Sinclair and the 'Sunrise' Technology: The Deconstruction of a Myth (4 page)

It was like something out of an Orson Welles drama. You went up and there was a girl sitting outside the door at the desk, a fierce young lady, an embryo dragon, and the doors were huge sliding doors. You slid them back, and there was Clive at the far side of the room beyond acres of carpet and nothing else - a marvellous setting, (ibid.)

After nine years of operation Radionics had, despite some ups and downs, grown fairly steadily. It had declared profits to the year ending May 1971 of £85,000 on a turnover of £563,000, was employing 54 people, and had acquired huge cheap premises in an area with low wage rates to help it continue in the same way.

[2] CALCULATING PROFITS

The first intrusion into daily life of the exponentially increasing density of components that could be packed on to a small slice of silicon was the calculator. The Sinclair Radionics success story hinges on the calculator. From a moderately successful electronics and hi-fi company Radionics was swept by the tide of cheap calculating power packaged for the pocket or the briefcase into the major league - briefly.

The electronic calculator, making use of logic circuits to perform the arithmetic operations required, first appeared in the late 1960s. These were desktop models, complex to produce and expensive. The key to the decreasing cost and size as the decade progressed was again the semiconductor transistor technology built up on the silicon chip. ‘Hard-wired’, i.e. fixed, logical pathways could be laid down on a chip, so that the pattern of voltages representing the numbers to be manipulated was changed in the appropriate ways. By 1970, the large-scale integration (LSI) technology for packing huge numbers of components on to a single sliver of silicon had advanced to the point where all the simple arithmetic operations (addition, subtraction, multiplication and division) could be achieved by the circuits formed on a single chip. Calculators became smaller, and portable, in the sense that they could be crammed into a briefcase or a poacher’s pocket if you didn’t mind the weight. The reductions in size pushed the price up rather than down, however.

It was not the chip itself that was the problem, but the ancillary devices that consumed space and added weight: a keypad for input, the visual display, and the ancillary electronics. The circuitry had to translate keypresses into the correct form of electrical information for the chip, and retranslate the output for display. Crucial to any portable machine was the fact that all this required a power source, and the size of battery required to give a reasonable period of use made the ‘portable’ calculator, independent of mains power, more of a backpack, rather than pocket, proposition.

There was an obvious market waiting for anyone who could produce a true pocket machine, and Sinclair identified this as a new area of endeavour for the Radionics R&D team. Again, the motivation was for a ‘one-per-person’ product. There was no reason why Radionics could not have produced calculators earlier, since they could have competed directly with the first imported versions, but this is not the sort of market that arouses Sinclair’s interest. His miniaturization mania was in this case appropriate to the product, enhancing utility and portability, rather than being a pursuit of smallness as its own justification.

The Radionics advertising manager of the day, Chris Fawkes, recalls Clive’s approach to him with the concept:

At first there was not really much more than an idea to sell. The meeting started by Clive showing us what was basically a drawing on the back of an envelope, of an idea he had had, which was for the Sinclair Executive, a very small calculator. He had this idea for a product, he had a conviction that it would sell in fairly high volume, but in answer to the standard questions as to how are you going to sell it, what sort of volumes are we talking about, and so on, he said, ‘Well, no, you tell me, this is why I’m talking to you.’ (BBC Radio 4, 18 January 1978.)

The Radionics R&D team, mainly Chris Curry and Jim Westwood, managed to package around the Texas Instruments GLS 1802 chip (which itself had some 7000 components built on to it) a circuit of 100 or so other components and the display. The package was of suitable size, but the battery was the problem. Most of the power was needed for the LED (light-emitting diode) display. Whereas the rest of the circuit drew its minimal power requirements only when actually calculating, the display had to be on all the time the calculator was in use. The clever solution was to use the persistence of the display diodes, which did not turn off immediately but continued to glow for what, in the millisecond units of electronics, was a fair while, and a similar persistence in the contents of the chip memory. By turning the power to the display and chip on and off rapidly enough the display would appear constant, and the chip would retain its memory, but power requirements would be minimized. Such a pulsed power supply to the display reduced power requirements by a factor of at least ten.

