Thursday 3 April 2014

The World of Work

My posts labelled 'Work' form an occasional series describing work-related events.You can find all the 'Work' posts here. This post goes right back to the start, over 50 years ago, when I left school and got a job.

For various reasons, I didn't enjoy secondary school. I obtained passes in ten subjects at G.C.E. 'O' Level and continued for two years in the Sixth Form (as it used to be called - we hadn't imported American 'Grades' then). I managed passes at G.C.E. 'A' Level in Physics, Chemistry and Maths for Science. The intention then was to join British Railways and undertake a 'Sandwich' Degree Course at a technical university, probably Southampton. Well, the Civil Engineers seemed quite keen to have me but I wanted a career in railway signalling design and the Signal and Telegraph Department didn't quite come up to my expectations - the posts on offer didn't match my expectations. At the time, I was very disappointed. I turned my back on the idea of a career on the railways and, instead, was interviewed and obtained a post as an Electronics Engineer Trainee at a local electrical/electronics firm.


This cast nameplate (in various sizes) was attached to panels made by Contactor Switchgear Limited (Photo: Wolverhampton History and Heritage Website).

Contactor Switchgear Limited had been founded in 1936 in Wolverhampton. As its name suggests, it built electrical control panels, manufacturing its own relays and contactors. A reputation for high quality meant that there were around 600 employees at the factory in Moorfield Road when I knew it. Industrial electronics was still in its infancy but, wishing to expand into that area, Contactor Switchgear had purchased an established company from Watford called Radio Aids. Renamed Contactor Switchgear (Electronics) Limited, the company moved to Wolverhampton and I joined about thirty employees. Contactor Switchgear (Electronics) divided its time between supplying electronic equipment to its parent which was incorporated into control gear for industrial automation and designing and building electronics on a sub-contract basis for various outside companies. Although I'd learnt a bit about electronics as a hobby, my job was all a bit of a challenge but I found the technical aspects fascinating and I became completely engrossed.

There's a brief history of Contactor Switchgear here, on the Wolverhampton History and Heritage Website.

One of the first jobs I worked on was a set of boiler controls using electronics required for modifications at the West Thurrock Power Station. This power station had opened in 1962 as a coal fired plant but Babcock and Wilcox had designed modifications to allow the boilers to burn oil as well.

West Thurrock Power Station in 1973 (Photo: Peter Land).

When I started at Contactor Switchgear (Electronics), the electronic boiler control equipment had already been built and it was set up in a combined laboratory/engineering/test department for factory testing. I was impressed - there were a number of six foot high 19-inch wide racks loaded with sub-racks of electronics. There were hundreds of adjustable potentiometers and indicator lights at the front of the sub-racks. At the rear, there were chunky multicore cables terminated in 'Plessey' military-style connectors, allowing sub-racks to be quickly exchanged in the event of a fault. Opening up a sub-rack revealed numerous plug-in printed circuit boards crammed with resistors and transistors. Well, Contactor Switchgear (Electronics) had built all this kit to the Client's drawings but when they came to test it, it didn't quite work. I arrived just as a strategy had been agreed with the Client for fixing the problem.

The transistor age was still developing and only a limited range of transistors was available. The Mullard OC71, first introduced around 1954, was used all over the new boiler controls. The OC71 was a small, sealed glass tube (painted black) with three thin lead-out wires (emitter, base, collector) emerging from the glass at one end. A few devices had suffered broken glass, offering the opportunity to examine the internal construction of these germanium alloy junction PNP transistors. The glass envelope was packed with silicone grease - carefully removing this revealed a small rectangle of germanium, supported on one of the lead-out wires, with a 'blob' on each side forming the two alloyed junctions. The two 'blobs' were connected to the other two lead out wires. It was hard to believe that such a construction could provide useful control of an electric current but that was the start of a revolution which has changed all our lives.

The designer of the boiler control equipment had been a little optimistic about just how good the Mullard OC71 germanium alloy junction transistors were. Manufacturers' Data Sheets usually give "typical characteristics" for parameters like gain or leakage current. However, in the manufacture of these early semiconductors, it was difficult to obtain consistent results over a batch and so, to allow the manufacturer to sell more devices rather than throw them away, they also published guaranteed limits. For instance, a device with a typical gain of 30 might have a guaranteed gain of only 20.

Rather than use electromechanical relays to provide the necessary interlocks, digital logic was used, with a large number of OC71 transistors arranged to form logic gates ('NOR' Gates) using a circuit arrangement called 'RTL' (Resistor Transistor Logic). This all worked well if the transistors provided the expected characteristics but where transistors had less gain or more leakage than expected, anarchy resulted. This wasn't the only problem - there were other specialised circuits associated with checking the brightness and 'flicker' of a boiler flame which also required modification.

So I was tasked with testing all the assembled printed circuit boards to identify transistors which needed changing. I also tested hundreds and hundreds of new transistors, measuring gain and leakage and 'selecting' the devices into groups. By the time all the modifications had been carried out and the equipment had been tested and shipped to site, I was starting to realise just how much I had to learn. But I had also embarked on a career that would be challenging and stimulating.

Related Posts on other sites.

History of the transistor (Wikipedia).
Mullard Semiconductors.
History of Philips’ Semiconductors in the 1950s
Bipolar junction transistor (Wikipedia).