原创 Saving the day with a scope

It was the early 1970s when I was sent overseas with three impressively sized racks of equipment controlled by a Honeywell H316 computer. Along with them came the orders to install the system and get it running. My adventure began when my boss sent me to a local one-week Honeywell computer course on machine language. That single course was the only training I received on this new system. Apparently, I couldn't be spared for the longer hardware courses, though my boss did allow a technician (who was going to the same location as the equipment) to attend. I only had to install the computer system, train everyone how to use it, and leave. That was the plan.

The peripheral devices included a paper-tape reader, teletype terminal, large reel-to-reel digital magnetic tape drive, high-speed printer, data interface, and a huge hard disc. The hard disc was about a foot high and completely sealed inside a cylindrical enclosure that was pressurized with an included nitrogen bottle and pressure regulator. As I later discovered, it was fortunate that all the hard disc electronics boards were mounted outside the pressurized enclosure.
"This is not me (or the system I installed), but it is a similar picture that I found on the Internet."—Steven Karty

This was back in the olden days before BIOS ROMs told the computer what to do after being turned on. So I had to "fat-finger" in around 30 16bit words of instruction, which told the computer how to read the punched paper-tape reader output. Then I had to load an ASCII punched paper-tape into the paper-tape reader, which told the computer how to read the magnetic tape drive's output. Then I had to make sure that the large magnetic tape reels, which contained the computer program, were mounted and rewound to their beginning.

Then, when I had everything ready, I would simply hit the start button, the computer would read the punched paper-tape, the magnetic tape reels would spin, and the whole system would start. This initialisation procedure had to be repeated each time the system was powered on.

Before the system could be shipped, it had to be packed. Before it could be packed, everything heavy had to be removed from the racks and packaged separately. Although other people did the packing and crating, I first had to disconnect and remove the equipment from the racks and make sure that I would remember how to reinstall it.

Everything went smoothly—at first. The equipment, the technician, and I arrived intact at our destination. I reinstalled and reconnected all the equipment, cued everything up, and hit start.

Then things got rough. The paper-tape reader ran, but the magnetic tape reels refused to budge. Most of the hardware peripheral interfaces were not only unique and custom-designed, but also poorly documented. I called the technician over and asked for his help. We single-stepped through the instructions where the computer was stuck and figured out that the computer was waiting for the hard disc interface. The computer could not go onto the next step and tell the magnetic tape reels to spin until this disc interface was ready. After using a Tektronix scope to trace through the disc interface, we concluded that the interface was waiting for a signal from the hard disc.

The technician then abandoned me, saying he had been trained only on the interface and not on the hard disc. As he slipped out, he mumbled that the "origin" signal from the hard disc seemed to be missing.

I realised that I would be blamed if I couldn't fix the system. That I hadn't been allowed to attend any hardware courses was irrelevant. Unfortunately, there were no replacement boards for the hard disc. Fortunately, the system documentation included schematic diagrams of the hard disc electronics boards.

Deserted by the technician and feeling very lonely, I picked up the scope probes and began tracing through every circuit where I thought the origin signal was supposed to go. I finally found the origin signal at the input to a potted delay line. But I didn't see anything at the delay line's output. In desperation, I decided to solder a jumper wire around the delay line. When I then repeated the initialisation procedure, everything worked perfectly!

The manufacturer of the hard disc later said the design had enough margin so it did not need a delay line, but it sent a replacement anyway. In the end, all it took was just a piece of wire to fix this computer system. But I never would have found the problem, and thus its solution, without an oscilloscope. And I would have lost interest long before finding the problem if using a Tektronix scope were not so much fun.

I met another technician (who had spent two years where I installed the equipment) before I left on this trip, and asked him for any hints about the site. He said it was nice and safe, so I wouldn't have any problems. That was only partly true, because I had stomach problems the entire time. I saw him after I returned and asked if he ever had any gastrointestinal issues. He said that while he always felt fine, his wife suffered from the same problem as I had described to him in such graphical detail. When I asked him what might account for the difference, he said his wife drank the local water but he drank nothing but beer. I didn't bother asking him why he forgot to tell me that before I left. But I still wonder if he brushed his teeth with beer.

I still get a kick out of using Tektronix scopes because their triggered sweep circuits always work perfectly—which is why I finally bought my own. But it's an old analogue scope with a CRT. The new Tektronix scopes with their digital displays are way more fun.

Steven Karty built an oscilloscope from an EICO kit in the mid-1950s when he was 10 years old. He started working in radio and TV repair shops at 13, became an amateur radio operator at 14. He has used Tektronix scopes almost exclusively for the last 48 years, and he has BSEE.

He submitted this article as part of Frankenstein's Fix, a design contest hosted by EE Times (US).