Push-Pull Basics.

[rowuk]

I think that we now have basic coverage of architecture, now sonics can be discussed.

We need a frame of reference. Especially in the DIY world where Beraneks law applies, we need a standard for playback and from there can make value judgements for the next step. Most all of us have had more than one amplifier, so lets look at the sonic requirements vs technology.

1) if we have a difficult speaker, we will only get superior results when the amplifier architecture can drive them. Extension and control for low frequencies depend on the type of speaker used. A low efficiency 12" speaker will not be happy with a SET with no feedback. The impedance and reactive element of the speaker will talk straight through the output transformer and "color" the playback in a serious way.
2) we read about "good" (2nd harmonic) and "bad" (odd harmonics) distortion with no reference. It is pure bullshit that 2nd harmonic distortion is OK and odd harmonics are generally bad. What is important is what they do to the sound. That means the quantity is more significant than the type. I play trumpet professionally. When I record a single tone and play it back with 10% 2nd harmonic distortion, the recording sounds more like a cornet or flugelhorn than a trumpet. When I listen to the same recording with 1% 3rd harmonic, it is hard to hear the difference. If anything, it sounds "crisper". This means that the quality of the OPT, the topology of the amp and the amount of power available simply has to match the speaker requirements. Then the distortion of a valve amp is negligible regardless of type of design.
3) Why do amps have different sonic signatures, well this is a BIG chapter and I am not the right person to explain it. It is safe to say that it has more to do with the entire package than each of the individual elements. From listening to a lot of DIY systems, I can report that what I hear has more to do with lack of synergy than faulty technology. To get superior playback, we do not need the best amp, source, valve or anything else material, we need a defined purpose, a goal that we define and follow step by step based on what we can currently identify as wrong. If for instance we have a low efficiency speaker with 2 12" woofers per side and an EL34 SET without feedback, we could get the idea that more power would be good, so we now build an 845 SET without feedback - we can now play more loudly, but is that better? Perhaps using the EL34 as a pentode with feedback would have solved the sonic issues for that particular woofer better. Does the EL34 as a pentode have the same finesse as a triode? Well, common knowledge says no - but that knowledge is universally not based on fair comparisons, rather agendas. If we live with that amp with the "fixed" bass for a while and are honest with ourselves, then we will know if we need to change the amp, get better OPTs or biamp.

I guess these two long posts have more to do with voicing being personal opinion based on rules that we do not invent on the fly, rather by increasing perception of the situation at hand. That is hard work for those only interested in soldering.

I guess in the context of this thread and amp we are talking about a special type of PP amp and the path to understanding it from a view of a Gary Pimm for instance. I can only stress that his view is based on his use case - music that he listens to, the speakers and his personal goals. Just as it should be. Our mileage will vary based on our use case.

[Cressy Snr]

Does the EL34 as a pentode have the same finesse as a triode? Well, common knowledge says no - but that knowledge is universally not based on fair comparisons, rather agendas.

An interesting post there Rowuk.
I picked out the above snippet, as it is extremely pertinent to what I am after with my playback. The push-pull, class AB pentode, fixed bias output stage is now, apart from its use in guitar amps, almost as forgotten as was the single ended triode output stage, in the early 90s. In fact the pentode pp class AB stage was forgotten by the mid 60s as everyone used ultralinear.

But enough of the history lesson. What I would like to talk about is this strange quality called "finesse" What do we mean by that? If the playback is meant to serve the music, then what defines finesse and how do we quantify it. How do we apply the finesse requirement to a piece of chamber music and at the same time a Sex Pistols record. Seems to me that finesse is a meaningless term in this context.

Saying that a push-pull, pentode class AB fixed bias output stage lacks finesse in comparison to a SE DHT output stage, when being used for playback simply does not make sense. The question has to be, does the amp/speaker combination serve the music, presenting it into the room so that the listener can appreciate and get lost in the production, playing, artistry, without becoming distracted, bored? If the playback allows the music to communicate, then, by definition, it is doing its job.

