Emo-Zine Discussion

[KeithL]

We're starting a little online magazine.

In this forum we will post "articles" about various subjects relevant to audio and home theater. Articles will cover a wide range of subjects, but all will be "general" in nature. In other words, nothing specific about individual products. (So, nothing about specific product features, product release dates, and stuff like that here.)

Since this forum will be arranged a little differently than most other forums, I'll explain that a little bit here.....

Emo-Zine, the magazine itself, will be LOCKED; which means that only the admin can post there. This is so people coming here can read the articles without having to pick through comments, replies, and such things.

There will be two SEPARATE threads associated with the main thread:

Emo-Zine Discussion will, as you might expect, be for discussion of articles in the zine (and related topics). You'll be able to post threads there, although I will ask you to stick to topics at least (vaguely) related to articles on the zine. This isn't a police state, but I probably will delete threads in that forum if they seem totally unrelated to the Zine.

Emo-Zine Suggestions will be a thread SPECIFICALLY dedicated to requests for future stories. Please note that I am NOT promising to do stories on all suggestions. I'm not even promising to read them all, especially if they're long. This is just a place where I can see, in a vague sort of way, what sort of stories folks want. I will NOT answer questions posted in this forum. I may also wipe posts in this forum from time to time (after "mentally noting" which subjects seem to be "generating interest").

[Deleted]

I see that Keith has already posted two excellent articles in the Emo-Zine Section. Article #0 is introductory and Article #1 is an explanation of the power amp classes.

I lack engineering type knowledge of the innards of amps. I find Keith's writing and content very clear, precise and informative. It has been mentioned that some folks don't want a rehash of the many articles on the web. I disagree because so many of those articles on varies subjects contain lots of contradictory information and opinions. My sense from reading Keith's post at the Lounge is that he is one I can trust to disseminate accurate information in a clear and precise tutorial style.

Thanks Keith.

[vcautokid]

Very nice job Keith. Very easy to understand. Finally the real deal on how Amplifiers classes work! I enjoyed reading it.

[mgbpuff]

Excellent articles, Keith. I really like the 'down to basics' approach. I don't mean to get ahead of the Emo-Zine discussion, but aren't all the Class A, B, A/B examples in Article 1 also being operated in what is called push-pull?

[cwt]

Mar 16, 2013 7:42:36 GMT -5 mgbpuff said:

Excellent articles, Keith. I really like the 'down to basics' approach. I don't mean to get ahead of the Emo-Zine discussion, but aren't all the Class A, B, A/B examples in Article 1 also being operated in what is called push-pull?

I think Keith was using the push pull 'template' to better compare the different classes mgbpuff . It would have unnecessarily complicated things to have say a single ended class A circuit being compared to the push pull transistors of the other classes I feel ..

It reads a lot more concisely this way [ and it ties in with the way the way the xpa 1l has its class a output transmitters at a guess - logical since its class a/b would be push/pull ]

[harri009]

I would actually enjoy an explanation of class A push-pull vs Class A SET in relation to SS amps. Also very nice article so far

[KeithL]

I received these questions (from garbulky... thanks) and I'm putting the answers here for everyone...

Q1. Could you talk about what kind of different circuitry/parts the XPA-1 has (if any) to enable the high bias Class A vs the XPA-1/another amp that doesnt have the stuff.

A: The actual difference between the bias levels is simply an adjustment - so, at most, a few parts values would have to be changed slightly. Since the power draw and efficiency at full power are about the same, it doesn't require a bigger power supply or bigger heat sinks. The XPA-1L in specific adds a feature that automatically lowers the bias if it gets too warm - and adding that feature required a few extra parts in the bias control circuits. (And, since there's a bit more going on, the design is a bit more critical.)

Q2. With dual rails with differring voltages for class H, is there additional strain on the torroidal power supply to provide two different voltages or any drawbacks.

A: In fact, since the overall amplifier is running more efficiently, the transformer itself is doing less work. The cost comes in terms of circuit complexity. Each supply rail is actually "fed" from two different power supply taps at different voltages, which are then controlled by additional circuitry to provide the single varying rail to the output stage. There are two sets of this circuitry, one for + and one for -. This takes a bunch more power supply parts and a more sophisticated design overall.

Q3. Is there anything in terms of sound quality that class H may have better over an amp without class H rails?

A: Not really. What Class H does is to reduce the amount of supply power that is wasted. In principle, because the amp is using less power on average, it may have a bit more for peaks when it needs it, which might give you slightly better dynamics. There's no particular reason other than that why it should sound different.

(It is, however, important to design a Class H amplifier correctly, since doing it improperly can negatively affect sound quality.)

