Memories of the 4101

FPQ-6 Radar

Back to Station Equipment

Contributed by Bob Hocking

Back to RCA Computer

The original RCA 4101 computer was a 30 bit machine with 2 parity bits. It had 4K of core memory, later increased to 8K, possibly prior to installation at CRO. The original clock rate was 1 MHz later increased to 2 MHz. The design was very elegant; all logic boards were identical consisting of eight 'nor' gates, as far as I remember.

Special boards were used for the memory interface and for analog-digital and digital-analog conversion. The control panel was vertically mounted and carried rows of gas discharge tubes in diagonal bunches of three displaying the data in all registers in easily read octal format. The computer could be stopped and stepped by 'Slot' or 'Cycle' buttons making fault finding straightforward. Data was entered from a row of octal thumb wheel switches or by paper tape, (see Flexowriter below). The software made provision for levelling, orthogonality and droop of the antenna pedestal and for data correction to improve angular data accuracy.

Bob loading the Program Recorder: Photo – Chris Goodlace
Bob loading the Program Recorder:
Photo – Chris Goodlace

Data was outputted to the Data Handling System to console displays; and to GSFC via a TTY machine (at one line per six seconds) and a High Speed Modem. Two 200 Bits Per Inch (BPI) data recorders were provided, one for the real time program and one for recording tracking data for shipment to Goddard Space Flight Centre (GSFC).

The principal input/output (I/O) device was a Fridan Flexowriter; an electric typewriter with an eight-hole paper tape punch and reader. It used its own 8-bit code. It was an early commercial word processor. A standard letter could be typed, leaving a question mark after for example "Dear". A question mark was eight holes; this stopped the reader, the operator then typed in a person's name. Another key was pressed and the printer continued to type the rest of the letter. The various letters were stored as paper tapes.

Pointing data for the radar was supplied by GSFC via a receive-only Teletype machine. The data was in the form of NORAD Elements, a set of parameters which the 4101 program converted into Range, Azimuth and Elevation to point the radar at the satellite. Once the radar had acquired the target, the 4101 generated its own data in the form of an Inter Range Vector (IRV) thus enabling the radar to follow the target in the event of loss of signal, for example, due to varying space craft attitude.

In 1970 the data handling system was improved to allow more interrupts and a new version of the computer was installed - the 4101C at a total cost of US$ 258,000.

NORAD Elements could now be read into the computer directly from the teletype machine.

The 4101C was a bit of a dog. It was of a completely different hardware design. Very few of the modules were alike. As I remember, there were about fifty different ones.

The new 16K core memory became faulty, possibly due to being uncased to make cooling easier, so it acted as an air filter and collected dust. Following unsuccessful attempts by the NASA programmer to reallocate the program to the good bits, the memory unit was hand carried to Hong Kong for repair with its own seat on the aircraft.

On the return journey, the heavy power supply, checked in as baggage, became lost; I think it turned up the next day. The computer worked fine for a time except that it occasionally stopped when the computer rack was opened. A notice was placed on the rack requesting that it not be opened unnecessarily. However, senior visitors still found it necessary to "Ooh and ah" at the maze of white wiring and eventually the fault became permanent. After several days the problem was found and fixed by Dave Gardiner. One shielded cable in the middle of several dozen in a tightly laced loom was too short and stressed at a point in a connector where it eventually fractured. It was re-terminated with great difficulty; from then on the computer worked well.

The original CRO 4101 was packed up and shipped to Tananarive (Madagascar) and installed in their "Capri" radar (an FPS-16 with solid state and I.C. electronics). During the installation some logistics support was provided from the CRO Q-6 basement "come in handy” cupboard. This was an informal arrangement between Central Stores, a certain Q-6 Senior Tech and the RCA engineer at TAN. A congratulatory TWX was received in due course.

Back to RCA Computer

Personal tools