This crucial breakthrough allowed the calculator to be driven by hearing-aid batteries, and in due course the Sinclair Executive was launched in June 1972. At £79.95, it had not only a low price tag as things were in those days (everything else with the same functions was way above £100), but it was also stylish, lightweight and a genuine breakthrough in size. Its design, or at least its external appearance, won it a place in the Museum of Modern Art in New York, as well as a London Design Centre Award. The aesthetic external case was designed by Richard Torrens. Presumably overriding political scruples with aesthetic considerations, the Czechoslovak Institute of Industrial Design also honoured this quintessentially elitist executive toy. The flexible plastic case was black, the Bowmar LED display was red, and the whole thing was 5 by 2 inches and the same thickness ‘as a cigarette packet’. The latter point was presumably of significance to Sinclair, who was a heavy smoker of Capstan Full Strength in pre-marathon days.

The Executive was a great success among those who could afford it. The advertising agency Primary Contact produced sophisticated ads in up-market magazines, and Harrods stocked it. The profit margin was high, even with a product that went increasingly into retail outlets and to those developed portions of the world where executives took to it like ducks to water. This was because component and manufacturing costs certainly didn’t exceed a quarter of the selling price. Some 80 per cent of the production was exported, and it reportedly earned some £2m in export sales during its eighteen months of active life. A significant coup, albeit a one-off, was a British reversal of the usual direction of technology flow when $1.5m worth of Executive calculators were sold to Japan. The oriental response to the reliability of the circuitry, which was not particularly impressive, and the service facilities being in St Ives if you wished to invoke the 5-year no-quibble guarantee, is not recorded!

As an impressive expense-account bauble, the Executive was correctly targeted, professionally marketed, highly profitable, and of limited use. That it was a simple four-function calculator, although with a percentage function allowing repeated calculations, was not a disadvantage, since this was state of the art. The problem was the battery life, which was a few hours only. If you wanted to bring it out just for a few swift calculations in a meeting, creating a good impression of being a businessman with the latest tools of hi-tech at your command, and remembered to turn it off, you could get it to last for a couple of weeks. Forgetting to turn it off not only ensured that it would be dead when you brought it out next, thus creating rather the wrong image, but such a constant drain (much higher than that for which they had been designed) on the batteries could cause them literally to explode!

The success of the Executive set Radionics off on a swift growth curve. The hi-fi products were still being produced, with the Project 60 kits doing good business despite their by then well-known characteristics of being ‘underspecified and over-rated’, as John Pemberton recollected it. (Rating referring to the power output, not its reputation.) System 2000 was followed by System 3000, with a new speaker, the Q30, but despite colour ads in the magazines the market penetration by means of retail sales didn’t take off. It was worthwhile, in sales terms, to persist with the products, but the production and quality-control problems were still there, even if design was better, as the Radionics design team was expanded with bright and enthusiastic graduates who occupied the top two floors of the Mill.

The other product line that started in 1972 was instruments. A small multimeter with a digital display, the DM1, designed by AIM (part of the Cambridge Consultants Ltd nexus) was put into production. Using a Nixie tube for display, and a polypropylene case with built-in pushbuttons, it was novel, cheap at £49 and produced an enthusiastic response when advertised. Too enthusiastic, in fact, since the delay between sending your money and getting the product was so long - a matter of months in some cases - as to make even Sinclair feel an apology was necessary. The ads incorporated a public statement of incapacity to meet demand in response. By the time production had geared up enough to satisfy the initial demand, the deficiencies in the product, especially the display, had become known and demand fell. The price was reduced later in the year, but it was not a great success.

However, with the DM2 digital multimeter, designed by John Nicholls and launched in February 1975, the instruments side started on a steady growth. It was not an area in which Sinclair played a part, since he thought the instruments uninteresting. The instruments were reliable, well engineered for production, the sub-contracted assembly unproblematic, and steady product development was shown over the years.

