, it would be
found that the scan limit target is finally out of the field of view of the scanner.
Returning to the diagram, note that the target on the scan
limit was recorded three times and appears to be larger than the target directly
below the aircraft. This is not caused by some deficiency in the system, but
by the fact that the ground resolution at the scan limit is three times larger
�than at the nadir because the target is three times farther away. Also of
interest is that the target, although enlarged, is placed properly.
_ -
Multichannel.. systems are somewhat different. The following
aragraphs discuss the degign:SOlUtions'effected in AN/AAS-10.. Similar
rriethOd'i 4i;lied'tO'iLre.e.;Channef 5 -; Figure':
illustrates a:zi:ANTAAS:-;.1 0 five-:channel scanner'flYing,,a similar target
CirriPlek as,hefOreAlS o shown is the five channel recorder. 1,-- The scan
attern of the: multichannel- sl'rstein' is considerabrattfeiVilf,f'koiri- that Of-the
16.3-Cliannel-syStenThe:center Channel' exhibits" bOW:7tie:'diatortion as did the,
e. single-channel systern;.however, the other.:fohr, Channels: perform differently.
round resolution in the flight direction of.-each!channel at the scan limit still three times largerthan:at:,the' nadir, :but at large., scan angles the out-
si sirice- each detector
3. scanning contiguous to the detectors On, each side of itself, no detector
tiring any one_sein.looki-rat terrain covered by another' detector. Therefore,
as: the center channel grows, it pushes the inner edge of the next channel
outward. As that channel growS:in2size, it must push the outer edge of its
envelope out still farther. This results in the scan pattern illustrated. As
the first scan is Made, channel 1 sees the target somewhat in front of the
aircraft. The recorder receives the signal and puts a dot on the film edge for
channel 1. On the next scan, the aircraft and film have moved forward five
*resolution elements, and the aircraft is abreast of the target which is seen by
channel 3. The recorder dutifully records the target at the edge under channel
3. On the next scan, the target is seen by channel 5, which is looking somewhat
behind the aircraft. Again, the recorder performs its function perfectly,
leaving a dot at the edge under channel 5. Now, for one target, there are
three targets separated by two resolution elements each. Obviously, this is
not a true representation. When the aircraft is flying at a V/H such that it
moves forward only one resolution element per scan, the recorder shows
the target to be 15 times as large as the target actually is. When the aircraft
moves one resolution element, the film moves one resolution element. It takes
15 scans before the target is removed from the field of view of all detectors.
On 15 successive scans, the recorder puts a target on the film, seldom putting
It at the same place. One small burning outhouse at the scan limit would
almost give the appearance of a forest fire on the film.
20
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 005752557
rIVII,MaaLINOD
FLIGHT
1 DON
SCAN LIMIT
THIRD SCAN
iii iii
SCAN
.4f.""011b4 TARGET
NO11414 ,. P. Fo 'agaiMillWA
1 1 I liebillii 0 iiii. k4 Orl�000/1 11000j110.11_ !..a.. ii i 0 p p p,11-III I pla.
14111114:N1111111111P111.1pli P IIP.1 . i . I. I I Pli;
lall
Svi- OPPIIIIIIMIftC T SCAN
11A
- '414111411 .411*110111111.111.110 Oil
14399
TARGET --
DIRECTION OF FILM FULL:
Figure 2-13. Printout Without FAR Effect Correction
0
t/i
s9d0-9z1p �oN t-esodoid
Approved for Release: 2020/12/28 005752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C1 2 6-CP6 5
Correction of this problem is possible. Figure 2-14 illustrates
the method of solution. In this figure, the same scanning pattern is� employed,
but the recording pattern is drastically altered. Instead of recording with
straight, constant-width lines, each channel on the film duplicates the pattern
of that channel on the ground. As the resolution element on the ground grows,
the spot image of the channel on the film grows at the same rate. As the outer
channels are pushed out on the ground, the outer images on the film are pushed
out at the same rate. In this way, there is a one-to-one correspondence between
channel location on the film and channel location on the ground.
