The certified flight instructor (CFI) stated that the purpose of the flight was instruction. He conducted a preflight of the helicopter and did not note any problems with the helicopter. Start up and run up was completed by the checklist with no problems. They lifted off at 1745, hovered for a few minutes and landed again. Shortly thereafter they departed to the Argyle Airport (NY02), Argyle, New York, for training. The take off and climb were normal, and they leveled off at 1,500 feet. Approximately 3 miles from NY02 the engine began to run rough and there was a drop in power. The dual student transferred the flight controls to the CFI. The CFI decided not to continue the flight to NY02 and initiated an approach to a field behind a house. On approach the engine stopped, and an autorotation was initiated to the field. The touchdown was "smooth" and level, with a slow forward speed. The helicopter slid approximately 10 feet and rolled over onto the left side. Examination of the helicopter by an FAA inspector revealed no flight control anomalies. The left skid sank into soft soil before rolling onto the left side. The pilot reported that the helicopter was fine until the engine power began decrease. The examination of the engine revealed that the number 4 rocker arm was unsecured and the spring retainer was damaged. Further examination revealed that the number 4 exhaust valve had stuck in the open position. In a review of the logbook it was discovered that the mechanic had made valve lash adjustments to all the cylinders on September 16, 2006, at an engine tachometer time of 17.4 hours. The engine tachometer time at the accident site was 18.1 hours. The engine had .7 hours before the exhaust valve failed. The cylinder heads were sent to Rotorway International for examination of the exhaust and intake valves. According to the director of R&D, examination of the heads revealed that the exhaust valves were found to be "sticky in their guides". The guide size measured at the low side of the tolerance before cleaning. After cleaning with a wire brush, the fit of the valves in the cylinder heads and the tolerance of sizes were acceptable.