Do people want pocket TVs? History says not but, back in the 1960s it seemed like a good idea to Sir Clive Sinclair.
"I had tried to get into pocket TV back in the Sixties because I missed the pocket radio market, which had been terrific, and I thought it would be fantastic to do the same thing with TV," recalls Sinclair, "in 1966 I nearly came out with a pocket TV - not a flat screen TV - but a pocket TV - using a lot of ASM transistors."
ASM - Associated Semiconductor Manufacturers - was a semiconductor jv between GEC and Philips which GEC pulled out of in 1971.
Sinclair then went for a bold innovation - to develop a flat CRT. It was to cost Sinclair Research £4 million.
"The flat cathode ray tube was the only one ever done, anywhere in the world," says Sinclair, "it was the only one which ever worked. It was a lovely thing. A lovely little tube, and we designed and built a factory to make it. We did everything ourselves. "
"And we did this chip for it which would be remarkable even nowadays, and really unbelievable for its time," continues Sinclair, "it was one chip and it had all the TV on it except for the tuner and a few high-voltage transistors at the output. It was automatic multi-standard - it worked on NTSC, European PAL or English PAL - quite automatically - so you could take it anywhere and it worked."
Sinclair got his pocket TV to market but there was a snag - no one wanted to watch TV on so small a screen..
"So it took years and years to get pocket TVs onto the market and, when they did get onto the market, it didn't turn out to be as big a business as expected," recalls Sinclair.
"I can rationalise that now because people want to look at a big screen, but you don't know that at the time, and I'm still surprised there was not more of a market," says Sinclair, "of course people still make pocket TVs now, but they don't sell in great quantities. It could be that people don't like to look at something so small - that it's a strain for our eyes to watch something so small."
"We lost quite a lot of money on that."
A lot of people did try to get Sinclair to supply these flat CRTs for numerous niche applications - oscilloscopes, medical monitors and anywhere where we would not use a LCD - but apparently he was never that interested. Rumour at the time was yields were so poor they could never have made enough of them if the product had taken off.
I was at Sinclair Research at the end, but the CRT research was still going on. There were various colour versions - they had one working with red and green, I remember, but never managed blue as well. Tennis at Wimbledon came out as this rather fetching antique sepia. The bigger effort was in making larger versions, and there was an experimental production line: the trouble was that the 4" flat faceplate had a nasty habit of imploding, and it was seen as rather a dangerous posting. Yields in general were very poor, and I have no doubt that this, combined with Sinclair's fixation with magic technology as the crown jewels of company success, meant that no deals would be done.
And the QL was supposed to have one of the larger flat CRTS built into it. Although it quickly became apparent that this was not going to happen, there was no money to redo the bit of the ULA that handled video and so the timings of the video out were always just that bit out of spec - so lots of monitors and some TVs just never worked with the thing. (There was also going to be a real-time clock with battery backup, which never worked: some time later, we got an email asking if anyone had any idea what to do with a warehouse full of zinc-air batteries that were now surplus to requirements.)
The Pandora business portable was also going to use the flat-screen CRT. By then, of course, the bigger ones were really not going to happen, so the two mechanical prototypes that were made had an extraordinary optical path of mirrors and lenses that flipped down from the lid when you opened it up, a bit like pre-war TVs with vertical CRTs and mirrors in the lid. This gave a virtual image of a 14" screen floating somewhere beneath your knees...
Various bits of company folklore attached to the tube. One was that the waveform needed to create a linear picture from the geometry of the deflected beam was very complex, and that it started off being modelled in BASIC on Spectrums. This worked, up to a point, but was excruciatingly slow - so a quick calculation was done about what was needed to speed it up, time booked on a Cray, the code ported to FORTRAN and job done. Very Sinclair.
Thanks very much indeed, Rupert, that's a cracking account. You must have had an awful lot of fun at Sinclair