Robinson Helicopter Company R44 II N3254E accident near Van Nuys, California

Probable cause

An improper repair of the right magneto's tachometer and governor signal wire, which failed during cruise flight and initiated a simultaneous malfunction of both the engine tachometer and governor control system, which resulted in erroneous and contradictory information between the engine tachometer and main rotor tachometer that led the pilot to perform a forced landing. Contributing to the accident was the lack of guidance on how to address the simultaneous failure or malfunction of both the engine tachometer and governor control system.

NTSB analysis

**This report was modified on July 26, 2022. Please see the public docket for this accident to view the original report.**

The commercial pilot reported that, during a local sightseeing flight over a densely populated area, he heard an unusual sound and saw that the engine tachometer was indicating a higher-than-normal engine rpm. He reduced engine power and attempted to manipulate the helicopter controls to align the engine and rotor speeds to no avail. Subsequently, the rotor rpm began decaying, and the pilot assumed the helicopter had experienced a catastrophic drive failure and chose to conduct an autorotation, during which the helicopter landed hard. The airframe sustained substantial damage.

Postaccident examination of the helicopter and engine did not reveal evidence of any mechanical malfunctions or failures that would have precluded normal operation, and witness signatures indicated that the engine was producing power at the time of ground impact. However, a wire within the right magneto, which provided a signal to both the engine governor and engine tachometer on the instrument panel, was found detached at a solder joint where it had been soldered to the connector rather than crimped. Further, neither the wire nor its connector was aviation-grade. The investigation could not determine when the modification to the wire was made.

Review of onboard video and a spectrographic analysis of the recorded audio revealed that the engine rpm initially decreased but then appeared to be restored a few seconds later while the helicopter slowed. For about the next 40 seconds, the engine continued to operate at the same rpm and the helicopter continued to slow and gradually descend. The engine rpm then quickly decreased, and the helicopter rapidly descended in a manner consistent with the pilot initiating an autorotation.

The detached wire would have resulted in both the tachometer and governor not receiving a correct engine rpm signal. As a result, the engine governor would not have been able to make appropriate adjustments to engine power as demand changed, and the pilot would not have had a way to gauge engine power and make corrective manual throttle inputs. The pilot's statement that the tachometer rpm increased does not match a scenario where the wire completely detached. However, it is possible that the event was initiated when the wire remained partially connected and thus created a spurious increase in signal frequency to the tachometer and governor due to vibrations. Thus, in this scenario, the tachometer would read higher than normal and the engine rpm would decrease as the governor tried to compensate for the high engine rpm signal.

The Helicopter's Operating Handbook provided separate procedures for the failure of the engine governor and tachometer but not for the failure of both. Because the engine governor control system and engine tachometer both share a common speed signal source and thereby a common failure point, this signal wire failure left the pilot with a confusing set of cues and no definitive procedures for corrective action. In addition, the helicopter's low rotor rpm warning horn was not working, but the low rotor lamp was working. This failure would have resulted in another set of contradictory cues for the pilot to process, further compounding the confusion he was likely experiencing during the high-stress event.

Text from the NTSB. Open the NTSB docket for WPR17LA159 for the full record.