The scanning area was illuminated by means of six car 48 watt bayonet fit headlight bulbs and so became very hot and rather surprisingly no forced air cooling had been provided.There were four narrow chimmneys mounted in the roof of the scanning area but these were insufficient in effect. The bulbs were in brass holders and the tiny springs in the sping loaded contact plungers couldn't withstand the daily eight hours of heat for very many months and one by one they would collapse. The last thing needed was a flickering light source so eventually they all had to be replaced by a different and simpler type of holder. These were more or less just clamps and wires had to be soldered to the bulb contacts. This completely solved the problem but of course the bulbs did tend to blacken with age.

Above the lens was a small mirror set at forty five degrees to deflect the image sideways. Cemented to it was a metal plate with a U shaped phospher bronze finger for fixing it to a post protruding from a transducer. This looked exactly like, and was the driving part of a loud speaker. Its voice coil was driven with a 100c/s supply and hence the mirror spent its time vibrating, this being done to cause the otherwise static mark image to be swept back and forth over the photomultiplier aperture, causing that device to give out not a steady level but several pulses. Due to the heat from the lamps the cement could become partially defective and the mirror would "flap" on its mount and this could cause "over scanning" resulting in the image to be projected on to the aperture of a neighbouring photomultiplier with the end result of an unwanted hole being punched in the card. There was no cure for this, new mirrors had to be fitted as required.

The heat also caused bubbles in the cement fixing the composite lens elements together. This rather surprisingly had little effect until it became very bad when the F2 lens had to be replaced.

Immediately over the row of photomultipliers within their darkened cabinet a red filter was fitted. This prevented any chance of the preprinting from the card getting into the photomultipliers as the colour red could not pass through it. The red filter was in fact a modification added to the original design which had used black printed cards when it appeared that the preprinting was indeed giving that sort of problem. In hindsight one wonders if it was the defective mirrors causing such "overscanning?"

The checking unit worked very well but after a few years of operation the brass plungers in popular positions became worn and would become trapped by the phosphor bronze contact fingers pushing on them. However they were fairly easily replaced. Incorrect wreck removal by the Operator could cause the tips of the sensing pins to become bent and this was a longer repair.

The card receiver had a designed-in problem which was only fully overcome by an unauthorised modification. When the receiver became full the weight of the cards was supposed to make a platform at the bottom drop about a quarter of an inch to operate a microswitch in the "knock off" circuit. An adjustable spring loaded ball bearing detent was supposed to gauge the weight of the cards but it was totally unreliable. The Operators would complain if the machine stopped and they found only a few cards in the receiver and also complain if the receiver overfilled spilling cards on to the floor but it was almost impossible to achieve a really reliable setting. On the worst example I constructed a very simple photoelectric system to detect the height of the stack requiring no adjustment and I am told it worked perfectly for the rest of the life of the machine.

Everybody has heard of the "hanging chad" problem with the US election punched card vote counting system. "Chad" is the punchings from the cards but Powers Samas called them "chips". After several years it was noticed that the perimeters of the holes in some popular nought positions were becoming furry.This was because the die had become blunt but it wasn't a huge problem as when those cards were sensed later in either the sorter or tabulators metal pins were used. Furry holes did create dust though. Initially, merely swapping the punch with one from an unused position was a sufficient enough fix. However, in the final years and long past the machines "use by" dates, new very high speed sorting machines with photoelectric sensing were introduced and they didn't like furry holes. Changing the die assembly was a "dockyard" job and in any case new assemblies became unavailable. Drastic measures were called for the only solution was to file down the surface of the die block until the indentations around the affected holes were eliminated. Then a thorough carborundum stoning was needed for to remove the file marks and to give a sharp edge to the holes. Desperate measures indeed and carried out by better mechanical craftsmen than me, as by then I had moved on to other things.

Now who would have thought that the red mylar filter material would deteriorate, clamped as it was between two thin sheets of glass? But Peter Preston has told me that it did and that no replacement material was available. However a resourceful local manager had a brainwave and contacted the stage manager of the Oxford Playhouse and they helped ICT out of its difficulty as they used it for their spotlights..