MrR
I think you need to read my previous expanation again - the problem is not with high density altitude (ie the Lama pilot) its quite the opposite, low density altitude below -3500'. (note that low density altitude = high air density! confusing!)
The problem is an old one that surfaced on the AS332L back in 1982. I think the primary cause is that the flight controls are "anchored" at one end (the pilot's end) by the AP Hydraulics etc, at the other end they are attached to the servos that are mounted on the main transmission. The main transmission is mounted on the fuselage by a slightly flexible mounting system. The problem is that forces generated by changing the disc attitude (ie cyclic input) cause the transmission to move slightly, which changes the distance from the servo to the AP hydraulics thus effectively makes a slight cyclic control input, which in turn modifies the disc attitude, which in turn causes the transmission to move slightly etc etc. If this "control loop" has a gain of more than unity, it rapidly diverges (ie the forces and movements mentioned above rapidly increase in magnitude). The natural frequency of this (ie the frequency where the phase shift is 180 degrees) was about 12 Hz for the 332L and 14 Hz for the 225.
It was fixed on the 332L by adding a bellcrank and rod to the forward servo control run. The other end of the rod was attached to the MGB and had the effect of nulling out the effect of transmission movement on the forward servo position.
On the 225 this reappeared due to the more powerful rotor system. Initial attempts to add more bellcranks on the other servos etc didn't seem to work, but reducing the Nr did work and that is the line EC have taken, for the time being at least.
So with current FADEC version the Nr is reduced according to the measured density altitude. Only problem is that its not easy to accurately measure the density altitude at low speed IGE - exhaust gas recirculation can artificially raise the temperature around the probes and then the system could calculate that the density altitude is higher than it is, increase the Nr and go into the oscillation. So a large margin has to be built in that causes the Nr to be lower than optimal - so there is less thrust in the hover etc.
The new FADEC version will have additional temperature sensors in the tail - the system will use the lower of the sensors to calculate the DA and having the more appropriate Nr allows them to have a minimum Nr lower than 100%, hence moving the limit from -3500' to -6000'.
HC