Listening in to the Pascal

The work of an electronic computer in solving arithmetical problems consists in many cases of running through a succession of repeated cycles, and of cycles within cycles. This will perhaps have become clear from the foregoing articles in this number. The idea has occurred to many builders of electronie computers to listen in to these cycles by making audible through a loudspeaker the passing of numbers through the registers. To this end the loudspeaker is connected to one of the flip-flop circuits in a register of the arithmetical unit. The listener will hear the voltage variations in that flipflop as the numbers pass. Sometimes the pattern of the voltage variations is so arbitrary and changes so fast that only a hissing noise is heard. Often, however, th~ cycles mentioned produce a recognizable regularity in the sound, which may even result in a musical tone. In this way every programme, or part of a programme, produces a characteristic sound by which it can be recognized. The sounds thus offer a means of checking the operation of a computer: the programmer who has become familiar with these characteristic sounds while testing his programme, can later often tell by ear whether the computation is proceeding normally. Use is made of the same facility with the PASCAL, for which purpose the loudspeaker is connected to the last digit (the least significant one) of the S register 1). By way of illustration we have brought together on the attached gramophone record some fragments of the sounds which the PASCAL produces in performing the calculations discussed in four of the articles in this issue. The reader will find these fragments in four tracks on side 1 of the record, separated by visible margins; acoustically they are identified by an introductory morse signal. The first fragment consists of three sections, each announced by a morse sign. The fragments are taken from the following calculations: