|The UNIVAC 1219 (418-II)|
I was relatively late getting into computers. After High school, In the 1970s, I was in the Navy as a Fire Control Technician on a Guided Missile Destroyer (DDG-2). I worked on the Tartar Missile System. That system was run by Two Univac 1219 Computers. (Military Mk 152). Those computers served as my introduction to the joys of computing.
The Univac 1219 was a 'real' computer. Designed in the early 1960s', it was a militarized version of the Univac 418-II. It was big, and the front panel had lots of lights, and not a few switches. The 1219 was an 18-bit machine and could do double precision math at 36 bits. The instruction cycle was measured in micro seconds (memory cycle time was 2 Microseconds). It had a real-time clock which could be turned off and the machine could be stepped through a phase at a time, there was also the ability to do an instruction at a time. It had a whopping 128k words (18 bit) of magnetic core memory. The input/output section had eight channels, and the I/O was interrupt driven. Quite a piece for it's day. Power was 115volts, 400hz, at 2000 watts.
We had a separate I/O console which contained a high speed paper tape punch, a paper tape reader, three data Cassette drives, and a teleprinter/keyboard on top which was used as the (only) operator interface.
Image Courtesy vintagecomputer.net
The ship had two tracking radars (AN/SPG-51). Each radar system had a dedicated 1219 Computer. All of the signals to and from the radars, the missile launcher, the CIC, and everything else went through a companion unit called the Signal Data Converter. This box, as big as the computer, took all the various types of analog signals, and converted them to digital for the 1219 to process. And, it did the same thing in reverse to the computer outputs. There were no "operating systems". The software was completely custom written for the system. We loaded the "op-program" via the Data Cassette drives, although we had backup Mylar punched tapes (Very long tapes).
On the front of the computer were eight 'program control' switches. Four stop switches and four skip switches. Each had a dedicated instruction that checked the switch position and either halted the program, or skipped an instruction when the switch was up. These were VERY handy for debugging, and changing modes of the program. Additional front panel switches were for things such as Start/Stop, instruction stepping modes, etc. During normal operation, the computers ran completely autonomously, running the equipment, and driving dedicated displays in the various other pieces of equipment - Radar consoles, weapon control consoles, etc.
As an aside. The way the ship was laid out, the missile system was spread among many different compartments. The computers were in 'missile plot' a compartment way below decks, literally between the shaft alleys. The radars were in the aft superstructure. You can see their dishes in the picture below. Additionally there was equipment in the Combat Information Center behind the bridge, and, of course the missile launcher was at the aft end. Many miles of cable connected all this together!
The computer was entirely wirewrapped, and incorporated backplanes that literally thousands of small circuit cards plugged into (One is show in the photo at top). Each of these cards performed a single logic function, register, gate, flip-flop, (much like a 74xx TTL IC). These used transistors exclusively, no ICs here. The cards had about 15-20 pins on them for the backplane, and about the same amount of discrete components, all potted in plastic shellac. I do remember going through an overhaul period, and the shipyard removed all the cards from the computers - to test them. We then received the 'good' ones back in boxes, and spent no little amount of time figuring out where they all went, and putting them all back. Realize, these computers had performed flawlessly (mostly) for four years before this. After putting them all back, many weeks were spent fixing the computers......Hmm. I used to have a box full of 'rejects', but I can't find it anymore. Drat.
The lights on the front were, interestingly, Neon. Very cool little indicator lights (From diallite I believe). They had a 100v or so DC supply to them for the bulb, and were operated from the logic levels (3v?). Some of these indicators were push button, and that was how you entered data manually into the registers. Just push the button, and it lit up in a neat Neon Glow. Why don't computers have lights and buttons on them anymore? Seriously, the lights were a very significant troubleshooting tool. And, after a while, as the program was running, it was easy to tell what it was doing from the light patterns, and often a malfunction (which could be anywhere in the Missile system) would show up, obviously, as the wrong pattern. (What do these flashing lights do? We have no idea sir..... Well, just keep them flashing - Airplane II).
I mentioned earlier that this computer ran off of 400 cycle AC. That no doubt made the power supply smaller, but, since the ship used 60hz power, we had Large Motor/Generator sets in an adjacent compartment to make the power for the computers from the ships' 60hz power. If I recall, we had three M/G Sets. One for each system, and one spare. These were much bigger than a 60hz power supply would have been, and required maintenance, and their own control boxes. Ah Well, the mind of the Military. As an aside, the radars used rather large M/G sets also. Mainly to power one small hydraulic pump and some synchros (remember those?) in the directors, which are what physically moved the Dishes (Fodder for another page someday?). As I remember it, everything else in the Radar systems ran on 60hz.
Although it was not standard issue, we had snagged a copy of an assembler program called RCM -Raytheon Conversational Monitor (Raytheon was the prime contractor for the missile system). During normal (Peacetime) operations only one computer was on line so we could do 'maintenance' on the other. I spent many a spare hour underway programming this machine in Octal Assembler. Most of my programs were of the 'do nothing useful' variety. I wrote a few games, but my most popular program was a countdown calendar, with all kinds of bells & whistles. Anyone who has been in the service will understand a countdown calendar (Days till you get out). RCM did not have a way to print out the program listing other than one line at a time. A common sight was a scotch tape dispenser carefully propped on the return key, and the printer spewing out one line of program after another. I must say I went through a lot of Taxpayer dollars for paper! Sorry. But, I really enjoyed programming, and I seemed to have a natural knack for logical thinking.
This is a picture of the USS Charles F. Adams, DDG-2. This is the ship that I described above. You can see the missile launcher aft behind the Gun Mount. The missile Fire control radars can be seen directly behind the after Stack. The 1219 was somewhere below the waterline...
Here is an (amateur) video I made about the Adams... Charles F. Adams Memories
After leaving the Navy, I started working for the Wire Service...