ERA14LA311

The pilot took off from his departure airport on a cross-country helicopter flight. The pilot reported that, when the helicopter was about 1 mile north of the airport and near a group of high-powered transmission broadcast towers, the engine started running roughly and that he felt an uncommanded decrease in engine power and movement of the twist grip throttle. The pilot initiated an autorotation, and the helicopter landed hard, which resulted in substantial damage to the tailboom. Postaccident examination of the engine revealed no mechanical anomalies that could have precluded normal operation. According to a flight instructor based at the departure airport, he had encountered a similar uncommanded decrease in engine power on multiple occasions while flying east of the same broadcast towers in the same make and model helicopter. During each encounter, the pilot felt the movement of the electronic governor through the throttle twist grip and was able to override the governor by advancing the throttle. Further, Federal Aviation Administration research revealed that high-power broadcast AM radio transmitters likely affect the governor circuit on this helicopter and can result in governor control that reduces the engine speed regardless of the pilot's power input. Thus, it is likely that radio frequency interference increased the engine speed signal frequency as the pilot flew the helicopter near the transmission towers. The governor controller then engaged the governor motor to decrease engine power, which the pilot perceived as a loss of power. About 15 years before the accident, the manufacturer had issued a safety notice to address previous reports of radio interference. The safety notice stated that flying near broadcast towers may result in "erratic operation of the governor," which may cause the governor to roll the throttle to the idle or open position. The notice also instructed pilots to keep one hand on the collective and throttle and to refrain from flying near broadcast towers. The pilot reported that he was familiar with the safety notice but that he had not read it since he purchased the helicopter about 7 years before the accident. The pilot should have been able to recognize that the engine was still producing power and overridden the governor by moving the twist grip throttle, and he should have been able to maintain rotor rpm while executing the autorotation, which would have resulted in a softer landing; however, he did not do so.

Robinson Governor According to the helicopter manufacturer, "The Robinson R22, R44 and R44 Raven II series governor controller determines engine speed from unpowered magneto points which produce a signal that oscillates between high impedance (open circuit) and ground (0VDC). With the engine speed known the governor calculates the appropriate throttle response and engages a motor that is tied to the throttle control to the engine. The R44 Raven II (IO 540 engine) governor controller output to the governor motor is a proportional controlled pulse width modulated signal. It operates at 300 Hz to 500 Hz. When the engine rpm is within .875 percent rpm of the centerpoint, it operates on a 20 percent +- 3 percent duty cycle. When the engine rpm is more than 5 percent rpm from the centerpoint, it operates on a 100 percent duty cycle. Between .875 percent and 5 percent rpm from centerpoint, the duty cycle increases proportionally." FAA Spectrum Engineering The helicopter and governor manufacturer provided circuit diagrams to FAA Spectrum Engineering in Washington D.C. for further review. According to the FAA's interpretation of the circuit diagrams, the input signal that is destined for the governor picks up the engine rpm signal through a pair of dry contacts in one of the two magnetos. The signal is also applied to the tachometer circuit. A twisted pair connects the two contacts to the governor controller; one of the pair goes to ground and the other is routed to the controller. The controller converts the pulse train to a voltage that represents engine speed. This voltage is then compared with stored reference voltages, which generates an error signal. The signal subsequently drives a circuit that causes a friction clutch motor to increase or decrease the engine speed to maintain the governor's desired rpm range. The FAA's research revealed that the governor input circuitry was not designed to reject a high power external radiated signal. The input wiring's grounded twisted pair includes an overall braid that provides some attenuation to RF, but is not effective against certain electric fields. The assembly offers no protection from magnetic fields, which makes the wiring harness susceptible to transformer effect. Within the controller, there is a capacitor with a 1000pF (picoFarad) feedthrough that acts as an initial RF filter. This does not protect against medium and strong high frequency RF fields and, although low risk, is not effective at lower RF frequencies. The circuit acts as an amplitude demodulator when RF is introduced into the wiring of the governor controller. This results in a signal that artificially increases the signal frequency that represents engine speed (adds pulses to the pulse train). Once the signal increases to more than 5 percent rpm from the centerpoint, the governor calculates the appropriate throttle response and engages a motor to decrease engine power. At the time of the accident, there were four 100KW towers thattransmitted at a frequency greater than 3 MHz near JWN. A search that included towers with emission A3E amplitude modulation, power greater than 5KW, and a frequency between 2MHz and 30MHz revealed a total of 49 high power stations in 8 physical locations within the United States. RF Interference Guidance In April 1999, the manufacturer issued Safety Notice (SN-35) to address previous reports of radio interference. The Safety Notice stated that flying near broadcast towers may result in "erratic operation of the governor", which may cause the governor to roll the throttle to the idle or open position. The Safety Notice also instructed pilots to keep one hand on the collective and