REPLY TO(Sanitized)COMMENTS

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B04747A000400030049-2
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 5
• Document Creation Date: December 28, 2016
• Document Release Date: February 16, 2001
• Sequence Number: 49
• Content Type: REQ

Approved For Relea&e 2001/03/06 : CIA-RDP789047*7A000400030049-2
STATINTL
Reply to Comments
1. 1 Installations considerations are vaas descrie in the er red 3 May 1966 r1A finalacopy
tion Form. A preliminary copy of this will follow shortly. e
i Since it is beyond the bounds of our r siy ht. a It will probablyebesinrthe best
the customer will prepare the installation g
interest of both the vendor and customer that installation be a joint effort wher-
ever there is a customer-vendor interface (i. e. power hook up, exhaust duct
hookup, etc.).
The unit will be customer property at the conclusion of the acceptance test
called out in 3. 4. 1 of the "Specification of a Chip Printer", 7 July 1965.
Since the acceptance test includes a 30 day operational period, the customer
need only accept consignee responsibility until the acceptance testing and 30
day period are over.
2.6.2 This paragraph indicates that for any one particular target, the unit
will produce not less than six identical prints in one minute with a design
objective of ten identical exposures of the same target in one minute. Neither
rate refers to a target to target exposure rate.
The rates called out in 2. 6. 2 are based upon study phase results which indi-
cated that this range of rates is the highest that can be achieved without using
unnecessarily sophisticated nonlinear indexing mechanism and larger motors
and without having to perform stress investigations on all but critical compon-
ents.
2. 7 rParagraphs 2.7.3 and 2.7.4 do describe the positioning accuracy
of the X (0. 3 mm) and Y (0. 1 mm) respectively.
It is our understanding that this accuracy is required of our unit so that it
will be compatible with the rest of the system.
As far as economy is concerned, unless the specified accuracies become
so tight that the servos are driven to difficulties with nonlinear instabilities,
or on the other hand, so loose as to be in the realm of the unreasonable
a change in accuracy will only be significant when the position of the last
-1-
Approved For Release 2001/03/06 : CIA-RDP78BO4747A000400030049-2
 .Approved For Relea%w2001/03/06 : CIA-RDP78BO474ZA000400030049-2
decimal digit is altered. For example, to increase a specification from 0. 3mm
to 0. 9mm would not generate a significant price reduction since the BCD digital
circuitsmust be capable of handling tenths of millimeters in either case. If how-
ever, there was a change of from 0. 9 (or 0. 3) to 1. 1 millimeters (or 1. 3 or for
that matter 9. 9 millimeters) there would be a reduction in digital circuit costs
since one less binary coded decimal word (tenths of millimeter) would need be
handled by the control and servo comparator circuits.
Hence except where a specification change is large enough to remove a signifi-
cant decimal digit, the assumption that cost is inversely proportional to toler-
ance is not valid for 2.7.3 and 2. 7. 4.
Paragraph 2. 7.9 indicates that the Chip Printer will be able to be used as a
mensuration device with an accuracy of 0. 5 mm.
2, 9 The 120%o deviation referred to in 2. 9 represents the target resolution.
change per sub group of a standard USAF 1951 Resolving Power Test Target.
The resolution change per sub group is a factor 2 . This number 672
is a typographical error.
Resolution was not stated as based on knowledge of a particular film because
it is anticipated that more than one variety of film will be used in the unit.
In addition, tying the specification to a particular film (or number of films)
would exclude from the specification any superior films that may have been
in the preproduction stages (and therefore had unpublished data) at the time
of the specification or contract.
2.16 Ninety percent duty cycle means 90% up time. The capability will
be demonstrated during customer acceptance by recording unit "up time" and
"down time" (i. e. , time to clear malfunction, trouble shoot and repair).
