DEAR(Sanitized)

• Document Type: CREST
• Collection: NGA Records (Formerly NIMA)
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78B04747A001800060018-8
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 5
• Document Creation Date: December 28, 2016
• Document Release Date: May 30, 2001
• Sequence Number: 18
• Publication Date: September 25, 1961
• Content Type: LETTER

Approved For Release 2001/08/13 : CIA-RDP78BO4747AO01800060018-8
September 25, 1961
STATINTL Reference:
On September 12, 1y6?, we received your verbal communication to stop further
work under the referenced contract. We would like to review briefly the
sequence of events leading up to that time and the work which has beer- accom-
plished to date.
In July of this year, system test of the LARR and PIGMI it was started. It was
immediately determined that a compatibility problem existed between the LARR
and P I Gil I and between the P IGM I and its readout equipment. Some of the readily
apparent problems were random changes in the counter position, unreliable oper)-
tion of the present function, line transients caused by other equipment Inter-
acting with the system and random sign reversals. Some of the less apparent
pro iens occurred in Incorrect buffer transfers from the counters on a "COPY"
command and random changes of buffer store before and during readout.
Most of these major problem areas were well defined within this first four
weeks of systems test. Work on correcting these errors was started in August.
dowever, due to masked or hidden problems which were not readily apparent dur-
ing our first evaluation, progress was slow although positive. Some of the
corrective actions taken were as follows:
1. Shielding of all cabling within the LAM. This consisted of shielding
of digitizer Lunp leads; separation of signal and power grounds and the
reshielding of signal lines in associated twisted pairs.
Redesign of counter circuitry to reduce sensitivity. This consisted
principally of the addition of plus bias voltage to logic: gates.
Redesign of PRESET circuitry in PIGMI. Redesign of logic circuitry to
isolate circuitry from counter and addition of bias to redesigned gatinq
system,
4. Redesign of sign circuitry to reduce sensitivity. This was accomplished
by changes in circuitry and also the addition of bias to gating circuitry.
Since progress had not been commensurate with the time spent, near the and of
August an engineering-management meeting was held to re-evaluate the p rob l em
and to determine the most direct approach to a positive solution. It was
DECLASS REVIEW by NIMA/DOD
STATINTL 18-8
 Approved For Release 2001/08/13 : CIA-RDP78B04747A001800060018-8
September 2a, 1961
Page Two
decided that two major steps needed to be taken before any additional work in
minor areas was to be accomplished. These consisted of the following:
1. Adding the required bias voltage to the counter flip-flops.
2. Removal of the pro-amplifiers from the PIGMI unit and placing them
directly with their associated digitizers within LARR.
The first corrective measure was required for proper operation of the flip--flops
and to prevent noise transients from appearing at the base of the transistors.
Once accomplished, the counters were stabilized.
The second action, that of rebuilding the pre-amplifiers directly adjacent to
the optisyn units eliminated the noise pickup due to the long high impedance
leads from the LARR to P IStI 1.
Another review was made on 11 September 1961 at the completion of this work.
Due to the major improvements made, I t was dec i ded to complete these modifi-
cations for one complete counter. This decision was made just prior to your
stop action.
Since that time (September 12, 1961), we have completed this-work on our own
initiative. The results of these tests have been satisfactory. The counter
at this time accumulates and buffer transfers accurately. The preset and sign
change have been corrected. The noise problem has been reduced to less than
five per cent of the original level. The one remaining problem area exists
between the counter and readout drawer, more specifically between the stepping
switches and counter which can now be tested statically. No changes or modi-
fications have been made to the readout drawer to arc suppress relay or step
switch coils, or to the LARK to arc suppress the switches of clutch coils.
