Digital audio of a recorded interview with Tony Earnshaw, who worked as an engineer on Shell-Mex & BP's LEO III.

Interviewer: David Philips
Date of interview: 16 February 2017
Length of recording: 47m13s
Format: 3 original .mp3 recordings 65.11MB (transferred to .mov video for presentation on YouTube 148.66MB)
Copyright in recording content: Tony Earnshaw and LEO Computers Society

Transcript editor: unknown

Abstract:  Worked as untrained TV repair man in 1950s, spotted by passing LEO engineer, invited to apply to LEO Computers. Started work as assistant to site engineers on the Shell-Mex and BP LEO III in Hemel Hampstead. Prides himself on his successful career as an engineer without having acquired any qualifications. Rose to be chief site engineer at Minerva Road responsible for seven sites. Left ICL to become independent engineer.

Transcript :

LEO COMPUTERS LIMITED  -  Oral History Project
Interview with Tony Earnshaw by David Phillips

[David Phillips.  My name is David Phillips and I am interviewing Tony Earnshaw on the
16th of February 2017 on behalf of the LEO Archive.  
First of all do you mind telling me when were you born?

[Tony Earnshaw]:  8/11/1935.  That makes me 81.

 [Tony Earnshaw]:  Born Queen Charlotte’s Hospital in London.
 [David Phillips]:   Right.  What did your father do?

[Tony Earnshaw]:  Originally he owned a little garage which became Spurling’s and was a main Bedford agents. But then he got called up and he was a soldier.

[David Phillips]:  And he stayed a soldier?

[Tony Earnshaw]:  Well he got out as quick as he could.

[David Phillips]:  And what, went back to his garage?

[Tony Earnshaw]:  No, he decided to retire.

[David Phillips]:  You went to school where?

[Tony Earnshaw]:  I went to school at Chester College, Harrow. I went to Miss Parnell’s Primary 
 [Tony Earnshaw]:   Well, I was interested in various things.  The headmaster at my college was an atomic scientist and also a reverend, which is an unusual combination.  And the principal of the school was another reverend and ran a local church.  But, the one who was the atomic scientist taught me a heck of a lot that is outside of the box.  I was not a very good pupil, in fact I'd class myself as a dunce.  
 
[David Phillips]:  Yes.  So, did you go to a technical college or a university?  

[Tony Earnshaw]:   Only to give lectures.  

[David Phillips]:  So, so when school finished, what did you do?  Where did you go?

[Tony Earnshaw]:   The first job was with a television company where I was repairing televisions.

[David Phillips]:  They taught you how to repair?

[Tony Earnshaw]:   No, I taught myself.  

 [David Phillips]:  Right.  So you're a sort of a natural electronics engineer? Now, how did you get into computing?

[Tony Earnshaw]:   Through being head-hunted by a gentleman who came in to our television shop.
I don't know why he was in the shop.  He saw me repairing them and then asked me if I'd like a job in computers.  So I said ‘yes’.  

[David Phillips]:  But did you know what a computer was?

[Tony Earnshaw]:   No.  But I thought I'll find out.

[David Phillips]:  What company was he representing?

[Tony Earnshaw]:   He was with LEO. And the place I worked was at Hemel Hempstead. I was living in Aylesbury and travelling daily.  
 [David Phillips]:  So there you are in Hemel with a rented LEO, probably a LEO II.

[Tony Earnshaw]:   I had a site engineer. And he chucked me panels and said ‘they are faulty, you go and mend them’, and that was it.  

[David Phillips]:  So how did you go about it?

[Tony Earnshaw]:   Well you check the components over and found out what was wrong with them.

[David Phillips]:  Yes.  And, would you, these spare parts, would you get those from LEO or different..

[Tony Earnshaw]:   They had a quantity of spare bits and pieces. On site. But anything that was needed they ordered up, resistors, capacitors etc.  Now, they were using a few transistors but this was mainly a valve orientated machine. There was twenty four hour cover on all the computers.

[David Phillips]:  Yes.  Yes.

[Tony Earnshaw]:   If it wasn't one engineer it would have been another, always.  

[David Phillips]:  Right. 

[Tony Earnshaw]:   But they had twenty four hour cover on all sites.  

[David Phillips]:  There’d be a lot of peripheral equipment, etc., the printers and the paper tape feed system.

[Tony Earnshaw]:   Oh yes, yes.  Maintaining these was a doddle because it was mainly mechanical work problems like dropping a screwdriver in the works.

 [David Phillips]:  Yes.  Well, how did the things progress from LEO in Hemel Hempstead? You went to another LEO?  Or another LEO company?

[Tony Earnshaw]:   An off-shoot of LEO, which rapidly became known as English Electric Computers.

[David Phillips]:  Right.