[KeithL]

We'll definitely get to that eventually

It's actually a double topic though. A push-pull design can operate in any class (Class A, Class A/B, or Class B), but a single ended design MUST operate in Class A (at least for audio).

Here's a preview, though:

A single-ended amplifier (like a SET) MUST operate in Class A. However, especially with tubes, there are major drawbacks to operating a power amplifier single ended (these apply to all classes). In a push-pull design, the distortions from each "half" cancel out to a large degree, as do the DC idle currents in the transformer. This is why a single-ended tube amplifier has higher distortion, and requires a larger and more expensive output transformer, and a more powerful and more expensive power supply, than an equivalent push-pull version of the same amplifier.

Those who prefer SET designs insist that there are sonic advantages that justify the drawbacks. (Those who don't prefer them are quite convinced that the only reason some people like SET amplifiers is that they LIKE the way the extra second harmonic distortion made by the SET amps sounds.)

Mar 16, 2013 16:36:56 GMT -5 harri009 said:

I would actually enjoy an explanation of class A push-pull vs Class A SET in relation to SS amps. Also very nice article so far

[KeithL]

Thanks, and yes on both counts....

The wide majority of modern power amplifiers operate this way.... to the point where those that don't can be considered "special cases".

There area lot of variations. For example, you could run that circuit off a single power supply, and then capacitor or transformer couple the output (which was quite common thirty years ago). It would then still be a push-pull design, but also single ended in another sense. Today, however, most of those variations are quite uncommon (in power amps). You do see plenty of different circuit topologies in preamps - where efficiency doesn't matter so much.

Honestly, you don't see the term "push-pull" applied to solid state designs anymore. (Perhaps "double-ended" - in contrast to "single-ended" - would be appropriate, but I haven't seen that used much either.) You mostly see "push-pull" used in the context of tubes, and to differentiate from "single ended" (but both usually operate from a single ended power supply).

"Operating class" specifically refers to when current is flowing in the output devices (it can also apply to things like drivers). Real single-ended designs MUST be Class A (you can't have current flowing only part of the time in only a single device and NOT flowing the rest of the time somewhere else or there would be a gap in the audio signal.) There are, however, "single ended" designs where one device carries signal and another supplies steady current (check out First Watt for a few of those). However, they tend to have few benefits and be VERY inefficient - even less than normal Class A.

It's a bit tricky since I'm trying to present little pieces of a very large picture - which sometimes makes it difficult to decide where to divide things.....

Mar 16, 2013 12:57:54 GMT -5 cwt said:

Mar 16, 2013 7:42:36 GMT -5 mgbpuff said:

Excellent articles, Keith. I really like the 'down to basics' approach. I don't mean to get ahead of the Emo-Zine discussion, but aren't all the Class A, B, A/B examples in Article 1 also being operated in what is called push-pull?

I think Keith was using the push pull 'template' to better compare the different classes mgbpuff . It would have unnecessarily complicated things to have say a single ended class A circuit being compared to the push pull transistors of the other classes I feel ..

It reads a lot more concisely this way [ and it ties in with the way the way the xpa 1l has its class a output transmitters at a guess - logical since its class a/b would be push/pull ]

[mgbpuff]

Thanks, Keith! Another advantage to push-pull, I think, is better control of the speaker coil, resulting in what some describe as a tight bass vs a loose slightly sloppy bass.

[harri009]

Doing a fine job explaining things in a basic level for us. Very good reads, thanks again

[SticknStones]

Thought I would chime in as well. I like the PDF option as it is easier to read and nice job overall. High Value initiative and great idea.

[cwt]

Mar 17, 2013 14:25:03 GMT -5 KeithL said:

Thanks, and yes on both counts....

Today, however, most of those variations are quite uncommon (in power amps). You do see plenty of different circuit topologies in preamps - where efficiency doesn't matter so much.

Glad I didnt misconstrue Keith ;interesting and spot on what you say about push pull being an older term ; its conspicuous that the only company that still uses that term iirc is onkyo/integra with its ''push pull inverted darlington output ''

I remember having a txnr xxx decades ago with it ;they seem to be stuck in a rut [ aus slang for inertia ]

[nouagie]

Nice article, for my knowledge maybe still too technical (but that's my problem). In advance, sorry for my broken English (I'm Dutch-speaking).

Maybe this isn't the best place to put this, but I noticed that Class D only gets a few lines. I know that Class D is many times treated as "not that good" (or at least less than the others) in audiophiles groups. Still, I notice that class D made quite an evolution over the years (more than the others classes?). Think of Hypex or ICE, nowadays.
Why still remain on the A - A/B-train, specifically if you think about efficiency?