When the aircraft scans the first time, the target is seen by
channel 1 and recorded on film; However, the channel 1 image has been shoved
out considerably, and the dot is displaced considerably from where it was
placed in the five-channel recorder illustrated in Figure 2-13. On the second
scan, channel 3 sees the target, and it is recorded as shown. With this method
each recorded target is placed on top of the previously recorded target. The
target is recorded three times larger than it actually is but its location
accurate. This was the same cas-e with the single-channel recorder.
_
One undesirable feature of this method is the oversca.nning,
at large scan angles. As can be seen, the target has been scanned three times,
hence the exposure in that area is three times that experienced by the target
in the center of the map that has been scanned once. If no correction were
made, the result would be a map covered with shaded, overexposed areas.
�
The most straightforward correction is to keep FAR effect correction but
eliminate the overscanned areas.
Figure 2-15 illustrates the method of accomplishing this. The
straightforward method was used and the overscanning areas were lopped off.
Now, instead of each channel expanding over into a different scan area, they
are constrained to expand between two lines. These lines define the total line
width recorded each scan. At the V/H shown (contiguous), there is no over-
scanning between channels or scans. (The exposure problem at lower V/H
ranges will be dealt with later.) FAR effect correction is maintained since
only one target prints out. This indicates the recorder is not limiting the
resolution throughout the scan.
All that remains is to design a device that records in this
manner. The necessary components are the recorder drum, microscope
objectives, compensation mask, and fiber optic assembly.
The recorder drum is an aluminum cylinder with mounting
fixtures for the four microscope objectives and four compensation masks. The
microscope objectives mount to the exterior surface toward one end of the drum.
The compensation mask mounts on the interior surface at the other end. The
drum assembly, microscope objectives, and compensation masks are the only
recording components that rotate. All other components, particularly the
fiber optic assembly, are held rigidly and firmly. The microscope objectives
CONFIDENTIAL 2
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126 -CP6 5
Figure 2-14.
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
23
Approved for Release: 2020/12/28 005752557:,
FLIGHT
DIRECTION
CONSTANT TOTAL
RECORDING
LINE WIDTH
1,1374
TARGET
SCAN
LIMIT
THIRD SCAN
Atilis.SECOND SCAN
vnio 4�000"/IIII_I
\ TARGET
FIRST
SCAN
Figure 2-15. Target Printout With FAR Effect Correction and Constant Line Width Recording
59c13-9ZI3 "oN irsodoad
Approved for Release: 2020/12/28 005752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C 126 -CP6 5
Figure 2-16. Oblique View of Fiber Optic Assembly
image the exit end of the fiber optic on the film, the compensation mask covers
the entrance end of the fiber optic in a manner to yield the correct scan pattern.
These components will be taken up in detail later.
The fiber optic assembly is the heart of the recording system.
Figures 2-16, 2-17, and 2-18 are three views of the AN/AAS-1 0 fiber optic
assembly. Basically, it consists of five aligned fiber optic bundles of different
cross sections that are tapered 5:1 from entrance end (large end) to exit end.
The optic assembly mounts to a shaft inside the drum. The shaft exits the
drum through the back drum plate and is held rigidly by the recorder housing.