[Tony Earnshaw]:   These were rather larger machines and, I'm afraid to say a lot faster and we had a gentleman who was a programmer who was the first person ever to get music and speech out of a   computer.  His name was Gordon Robson.  He's never had any acknowledgement at all on this, but he was a paraplegic and couldn't speak and could only use one arm.  He became the chief programmer for Barrick, Barclays Bank Bureau.  He came in one day and started using the computer.  I thought he was an operator, he was only one handed but he was a genius.  I don't know what's happened to him.

[David Phillips]:  So, I'm still trying to work out how you move from fixing the valves to becoming the chief site engineer.  

[Tony Earnshaw]:   Well I can't understand that myself because I became chief site engineer of only two sites, that was at Minerva Road, that's near Park Royal.  
 [David Phillips]:  So, tell me about your move to Minerva Road, how did that come about?

[Tony Earnshaw]:   Well, they needed an engineer there. I made the machines run better.

[David Phillips]:  How?

[Tony Earnshaw]:   We speeded up the clocks, we adjusted the delay lines and, also there was a lot of programming.  At this time the machines ran on a wired ROM (Read Only Memory) and often we had to re-programme the ROM’s by hand, by feeding the wires through. Nowadays, this would either be put in a solid state device rather than hand worked. Some of wires were even wound round nails.  Instead, I suggested that they use ferrite cores which were much more efficient.  

[David Phillips]:   So, how did you learn about ferrite cores?

[Tony Earnshaw]:   That's a good question.  I had done work on televisions, etc., and transformers use different types of materials depending on the frequency at which they are used.  Ferrite cores are much more efficient at higher frequencies than iron cores because of the remanence in the magnetic substances.  

[David Phillips]:  So you were taking some of your television repair skills and knowledge and applying them to the computers.


[David Phillips]:  When you saw the computers at Minerva Road how were you able to speed them up or find that they might be able to work faster?

[Tony Earnshaw]:   Well mainly by speeding up the clocks, the faster the clock, up to a point, the better they ran.  But it, if you went too fast, obviously, you started getting errors.  The funny thing was they had been keeping them slow, thinking that they wouldn't get the errors, and they were getting more errors running slow than they were at the right speed.

[David Phillips]:  Why do you think that was?

[Tony Earnshaw]:   It's to do with skews in the wiring.  If you run at the wrong speed you also get ringing on the lines and this can cause cross line interference, from one line to another.  Even now computers are, even on motherboards and stuff they've got, sometimes there's, and one line is longer than the other to correct for skews and things like that.  

[David Phillips]:  So how did the engineers who were building these computers react to your suggestions.
[Tony Earnshaw]:   Very negatively. They didn't like me talking to them at all.  But when they weren't there we got it faster for them.  And in a lot of cases they didn't know what we’d done but they were saying ‘oh, look what we've done, it's running better’. 

[David Phillips]:  So how long were you at Minerva Road do you think?

[Tony Earnshaw]:   I should think it's about two years, or thereabouts.  So I was still covering the other sites.  Well, the way it worked was I was resident, more or less, at Minerva Road and then called out to the other sites as and when needed.  Since most of them were in the London area, except for a few, like Letchworth which was another place where there were some.  I didn't have too much trouble travelling.  But the further afield ones were covered by different engineers not myself,   because they had Bristol, Birmingham etc.  These were what we called ‘customers’ I would say, were actually being run by the customers.  They'd rented them under contract and they provided their 
The way they did this, they sent them for training prior to them acting as engineers. There was a training school being run at Letchworth, by the way, for engineers.

[

[David Phillips]:  But you had to move from, sort of valve technology, to transistors and now microchips?

[Tony Earnshaw]:   The move from to transistors was a bit peculiar because while I was with English Electric they more or less went over to transistors, but they also got involved with two other companies.  One of them was Hyva, who made dekatron tubes with the lights going round.  And the other one was with another company, I can't remember the name, who made backward diodes.  And these had the same sort of affect as a memory that was easy to use, in theory.  The only trouble is they tended to latch up and they weren't a lot of good over a period of time.  

[David Phillips]:  What does ‘latch up’ mean?

[Tony Earnshaw]:   Latching up is, instead of going backwards and forwards, they stuck in one position.  That's the best I can give you without going technical.

[David Phillips]:  Well what was your impression when you first saw the LEO computer because it must have been enormous compared to what you've been working with.

[Tony Earnshaw]:   Well it wasn't enormous, it was just the fact that it was vastly different, because I have worked with some large equipment in the past even so.  I mean engineering used for motor manufacturing etc. is quite large.  And since my father had had a garage I was used to be engines and things like that.
 [David Phillips]:  So when you moved on and you spent really most of your career associated with..

