LAUNCH OF THE AMSTRAD CPC Part 1

Zoë Kirk Robinson tells the story of how the Amstrad CPC home computer came to be.

In the early 1980s, Alan Sugar thought nobody needed a computer at home. When Clive Sinclair released the ZX80, everything changed. That simple machine, made to be cheap and simple to operate, opened the floodgates and out poured computers from every company that could throw a chip into a box with even the most rudimentary keyboard attached. Alan Sugar checked out the competition, and he spotted an opportunity. “There were hardly any electronics... just a few chips on a PCB,” he said in his autobiography. The computers may have been simple, but they were selling like hotcakes, and Amstrad wanted in. There was only one problem: Amstrad knew TVs and Hi-Fis, they hadn’t a clue about computers.

That wasn’t going to stop them, however. Alan and Bob Watkins sketched out and built a mock-up of what they wanted the Amstrad Computer to look like. Working from the same mindset as Clive Sinclair, Alan wanted a simple computer that anyone could set it up and get started — but unlike Clive’s machine, Amstrad’s would be an all-in-one system. It only needed one plug, it wouldn’t hog the family TV, and it didn’t need an external cassette deck. The Amstrad computer would be simple yet powerful. It was elegant, with a full-length keyboard “like you would see at an airport” and a sleek, modern colour scheme. It was almost perfect, except for one thing: Amstrad had no idea how to make it work.

A computer needs circuitry to make all the parts work and a CPU to control it all. Amstrad had

neither, and they didn’t know what to use. So Alan and Bob called on “two long-haired hippies”, as Alan calls them in his book. These hippies had solved a problem with some early Amstrad TVs, so they knew their way around circuit boards.

The hippies designed a circuit board based around the MOS 6502 CPU to run the Amstrad computer, and it would all focus on a proprietary ROM chip that Amstrad would own, so nobody else could make their computer. This was important to Alan because if only Amstrad could make an Amstrad computer, he would control the machine’s supply and quality.

Alan recalls in his autobiography his later realisation that creating the ROM “[would] have taken a team of fifty software engineers a year”. Amstrad’s problem was that the hippies assured them it could be done in a month, and both Alan and Bob accepted this because they had no experience making computers. In retrospect, expecting two guys who had fixed some TVs once to make a computer in a month was, at best, going to create a very high-pressure situation. The thing is, every company under the sun was bringing out a computer, so Amstrad needed theirs out soon, or they would be left behind, and it wasn’t as if Alan had come up with the onemonth deadline, was it? One month, lots of pressure from Bob and a hippy passing out from stress later, the code for the ROM arrived at Amstrad’s offices. Amstrad was on a deadline, so this code was immediately sent to Toshiba for manufacturing. Toshiba, confused, called Amstrad back and asked what was going on. The code was garbage; it would

never work. Had the hippies sent it to Amstrad to keep the company quiet for a while longer? As you can imagine, Alan was furious.

What could they do now? The computer needed to be out by 1984 if there was going to be a chance of it gaining a market share and without the ROM chip, the project was dead in the water. Thankfully Bob knew a guy who knew a guy who could put them in touch with some other guys. Maybe this could all get sorted in time after all because Bob had heard those other guys had most of what Amstrad needed already worked out.

The guy Bob’s guy knew was Roland Perry, and the other guys Roland knew were a small team called Locomotive Software.

Together, they took Alan and Bob’s original mock-up, threw out the

6502 chip and used the Z80A instead, as that’s what Locomotive Software was already working with. Making the switch to a new CPU this early in the design stage would save a lot of time re-coding Locomotive’s operating system. It meant switching from a machine running the same CPU as the BBC Micro to one running the same as Sinclair’s Spectrum, but it also meant getting the design and production back on track.

At some point, it seems the team also dropped a second joystick port from the machine. Early magazine chatter about Amstrad’s secret project, such as the report in Personal Computing Weekly from before the press conference announcing the machine’s existence, make a big thing about the computer’s two joystick ports.

Given photographs of Arnold, the prototype of what would become the CPC 464 only show one port; this was either a very early change that the magazine managed to get wind of, or they were confused about the second port being on the JY-1 joystick. Sadly, we can’t know for sure, but we can say many Amstrad fans would have loved an extra joystick port on the machine, so if this was a change, it wasn’t for the best.

Either way, Arnold was here, and he looked good. All that was left for the team to do now was build the machine and get it out to market.

But that’s a story for another time.