The calculator boom was on, however, and in true entrepreneurial fashion the Executive generated enough profit to finance the next products, moving down in price, and up in facilities offered. John Pemberton, who joined Radionics shortly after the Executive launch, and was responsible for the external design of all Sinclair products up to the ZX80, remembers the atmosphere among the design team as 
‘fun, because the technology changed so quickly, the market was volatile, and it was a team of clever young people in their early twenties’ (Interview, 22 October 1985).

The top two floors of the Mill at St Ives held the design and development team, Sinclair’s lavish office, and the full-size snooker table. With contracted-out assembly Radionics was employing only 70 people in early 1972, rising to 130 or so by the end of the year. Of these the R&D department was some 20 or 30 people, Sinclair’s hand-picked bright boys from university, Cambridge Consultants or Mensa contacts. The number of employees expanded in 1973, when component and finished-product quality-control departments were set up, in response to poor product reliability. Still, turnover per employee was £14,000 in 1972-3, and £17,000 in 1973-4.

The success of the Executive generated a turnover of £1.8m in the year ending April 1973, using a capital base of just over £300,000. The follow-up product was the Cambridge, incorporating some mathematical functions and a constant facility, which was launched in August 1973 at £24.95 exc. VAT. This had a notoriously unreliable on/off switch, which was a minimal-cost design solution, fine for production, but with the problem that after some use it would be impossible to switch the calculator off. A few months later the Executive Memory, with the additional function the name implies, and ‘a new black and white styling’ came out at £24.95 plus VAT. This was the way the market continued, with prices for comparable calculators roughly halving each year, and more functions being packed into the machines occupying the top end. Texas Instruments was the first chip producer to manufacture calculators, cutting off the likes of Sinclair from leading-edge chips, but the main threat came from the East. The Japanese, especially Casio and Sharp, with high-quality standards and a flair for creative plagiarization and price competition were the threat at the lower end.

Radionics kept ahead of the game for a while, with the Sinclair Scientific in 1974, which had reverse Polish notation, which meant, for example, that in order to add 2 and 4 you had to enter 2, then 4, then the plus symbol. This perversity, which was easier to present to the chip, was not the most attractive method to those who had to use it, however powerful its facilities. It was noteworthy for having a keyboard the same as the Cambridge, but packing twelve functions on four keys, a Sinclair trait that would later be epitomized by the Spectrum keyboard, and for being a semi-custom chip (a modified Texas Instruments chip).

In early 1975 the Oxford range was produced, with a larger, non-pocketable size. The size was due mainly to the 9-volt PP3 batteries, which were only just adequate to their assigned task and gave a short life. The 100 was a simple four-function machine, at £12.95 plus VAT; the 200 gave you a percent key and a memory for your extra seven quid, and the 300 was the scientific model, with an impressive range of functions, at £29.95. They all used the same case, 6 by 3 by 1 inches, but were distinguished by their ‘click-action’ keys being ‘colour-coded yellow, blue and avocado’. That they all used the same case was a handy by-product of a deal with Gillette, who had enlisted Radionics’ aid in designing a desktop calculator, but then pulled out, probably wisely, from this new market area. As the Financial Times reported on 10 March 1976, the ‘abortive US venture did not cost [Radionics] real money since the tooling costs were covered’. Never one to miss out on the recycling possibilities of either components or cases, the outcome of the ‘unsuccessful courtship’ was put to good use by Sinclair. The other event connected with the Gillette deal was that Michael Pie, then Radionics’ technical director, left to take a job with the US shaver giant. He will return in our next chapter.

In 1975 Radionics was Europe’s biggest manufacturer of calculators - not that there were many competitors. The success, attributed to ‘control of chip design, but not in-house’ and ‘styling, aggressive marketing and easy reputation in the form of design awards’ (a Radionics spokesman quoted in the Financial Times, 10 March 1976), was at its height, and the fall would not be long delayed. In March 1975 the monthly throughput of calculators was raised to 100,000, compared with 40,000 in September 1974, and 30,000 in March of the same year. The decline in sales forced a cutback to 60,000 a month later in 1975, and the boom was over. Receiving the Queen’s Award to Industry, for both exports and technological innovation, in April must have seemed a bit of a hollow honour when the accounts showed that although the 1974-5 turnover was up 50 per cent, to £6.3m, profits were minuscule at £45,000.