Edge detection

Detections in a Tausand Abacus occur by reading the instant where a rising edge of an input pulse arrives, since this is the signature of a valid detection. At this moment an internal clock starts ticking at the resolution of the device, namely 5ns, to wait for its partner pulse in the other port to arrive before the coincidence window falls, and taking it as a coincidence.

This implies that the pulse width of your detector might not be equal and the Abacus will read them correctly as far as the pulse width is larger than the resolution of the device. If your pulse width is shorter than the resolution, there is a chance that the pulse is not detected.

A major advantage of this method of coincidence detection is that avoids double counting a coincidence due to different pulse widths between detectors. Imagine detector A producing pulses of 10ns and detector B producing pulses of 50ns. If you apply a simple ‘and’ gate of these inputs, you might find that the long pulse of B is in coincidence with more than one pulse of A, erroneously counting them as several coincidences, but, how is it possible with a single detection in B to find more than one coincidence between A and B?