If you imagine that all pencil mark scanners worked by electrically sensing the graphite marks with sensing brushes, then you are wrong because this machine worked by optical methods.

The Card Feeding Mechanism

The cards were fed into a reading stage one by one at a rate of about sixty per minute from a magazine into which a stack of a couple of hundred or so had been placed. The machine was to 'scan' for nine pencil or biro marks about a quarter of an inch wide which had been put there by the switchboard operator. These marks were in strictly defined preprinted columns (a picture of a card is provided on the previous page) along about half the width of the card, and the positioning of these bore no relationship at all with where the holes would be eventually punched. The bottom card of the pack rested on a plate which had a raised edge half the thickness of the card. When the machine started this plate moved outwards then inwards and the bottom card was 'picked' i.e. scraped from the bottom of the pack and pushed through a narrow slot at the bottom of the inboard edge of the magazine. This slot was only one and half times the thickness of the card, so only the lowermost card could progress. Immediately it was gripped by rotating rollers, mangle fashion, with the upper roller being spring loaded. The card moved on until it met an obstruction, the 'card stop' whereupon although the underneath rollers continued to turn, the upper ones which were not driven ceased rotating and the lower rollers 'skidded' on the underneath of the card.

The Sensing Station

The card was now resting in a brightly illuminated area above which was a large lens, and the first 'marked' area was centrally aligned below it. Above the lens a small silvered mirror diverted the image at rightangles into a shaped metal box. The image encountered other mirrors and was projected downwards through a red filter onto a row of photomultiplier valves, one for each digit. The red filter was necessary to eliminate the red printing on the card. Only the photomultiplier receiving the image of the pencil mark would give an output, and this indirectly activated a 'punch magnet', which set down an appropriate pin in the entirely mechanical 'set-bar'unit in the heart of the punching station. The 'Set Bar Unit' is the basis of all Powers Samas punching machines, being a mechanical storage device which accepts an input incrementally for each column to be punched.

Back at the sensing station the card was allowed to proceed forwards and then halted again for the second marked row to be sensed. This went on until all nine possible marked rows had been sensed and their values mechanically stored, then the cardstop dropped down to finally release the card into the mouth of the punch unit. When it was fully in the correct position its passage was blocked by another cardstop.

The Punching Stage

Here the card was lifted against all the cutting edges of the 480 punches in the punch & die unit. The set bar unit was positioned above these and the set pins protruded downwards from it to correspond to the values of the sensed marks. The raised card lifted all the punches apart from those which had a pin above them and thus only those punches perforated the card. The card was lowered again, the cardstop went down and the card progressed into the 'checking unit'.

The Checking Unit

Again the card halted then was raised against 480 'sensing pins'. Where there were holes the pins remained in a lowered position, but where the card was blank the pins were raised. Above the sensing pins were an array of bullet shaped brass and tufnol 'sensing plungers', each consisting of rings of brass insulated from each other. Springy phosphor bronze fingers pressed on either side of these rings. Various electrical circuits were formed by this switch action, and it was possible to check for the following. That there was in fact a hole in each of the columns that should be punched, but also that there were not two or more. The card was lowered again and moved out to be deposited into the 'card receiver', passing over the 'rejects' box whose diverting flap would only raise to collect a card not conforming to the rule previously described.