First jobs after University
I left Southampton University with 2:2 Hons in Electronic engineering in 1975
Kent Automation
My first job was as an Electronic Design Engineer with Kent Automation (part of Brown Broveri group) in Hitchin, about 40 miles north of London. The projects were to design interface boards for the new P4000 process control system that used a PDP-8 or PDP-11 as the main computing power.
It was an interestig time to be in development as everything was moving very quickly. The Intel 8080 micro prcessor had been launched in 1974. Integrated circuits were being launched on a regular basis with the core product being the Texas 7400 series, in 14 and 16 pin DIL packages and 0.1" pin spacing. At university, I had used the equivalent 4000 series CMOS chips - low power, but also slow and prone to damage due to electro static shock. About 1976, low power schottky was introduced. This was fast, but much lower power than standard TTL. It was always a delight to read commercial magazines and discover new chips that matched the required application. One controller I disgned featured in effect a state processor where we used a PROM to store several configurations, a bit like a PLA (Programmable Logic Array).
What I never saw when working at Kents was a system on site. I also never saw the P4000 working - I was totally in the development stage. The closest I got was about 20 years later when I saw a reference to it in the London Science Museum
Technicon Isca
In 1978 I moved to South Wales (near Newport) and worked for Technicon Isca. This is a company that specalised in continuous weighing systems for the food and flour industry. Again based on the PDP 8, but also using the Intersil 6100 processor. This was a CMOS single chip PDP 8.
My job was intially to design interface boards to operate in the PDP 8. I also started to do more programming, partly for testing, and also for fun. There was free use of the PDP 8 systems after office hours. A big difference with this job was that I was also involved at times with system maintenance and repairs. This meant being given the keys, typically to a Ford Cortina, and heading around the country. For the first time, I came in to contact with customers, and systems being used. This did not always go smoothly. Once I was having touble with a PDP 8 which had an extended interface bus. To try and isolate the issue I unplugged the extension bus and repowered the system up. In doing so, the unplugged extension bus opened many control valves and it dumped 1-2 tons of grain into a hopper. I was greeted by my manager on return to the office. The cause of the problem was dust (flour dust) clogging up the electronics, and issue as a company we hit many times. Computers in the 1970s ran hot and so needed cooling by powerful, noisy fans. Filters would also block up causing overheating.
On the day I joined Isca (April 1st 1978) it was taken over by an American company, Technicon who made medical analysis equipment. After about 1 year it was decided to convert our Intersil 6100 product (an IscaFlo) to run on joint Technicon equipment, based on an Intel 8048 processor (actually a 8035 which used external EPROM). I was given the task to program this, in assembler. The product was the next generation of a system to weigh flour, or grain on a continous basis. It worked by having two hoppers. One was filled while the other was weighed and then discharged. A divertor would then reverse the process, so at any one time one hopper would be filling. The controller had to monitor the weight in the filling hopper and then decide when to flip the divertor, wait a few seconds, measure the static weight and discharge. At the same time, display the current total weight, and communicate to other systems via an RS232 link. An advanced adaptation of this was for the controller to batch process upto about 12 types of input grain to make up a feed formula.
All of the above, multi-tasking program was performed with 6KB or ROM and 256 bytes of memory. Development was using an Intel development system and an In Circuit Emulator (ICE) which cost about £12,000 in 1980. The ICE was a module that was placed in the development board and behaved the same was as the 8048 would, but one could view registers etc for development and debugging. It included 2 x 8" floppy disks.
A second development I undertook was a control system to monitor upto 15 IscaFlos at the same time. For this I designed the controller board based on an Intel 8085 and programmed it using PL/M - an Intel Language.
Part of the development process meant working with our American parent company - based in New York, about 1 hour north of Manhatten. In my 5 years with Techicon Isca I had 5 trips to USA. One was for 6 weeks which I moaned was rather a long time. I was suddenly offered a free ticket for the 'new girl friend' which she took with delight, and has shared my life for the past 46 years - eventually becoming my wife.