Signature Analysis has long been used to detect faults in a wide variety of machines. Recent efforts to locate faults by acoustic methods is illustrated by the location of corona in large power transformers. Excessive corona can damage insulation and require costly shutdown and repair. Corona generates both high-velocity electrical signals and ultrasonic signals which travel at the velocity of sound in the insulating oil. The acoustic signals are sensed by accelerometers attached to the exterior of the transformer tank. By triggering a signal average from the electrical signals, measurements of the acoustic travel time from the corona source to several transducers can be obtained. In simple cases these can be transformed to distance, and triangulation calculations yield a corona source location. More generally, the acoustic waves encounter obstacles, are reflected, and the triangulation techniques fail. A new technique called Space Interrogation systematically assumes a series of allowed corona locations and calculates the mean-square fit of the data to each assumed source location and set of permitted acoustic paths. The result is a more powerful means of locating corona sources.