The optic is designed such that when mounted in the drum, the exit is on the
axis of rotation of the drum in the plane of rotation of the objectives, the en-
trance end is very close to the compensation mask, and the individual channel
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
25
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
22780
Figure 2-18. Exit End of Fiber Optic Assembly
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
EXIT END
26
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
bundles are coincident with the compen-
sation mask slits. With the exit end on
the axis of rotation, the microscope ob-
jectives are maintained at a constant
distance from the exit end which serves
as an object for the objective. This
allows the objectives to image the exit
and on the circular film plane and stay
in focus throughout the entire recording
angle. As mentioned before, the fiber
optic also 'allows several separat.e light
sources to bereorganized such that each
source appe ke, '66,ntiguOus.-��.-71"11,e
taper in the fibers accomplishes, two
objectives. The taper allows light
entering at a small cone angle to be
dispensed through a large cone angle,
thus allowing light to enter the micro-
2 size-,,required
s COI; 'c;bj Second, the small
of�eacn'channel-at the exit
ir-iii.i'squa:re) would be
difficult to work with at the entrance
end where other functions such as FAR
effect compensation and banding control are taking place. The image trans-
mitting ability of the fiber optic allows compensation for FAR effect and banding
to be accomplished. FAR effect is achieved by transmitting an image of the
compensation mask to the exit end. The same is true of the banding filter. The
technical details of the fiber optic are discussed later.
The AN/AAS-10 compensation mask is shown in Figure 2-19.
The white areas are slits and the rest is opaque. The mask is photo-etched
from 0.006-inch-thick steel and then mounted to a curved bracket which con-
forms to the inside diameter of the drum. Of interest is the slit pattern. This
is the same pattern as shown necessary in Figure 2-15 for the recording traces,
the only difference being that each channel pattern is now separated. This is
accomplished by imaging the separated patterns on the entrance ends of the
fiber, optics, transmitting this to the exit end where the patterns are again put
adjacent to each other, and imaging this on the film. �
Figure 2-20 is an illustration showing the sequence of events
through one scan that allows proper recorder channel traces for FAR effect
correction. The top figure shows the mask fiber optic orientation when the
angle of scan is zero. Note that channel 3 fiber optic element is three times as
large as either channel 1 or channel 5's element. Channel 2 and channel 4 are
twice as large as channel 1 and channel 5. However, at zero scan, the mask
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
27
Approved for Release: 2020/12/28 0057525571s,
MASK AT ENTRANCE END
SCAN
ANGLE Ar"-^"-'
o�
45�
\
55�
60�
650
22783
r.
L _J
A,A-ANA-t.
L_.
-
___1
FIBER OPT IC
�
r -
IMAGE OF MASK
TRANSFEREDTO EXIT', END
Figure 2-20. Compense..tionlMask-Operation
FILM PRINT-OUT
45�
550
60�
700
viiNa CILINOO
s 9cio- 9z To �oN Fesodozci
Approved for Release: 2020/12/28 005752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
has allowed channel 1 and channel 5 to be completely open. The outside ends
of channel 2 and channel 4 are masked off as is the outside end of channel 3.
At this position, each channel has the same amount of area capable of being
illuminated. The light impinging on the open areas is transmitted to the exit
end as shown to the right (Figure 2-20). The two far outside channels (channel
1 and channel 5) are on each side of the middle channel (channel 3). Channel 2
and channel 4 are below. The light areas of the array are the areas of each
channel that are not masked off. Hence, there are five small, square, equally
sized light areas whose edges are adjacent but offset. This is imaged on the
film by the microscope objective which leaves a spacing of squares on the film.
The lines on the schematic of the film represent the edges of the various light)
areas. The next sequence shows the mask as it has rotated past the fiber
optic to where it corresponds to a scan angle of 45 degrees. Here, channel 3
has opened up slightly on each side The inside .edges of channel 2 and channel
4 have been covered up by an amount equal to the amount of opening of channel 3.