This is not, for me, a question of subjectivity, or something that lends itself to glib statements like "well If you like it, then that's all that matters"

Now if we remove finesse from the equation we can simply say that for me, push-pull pentode, class AB amplifiers, using EL34s in partial feedback operation driven by small signal pentodes, allow the music I play and am most interested in, to communicate better than 4W of single-ended 2A3s, when being played through quadratic, tapered quarter wave Voigt pipe loudspeakers, using Fostex FF225WK drivers married to Monacor ribbon tweeters.

Now try to apply the statement above to anything other than the context in which it has been uttered and we run slap bang into Pirsig's "Metaphysics of Quality" and there we have to leave it or go insane like his narrator, Phaedrus did. The problem is unsolvable, there is no common language that can be developed, the phenomena are uncommunicable, but that's OK.

[rowuk]

Steve, this is exactly what I was hoping to hear - a description of the sonic result - lack of boredom, drawn into the music, ...... instead of descriptions of "imaging", frequency response, distortion, slam, .......

To me, this is the real reason that I do audio. Still, like with the trumpet that I play professionally, success is not measured in days or weeks, rather months and years. It simply takes time to (re)train our ears/brains and then we know if the net result is "less work" when listening which ultimately frees up resources for things like understanding what the musicians really meant.

Please keep us informed of the things that open up as you continue your journey.

[Cressy Snr]

Yes....
I am moving away from talking about soundstaging, dynamics, bass, treble midrange and slam, and towards talking about "sound"
What I mean by this is, I appear to be making a few first, faltering steps towards having an opinion about my playback rather than just rehashing the same tired old dross. A music system needs to be capable of providing the sound of music in the room or else it is inadequate, therefore should be dismantled and thrown in the trash.

Note I said "the sound of music in the room" not the illusion of performers in the room. It is easy for a hi-fi to set up an illusion of a performer or performers in your room. But it is no good having an illusion of Jacqueline Du Pre in your room if she is playing her cello so badly, it sounds like a buzz saw.
In order for Ms Du Pre to be playing music in your room, she has to be producing that lovely mellow full-bodied sadness of tone and be heard to be interacting with the backing of the other players, feeding off both the intention of the composer and being inspired by the conversation she is having with her fellow players.
If that is happening, then the system is doing its job properly.
Whether it is putting the performers in your lap or spreading them out fifty feet wide and ten feet on front of and behind the speakers, is an irrelevancy.

For me, the sound of music in the room is not sitting there with the performers in front of you, it is a sound that is easy to listen to; a sound that allows the artist to speak to the listener on a deeper level than mere surface and superficiality. It is totally and completely a sound, not an illusion we need to search for.

That "sound" that makes us hear music in our living room is unique to each of us and is produced for us by the playback system, which starts with the room and ends with the speaker/amp interface and how it reacts electrically and acoustically to the environment in which it is working. Get these characteristics right and the sound of music being played pours out into your room and entertains and enthrals in equal measure.

The engineering process of applying fixed bias to a high voltage, class AB, push-pull EL34 output stage with plate-to-grid feedback has, by pure fluke, I have to admit, brought the sound of music into my listening space. It has given me a "sound" unique to my own neurological and psychological and physiological needs and wants. I can tell this has happened by the way the playback keeps me up into the wee small hours of the morning.

Now whether it is accurate is a useful question to ask. Yes it is; it has to be or I would not be hearing the sound of music. Would it be accurate if I upped and shifted the system into another venue. No it wouldn't be because the room would be different, and that is fundamental to understanding how domestic playback of records, CDs sound files or radio broadcasts can be made to give the sound of music being played in the home environment, or not, as the case may be.

When you start to think of it like that the scales fall from your eyes and amp topologies, fads and fashions become really rather irrelevant. You gain clear insight into what is actually important and what is mere window dressing. This gives you a sense purpose and something concrete to aim for rather than build and hope for another piece of happenstance. This clarity makes the hobby so much more rewarding it beggars belief.

