What is wrong with negative feedback?

"These amps are high damping factor which matches well to the OHMs, so I suspect there is NF applied, but not sure"

Mapman - Not only NFB is applied but it is mulitiple (two) called Mulivariable Enhanced Cascade Control. It is in the Karsten Nielsen doctorate work available on the web (Icepower). Feedback doesn't have to be deep since class D amp has inherently low output impedance even without feedback. Speaker is always connected to +Vs and GND and only direction changes (very low resistance Mosfet Bridge). Output impedance increases with frequency but DF is still about 1000 at 1kHz.

A few years ago, we had a regular and good contributor (be nice if he came back) Ar t an amp designer, suggested that in his typical ss designs he would try to keep NF to a minimum, but in his switching amps though there was much more NF, it didn't seem to be an issue. I wonder if it was due to the NF in the switching amps occurring at a faster rate?

Jamesgarvin I had a pair of Manley ref 440/200 monos that had adjustable slopes and feedback to custom taylor the sound;there were noticable differences but I did not consider it annoying just different.

Unsound - Feedback in Icepower amps is called Multivariable because it consists of two feedbacks controling voltage and time.

I suspect that feedback in class D is less evil for few reasons:

Amp cannot become unstable (oscillate) since it is already oscillating.

Response can be very fast limited only by Mosfet's max current and resistance (sort of Hysteretic converter)

Voltage feedback is shallow because duty cycle is more linear than transistor characteristic(less feedback required) and also because "time feedback" already corrected most of nonlinearities. I suspect that voltage feedback is helping to deal with load regulation. Early class D amps (Tact) had no voltage feedback at all and were sensitive to load conditions.

Momentary saturation of output stage (charge trapped at the junction) that happens in bad cases of class AB is irrelevant since time (duty cycle) and not the voltage is analog quantity.

When I first heard the IcePower amps in my setup, the difference in the nature of teh sound from anything I had heard prior was most startling. It was most apparent in that major volume adjustments seemed to have comparably little effect on volume, though the SPLs were clearly increasing as expected.

My observation of how the IcePower amp sounds as volume is increased seems consistent to me with what Ralph describes as the benefit of not using feedback and loudness cues. Most really good systems I have heard over the years tend to behave this way more or less I believe.

At the same time, Icepower applies negative fedback clearly. So I have to conclude that the IcePower amp has achieved a lot of teh benefits that Ralph attributes to no feedback in its specific implementation.

If so, another reason why Icepower/Class D can be considered a major innovation in home audio along with the other advantages.