However, channel 2 and channel 4 have opened up more area to the outside edges
of the fiber-optic, and they are of equal si ze to channel- 3. � Channel 1 and channel
5 have been closed considerably. This,im,age.ii then transmitted to'..theJekit
� end as shown. As before, the edges of each channel are contiguous, but their
relative position and size in the array have changed: This is imaged on the
film, and the recording pattern begins to emerge On the other three scan
angles, the same thing is happening. As the microscope objective rotates past
the film and the FAR effect compensation mask rotates past the film and the
FAR effect compensation mask rotates past the entrance end of the fiber optic
the proper rec-ording. pattern is- ipean on the film. This Method is completely
satisfactory in that only the microscope objective and compensation mask move
and then only in a smooth rotational motion. All other components are held
stationary. Once the compensation mask is adjusted to the proper location with
respect to the fiber optic, no further maintenance or adjustment is necessary
to keep the FAR effect compensation system operating. Figure 2-21 is a
section of a map made during an AN/AAS-10 flight test. The area shown in
the map is Carswell AFB in Fort Worth, Texas. The mapping aircraft altitude
is 10,000 feet above the ground. The section shown is between 60 and 70
degrees off the nadir. The distance to the parked aircraft on the map is
approximately 5-1/2 miles. At this distance, the ground resolution on the
direction of flight is 15 leet. Examination of tie map reveals the swept wings,
tail assemblies, and other details of the aircraft. This degree of resolution ,
and detail would be impossible at this large scan angle without FAR effect
correction. Examination of other maps made in the course of the AN/AAS-1 0
program reveals that FAR effect correction does nothing to deteriorate any
portion of the map but does indeed enhance the information at the edge.
CONFIDENTIAL 29
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
Figure 2-21. Infrared Map of Carswell Air Force Base
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
30
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. G126 -CP65
d. Recorder Resolution
In general, there are four major factors that determine the
resolution of a recorder of this type. These are:
Angle alignment
Track alignment
Focus
Jitter.
The first two are concerned with the location of each micro-
scope objective with all other objectives, The third is concerned with the
location of all objectives with respect to the film, and the last is concerned
with the short-term phase difference between the scan mirror and the recording
objectives.
Angle alignment places each microscope objective precisely
90 degrees from each adjacent objective ..This is required because the scan
mirror will be looking at the Same point on the ground every 90 degrees of its
rotation.: If the scanner detects two targets 1-; 0 milliridian-apart,,one objec-
tive will record these targets with 1.0 milliradian sepaiating them. When the
scanner rotates exactly 90 degrees, it can see these same ta:rgets again. If
the objective that records for that mirror face pair is out of alignment, 1. 0
milliradian, it will displace the location of the targets that amount. Hence, it
illcod between the two' targets-alreadY re-corded, and only one large
target can be discerned. This would ruin the recorder resolution in the
direction of scan.
The recorder can be aligned easily to 0.1 milliradian in angle
and, with extra patience, to 0.05 milliradian. This degree of alignment is
Important because all other factors such as film resolution, slight variation
In focus, etc., tend to decrease the ultimate resolution of any recorder. The
additional error of as much as 0.4 milliradian in angle alignment would pre-
clude the_resolution of 1. 0-milliradian targets.
Track alignment places each microscope objective such that,
if the film did not move, each objective would place its image on film directly
on top of the previously recorded image. Error in track deteriorates recorder
resolution along the direction of flight. Alignment of track can be made with at
least as much accuracy as the angle alignment and is considerably easier to
achieve.
Other than static track alignment (moving each objective into
alignment relative to the drum) precautions have been taken to ensure that the
drum does not move along its axis relative to the film. The drum support
bearings are preloaded, thereby removing their axial play.
Focus in the multichannel recorder is the most critical of all
adjustments. It affects resolution in both scan and flight directions. Because
CONFIDENTIAL 31
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C126-CP65
of the necessity of gathering as much light as possible, an objective with a high
numerical aperture was used. This type of lens has a very small depth of
focus (distance film can be away from lens image plane and still be in adequate
focus). Figure 2-22 schematically illustrates the problem. It can be seen
that
x = 2AS1(tan 0) + x'
where
x is the defocused spot size
x' is the desired image size
o is the half angle of the light out of the lens
ZS' is the distance the film plane is removed from
e image plane
For the objective system anticipate .88 and x is about 0.'0008 inch.
therefore, for x to be. twice,the.-Sizeof x, the film plane' peed only be displaced
0.0005 inch. This means, assuming the film is formed in ,a perfect circle,
that each objective must be adjusted radially on the drum so that all are the
proper distance from the film within �0. 0003 inch to ensure 1. 0-milliradian
resolution.