[Mike H]

Does the EL34 as a pentode have the same finesse as a triode? Well, common knowledge says no - but that knowledge is universally not based on fair comparisons, rather agendas.

If the amplifier has high open-loop gain and negative feedback, it ought to be difficult to tell any difference.

I think all the latter p-p valve amps were made this way, must have been a good reason. Well, one cited reason was in order to get a good speaker damping factor. That means, low output impedance, achieved by high gain and negative feedback.

[Cressy Snr]

Does the EL34 as a pentode have the same finesse as a triode? Well, common knowledge says no - but that knowledge is universally not based on fair comparisons, rather agendas.

If the amplifier has high open-loop gain and negative feedback, it ought to be difficult to tell any difference.

I think all the latter p-p valve amps were made this way, must have been a good reason. Well, one cited reason was in order to get a good speaker damping factor. That means, low output impedance, achieved by high gain and negative feedback.

Well further to this Mike,
As you are talking in engineering terms, we have to ask ourselves what part engineering has to play in getting the sound of musicians playing, to appear in your listening space.
I would say that engineering of the bits that amplify and reproduce the sound is incredibly important and absolutely not to be dismissed.

If a designer/engineer is going to produce a piece of equipment that is destined to be part of a playback system, then that equipment has to be built so that as far as is humanly possible, what goes in at one end, comes out of the other, unmolested. IOW the designer has to use some kind of benchmark upon which he/she can judge whether their design is fit for purpose and fulfils its specification. That is what designers do. And furthermore, often, in order to get what goes in to come out bigger but otherwise unchanged requires a great deal of hard work, testing evaluating and measuring before it satisfies the designer. Again, the metaphysics of Quality rears its head. Any designer with any sort of conscience, who wants to remain credible, is not going to deliberately put out gear that distorts; it's a matter of being professional about what you are doing. Agenda has nothing to do with any of it, self respect and the satisfaction of a job well done on the other hand is very important to most people. Building gear that distorts is either the unfortunate province of the wet-behind-the-ears amateur like me, or far worse, the flim-flam peddling snake oil selling charlatan.
In fact using gear that puts out what goes in, makes it easier for the end user to put together a system that puts music into the living space, otherwise it becomes nothing more than a crap shoot dictated by whatever is flavour of the month. Designers in it for the long term don't do flavour of the month.
We need a bit more producer driven philosophy and deference/ respect to return to all areas rather than the consumer driven, slick- marketed, business suited, fashion audio madhouse, we have at the moment.

It pleases me no end that people like Nick and Andrew for example, continue to work quietly and steadily away to further the cause of decent reproduction of music in the home; no BS, no mysticism, no buzzwords, just great sounding gear that measures impeccably, with no excuses necessary.

[Mike H]

To head off possible misinterpretation, I am not dismissing anything.

Just that I've been in a state of hiatus last couple of years, wondering if I've been wasting my time. Because it's all boiled down to the startling (but predictable) conclusion that what works best for me actually is anything that uses negative feedback.

Historic examples, Mullard's 510, 520, Mr. Leak's series of p-p amps, etcetera etcetera ad nauseum ~ even Westen Electric's venerable SE 300B cinema amp, all based on the basic design concept of high open-loop gain, and negative feedback. (The WE cinema amp had TWO pentodes as input and driver.)

My still used active unity gain op-amp buffered selector switch box pre-amp kicks the arse of any hitherto emitter/source/cathode follower versions I've made.

So I am feeling like I have just been reinventing the wheel, only to come to the same conclusions as all those other earlier designers. Everything I'm thinking about building nowadays is along the same lines.

Ironic, it's like back to Maplin Millennium.... kind of thing...

Just IMHO.

[Cressy Snr]

To head off possible misinterpretation, I am not dismissing anything.

Just that I've been in a state of hiatus last couple of years, wondering if I've been wasting my time. Because it's all boiled down to the startling (but predictable) conclusion that what works best for me actually is anything that uses negative feedback.

