Sinclair ZX Spectrum at 40: Interviews with Crispin Sinclair, Richard Altwasser, and Steven Vickers

On April 23, 2022, the 40th anniversary of the release of the Sinclair ZX Spectrum home computer, the National Museum of Computing in the UK organized an event that included interviews with Sir Clive Sinclair's son, Crispin Sinclair, and with Steven Vickers and Richard Altwasser who were key in designing the ZX Spectrum at Sinclair Research.

These are my personal notes taken during the interviews that the museum conducted during this hybrid in-person and online event.

40 Years Celebratory ZX Spectrum Event — Notes

TNMOC via Zoom, 23.04.2022

Geoffrey Wearmouth is logged on (wrote Gosh Wonderful ROM for the ZX Spectrum)

Crispin Sinclair (Sir Clive Sinclair’s son)

Clive was always talking about inventions, not so much about business
Crispin had one of the first Black Watches and QL’s
Clive didn’t want to hear about the bad sides of things, like the small wheels of the ABike
When Crispin had an endlessly spinning Microdrive, Clive just said, “It’s an early one….” Crispin lost a few university assignments that way.
Most early inventions like radios and calculator were already there but he made them much smaller and cheaper than anyone else on the planet at that time.
In 1977, Crispin was the first child at school with a TRS-80. Crispin got into it, which inspired Clive to create a computer for education.
When Clive was 10, his school couldn’t teach him any more maths.
Clive didn’t have a soldering iron at home, he worked using just pencil and paper (and his sliderule)
Crispin cannot remember his father ever sitting down in front of a computer. He definitely did not have one at home, and possibly not even at the office.
Clive’s heart was at inventing; business was a necessary evil to get inventions out the door
At some point when the Spectrum did so well they had a Sinclair twin-engine company plane
The C5 chassis was made by Lotus cars, Crispin joined Clive when they flew up to the plant
Crispin got a sense of Sinclair’s success when everyone in 6th form had a ZX81 or a Spectrum.
They ended up moving house because of the C5 disaster; Clive had to sell the family house, but it didn’t really bother the family.
Crispin was about 16 when the Spectrum came out. They flew up to the Dundee factory with his father; he saw the soldering baths where the Spectrums were made.
The Spectrum was the biggest heritage of Sir Clive. Crispin wasn’t aware of the Eastern European clone scene until recently. Last year saw 200 new games for the Spectrum made.
Clive was very frustrated not getting the BBC contract; he thought it should have been the Spectrum. Other than that, competition was just part of the game.
Crispin thought Micro Men was quite good, but Clive didn’t want to see it. Crispin thought Clive’s actor was spot on, while his brother said the complete opposite.
Clive went to a few retro events about 15 years ago (Brighton library, also met with retro fans from Poland in person).
The Z88 worked well; Crispin had one later during university, used it for writing essays and assignments
Polish scientists used a Spectrum to hack the TV system to broadcast “Solidarnosc” during the Polish democratic revolution in the 80s
Have any of your own Sinclair products survived to this day? No, were in a cupboard at home, but when he went for university, his mother binned his entire collection. :(

Steven Vickers / Richard Altwasser (abbreviated V/A below):

A first was apprentice for Jim Westwood on the ZX81, designed the PCB for the 81, then became responsible for the Spectrum hardware.
From 100,000 to 1,5 million was a huge step (ZX80 to ZX81). Computer magazines were outselling women’s magazines at WHSMITH during the ZX81.
A asked Sinclair to let him publish a simple ZX81 programming book, and he sold 30,000 copies.
The world expected the ZX82 to be launched in April 1982 because of the annual convention at which the ZX80 and ZX81 had been released.
V on how ROM development worked: Write in on separate computer, burn the EPROM, find that it doesn’t work, erase the EPROM under UV, and repeat.
A on color clash: Of course they could have done full color and hardware sprites. But Clive’s rule was to have a fixed price at 4 x the component cost. Clive would walk around asking, “What would happen if this component was removed?”, keeping engineers on their toes.
A and V were together on the stand at Earl’s Court, where Clive announced that the first 1000 customers paying in cash would get a machine within a certain short time frame. At the fair, V&A overheard school boys talking about every fine detail of the hardware, the BASIC, etc. that they had just heard about an hour ago.
A sold 50,000 copies of his Spectrum book. He wonders: “Let’s assume all rumours about bribery were myths. Let’s also assume it’s legal for a public institution to back a commercial product. Why did the Spectrum still end up selling 3x as many units as the BBC Micro?”
V: When writing the Spectrum manual, he wanted to avoid readers hitting a brick wall at any point during learning. Example: He did not just introduce PEEK and POKE, but talked about memory and system variables so there was something to do with the knowledge.
John Grant created the function key system for the ZX80. The keyword system allowed users to quickly type something and get a reward.
A has worked in industry and hired lots of people who cut their teeth on the Spectrum.
V: He is proud of the BEEP command. He looked at how it worked on the Acorn, where you had to provide frequency. V did the math to do it using semitone pitches instead, a much more user-friendly format. Since floating point calculations were too slow for this, he added a few tables that were faster and precise enough.
The manual jokingly suggested to enter the rest of Mahler’s symphony as an exercise. In 2014, Matt Westcott actually did this using networked spectrums, with a Raspberry Pi for synchronization, for an event at the History of Science Museum at the University of Oxford (it’s also on YouTube).
A (at 4:06pm): Challenge of having to use an asynchronous counter.
V: Continuing challenge of fitting everything into 16K. With one bug, probed the circuitry with a logic probe. Found lots of memory with the same content could create extraneous fields.
They did little analysis of competitors’ products.
V: Clive liked demos set up so he just had to press enter.
A: Clive cultivated the persona of slightly eccentric inventor: “Sinclair Research exists to invent. It only sells products in order to fund the research.” While some criticized this, this egocentric image ultimately helped the company and its business.
A is completely surprised and shocked at the Spectrum’s continuing success, and does not understand it.
V&A left Sinclair. A came with his Ace design and suggested to V to start their own company around it. V had gotten a tip from a friend about Forth. They quickly sold 300 Jupiter Aces, then had saturated the market of Forth enthusiasts. :) V wanted to do something less ephemeral than computers :) so went back to academia.
A: Rick secretly helped them design the case for the Jupiter Ace. It was very much faster than a BASIC-using computer with equivalent CPU. It was produced and launched successfully, but not a commercial success.
The Sinclair computers seemed to follow a philosophy of being very accessible on all levels, from hardware to BASIC to manual. But this was largely a team effect, not from Clive’s primal design guidelines.
A submitted hand-drawn circuit designs to Ferranti on Mylar sheet. A Ferranti engineer then turned this into a design on their circuit grid. Ferranti allowed them to use an uncut wafer. A went there, and none of the ICs worked due to a missing trace in the counter — except one where a speck of dust had fallen to bridge this connection, so A could run his tests.
A has seen 5-6y old kids getting stuff to happen on the screen using Scratch. So the lesson is: Make It Easy!