Figure 2-22. Focal Depth of
Microscope Objective
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
32
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C1 26-CP6 5
It is comparatively easy to perform this adjustment, but it is
difficult to maintain the film plane location, with this accuracy throughout
the recording angle. Any buckle or dimpling of the film would ruin resolution.
If the film plane were out of concentricity with axis of rotation of the micro-
scope, objectives of only one small area of the map will be in focus.
This problem was solved in the D-series equipments by
machining the inner and outer platen as a single unit. The final process includes
machining the film-forming surfaces of each platen, concentric to the mounting
base of the inner platen. This mounting base is then located in the drum bearing
base of the recorder. In this way, the film plane is referenced to the axis of
spin of the microscope objectives which ensures that the film surface is con-
centric with the axis of rotation. t�
Many systems of this type have been plagued with the problem
of jitter. This problem arises when the scan mirror, and recorder are allowed
o move relative to each other during a scan. It manifests itself as a variable
angle alignment error, and the, resolution along the direction of scan can be no
etter_than the maximum play between the two components. To be able to
resolve 1.0 milliradian, the jitter must be maintained less than �40 seconds
- of arc-. The system uses a fitted coupling as shown in Figure 2-23. This
coupling has a precision semicylinder base that fits both shafts very snugly
and allows no lateral play. The caps are drawn down on each shaft flat,
Figure 2-23. Motor-Recorder Coupling to Eliminate Jitter
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
33
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. Cl 26 -CP6 5
thereby aligning each flat and retaining the shafts. The recorder drum shaft
is coupled to one end of the motor shaft and the scan mirror to the other end
of the motor shaft. This, in effect, creates one solid shaft from scan mirror
to recorder drum and limits angular jitter to torsion in the shafts. High pre-
cision ball bearings are used in the shaft line to reduce radial jitter. All
bearing bores are line bored to reduce wear and to permit assembly.
e. Photographic Film
No less important than any other component in the system is
the photographic film used. The film used in the system depends on the light,
'source used - glow tube or gallium arsenide diodes'. For glow tube use, Tri-X
lin is desirable because of its high photographic 'speed and reasonable resolu-
p sensitivity closeIk matches the spectral output of the glow modulator
ri.=X is capable of resolving between50 and. 0 line pairs, per,rnillimeter.
ending On:proce Ss ing. This or the. system requirements
,
e,use of an infraredemitting diode as a light source ne,ces-
sit'a_tes-Kodakhigh speed film. This film
is quite 7en.sftive, to the
spectral 'OutpUt, of the dio4e,-�ren.ithe, 1:od,ei2gcool'e'd-to- liquid nitrogen tempera-
re..
has..'resolation.compa.rable to 'Tri-X and exhibits good
'stOi-'ige:characteristios.
At this time, Kodak only supplies the polyester base infrared
film in special orders of large quantities.
Automatic Exposure Control
The purpose of an exposure control is to ensure that the film
does not vary in darkness over the entire V/H range. This is necessary because
at low V/H ranges overscanning of the map occurs in the recorder. For instance,
'at a V/H of 0.1 radian per second, each area of the map is scanned by the
recorder spot 10 times. At V/H = 0.02 radian per second, the overscanning
factor is 50. Obviously, this level of exposure difference requires compensation.
As mentioned before, multichannel recorders have two ex-
posure control problems:
Banding - Overexposure on one area of the film but
not on others due to overscanning of one
channel by another
Gross overexposure - Overexposure on all areas
of the film due to many overscans.
Each of these problems has a V/H range where they are objectionable. Banding
Is a problem from V/H = 1 radian per second to V/H = 0.2 5 radian per second..