Proper arc suppression of these components should eliminate the few random
counter changes now occurring. In operation with our standard product line,
the back Inductive voltage generated by these coils located in our standard
readout drawer is not sufficient to cause adverse effects or, other sections
of that equipment. This equipment generally can be described as operating
from low Impedance sources with large signals, whereas the transistorized
version of the PIGIl has high Impedance sources with small signals. There-
fore, evert small induced voltages from these coils appear as signal levels
(from 3 to TO volts) and are processed in the logic circuitry in the same
manner as signals from the digitizers. In the case of our vacuum tube ver-
sion of the PICt41., the signal levels are considerably larger than the in-
duced voltages, of the order of 50 volts, thus not presenting the present
compatibility problem.
Approved For Release 2001/08/13 : CIA-RDP78B04747A001800060018-8
 ILLEGIB
Approved For Release 2001/08/13 : CIA-RDP78BO4747AO01800060018-8
September 25, 1961
Page Three
USN, visited our facilities on 21 September 196-1 to
determine the status of the equipment. We reviewed with him the work which
has been done over the past three months and demonstrated the equipment. For
your convenience, we have attempted to summarize his findings:
1. The main problem has been one of a
lc pickup.
2. The remaining problem appears to be localized in the readout drawer an
more specifically, to the stepping switches within that drawer.
There exists at le:?:st a 90% probability of success if the following ac-
tions are taken:
:a. Arc suppress all relays and stepping switches within the readout
d rawe r.
Arc suppress, where possible, clutches, relays and switches within
the LARR unit.
Make necessary modification on the second counter as has been accom-
plished in. the first unit.
d. i er?oute and change as necessary caul ing within the LARK.
e. Rebuild the pre-ary l iffier on the second optisyn unit.
f. Provide additional shields as necessary on interconnecting cables.
Recommend the use of line filters or shielded enc t ease res as requ i red
for proper operation of the system at your facility.
Based on the above factors, as well as the discussions on Septeimber
STATINTL both you and we are recommending the following cou
action:
Alternatl
That we proceed as quickly as possible with the completion
of the P1G1M1 I t called for by the existing contract. As Indicated, the pro;
bliity of success for this course of action is at least 90%. As regards the
time schedule, the modifications could be completed in approximately two weeks
with final systems tests requiring an additional two to three weeks. Accord
assuming the present schedules and technical assumptions hold,, shipment
of the completed system could be made by or before the end of October. For
your reference, we are su narizing in the attached schedule the status of costs
at the several most recent check points as well as the current estimate of costs
required to complete.
Approved For Release 2001/08/13 : CIA-RDP78BO4747AO01800060018-8
 Approved For Release 2001/08/13 : CIA-RDP78BO4747AO01800060018-8
September 25, 19b
Page Four
Alternative No. 2. Stop all further work on the PiGHt 11 and proceed with
the necessary modifications to the LA RR so as to make I t compatible with
the PIGMI unit presently installed at your facility. These modifications
will require essentially mounting of two rotoverters in place of the present STATINTL
optisyn unit as well as providing the necessary interconnecting cables and
mechanical hardware required to attach the rotoverters. The costs required
to accomplish these modifications to the LARR are approximately In-
cluding installation of the equipment at your facility. It should be noted
that choice of Alternate No. 2 will degrade the currently designed operating
speed of the LARR by about 40% because of the fact that the vacuum tube ver-
sion of the PIGMI which you currently have has an inherent counting rate
limitation and, also, due to the fact that it requires a greater amount: of
time to transfer data to the buffer, thus precluding readout "on the fly".
As indicated earlier In this letter, we are at present continuing with work
on the PIGMI 11 on our own cognizance. In the event you wish to select
Alternative No. 1, the estimates indicated above assume no d scontinuiE:y
in the work and, hence, a prompt indication from you as to your wishing to
proceed. I shall be in Washington, G.C. during the latter part of thi week
and will plan to call you on this matter at that time.
Sincerely yours,
HJR;pg
Approved For Release 2001/08/13 : CIA-RDP78BO4747AO01800060018-8
 STATINTL Approved For Release 2001/08/13 : CIA-RDP78BO4747AO01800060018-8
Approved For Release 2001/08/13 : CIA-RDP78BO4747AO01800060018-8