www.wavebourn.com

Well;

it was the time to make some SE amp finally, to test my theory that SE has minimum of possible audio distortions on decaying sound, hence sounds more "airy". However, there is another problem wit them; usually as soon as they sound nice on low volume, people tend to build them with triodes. Pretty low output power forces to use them on higher volume that would drive them out of headroom on peaks. Since vacuum triodes without additional feedback cause pretty soft clipping without nasty sounding distortions, it is the usual way to build them: 2 or 3 triode stages depending on tubes, no or little feedback. As the result, they have reputation of amps good for audiophile recordings with one solo and few instruments, but not good for orchestra and other music that does not tolerate pretty high intermodulation distortions caused by relatively high low order errors of triodes with no feedback.

One more thing that adds to negative reputation of such amplifiers, it pretty high output impedance. A first, it causes sloppy bass, since can't damp modern speakers, and requires enormously expensive output transformers to reproduce the whole range of audio frequencies.

Can it be done differently, first of all, to preserve "airy" sound of SE amplifier, but without drawbacks, like high intermodulation distortions of loud sounds and weak speaker control?

Yes!

So, I after extensive research, I selected tubes. GK-71 for the output stage, and 12HL7 as a driver.

GK-71 is a quite powerful directly heated pentode, with graphite anode, that can dissipate 135W according to datasheet. It was manufactured for Soviet military transmitters, used in critical application, and is quite reliable. Here is the picture attached that shows how linear it is!

For the driver I selected 12HL7 tube. It was designed for video amplifiers, so is very linear, and has high Gm.

21 mA/V versus 15 of 6P15P.

As the result, only 2 stages were needed for 45W output driven by 0.7V input.

I am still getting questions, why pentodes, why nested feedbacks, why not triodes with zero feedback?

Here you go: 2-stage pentode amp with nested feedback loops
Drives 4-way speakers with complex crossovers, with heavy woofers that rely on electrical damping.
https://www.youtube.com/watch?v=skCCd4O-fVQ&t=16s

Similar boutique amps with GM-70 output (the same like GK-71, but castrated down to triode) produce 20W usually (twice lower), and can't drive such speakers properly. They require several times more expensive horn speakers.

Here is an example to compare:
http://lampizator.eu/Fikus/GM70_AMP.html

Here is some theory:
http://www.clarisonus.com/Archives/Tube ... 0Tubes.pdf

Otto Schade back in 1938 wrote an article about how to use in audio power pentodes and beam tetrodes. He explained the theory and proposed one practical example how to apply local feedback to output tube, through secondary winding of an interstage transformer. You can't use Schade feedback (like in his article) with nested loops, because each transformer has parasitic capacitances and loss inductance that limit feedbacks due to phase shifts and frequency dependence. But the principle is the same, and there are several ways to apply local feedback around output tubes.

If you want an example of nested feedback loops, look at SpaceX landing. It cant be possible with a single global FB loop. In order to make it happen, you need to solve differential equations with multiple variables, to optimize resulting function according to given criteria. One constant's value is wrong, and you get either oscillation, or insufficient end result. But proper balance of nested loops allows to control end results fast and precise.

What is common between my pentode amps with nested feedbacks and triode amps with zero feedback? They are both single ended, and both preserve "air" due to the essence of SE amps the best.
What is the difference?
Zero feedback triode amps rely on internal resistance of tubes and their distortions. They can't drive properly speakers with complex crossovers, they can't drive woofers that need electrical damping. Their distortions even though of low order, raise constantly, so they sound nice when you listening to some solo on background of few instruments, in the style of Diana Krall, but when listening to bands, high level of distortions, even when they are low order, cause unbearable intermodulations that turn sounds of orchestra into some sound porridge.
My amps deliver more power from tubes of similar size, have low distortions up to full power, so can reproduce any music, with more power, more headroom, drive any speakers, but due to properly balanced feedback loops, clipping if happens is soft, like in case of no feedback triode amps.

The prototype during construction...