Gross overexposure is objectionable from V/H = 0. 2 5down to zero.
ONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
34
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No, C126-CP65
The automatic exposure control developed for the AN/AAS-10
system will be utilized in the D-5 recorder.
g. Banding Filter
With FAR effect correction incorporated in the system,
banding control is difficult. Figure 2-24 shows the FAR effect scan pattern
(5 channel AN/AAS-10) on the film at a V/H of 1 radian per second. Three
scans are illustrated. Note that no area is overscanned at this V/H, and
banding control is not necessary. If the V/H is decreased to 0.8 radian per
second, the film pull rate will be decreased by one channel width per scan.
Figure 2-25 illustrates the result of this reduction as found in the AN/AAS-10)
here is now a band across the full Width of the film. This
and is one channel width wide-. The pattern is not as simple as if only-
channel,,1,and ,channel, 5, were OVerS_canning,each�Other. There are four separate
reaS(�Of::,oversc'an where one channel' is bvSis6anning- some other 'channel.
r instance, only channel 1 d.ns;3fehannel- 5 WeieattenUated_,5,0�perceni,:i.this
would cause thel.,,eigar:7:shapect:arain the center, to:',.haVe.prsoPerexpOsure; ,bu
,e:-.,-:'ai,ed.:6:Whee:Ahalnnel,--2.;'�hann...el' 5, channel 4, and channel 1 are over -
464:nr4icv:4.1iiiciii-i;;:,i.:er'eXPbsea 15150 percent. 1,-AlSo:;.'the'erei'' Where channel
nd cli:d.nnet-,4�.6yei!se-a."nsWoUld be overexposed by 100 peiCent.t�::::,-A','sOlution to this -.-
wOUld,b0to.'atienuate channel 2 and channel 4 50 percent. This Would eliminate
oveieiipOg.tiie in the area over scanning, but in the areas where channel' 2 and
efidrinel-4;are'not,Oversc'a:rinini,' a'56=percent underexposed area would result.
This indicates that not only is it necessary to attenuate certain channels, but
also parts of certain channels, and this attenuation must be a function of scan
angle.
Figure 2-26 shows the result of decreasing the V/H to 0.6
radian per second. A more complicated overscan pattern results. Now, instead
of four different overscan areas, there are six.
Figure 2-27 is a flat view of the banding control filter. Each
vertical line accommodates one channel and is placed between the glow modu-
lator and the entrance end of the tapered fiber optic. This allows the light
coming through the banding filter to cast a shadow image of the banding filter
on the entrance end of the tapered fiber optic. The quantity of light reaching
the fiber optic is a function of the transmission through the banding filter.
The numbers on the right indicate that the area of the filter that will cast its
shadow image of the banding filter on the fiber optic and that particular V/H.
If, for instance, the V/H were 0. 8 radian per second, the area lying on the
0.8 line would be shadow imaged on the fiber optic.
Figure 2-28 illustrates the banding control action of the
banding filter, compensation mask, and fiber optic. The top set shows the
compensation mask position at a scan angle of zero and an over scan pattern at
the right corresponding to V/H = 1. As mentioned before, there is no over-
scanning at V/H = 1 radian per second; therefore, the edges of scan 1, 2, and 3
CONFIDENTIAL
35
Approved for Release: 2020/12/28 C05752557
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. C1 26 -CP6 5
Figure 2-24, Scan Pattern
CONFIDENTIAL
Approved for Release: 2020/12/28 C05752557
36
Approved for Release: 2020/12/28 C05752557
CONFIDENTIAL Proposal No. 0126-CP65
70
60
50
�
10ID
20
30
40
50
60
70
22793
V/H = 0.
SCAN 1
OVERSCAN AREA
CHANNEL I AND 4
OVERSCAN AR EA
CHANNEL 2 AND 5
SCAN 2
\
0
ZZ
11.