HISTORY OF FLIGHTOn October 17, 2022, about 1140 eastern daylight time, an experimental, amateur-built Aventura II airplane, N32856, was substantially damaged when it was involved in an accident near Miramar, Florida. The commercial pilot and passenger were fatally injured. The flight was operated under Title 14 Code of Federal Regulations Part 91 as a test flight. According to air traffic control information obtained from the Federal Aviation Administration, the pilot contacted North Perry Airport (HWO), Hollywood Florida, ground control at 1135 for taxi instructions to the active runway. The ground controller provided instructions to taxi to runway 10R and a subsequent takeoff clearance for right traffic off runway 10R, which the pilot acknowledged. At 1138, the local controller instructed the pilot to extend on the downwind leg, and the pilot acknowledged. There were no further communications from the pilot. At 1139, the controller advised the pilot that the airplane’s transponder was not working and subsequently requested radio checks at 1140 and 1141, but received no response. A witness who was at HWO, stated that the pilot and passenger visited his hangar on the day of the accident to borrow a screwdriver. He reported that their airplane had experienced problems in the days before the accident, but he was unsure of the exact issues. On the day of the accident, he observed that the airplane’s engine "did not sound right" before departure. A representative of the airplane manufacturer reported that the pilot and passenger were employees of the company. The passenger had contacted him several days before the accident, stating that the engine control unit (ECU) malfunctioned during flight, causing a power loss. The power returned shortly after, and the airplane landed safely. The incident was reported to the engine manufacturer, and troubleshooting began. On the day of the accident, the pilot and passenger were still troubleshooting the engine and the flight was intended as a test flight. The engine manufacturer confirmed that the passenger had contacted him regarding the engine power loss and provided a video showing the engine shutting down and restarting in flight. The engine manufacturer suggested several checks and noted that the airplane required updates. WRECKAGE AND IMPACT INFORMATIONThe accident site was located about 1 mile south of HWO, with the airplane coming to rest partially on the roof of a residence. All major flight components were located at the main wreckage site. Flight control continuity was confirmed to all primary control surfaces. Both wings were impact-damaged but remained partially attached to the airframe. The engine, which showed no signs of impact damage, remained attached to its mounts. Although the engine controls were present, they were shifted due to impact forces. The instrument cluster was found separated within the wreckage. There was about a half tank of fuel in the fuel tank, with no breaches observed in the fuel lines or fuel tank. Engine Examination and Test Run Postaccident examination of the engine revealed no visible damage; however, all wiring connections were found to be loose consistent with pulling during the accident sequence. The battery terminal wires were also loose but observed to be stretched consistent with damage from the impact. The operator had previously communicated that the passenger was troubleshooting engine issues and planned to inspect the wiring connections before the flight. A subsequent engine test run was conducted to replicate the reported power loss. The engine initially started and ran normally but sustained a power loss after several minutes of operation. Further inspection revealed that the ECU wiring was not in compliance with the engine manufacturer’s Service Bulletin Viking 110 Wiring Upgrade, which recommended removal of the "computer on" selector switch to allow operation on a single ECU. The service bulletin specified, "Remove ECU select switch to operate on a single computer only. Locate the select switch on the panel, remove the three wires labeled ECU 1, ECU 2, and SELECT. Disregard and tape the ECU 2 wire. Solder ECU 1 and SELECT together and protect with shrink tubing." After separating the ECU wiring to isolate each ECU, the engine operated normally. MEDICAL AND PATHOLOGICAL INFORMATIONAn autopsy of the pilot was performed by the Boward County Medical Examiner. According to the autopsy report, the cause of death was blunt force injuries, and the manner of death was accident. Toxicology testing performed at the FAA Forensic Sciences Laboratory found no ethanol or drugs of abuse. TESTS AND RESEARCHOnboard Image Recorders A GoPro Max and GoPro Hero7 compact digital camera were recovered from the accident site and forwarded to the NTSB Vehicle Recorders Laboratory, Washington, DC. No data was recovered from the microSD card from the GoPro Max camera. The microSD card from the GoPro Hero7 did not capture that accident flight; however, it captured a flight on October 11, 2022, when the airplane experienced a sudden loss of engine power on final approach while the throttle was stationary. The throttle was briefly moved to idle and moved forward again. As the throttle moved forward, the engine power returned. The airplane landed on runway 10R at HWO without further incident. The GoPro Hero 7 also captured a flight on the day of the accident, which began at approximately 1128 and lasted 7 minutes and 50 seconds. The airplane was airborne and on final approach for runway 10R at HWO. A red light to the left of the Viking View Engine Information System (EIS) was flashing for the duration of the recording, but the label for the light was not legible. The engine sounded normal and appeared to respond appropriately to throttle lever movements. The airplane landed about 1130, the airplane subsequently stopped on a taxiway and the pilot and passenger had a discussion for about 2 minutes as they both pointed at the Viking View EIS. Their conversation was not audible due to the ambient engine noise and the values displayed on the screen were not visible. At 1131:45, the pilot started to taxi back to runway 10R for departure. At 1133:55, the airplane was holding short of runway 10R. At 1135:43, the pilot turned off the camera and the video ended.