Rodney Senior

I was good at physics in school in England, so I studied it at Trinity College Dublin, where the lecturers included Ernest Walton and Jackie Poole. During my time at Trinity I did summer work for two years at Mullard Research Laboratories, a company that produced very precise tuning circuits. I got through the physics course, although I had to resort to rote learning for some of the maths, and graduated in 1963. Away from the lectures, my main interests in Trinity were athletics, 16mm film and sound recording.

Seismology was another form of sound recording, although the frequencies were obviously very different. After leaving the university I joined Seismograph Services Ltd, a company that specialised in vibration seismology. My initial work was in Holland and the UK. Then they sent me to Australia as an observer on vibration seismology projects in oil exploration. This was a laborious task and the company expected us to work seven days a week. I left Seismograph Services and started teaching science and maths instead. I continued to teach when I returned from Australia in 1967, working for one term at a school in Northern Ireland.

Then I saw an advertisement in Wireless World magazine. IBM Ireland was recruiting staff through this British publication. I had always been keen on amateur radio, but I would describe the readership of Wireless World as semi-professional rather than hobbyist.

It was Alan Perkins’s idea to put an ad there. He was manager of the customer engineering group at IBM and chose the magazine because he was looking for people who needed to be really technical. I applied and was accepted by the company. Gerry Donegan and John Millea joined as engineers at around the same time.

Our main office was in 28 Fitzwilliam Place, which also housed the bureau service. The customer engineering group, however, was in number 24 and the office products and general spares departments were in number 1. We went over every morning at 11:00 to number 28 and there would usually be between fifteen and twenty people present. IBM was still a small organisation in Ireland.

Engineers had to comply with the company’s strict dress code in the late 1960s – unlike ICT’s engineers who we regarded as a ragamuffin group. We were told to wear suits in order to conform with an office environment. Another aspect of IBM’s approach to customer engineering was that we aimed to deliver Rolls Royce quality, even if the parts were more like Ford products.  Hardware adjustments must be precise. All the timings, even on punch and verification equipment, had to be correct within one degree out of 360.

Take, for example, the punched card handling equipment. IBM used a check digit in column 81 of its cards and the machinery had to be adjusted very finely to place holes in that column. We used a skew wheel to hold every card in position with absolute accuracy. The 81 columns also required a superior card with a precise thickness. Adjusting the clutch in the handling machinery was also vital. We were expected to give it as much tension as we could in order to move the cards quickly.

The success of IBM in those years was due to the quality of its input-output equipment rather than its CPUs. The I/O technology also ensured that benchmarks ran well on IBM systems.

The first training that the company gave me was on its card punches and verifiers. In 1967 I started working for customers with unit record equipment, including a subset of the IBM 400 tabulator range that was adapted for sterling transactions. Among others I looked after Cavan County Council and Rowntree Mackintosh, Colgate Palmolive and Caltex, Veha Radiators and Hughes Brothers.

The most unusual punched card installation that I worked in was the GPO, where telephone operators used a mark sensing technique to record the calls that they connected for customers. The cards were the same as everywhere else and the card holes were positioned in the same places. But the operators in the GPO wrote on these cards with lead pencils, putting individual marks in the appropriate column for each of the digits in a phone number. These marks were subsequently replaced with holes and the information on the card became part of a customer’s record.

In 1968 I was sent on a course to learn about the IBM 1401 computer. The company was still selling this product four years after it had introduced the more advanced System/360. It continued to sell the IBM 1401H, a reconditioned model, as a cheap and cheerful version that ran more slowly than the original. Irish Ropes had bought one for its manufacturing facility in Newbridge. I was given training so that I could support this machine.

The Newbridge premises had a water sprinkling system. One of my first tasks there was to place a hood over the computer equipment so that it would stay dry if the sprinkler went off.

The biggest 1401 installation in Ireland was at the ESB offices behind Fitzwilliam Street. Harold O’Hare was IBM’s engineer on the site and I began to take over from him during holidays. This gave me hands-on experience of the hardware. Then Harold moved on to the System/360 and I found myself working full time on the ESB’s two 1401s.

The company’s whole business hung on those 1401s. The ESB had a practice of selecting the final model in a computer range before it was replaced by a new series, based on an assumption that this technology would be relatively stable. Its did not upgrade to a System/360 until March 1971 – nine months after IBM had announced its new System/370.

In the late 1960s the two IBM 1401s produced all of the bills for ESB customers, using data from punched cards that were fed into the computers from high-speed IBM 084 sorters and IBM 088 collators. These sorters were the only models of their type in the country at time time, running at 2,000 cards a minute and they needed a stroboscope for maintenance purposes. The company ran those systems on a single shift from 9:00 to 5:00. Staff worked one evening a month to catch up with any unfinished work. These operations were overseen by two managers in the computer room who, in turn, were overseen by three layers of more senior management elsewhere in the Fitzwilliam Street complex.

Having two systems meant that, if a fault arose, we could switch all the punched card equipment from one computer to another and see whether the problem moved. There were very few occasions, however, when this was necessary. We never had real problems with the CPUs. If something went wrong, the cause was probably in the card feeds.

Preventive maintenance was performed during the prime shift – this arrangement was part of IBM’s agreed policy. The main things that we had to do was to keep the rollers rolling and to keep checking the quality of the power supply. Even the company that supplied electricity experienced an occasional loss of power.

Souvenir from the life of an IBM engineer: An overnight job that did not disturb the customers.

Security at the ESB control centre consisted of one man who clocked in at different locations on the premises to show that he had been there. We computer engineers had a key that gave us access to the whole place. It was amazing how much freedom we had.

IBM had an edict that no one should ever work on their own at night on a computer with the power on. This rule was bent occasionally. Once, after cracking a thick metal shaft in the 1401, I failed to persuade anyone else from IBM or the ESB to stay overnight with me and repair it. This created an incident, which was later resolved with assistance from Sean Queenan, who had previously worked on 1401s in New York.

Our customer engineering group expanded in the late 1960s. Liam Ryan succeeded Alan Perkins as its manager and Bernie Trench took charge of engineering at the Cork office. Sean McCluskey and Joe Ryan joined. Another colleague, Nick Spalding, moved from hardware engineering to software development. He was a brilliant customer engineer who had installed one of the first System/360s in the country at Irish Life.

My training and hands-on experience in the unit record days was really valuable. The mechanical adjustments always had to be meticulous. So you learned how to be meticulous. As a former colleague, Ron Weyda from IBM UK, used to say, this form of engineering was like a vocation. Computing in the 1960s was like a brotherhood.

Fault finding was what made the job really interesting. If the phone rang beside your bed, you got five pounds for picking up the call. But you never knew what you were letting yourself in for. One job required me to get up at 1 AM and to remove a sonic delay line that was stored in an oven at the Model 44 installation in TCD. Then I drove to Tuam, Co Galway, delivered the part to the Dart call centre there and returned to Dublin straight away.

I left IBM in 1974. I later worked for an electrical supplies company and in petrol pump engineering. I also started the New Irish Recording Company in the 1970s. I don’t have any punched cards around the house now, but sound recording has always been part of my life.

Last edit: December 2016

© Rodney Senior 2016