Sinclair and the 'Sunrise' Technology: The Deconstruction of a Myth (20 page)

Sinclair made no attempt to hide his rage over the Timex dispute. He maintained that the strike had set back the Microvision launch by at least six weeks, and that each week’s delay cost his company £1m in lost turnover. However, when the £79.95 television was finally launched in September 1983, Sinclair made it clear that unforeseen manufacturing problems were largely to blame for the delays. Certainly, the original launch date of autumn 1982 had been hopelessly unrealistic. By way of explanation for the delays, Sinclair told the Sunday Times (22 May 1983), ‘
You must put that down to our pure ignorance about the problems of setting up a totally new technology.
’ A few months later, however, company policy seemed more inclined to forge virtue out of necessity:

‘Unusually for us we’re second in the field ... But we have learnt a lesson from calculators.’ There, he explained, Sinclair products were first, but were overtaken by low-cost, mass-produced calculators from Japan which commanded the market. This time Sir Clive waited until a plant was ready for volume production. (Daily Telegraph, 17 September 1983.)

However, a little digging into the development of the manufacturing environment in which the flat-screen television was produced enables us to be slightly more precise about Sir Clive’s problems. Before going into detail, it should be emphasized that Timex was by no means a natural choice as manufacturer for the project. The company had never built a television tube evacuation plant, and indeed had little or no experience of the technology. Since a number of UK companies had been developing such facilities for many decades - for example, Thorn-EMI and Ferranti - one can only assume that Sinclair’s choice had more to do with the development grants associated with the Timex production than the expertise and capital equipment offered by the deal.

One of the laudable characteristics of Sinclair’s approach to production has been that no manufacturing convention is sacred. Unfortunately, his companies’ mode and motivation when bucking convention has always had more to do with corner-cutting economies than the exploration of the potential of genuine technical innovation (the new wafer-scale project being an honourable exception). The decision to create from scratch tube production in Dundee precipitated a host of problems which eventually became so serious as to require the recruitment of a specialist to sort them out.

The conventional approach to tube production involves preliminary evacuation of each set of TV tubes by means of a roughing pump, and then completion of evacuation by a diffusion pump, backed up by another pump. Sinclair’s cost-cutting tendency led them to reduce the number of pumps, swapping the pump action by means of high-speed switches between different sets of tubes. While fine and economical in concept, in practice the lubrication of the switches released methane, a light gas which it is difficult to evacuate. When you get methane inside a television tube it is not only harmful to the phosphors coated on the inside of the glass, but produces carbon deposits on the cathode when the stream of electrons is turned on to produce a picture, which degrades the picture quality.

Early televisions, produced when evacuation was a less-developed art, avoided such problems by using a bent electron gun, with a magnet to correct the electron beam. With such a system any undesirable positively charged particles get deposited harmlessly on the side of the electron gun. With this arrangement, Sinclair would not have experienced so many problems. However, magnets are expensive, and the flat-screen tube had been designed with a straight gun. The combination led to severe delays in achieving tubes of sufficient quality. Sinclair’s tendency to pursue idiosyncratic solutions had once again led the company into difficulties.

Even when the new-look, one-per-person, flat-screen TV arrived, however, the initial publicity for this technological marvel failed to stir up the enthusiasm Clive no doubt expected.

Implicitly conceding defeat on the American retail market, Timex made a brave attempt to minimize the effects of its home-computer failure by signing a deal with a major US bank that wanted to give away TS1000s to new customers. The contract disposed of a couple of thousand machines, but did nothing to fade the writing on the wall. By August 1983, Timex-Sinclair finally got around to responding to complaints that the TS1000 had too little memory (RAM) to be of any practical use. The TS1500 was essentially an Americanized ZX81 with a built-in 16K RAM pack. Its launch did little to overcome the growing conviction that Sinclair products were intrinsically unsuited to the demands of the US home-computer market.

