The aircraft was being flown from Adelaide to Dubbo in accordance with instrument flight rules. At 1921 EST, the pilot reported over Mildura maintaining 9,000 ft, and estimating abeam Griffith at 2026 on a direct track to Dubbo.
The weather in the area was clear, with no restrictions to visibility, and scattered cloud at 30,000 ft. Sky conditions were dark, with no moon.
At 1958 the pilot made a routine frequency change, and 8 minutes later advised Melbourne Flight Service (FS) there had been a loss of vacuum and that he was returning to Mildura. In response to enquiries from FS the pilot advised that he had 'electric backups' and felt it safer to return to Mildura. He also confirmed his approach and landing would not be affected, and that his estimated time of arrival would be 2029. At 2007 FS asked the pilot for his approximate distance from Mildura. The pilot asked FS to repeat the request, but subsequently failed to reply. Further attempts to contact the pilot were unsuccessful. The last recorded radio transmission from the pilot was at 2007:49.
An uncertainty phase was declared after communication and ground checks failed to establish the location of the aircraft. A local resident reported seeing the lights of an aircraft shortly after 2000, and then hearing the sounds of an impact. The wreckage of the aircraft was subsequently found some hours later. The accident was non-survivable.
An examination of the wreckage indicated the aircraft had impacted the ground at high speed, in a steep nose-down attitude, consistent with loss of control. With the exception of the vacuum system, the aircraft was considered to have been capable of normal operation prior to impact.
The aircraft was equipped with an attitude indicator and a directional indicator, each reliant on air-driven gyroscopes. An electrically powered turn co-ordinator was also fitted. Examination of the attitude indicator showed evidence of a witness mark consistent with the gyro-rotor being stationary at impact. The turn coordinator gyro-rotor was recovered and showed evidence of rotation at the time of impact.
The engine-driven vacuum pump and drive coupling were recovered from the wreckage. The impact resulted in separation of the pump body from its base. Only a few large pieces of the rotor and vanes were recovered. The frangible drive shaft coupling had sheared at some time prior to impact. A specialist examination considered that either the carbon rotor, or one or more of its vanes, had failed, resulting in pump seizure and consequent shearing of the drive coupling.
An entry in the aircraft logbook indicated that the vacuum pump was installed in September 1991. From that date, until the last periodic inspection in August 1997, the pump had operated for some 1,248 hours. No evidence was found of vacuum pump replacement during that period.
The Mooney 20J Service and Maintenance Manual recommends that the schedule for the primary vacuum pump replacement be either on condition or at 500 hours, and at engine overhaul. The Civil Aviation Safety Authority provides no additional requirements regarding maintenance of the vacuum pump.
The pilot held a Private Pilot licence with a valid medical certificate. His command instrument rating had expired 3 days prior to the accident. Although no evidence could be found of the pilot having flown in instrument meteorological conditions in the previous 12 months, the pilot had conducted a night flight six weeks prior to the accident. A passenger on that flight reported they did not encounter cloud.