I've never had an engine failure on a single engined helicopter, thank god, but one of my instructors did once. I see that there has been some discussion as to the best method of entering autorotation. His biggest point was that it will not happen the way you practice it.
When the nasty instructor says "practice engine failure, GO", you smoothly lower the lever and keep the aircraft straight with pedal. In reality, he didn't immediately realise that the engine had failed. The first thing he noticed was that his Huey was flying sideways due the pedal he was using to counteract the tourque that the engine wasn't producing anymore. He was still trying to figure that one out when the low NR warner went off. Luckily the Huey has legendary amounts of inertia in the blades and he was able to recover.
I accept that he was flying a turbine and that pistons are less likely to cut suddenly, but even so, it is impossible to truly simulate an engine failure into autorotation.