The TS1500 was a belated and decidedly flabby response to Timex’s rapid fall from grace. This definitively interim product was intended as a stop-gap between the TS1000 and the American version of the Spectrum (TS2000), but it was hardly worth the effort. By the time Timex got around to shipping the TS1500 in August 1983, Sinclair was promoting the Spectrum to a largely indifferent America. According to the Wall Street Journal (17 August 1983), even Nigel Searle, who was looking after the U S side of the company’s business, had misgivings about the applicability of the TS2000 for the US market:

The Timex Sinclair 2000, an adaptation of Sinclair’s ZX Spectrum computer, faces tough competition from the Commodore 64. Both sell for $200, but the Commodore unit has a superior keyboard and much more available software. Even Nigel Searle, Sinclair’s managing director, doubts the appeal of the model 2000. ‘By the standards of six or nine months ago the 2000 would offer a hell of a lot,’ he says. ‘But now I don’t know what the consumer’s preference will be.’

According to the Wall Street Journal, it was unsaleable stocks of the TS 1000 that scuppered the Spectrum’s chances in the States. By the time the TS2000 made an appearance, retailers in the US had come to regard the TS 1000 as a white elephant. No one wanted to know about its successors. The chainstores dropped the TS line, and Timex found itself out of its depth without the comfort of a distributive float. In the same month the TS2000 was offered to the American people, a senior vice-president at Timex inspired his sales force by proclaiming that, ‘
I don’t think Timex can make it in computers.
’

Smothered by competition from slicker home-based machines with fortunes to promote them and substantial software bases to support them, the TS2000 disappeared without a trace. The last recorded sighting of a TS1000 was a $10 gizmo at a supermarket checkout. Production of the TS line was discontinued in the summer of 1983, and Timex formally announced its withdrawal from the microcomputer market in February 1984. Vice-president C. M. Jacobi was landed with the thankless task of composing the meaningless newspeak that purported to justify Timex’s acknowledgement of defeat:

We believe that instability in the market will cause the value of inventories to decline, making it difficult to make a reasonable profit. Further, we are concerned that those conditions will strain trade relations between manufacturers and retailers, a relationship which the company values very highly. (New York Times, 22 February 1984.)

While in the States it would soon become a struggle to give away a Timex-Sinclair creation, in the UK Sinclair Research had established a market dominance that would be the envy of a multinational. Continuing production delays had inhibited Sinclair’s effective exploitation of the 1982-3 Christmas market, but by the spring of 1983 supply finally fell into step with demand. In February, Sinclair contracted Prism to handle wholesale trade of the Spectrum, and extended his assault on the high street to include the Boots, John Menzies and Curry chains. At the beginning of the year, the Sinclair machine looked unstoppable. Clive had sold off 10% of his personal shareholding, which valued the company at around £136m, of which Clive Sinclair still owned 85 per cent. The company’s triumphant computer division moved into spacious new offices in Willis Road, Cambridge, and in March Nigel Searle was appointed managing director of the Advance Products Division, which at this time meant that he was responsible for the development of any non-computing product. By the end of March, Sinclair Research had sold 200,000 Spectrums, and was announcing a £13.8m profit on a turnover of £54.53m A remarkable achievement for a new company which employed a staff of only 55.

Although 1983 will undoubtedly go down in history as the halcyon days of Sinclair Research, it would be remiss to pass over this period without reference to the occasional problem. For example, in March the company discovered that an entire shipment of Spectrum power-packs was faulty. It seems that 14,000 units were capable of flooring their owners with an electric shock. A massive recall operation was initiated in an attempt to replace the defectives before a hacker bit the dust. Fortunately, none of the Spectrum fraternity came to harm. As if to compensate Sinclair for an unexpected headache, fate delivered him the Guardian’s Young Businessman of the Year Award (at the age of 44).