MTBF values only evaluate reliability from the standpoint of statistical failure
rate of known components under certain stress conditions. It will predict fail-
ure due to component failure in a large number of units over a long period of
time. However, it does not consider good (or bad) design, engineering or fab-
rication. Hence, it is possible to have a unit with a very high statisically de-
rived MTBF and with a very high malfunction rate (caused by poor design,
engineering, or fabrication). Particularly in prototype equipment, high
Approved For Release 2001/03/06 : CIA-RDP78BO4747A000400030049-2
 Approved For Releas&2001/03/06 : CIA-RDP78BO474 x000400030049-2
MTBF does not necessarily insure low down time.
2. 17 Assuming a forty hour work week, service life will be two and one
half years. :Performance will be considered degraded when the unit falls
below the performance requirements of the specification.
Normal mortality in 2. 17 refers to the anticipated expenditure or failure of
components and material that cannot reasonably be expected to survive through
the service life of the unit without going to huge internal storage areas, com-
ponent redundancy and ex.orbitaritly costly components (e. g. , gate liquid, cron-
apress, indicator lights, exposure lamps and relays that have high operation
rate per target).
Cost per unit of service time is a very good measure of dollar value for pro-
duction item purchases, for example desk typewriters or automobiles. How-
ever, development or prototype items might perhaps be better evaluated on
the basis of cost as a function of utility, cost as a function of urgency or cost
as a function of complexity. (The fallacy of using the cost per unit of service
life to judge prototype or small lot production can be illustrated by applying
it to something like a Saturn rocket, where the cost per unit of service life
is astronomical. )
3.4 The specification places no limitation on the operational hours during
the thirty day acceptance period. The choice of numbers of hours and schedule
is the customers. Some absolute minimum of hours will have to be run how-
ever for operation training at the delivery destination.
7.0 All of the points indicated will be covered by the Operational and
Maintenance Manual.
9. 17 The chip cassette holds 500 chips. It is the developing magazine
that holds 36. There is no reason to limit the number of print copies since
more than one developing magazine can be used.
Approved For'Release 2001/03/06 : CIA-RDP78BO4747A000400030049-2
 Approved For Reles a 2001/03/06 : CIA-RDP78BO474JA000400030049-2
Amendment I
3. The environmental conditions called out in this section are those
of a typical office.
4.3.3 "Manufacturers published resolution" does refer to the film manu-
facturer. The number 6 2 should be
4.3.4 Section 4 of the Amendment is only an outline or basic plan of the
areas to be covered in acceptance testing. It is by no means intended to
represent a detailed acceptance test specification. An Acceptance Test
Specification which will cover 1) a detailed procedure for each test 2) the
number of times it is to be performed,. 3) type of equipment necessary to
perform the test 4) equipment calibration data, (including mensuration tar -
gets) 5) and data sheets with indications of allowable tolerances for each
test will be submitted for customer approval on 1 November 1966.
To specify positioning accuracy as a percentage per unit travel is imprac-
tical unless a limit is placed on the smallest increment to be tested, or
unless it is accompanied by a specification for resolution. Any practical
system has a lower limit of resolution set by a potentiometer (even film
pots have finite resolution contrary to manufacturers claims) or backlash
or a stiction to servo gain ratio or a electromagnetic interference threshold
or, as in our case, a least significant digital bit. Therefore, no matter how
low a system resolution is, it. is always possible to pick a unit of travel so
small that the unit of travel times the specified percent error per unit of
travel yields an allowable error smaller than the system resolution and hence
an out of spec condition.
Functional Tests - The capability of the unit to be used as a mensuration
device (2. 7. 9) will be tested. The print rate (2. 6. 2) will be tested. These
tests will be described in detail in the Acceptance Test Specification.
Approved For Release 2001/03/06 : CIA-RDP78B04747A000400030049-2
 Approved For Release 2001/03/06: CIA-RDP78B047A000400030049-2
5. and Remark 2 - . The ability of theunit to perform hour li life of dthe ictated.
by the requirement of a 90% up time during the
Approved For Release 2001/03/06 : CIA-RDP78BO4747A000400030049-2