Abstract: A process used to locate cable faults, in particular to locate non-burnout cable faults in the vicinity of branched cables. The pulse reflection method is used and a capacitor is charged and then discharged across a spark gap by means of spark-over at a fault in the cable. Simultaneously with a main capacitor in an impulse wave generator, a second pulse capacitor is discharged into the cable by means of a short circuit switch. The resulting reflection wave is input via a decoupler into a first channel of a digital storage unit. After opening the short circuit switch, the main capacitor in the impulse wave generator and the second pulse capacitor are again charged and then again discharged into the faulty cable by closing the main switch. The pulse capacitor is discharged by means of a plasma-triggered spark gap. The spark gap is triggered by the charge of the impulse wave generator.