Things I remember about IBM the year 1963 [The Television antenna caper]

The Problem:
In 1963, on my first day at IBM in Endicott, NY., my new manager, and I were taking a familiarization walk around the North Street Plant, in Endicott, NY. We noticed several engineers trying to lay triple shielded wire in the ceiling of the North Street plant. (This wire I knew, was costing well over $35 per foot. And this cable run was at least a 350 foot long.) I was curious and asked what they were doing. They told me they were trying to make a data link between the IBM 7030 data collection computer at one end of the building, to the control interface on the computer manufacturing production floor, but they could not get the noise out of the network with double shielding and now they were now installing triple shielded wire for the data transmission. They told me that they had been working to solve the problem for several weeks, and the data was still very unreliable.

I observed that they were using a Simpson 260 multi-meter to measure from the center conductor to the outside shield to show me. They had about 8 volts rms.

I asked if they had used an Oscilloscope such as a Tektronix Model 535. They said that there were some oscilloscopes in the plant, but they had not used one.

I had seen such an Oscilloscope on the production floor and I asked if I could get it and try to see what the noise looked like on the Oscilloscope. They told me I was wasting my time, since they were experienced engineers and they knew that a new guy did not have their collective experience. I said I know that, but I might have seen this problem before, and I might be able to help.

The Investigation:
I went and got the Oscilloscope and extension cord for power. On the floor, near a work bench, I also saw a diode and picked it up. I connected it to the Oscilloscope probe to make a rectifier and touched the center of the coax signal. The problem was obvious when up on the screen was the horizontal and vertical sync pulse of that Television Station at some 18 volts peak to peak.

[ pictures in photo section since pictures can not be presented here]
[ Note: The horizontal sync* signal is a short pulse at the beginning of each video line which keeps the horizontal scanning of the monitor exactly in step with the transmission of each new line. The vertical sync signal is a pulse transmitted at the beginning of each field and frame. Its purpose is to maintain the monitor in field-by-field synchronization with transmission of the next frame pulse. Sync signals reside in a part of the video signal in which no visual picture information is transmitted. During that particular part of the transmission (the blanking retrace period* or vertical interval*), the electronic beam is blanked and retraces back to the other side of the screen to start a new line or a new frame. Both horizontal and vertical sync are needed to create a fully stable picture. ]

I asked them if they had a Radio or Television Antenna anywhere in Endicott and they said no. But, we had one on the mountain in Vestal, New York, just south of the river south of the IBM North Street plant. I went to see if I could locate it, and I saw the antenna through one of the cross walks that ran between the several buildings. There it was, less than a mile away. A 500,000 watt Television station, that broadcast its signal on Channel 6. I smiled and thought, “URICA I have found it”
.
I realized that the triple shielded cable was acting as a TV reception antenna and was picking up standing waves from the Vestal TV transmitting antenna.



The Challenge:
When I told Doug Hertzig, my manager what I planned to do, he said “I do not think that you can do it. These guys have been working to solve the problem for weeks”. I told them that I would bet any of you a cup of coffee, -that I can fix it in 30 minutes. My manager did not have a problem with at cup of coffee as the wager. They all wanted to prove me, the new guy wrong, so they accepted my wager.

The Solution:
We obtained a spool of 26 gage twisted pair wire, from the parts crib and I started to run the wire from the test floor console, up through the ceiling. I had borrowed a stapler from a desk and a roll of tape from one of the tool boxes. Using the step ladder, we started to run the twisted pair wires in an unusual rectangular path that they could not understand and they said that I was wasting the wire and my time, but they wanted to get a laugh at my weird wiring. In about 20 minutes I had gathered a crowd at my weird work habits.

After I ran the wire through out the ceiling from one side to the other, I dropped the spool down in the computer room by the IBM 7030 interface and asked if it was all right for me to connect it to the terminal where the Coaxial cable had been.

When we were through, I connected the unshielded twisted pair to the terminal block on the IBM 7030 interface and said you problem is fixed.

The Verification:
Of course no one believed me, so they started to check it out with the Simpson 260 multi-meter and they could not find any noise.

I told them to try to send a data signal from the production instrument to the computer interface. They did and it worked like a champ.

[pictures in photo section since pictures can not be presented here]

Of course they wanted to know what I had done to fix the problem. I said that I killed the TV station signal they had riding on their data signal, by nulling or canceling it out. I did this by running the 26 gage twisted pair wire in a cross pattern, so that the signal picked up by one half, was cancelled out by the other. Now nothing but the data was now being received at the IBM 7030 data computer.

I suggested that they try and learn how the oscilloscope worked, since it was the much better tool for trouble shooting this kind of problem.

I did not have to buy coffee for the rest of the week.

Russell E. Theisen
Life Senior Member IEEE