Cirrus SR22 N110EB

About 5 minutes before the accident, when the airplane was in cruise flight in instrument meteorological conditions (IMC), at 8,000 ft mean sea level (msl), and at 156 knots ground speed, an air traffic controller updated the altimeter setting via the radio. The pilot acknowledged the call, and his voice sounded normal. No other radio transmissions were received from the pilot, and he made no distress calls. Radar data showed that the airplane then began a descending right turn, that the airspeed increased slightly through the turn, and that the airspeed then suddenly decreased to 61 knots as the turn radius decreased. The airplane had turned about 270 degrees and descended to an altitude of 4,900 ft msl before radar contact was lost. A witness reported seeing the airplane coming out of the low clouds (the cloud ceiling was about 700 to 800 ft above ground level) in a nose-down descent with the wings about level before it disappeared behind a tree line. Examination of the airframe and engine did not reveal any preimpact anomalies. Data from the on-board recoverable data module (RDM) showed an anomaly in the electrical bus voltages, autopilot mode changes, and a momentary dropout of the pilot transmission frequency. Specifically, the RDM data showed that the airplane was flying at 8,000 ft pressure altitude at an airspeed of 148 knots with the autopilot on when alternator 1 dropped offline followed by alternator 2 dropping offline; the airplane is equipped with a two-alternator, two-battery, 28-volt direct current electrical system. About 3 seconds later, the autopilot disengaged. The RDM data confirmed that the airplane then entered a descending right turn with the airspeed increasing slightly and then suddenly decreasing. The data further showed that the airplane then rolled right and that the pitch attitude sharply decreased. The airspeed then increased, and both the roll and pitch reversed back toward wings level before the data ended. The bolster switches for alternator 1, alternator 2, battery 1, battery 2, and the avionics were cycled "on" and "off" on an exemplar airplane to determine what may occur in the cockpit when the bolster switches are manually moved. Seven sets of various electrical bolster switch activations and deactivations were documented. During the tests, bolster switch operation on the exemplar airplane yielded recorded data similar to the accident flight data; however, the reason for the recorded electrical anomalies could not be determined because the RDM does not record the physical position of either the bolster switches or circuit breakers. The data did reveal that the airplane began its deviation off course and its subsequent descending right turn in IMC conditions a few seconds after the alternators went offline and that the pilot was likely attempting to troubleshoot the electrical anomaly with bolster switch activations and deactivations as the airplane descended and turned through the clouds. The electrical anomalies likely distracted the pilot and led to his subsequent loss of airplane control. The witness's statement that the airplane came out of the low clouds with the wings about level and the RDM data indicate that the pilot may have been able to recover from the turn and rapid descent to some degree but that there was insufficient altitude for a full recovery.

Quoted verbatim from the NTSB record.