Forgotten Voice

 In 1965, when I was sent on my first IBM 6400 service call, the Rolling Stones were singing Maybe the last time I don’t know-oh-oh. Unlikely, however, there would ever be a last time for an OPCE when it came to keeping a 6400 accounting machine ticking over. Not that there was anything wrong with the build quality, far from it. Unlike its DP cousins safely ensconced in climate controlled environments with trained professional operators, the 6400 was subject to spilt coffee, fag ash, the dust and muck of a regular office and, often, ham-fisted operators. It was, after all, an Office Product.

 These days, whenever I find myself holidaying in warmer climes, the call of cicadas is reminiscent of the chatter coming from the 6400’s bank of relays. A hinged relay panel was accessed via a door at the back of the main unit. The relays took care of the program stepping through the maximum of 160 program steps and called up the arithmetical and other machine functions. Behind the relay panel were the PCBs that took care of the calculations. The electronic panel was augmented by a magnetic core memory.

I was one of the first engineers in the UK to be trained on the 6400. The chap who trained me, whose name I have unfortunately forgotten, had previously been to the States to help modify the electronics to accommodate sterling calculations for the UK market. His claim to fame was the “two-penny” latch; our pesky shilling being composed of twelve pennies. 

 Programming the 6400 was a massively time-consuming activity. It was done manually by plugging colour-coded patch panel leads into a removable plug board. The front and rear views of such a panel is illustrated below.

There was a hinged door at the end of the main unit where these panels could be dropped in. There would be individual panels for each accounting process. When the required panel was inserted and the door closed, the plug leads that protruded through the back of the board made contact with a forest of sprung-loaded fingers that, in turn, made contact with the machine’s circuitry. There was always a spare board for the benefit of the OPCEs to set up diagnostics. 

There were several additions to the 6400 engineers’ toolkit. I remember being issued with tongs to extract relays from their housings, a file to clean the contact points, a supply of silver replacement contact wires that were kept in a plastic vial and a jig to insert them. The most bizarre servicing aid was a bottle of carbon tetrachloride (or some similar volatile liquid) containing a suspension of fine metallic particles, the purpose of which I will come to in a moment. 

The magnetic ledger card unit accepted file cards just slightly larger than A4 size. A printer inside the unit printed transaction details on the front, while a magnetic stripe on the back was used to record the same information.  It was essential to properly align the read/write head as it tracked the stripe, especially if a card was to be used across several different machines. In order to do this, fluid from the bottle was wiped on the stripe. When it evaporated, the metallic particles clung to the stripe revealing the recorded bits, leaving a visual record that could be used to make adjustments.

The golfball printer in the magnetic ledger unit was keyboardless. The printer on the main console was an I/O device with a specially widened carriage to accommodate the sprocket-fed computer print-out paper. Both printers had a sub-frame of electronics to decode the ASCII alpha-numeric codes and solenoids to operate the printers in output mode.  

It was the nature of these modifications to the Selectric typewriter that led me on to life after the 6400. In the early ’70s, word processing in the form of word processing machines were becoming widely available. IBM’s offerings at the time were the MC (magnetic card) word processor, the successor to the MT (magnetic tape) machine.  

Competition to IBM came from a number of companies who tried, with varying degrees of success, to build word processing units based on their own modifications to a standard Selectric typewriter. One such company was the short-lived Kalle Infotec Ltd. Their ambition was to be the “third force” in the office products market, pitting themselves against IBM and Rank Zerox. Kalle Infotec was launched in 1973 with a WP offering called the Infotec 7000.  

Problems arose from day one with a dearth of engineering support for the Selectric machine, let alone the decidedly iffy electro-mechanical sub-frame that Kalle Infotec had had built for it. Shortly after being awarded my IBM ten year clock, Kalle Infotec made me an offer I couldn’t refuse to dig them out of a hole - thereby ending my time at the Lodge Road training centre in St Johns Wood. My clock is now more than fifty years old and is still working - which is more than can be said for the Infotec 7000.  

On to Wang Electronics who embarked on a similar misadventure into the world of word processing. At least Wang had some notable clients, including the film director, Stanley Kubrick, who foolishly relied on the reliability of his Wang unit to get his screenplays written up on time. After that, I served my time with A. B. Dick Ltd. The A. B. Dick Magna I was in direct competition to the IBM MC and sunk without trace, as did the now defunct A.B. Dick company. 

The rate that technology was advancing at the time was the ultimate nemesis, not only of the word processing machine, but also the long forgotten IBM 6400 that was discontinued in 1976 (although service agreements continued for a few years up until 1983). I have often wondered if IBM were pleased to see the back of it because, unusually, the product number for the IBM 6400 accounting machine was recycled to a brand new product, the 6400 line matrix printer, thereby committing the machine I had a great fondness for to the technological dustbin. Hey-ho.