It is obvious. I explained it many times, but in short: SE can sound cleaner per output power, with the same dissipation. PP tube amp is optimal when works in AB (AB1/2 exist in textbooks only, anyway grid current is non-linear, and real music has wide dynamic range, and significant crest-factor).
For SS, I prefer either class A, or class A+C. For class A I use augmented source follower loaded on counter-modulated current source with no global feedback, for class A+C I use nested feedbacks.
By the way, speaking of feedbacks: I would recommend for your SE amp with parallel feedback across output tube to use pentode driver. Triode with no cathode degradation is the worst case: you are getting huge 2'nd order error such a way.
"Arcane" means hidden or secret. I think the articles posted so far are models of clear writing.
I have just uploaded the Baxandall preamplifier design referred to in the March 1957 power amplifier article. As before, I would be glad to hear any comments you have, especially on the triode/pentode mode switching, which is a new one on me.
The Wireless World Archive
Enjoy.
The Wireless World Archive
Enjoy.
I like too reading vintage mags, here are some more. Maybe someone finds them interesting. No selected articles, just big files.
Vecchie riviste
It is silly question today, what to avoid: tubes or transistors. The proper question is, how to use all available technologies to get best results. We speak here of high end, but not of ancient transformerless All-American-Five radios. That days everything was made on tubes: from garbage to jewels. Like, today everything from garbage to jewels is made on ICs.
What works the best, combination of SS and tubes. For example, triode stage is a good stage, but triode stage loaded on SS gyrator is perfect stage. The same stage powered by voltages stabilized using SS devices is even better.
Or, folded stage with one vacuum tube and couple of MOSFETs can have huge open loop gain and close to zero DC shift between input and output; if to add OpAmp servo you can get zero output DC, so it would be like opamp stage, but very wideband, fast, and linear, with very low noise and enormous headroom.
Back to SS power amps, I have to admit that you explored opamp topology thoroughly, carefully, perfectly. But it is not the single way to skin the cat, and not the optimal one, if to speak of sound reproduction.
What works the best, combination of SS and tubes. For example, triode stage is a good stage, but triode stage loaded on SS gyrator is perfect stage. The same stage powered by voltages stabilized using SS devices is even better.
Or, folded stage with one vacuum tube and couple of MOSFETs can have huge open loop gain and close to zero DC shift between input and output; if to add OpAmp servo you can get zero output DC, so it would be like opamp stage, but very wideband, fast, and linear, with very low noise and enormous headroom.
Back to SS power amps, I have to admit that you explored opamp topology thoroughly, carefully, perfectly. But it is not the single way to skin the cat, and not the optimal one, if to speak of sound reproduction.
Last edited:
OK, I think I see what you are saying. If I understand correctly you are saying that PSE would give you the same power output as PP Class A with better sound and that if you are going PP you might as well get the improved output power of Class AB. Am I close?
WRT the SE feedback, are you referring to the KT-88 with plate to plate feedback for the mini-console? Is the problem related to changes in rp with changes in plate current? I have wondered about that (I have gotten conflicting opinions on that - surprise)and it seems that at least in theory triode with resistive cathode bias (un-bypassed) should be better than the LED bias or by-passed resistive bias and that the pentodes have a more constant rp than either and thus the feedback ratio remains more constant. Is that the gist of it?
I have not studied it but have do we any data indicating that this effect more than makes up for the inherently less linear behavior of the pentode? The other question is whether there is any possibility that the variability in feedback ratio might in some situations be beneficial in compensating for some other non-linear effects? I am not indicating any belief that it does. I am just wondering if the possibility has been considered.
On my mini-console project I am not sure I have the time to investigate this fully especially since I have no access to proper measuring equipment for determining the distortion spectrum of different implementations. The design (which I shamelessly stole) seems to get reasonably good reviews. It would not be hard for me to try it with both the LED bias and resistive and see if I can hear any difference (hard to do immediate A-B tests though). Redoing it for pentode might be pretty complicated though.
WRT the SE feedback, are you referring to the KT-88 with plate to plate feedback for the mini-console? Is the problem related to changes in rp with changes in plate current? I have wondered about that (I have gotten conflicting opinions on that - surprise
)and it seems that at least in theory triode with resistive cathode bias (un-bypassed) should be better than the LED bias or by-passed resistive bias and that the pentodes have a more constant rp than either and thus the feedback ratio remains more constant. Is that the gist of it?I have not studied it but have do we any data indicating that this effect more than makes up for the inherently less linear behavior of the pentode? The other question is whether there is any possibility that the variability in feedback ratio might in some situations be beneficial in compensating for some other non-linear effects? I am not indicating any belief that it does. I am just wondering if the possibility has been considered.
On my mini-console project I am not sure I have the time to investigate this fully especially since I have no access to proper measuring equipment for determining the distortion spectrum of different implementations. The design (which I shamelessly stole) seems to get reasonably good reviews. It would not be hard for me to try it with both the LED bias and resistive and see if I can hear any difference (hard to do immediate A-B tests though). Redoing it for pentode might be pretty complicated though.
Only if you like higher output impedance and higher order distortion.
If you compare grounding the cathode at AC (via LED bias or bypassed cathode resistor) with unbypassed cathode resistor, I will guarantee you'll hear a difference- the gain will not be the same, nor (in all probability) the frequency response.
sorry for the typo on arcane, but for clarification:
ar·cane
http://app.dictionary.com/signup/po...lbackAction=addToFav&domaindest=reference.com/ɑrˈkeɪn/
Show Spelled[ahr-keyn]
Show IPA –adjective known or understood by very few; mysterious; secret; obscure; esoteric: She knew a lot about sanskritgrammar and other arcane matters.
To you Douglas it may be clear and unambiguous, but it is also steeped in a culture of verbage and tone that to me is esoteric and mysterious. I truly love it - we use language so imprecisely these days its made us all the poorer.
And thanks for allowing the bandwidth for these articles to be available - its a fantastic resource.
The Class A PP EL84 amp we have i'd rate above the SE EL84 amps... and better than the same amp biased into AB. Class A has sonic benefits.
dave
Yes.
Yes.
I have wondered about that (I have gotten conflicting opinions on that - surprise
Yes.
Yes.
No.
Why?
One Zener, one capacitor, one resistor additionally. Is it complicated?
Yes, 2 tubes are more powerful than one.
No, class A PP has no sonic benefit, if to compare on the same power.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.