Historic examples, Mullard's 510, 520, Mr. Leak's series of p-p amps, etcetera etcetera ad nauseum ~ even Westen Electric's venerable SE 300B cinema amp, all based on the basic design concept of high open-loop gain, and negative feedback. (The WE cinema amp had TWO pentodes as input and driver.)

My still used active unity gain op-amp buffered selector switch box pre-amp kicks the arse of any hitherto emitter/source/cathode follower versions I've made.

So I am feeling like I have just been reinventing the wheel, only to come to the same conclusions as all those other earlier designers. Everything I'm thinking about building nowadays is along the same lines.

Ironic, it's like back to Maplin Millennium.... kind of thing...

Just IMHO.

Well I'm glad that someone else feels the same as me.

I was feeling quite out on a limb; wondering if I had cloth ears, or was genuinely missing something. I'm pleased you posted that Mike.
I don't feel quite as isolated as before.

[IslandPink]

Yeh, but .. Steve - your amps use local feedback on the pentodes to 'tailor' the characteristics of those valves . That's not the same as a Williamson-style pentode PP amp with global feedback from the output signal . These two approaches definitely sound different to my ears . There are cases where the global feedback PP amps can sound better - eg. when driving very difficult speakers like electrostats, but generally I much prefer the sound of amps only using local feedback around individual devices.

[Cressy Snr]

Yeh, but .. Steve - your amps use local feedback on the pentodes to 'tailor' the characteristics of those valves . That's not the same as a Williamson-style pentode PP amp with global feedback from the output signal . These two approaches definitely sound different to my ears . There are cases where the global feedback PP amps can sound better - eg. when driving very difficult speakers like electrostats, but generally I much prefer the sound of amps only using local feedback around individual devices.

It wasn't that I was commenting on Mark. I was just saying it is nice that there is someone I know, working in a similar area to me, liking what he is doing but somehow feeling a bit disconcerted by it.
I don't know anywhere near as much about electronics as Mike, so unlike him, I don't have the feeling I have been wasting my time, with all the previous explorations into SETs; I've learned a lot doing it but on the other hand, I fully understand his sense of disconnection, hiatus or whatever we like to call it.
It has felt, for me rather uncomfortable doing these pentode push-pull things, but recently, I have started to feel a bit better about it, as I no longer have any case to argue against myself. My ears tell me that this way of doing things gives me music at home, so I see no point in fighting it anymore

[rowuk]

I think that we are comparing apples and oranges.

SE Triodes without feedback require the speaker to be "very" suitable for that type of an amp as there is NOTHING helping to control the load on the output tube. The more difficult the speaker, the worse the marriage.

SE Triodes with feedback make the match to speaker X less difficult, the amount of power available goes down.

SE Pentodes are generally about feedback by design.

PP amplifiers are a completely different discipline altogether. There is much more power generally available, distortion can be much lower but in different proportions than SE amplifiers.

Why are SE feedbackless amps still interesting at all? Well, if you have the right driver there is a certain type of magic that is very unique.

So if the speakers are not chosen based on the amp, if the speakers are not >high 90s in efficiency, I am convinced that a PP amplifier with feedback will most likely give the best results.

I went the other way last year. I had a EL34 PP amplifier and now I am using one of Alex's SEP and feedback RH307A amps. The current evolutionary stage of my speakers are high 90s efficient. The difference in my case is not color, depth, slam or the sound of the instruments. It is more like texture. Certain things just "explode" with flavour (brass for instance), other things are like floating on clouds. The 81dB efficient speakers in my office "work" with the RH amp, but it is not an optimal match - way too little power, the speakers with polypropylene cones need something with a lot more power and control. For sure >25 Watts better >50 watts.

Steve's amp is a real design statement - technically and cosmetically. I am sure that with the power and control available, it will harmonize with a lot of speakers.

[pre65]

A question if I may.

If a push pull output transformer is rated at (as an example) 160ma, is that the combined current of both valves, or just one ?