This mainframe computer, though considered enormously powerful in those days, and certainly hugely expensive, was laughably slow by the standards of today's personal computers, even smart phones. Memory was less than 10,000 36 bit words. Data and program storage was stored on large tape units. It was programmed in Fortran with Holerith cards that were prepared on noisy desktop machine that punched "chads" in the cards. A stack of cards, with special control cords front and back, formed your program that was submitted for a daily run in the morning. Collect the printout and cards in the afternoon, correct the errors and submit the pack of cards again next morning. It was slow going!
Here's the output: |
When planning for the flight of AO5, we realised that we would not necessarily know the time of the launch or indeed the time that AO5 was inserted into orbit. Therefore we could not predict when the satellite could be tracked or where to point antennas.
We therefore needed a means of providing amateur radio operators with antenna pointing information quickly after launch and so we devised the "Standard Orbits" method. First, we assumed that the orbit would be circular at about 800 km altitude and this greatly simplified the whole process. We recommended an antenna of 10 dB or more gain. Such antennas have a beam width of around 40° and so a pointing accuracy of 15° was more than adequate. The Standard Orbits method predicted the azimuth and elevation of AO5 in its assumed orbit. It was only necessary to provide tables of azimuth, elevation and time after a northbound equator crossing (called the Ascending Node) for a range of latitudes and longitudes of receiving stations. Given the allowable antenna pointing errors, this amounted to just a few pages of computer output. So, computer programs were written and keyed onto punched cards which were then delivered to the University of Melbourne's IBM 7044 computer to be run. After correcting errors, we finally got printouts that we thought were correct. Prior to launch we were given the nominal orbit and so we were able to run the program and distribute the Standard Orbits information throughout the world. As a further convenience, we planned to connect the IBM computer to an amateur radio transmitter so that the Standard Orbits data could be transmitted around the world soon after launch. The computer was programmed in assembler code to toggle the "Sense Lines" to form characters that were transmitted to a nearby teleprinter and thence to an operator's radio shack for broadcast. As it turned out, the launch was so accurate that the pre-launch data was good enough for three weeks of flight and it was barely necessary to update Standards Orbits for the remainder of the six week flight. |