throttle and to refrain from flying near broadcast towers. According to the pilot, he was familiar with SN-35 and added that he had not read the notice since he purchased the helicopter in 2007. According to the pilot, his preflight research did not produce any relevant NOTAMS on the day of the accident flight. The four-seat, single main rotor, single-engine helicopter ,was constructed primarily of metal, and manufactured in 2003. The helicopter was powered by a 205-horsepower Lycoming IO-540-AE1A5 series reciprocating engine. The collective control was equipped with a "twist grip" throttle, and a 24 volt, dual wound, DC reversing motor electronic governor system which was installed to maintain engine rpm by adjusting the throttle when the collective control was moved. The Pilots Operating Handbook (POH) stated that the governor was "easily overridden by the pilot," was only active above 80 percent engine rpm, and could be switched on or off via a toggle switch on the end of the right-seat collective control. Review of maintenance records revealed that the helicopter had been operated for about 31.2 hours since its most recent 100-hour inspection, which was completed on April 1, 2014. At the time of the inspection, the airframe and engine had accumulated about 1,658 hours of operation. On June 24, 2014, about 1235 central daylight time, a Robinson R44 II, N544BS, was substantially damaged during an autorotation near John C. Tune Airport (JWN) Nashville, Tennessee. The commercial pilot and passenger were uninjured. Visual meteorological conditions prevailed, and no flight plan was filed for flight, which originated from JWN and was destined for Decatur Airport (DEC), Decatur, Illinois. The personal flight was conducted under the provisions of Title 14 Code of Federal Regulations Part 91. The pilot made an intermediate stop at JWN to obtain fuel during a flight from Florida to Wisconsin. After departure, the pilot turned the helicopter north on course to DEC, towards a group of broadcast towers. About 1.5 nautical miles from the airport and about 400 feet above ground level, the pilot initiated a climb to clear a hill covered by tall trees. The helicopter was traveling about 70 knots. The engine began running "very rough" and the pilot felt an uncommanded movement of thethrottle twist grip, which was followed by an abrupt loss of altitude. He lowered the collective, the low rotor rpm horn activated, and the low rotor rpm light illuminated. The pilot then turned west to initiate an autorotation. He glanced at the engine/rotor rpm gauge and noticed the engine was producing about 70% rpm and the rotor rpm was in the green; however, the engine continued to run rough. The pilot initiated an autorotation to a nearby highway. The helicopter landed hard in a level attitude and the main rotor contacted the tail boom, resulting in substantial damage. The pilot reported that he "didn't have enough rotor energy left to put it down gently." He added that he did not increase throttle at any time during his descent, but did decrease engine throttle when he initiated the auto rotation. According to a helicopter flight instructor who taught at JWN, he encountered a similar uncommanded throttle movement about two months before the accident while flying a Robinson R44 helicopter east of the same group of broadcast towers. The instructor added that during this event, the low rotor rpm light activated, and he lowered the collective and entered an autorotation. Moments later he noticed that he still had engine power, increased throttle, and was able to return safely to the airport. The instructor believed the event was the result of sporadic governor operation and decided to fly near the towers again in an attempt to replicate the same uncommanded throttle motion. The instructor was accompanied by another pilot who assisted him while he repeatedly flew the same route. During each pass near the broadcast towers, the low rotor rpm light activated and the governor-actuated twist grip throttle was felt moving in the throttle off direction. The pilots were able to override the governor each time by increasing throttle. The 1253 automated weather observation at Nashville International Airport (BNA), Nashville, Tennessee, located about 11 nautical miles east of the accident site, included winds from 190 degrees at 13 knots and gusts to 21 knots; visibility 10 statute miles; few clouds at 2,900 feet, broken ceiling at 3,500 feet; temperature 29 degrees Celsius (C); dew point 21 degrees C, and an altimeter setting of 30.16 inches of mercury. The pilot held a commercial pilot certificate with ratings for helicopter, airplane multi engine land, and instrument airplane. He had accumulated about 10,197 hours of total flight time, of which 1,678 hours were in the accident helicopter make and model. His most recent second class Federal Aviation Administration (FAA) medical certificate was issued on January 23, 2014. According to information provided by an FAA inspector, the helicopter came to rest about 1.5 miles northeast of JWN near a high-power radio station. The helicopter remained intact with the exception of the tailboom, which was fractured about 3 feet forward of the tail rotor. Control continuity was traced from the flight controls to thecyclic and collective pitch links. Tail rotor continuity was traced from the control pedals through the fracture of the tail boom, to the tail rotor pitch change links, which remained attached to the tail rotor. The pilot did not report any preimpact anomalies or malfunctions with the control system that could have precluded normal operation. Postaccident examination of the engine was performed at a recovery facility under the supervision of an FAA inspector. The engine ran smoothly and continuously during an initial test run. A compression check revealed that all cylinders were operating in the normal range. The lower spark plugs were examined and found to display normal wear. Both magnetos were rotated by hand and produced spark at their terminal leads. A cylinder borescope inspection revealed normal cylinder wear and a uniform heat pattern on the exhaust valve faces.