Tales of collecting 4

My introduction to vacuum tubes dates back to 1934 when we got a new Philco table radio (one of the famous round-topped types). We were depression-poor and my mother was going slowly crazy with the frustration of living on $90 a month for a family of five. My father saved for a year and got the radio for her for Christmas. That radio lasted until I took it apart in 1949 to learn something about it. I didn’t learn much but it got me interested in electronics and here I am, still hooked to this day! In 1956 I got a job at Brookhaven National Laboratory, near our home on Long Island, as a "Technician BE Trainee". My first project was to build a device to measure the beam current in a Cyclotron. Beam current was on the range of a few nanoamps, and since this was before the day of semiconductors, etc., the key to the measurement was to use the current run through a resistor to develop a voltage and measure it. Simple Ohm’s Law, except it took a resistance measured in kilomegohms to get a voltage with the minute currents involved. With a resistor like that, just leaving a flinger print on it would change it’s resistance by 15 to 20 percent. It was a tough fight but I finally got it to work.

From there I went on to construct nuclear pulse height analyzers and similar devices. These were really digital computers with a hard-wired program. They typically used about 100 to 125 tubes and sometimes a digital display using a modified 5-inch oscilloscope display and various printers, such as an adding machine with solenoids over the manual keys, or a standard numeric key punch similarly modified.

After a couple of years of that I enlisted in the U.S. Air Force and my experience with "pulse and digital" circuits got me into a 40-week course in the theory and maintenance of a big computer used in the SAGE system. I passed the course with the highest grade yet achieved for that course so they made an instructor out of me. I went through their Technical Instructor Course and did it again.

The computer I taught had about 7,400 tubes and about 75,000 diodes in the entire system.

After teaching one class through the course I got out of the Air Force on a hardship discharge. I had one child and when it became apparent that another was on the way I couldn’t live on my meager pay so I got out.

My first commercial computer job was for Royal McBee, servicing LGP-30 and RPC-4000 computers. In my opinion, the LGP-30 was the first true minicomputer. It was delivered in 1956 and by 1960 the 200th was delivered to the X-ray crystallography department at Johns Hopkins University and became one of my charges.

The machine had 139 vacuum tubes, a rotating drum memory and its basic I/O device was a Frieden Flexowriter. One could also opt for the "High Speed Paper Tape Reader and Punch" which read paper tape at the tremendous rate of 60 characters per second and punched tape at 20cps (the Flexowriter did everything at 10cps). The computer cost about $39,000 and was packaged to look like a standard office desk. It used a fixed point 32-bit word and due to a coincidental quirk the hexadecimal numbering system used in coding the machine went from 0 to 10 normally, but then instead of the conventional "abcdef" for 11 through 15 used something like "fgkrqw".

That was the last of the "valve" machines. I went on to accomplish quite a career for myself, but starting with the RPC-4000 everything was transistors or integrated circuits.

- Ray B.