The first type of event occurs upon detection by the isolator's detection circuitry of an anomaly in the device network, such as an abnormal change in current or voltages in the detector network that may be caused by, for example, a short-circuit on the network 130. When such an event occurs, the controller wakes up and, in step 411, measures the current and voltages on the line 200. A starting delay duration is determined based on the measured values (step 411), and loaded into a timer (ShortConfTmr) at step 412.

It may be desirable to limit the maximum delay to, for example, 6 ms. If the controller uses a 10-bit ADC, for example, to sample voltage, then the 2¹⁰=1,024 output values of the ADC may be spread evenly between 0 ms and 6 ms, so that each value step represents 6 ms/1024, or approximately 6 μs per voltage step. If the maximum voltage represented by an ADC output value of 1,023 is about 30 volts, then each 6 μs delay corresponds with a voltage drop of about 6 μs*1024/30 v≈200 μs per volt.

Meanwhile, steps 414, 416 and 418 act to shorten the duration of the delay based on the number of false positives that have already been registered. If the number of false positives registered (in FalsePosCntr) exceeds some predetermined threshold ZYX, say, for example, five false positive events, i.e., there is a history of a large number of false positives within a predetermined period, as determined in step 414, then another timer, TimerAdj is increased by XX microseconds (step 416). At step 418 the ShortConfTmr value is reduced by the value in TimerAdj (but not below zero).