[Tony Earnshaw]:   In electronics.

[David Phillips]:  Electronics and that's put you in touch, or in contact, with some fairly sophisticated computing equipment.  Anything particularly memorable stands out?

[Tony Earnshaw]:   Yes. You're going to chuckle.  Whilst working at Hartree House, which was part of Barclays Bank Bureau, we found that the distance between the engineers and the operators’ room was a long way from the actual computer room.  So having had quite a bit of thought on this we decided to use roller skates.  Now I'm afraid I got a letter from the people owning Hartree House, ‘you must not use roller skates going to the computer room as it leaves black marks on the floor’.  So I got banned from using roller skates.  But the efficiency was, we were all using them, went up twenty per cent.  The time taken to get from one end of Hartree House to the other was twenty minutes, we timed it.  So, just to get a box of paper for the operators, for the printer, it was taking them twenty minutes of down time.
 [Tony Earnshaw]:   Something else. To run the cables or run a new printer or something you had to get under the floorboards and run the cable physically.  You got nice and dirty that way. We did have one bloke who was jammed under the floor and we had to cut the floor open. Most of the wiring was run, whenever possible, under the floor because otherwise you'd be tripping over it.

[David Phillips]: And did you have experience with delay tubes?

[Tony Earnshaw]:   The stuff we were using was glass delay lines.  They worked it out roughly a Nano second per metre of wire.  So you could make up your own delay lines if needed by just lengthening the wire.
[Tony Earnshaw]:   Well the LEO stuff I realise it's what you're after.  But it's a peculiar thing because they didn't do an actual merger and they didn't do a true change of ownership or anything.  It was most peculiar the way that it progressed.

[David Phillips]:  How did it impact on you personally?

[Tony Earnshaw]:   Well I was happy until it became involved with ICT, because as soon as that happened they wanted everybody out except their own people.
 [Tony Earnshaw]:   When it became ICL it rapidly went downhill.  The equipment that ICT made, like printers, scanners, etc., etc., they were good, don't get me wrong, and they worked well with the equipment.  But when they started trying to change the computers it didn't go.
And of course then IBM jumped in.  And it wasn't that the IBM computers were good, they're still not really good.  It's the fact that they had the money behind them.  But there was another independent company over here and there are also in America, people don't realise it was a dual country thing, and that was Cray.

[David Phillips]:  Super powerful, super computers.

[Tony Earnshaw]:   You see the door there?  To about here, to about a wall, that was it.

[David Phillips]:  That's about six foot by two foot, three.

[Tony Earnshaw]:   Outside there's about eighteen tanks of liquid nitrogen to keep it cool.  And the Cray’s were running eighteen thousand times faster than the fastest IBM.  And the last Cray, and this is where people get muddled up because Mrs Cray too over when Cray died.  They built one more machine, the Cray 2, and nobody’s heard about it hardly, it's American military.
And that is the most powerful computer in existence.  The Japanese and the others have built really powerful machines but they can't touch a Cray 2.  And they're gonna have to go a long way till you get up to the speed of a Cray 2 because it's running super conductively.  In other words, you put a signal at one end, it's at the other end before you get any delay.  But I did design something and I tried to get it off the ground a while back but they pinched my idea and, although it's not gone into production.

[David Phillips]:  What's that?

[Tony Earnshaw]:   A light computer, works entirely on light, internally.  Not in the normal sense.  
 But anyway, the way it works is there's no clocks.  And up to now all computers have had a clock.  The way it works is an instruction goes out, say they had, and then the information is fed in and then it's gonna be output.  What it does is as soon as it's finished that instruction the next one goes, and the next one.  There is no actual delay between instructions, therefore it runs as fast as the components will allow.  It's an unclocked computer.  And that is the fastest way you can do it, especially when you're using light.  

[David Phillips]:  So why don't you develop it?

[Tony Earnshaw]:   Sixteen billion pounds to develop it.

[David Phillips]:  Okay.  Alright, well Tony, we’ll leave it at that if we may and I'd just like to conclude by saying that...
This interview with Tony Earnshaw has been recorded by the LEO Computer Society.  The Society would like to thank Tony very much for his time and reminiscences.  The interview and the transcript form part of an Oral History Project to document the early use of electronic computers in business and other applications, but particularly in business.  And any opinions expressed are those of the interviewee, that's Tony, and not of the Society.  And the copyright of this interview is recorded, in recorded form and in transcript remains the property of LEO Computer Society.  Thank you very much.


[End]

Provenance :
Recording made by the LEO Computers Society as part of their ongoing oral history project.

Copyright
The video above may be subject to copyright of the original author and is present only on this website for non-profit educational purposes only. If you are the original copyright owner and wish to have the media removed from this website please contact us using the relevant contact email address listed on our contact page.