Another hardware problem precipitated crisis and correspondence in the autumn. Minor chip-tampering prior to the release of the Issue 3 Spectrum proved to have radical consequences for the user. It seems that in August 1983 Research introduced a new chip into the machine’s design which changed the entry point for the Spectrum’s cassette-loaded programs. So what? Well, as a result of this minor change, a significant percentage of commercial software for the Spectrum simply wouldn’t run on the Issue 3. Sinclair Research pointed the finger at the ‘unprofessional programming practices’ of the software houses. Software companies complained that Sinclair hadn’t bothered to offer advance warning of the change. The computer press made a mountain out of a molehill, and offered untested suggestions as to how the Issue 3 problem could be circumvented. In the end, it seems the crisis simply disappeared of its own accord.

In May, Sinclair Research initiated a series of price cuts which are worthy of note simply because they exemplify an approach to marketing that has proved extremely effective for the company. At the height of the Spectrum’s popularity, Sinclair reduced its price - £99.95 for 16K, £129.95 for 48K - and cut the prices of both the ZX81 and the ZX Printer to £39.95. The effect of this move was to secure and extend Sinclair’s market lead and panic the competition. While most companies reduce prices when their products are in steep decline, Sinclair tends to discount shortly after sales have peaked. The advantage of his approach is that vacillating consumers are drawn into the fold while the product’s promotion retains a commercial urgency, and the costings of the competition are thrown into utter disarray.

The year 1983 was the one in which Sinclair made his dreams come true. In June, Margaret Thatcher’s ‘favourite entrepreneur’ was honoured with a knighthood. It seems that Sinclair was surprised and mildly embarrassed by the title, but in the context of the times the Prime Minister’s decision was hardly controversial. For Thatcher’s Conservatives, Sinclair symbolized the backbone of the nation and the country’s hope for economic recovery. As the personification of individual enterprise and endeavour, his example was meant to demonstrate that from the accumulation of personal wealth could spring employment and technological progress. The fallacy of this position would soon become as clear as the mounting unemployment figures.

Having recently emerged from a period in which his vision was curbed by the budgets of bureaucrats, Sinclair decided that his cash-rich company would never again suffer from constraints on its development resources. In May 1983 he announced that he had acquired the idyllic Milton Hall, situated in Milton village just outside Cambridge, at a cost of around £2m The building was to become the company’s MetaLab, a new kind of laboratory housing a brilliant team of researchers devoted to the pursuit of excellence in technological innovation. Sir Clive explained his plans for the centre in an interview with The Times (15 July 1983):

The MetaLab will bring together those parts of the company that might loosely be called ‘blue-skies’ research but are not really quite as blue skies as all that. It is the leading edge sort of work but also very much dedicated to a definite product - generally high-risk or very difficult work ... Freedom is the password. More freedom than researchers have been accustomed to ... In particular, [researchers] tend to be constrained in this country by their inability to purchase the goods they need. They have the odd business in Britain of employing a man at £20,000 or £30,000 a year who has to argue if he wants £100 worth of equipment. We will be employing people at very high salaries. They will be the crème de la crème, and if they think they want equipment they are going to have it.

Heading the centre would be two stalwarts from within Research, David Southward and Jim Westwood, and one ‘known quantity’ from outside the company in the shape of Richard Cutting. Before throwing in his lot with Sir Clive, Cutting had been managing director of Cambridge Consultants (since 1970), and was certainly well qualified to run a large research team. However, financial crisis was to hit the Sinclair empire before Sir Clive’s dreams for the MetaLab could be realized, although before the storm clouds gathered he managed to transform Milton Hall into one of the best-equipped research centres in the private sector. Another positive effect of the MetaLab was that it attracted the leading members of what would become Sinclair’s ‘wafer-scale’ team, the significance of which we will examine in our penultimate chapter.

As we have seen with the earlier products in the ZX range, Sinclair’s policy regarding the production of peripherals has been that such developments are best left to others. The company has felt disinclined to compete with the low profit margins tolerated by the garage entrepreneurs who emerged to support the Sinclair micros. The only exception to this rule was the Spectrum, which, from the start of its development, was always intended to be the central component of a complete microcomputer system. Unfortunately, as our account of the machine’s software development has revealed, the design of the Spectrum’s peripherals proved considerably more problematic than that of the machine itself.