New criteria for estimating the power
Measurement and analysis techniques in RAA are constantly evolving.The main task of numerous measurements is not to find the best product in the world. We are developing the ability to compare products and are searching for parameters that can directly tell about the quality or features of the sound of the product.
One of the important parameters of the amplifier is its output power.
First, the power was determined from the point of the amplifier's transition into clipping. But we are interested in comparing amplifiers not only in terms of level when we hear an obvious wheeze. We are interested in comparing the power of amplifiers when they operate in high quality mode.
In many amplifiers, the transition point to clipping is the change in the character of the spectrum in the form of rapid growth of high order harmonics. There are amplifiers whose harmonics of high order grow with increasing signal level not too fast. To differentiate between different modes at the end of 2015, three variants were used: "clean", "optimal" and "maximum" on the verge of clipping.
When there are numerous tests of various amplifiers, different dependencies become apparent. from the pure mode to the optimal mode and the clipping mode can be described by several values of the voltage and current at which the spectra change in character.
Previously, voltage, current and power curves were constructed at different times with an approximate representation of what kind of spectrum belongs to the pure, optimal and maximum mode. The power results were not quite comparable for comparisons.
Reports in the RAA contain full data on the spectra of the amplitudes. This allowed us to determine more precise criteria for determining the types of spectra characters and to update the graphs of the maximum voltage, current and power. This is a big plus of the RAA approach with the availability of detailed reports.
In the new methodology for plotting, instead of the three modes, four modes are defined.
Clean Mode - Class A
Previously, it was a mode with a low level of distortion and, if possible, no harmonics of high orders. Now, when determining the pure mode, a mask is used, which strictly limits the amplitudes of the harmonics. The pure mode is now such a mode, according to the spectrum of which we can say that the amplifier works as in class A. This is the absence of high-order harmonics and a significant decrease in the amplitudes of the harmonics toward higher orders.
An exception to this rule can be devices with their own DAC and analog volume control, where higher-order harmonics are produced not by an amplifier, but by a DAC.
Some amplifiers may not have a clean mode. In this case, the minimum values are fixed. Our goal is to compare the amplifiers, and for the general standardization of the reports, the amplifier must have all modes defined.
With standardization, it will be possible to open a section with direct parameter comparisons.
AB class mode
This mode is the development of an optimal mode close to the pure mode. In this mode, the values of the harmonics are small and decrease in amplitude toward higher orders. Spectra at the same time look like in AB mode.
This mode is also the development of the optimal mode, but already close to the clipping mode. This is a higher level of harmonics, but similarly decreasing in amplitude in the direction of higher order.
This is the mode before the point of the amplifier's transition into clipping. In this mode, the harmonics do not decrease or reach a high level.
The names of modes A, AB and B do not mean that the amplifier works in this mode. They give an analogy with modes for assessing subjective quality
When you separate each amplifier into four modes, you can compare the power in comparable quality levels. This does not mean that the loudest amplifier in pure mode is the best. Each amplifier has its own peculiarities of sound. Additional information can be given by the spectra of intermodulation distortions and distortions on a multitone signal.
We are trying to find such criteria for determining the power, which will allow the most correct comparison of devices among themselves. Our goal is not to convince any manufacturer of overvalued data.
Should the power characteristics coincide with the manufacturer's passport data?
No, because each manufacturer outputs the data in accordance with their criteria.
What is the value of the power of the four "standard" or "right"?
They are all right, each for their own mode. There is no single power value for the device. Focus on the one that suits your tasks and hardware class.
I believe that power should be determined by other criteria!
All the spectra are laid out in the reports and you can determine the power according to your criteria. You can suggest and discuss your methodology on the forum
or social networks facebook