cbdb>Is there a schematic of the OPS?
Yes.
This circuit I am about to show you is an older one, created the above plot.
Certainly not the simplest, nor best way I could have drawn for you today.
Magic is simply two independent single ended class A amplifiers, any type will do.
Offset one amplifier to track one emitter drop above the desired output voltage.
Offset the other one to track one emitter drop below the desired output voltage.
A quiescent current is set by resistors spanning the gap between these two offsets.
If that resistance is linearly resistive, then only linear class A crossing can be the
result... Power wasted at quiescent is exactly the same as at full output swing.
But if that resistance rides on the curve of a non-linear diode.... You can shape it.
The Quiescent current can be far less than full output swing. But still no transistor
is ever turning off.
This is my own circuit, nothing to do with Yamaha's. And possibly irrelevant.
We still at this point know nothing about "Hyperbolic" to guess what that means?
But the Yamaha advertised description (so far) fits how mine behaves to a tee.
Is Yamaha using the same trick, or I've spun completely off a different tangent???
The MX-10000 circuit looks like a bridge amplifier without overall NFB and consists basicly two operational amplifiers and two power buffers - one pure class A (low voltage supply) and one class AB (high voltage supply). The first stage is a complete independend voltage gain amplifier (55 times).
The output stage of this one drives
1) the class AB power buffer (the output of class AB power buffer goes to the positive speaker terminal) and
2) the inverted input of a summing amp about a 6K8 resistor (non inverted input of this summing amp goes to GND)
The output of this summing amp (ERCO amp = Error Correction amp) drives the second power buffer, that is the pure class A buffer (the output of this buffer goes to the negative speaker terminal and also to the inverted input of a summing amp about a second 6K8 resistor
So I have a mixed signal between an (more or few) undistorted signal version from the voltage gain stage output and a distorted (and inverted) signal version from the neg. speaker terminal at the inverted input from the summing amp.
I cannot explain, why Yamaha calls this technology so than the head line of this thread.
If I replace the pure class A buffer through a "ZEN" or other simple single ended topology and the class AB buffer through an simple power buffer with 30-50 mA idle current, I get interest possibilities for diy projects (together with appropriate OP amp models for the voltage gain and ERCO amplifier stages).
This goal from Yamaha was the same than that of Quad's current dumping topology and the main different to Quad 405 power amplifier is the serial mode of the correction amp - Quad choises a parallel mode like the EDWIN amp projects from the Geman magazine "elektor" arround 1970 - 1975.
If I have remove the WIN32/ConflickerC worm from my own personal computer I will upload the basic schematic of Yamaha MX-10000 for better understand.