SE distortion

[andyjevans]

You are free to own and listen to as many SE amps as you like. You are free to describe them as 'better' than PP. I am free to tell you that you are mistaken.

Psychologists use a device to detect hidden bias which consists of turning both sides around and seeing if the effect is the same. For instance:

"My wife and I both qualified with similar degrees. We moved to a university where I got a great job, and she was also able to find work. I help as much as I can in the home and feel we have a good relationship"

"My husband and I both qualified with similar degrees. We moved to a university where I got a great job, and he was also able to find work. I help as much as I can in the home and feel we have a good relationship"

So let's turn it round to the perspective of a SE person:
"You are free to own and listen to as many PP amps as you like. You are free to describe them as 'better' than SE. I am free to tell you that you are mistaken.."

I don't believe you would accept this. So is what you call "freedom" actually the same thing from both sides of the argument?

 

[M Gregg]

Well as I sit here,

listening to my RH84/EF86 and relax....I think its quite good..input from my Cyrus CD..
And the thought of other a bigger louder SE drifts through my mind..
Why should I go PP I have one (not in use at the moment) 8X EL34...

I also have a SE EL34....Its very nice...But I wish I had used a tube rectifier..

Then the thought of another SE..drifts through my mind

Ah well........its all about the sound...

Regards
M. Gregg

 

[gsd]

Talk of 'underdogs' and 'tyranny' suggest some form of coercion is taking place. Last time I checked there were not vicious hordes of engineers patrolling around wresting cherished SE amps from the hands of their secretive owners and then publicly smashing them up.

All that is happening is that some if us are trying to maintain the meanings of words and the reality of physics/maths. Is that too challenging for some people?

You are free to own and listen to as many SE amps as you like. You are free to describe them as 'better' than PP. I am free to tell you that you are mistaken.

I completely support your right to tell me I am mistaken. You are wrong of course but then I am equally free to point that out.
In which thread did I describe S.E. as 'better'?
I think you may protest a little too much!

Regards,

gsd

 

[Miles Prower]

Why because I don't like crossover distortion, odd-order harmonics and the hassle of matching tubes, or mythical class A PP amps?

What earthly benefit would there be for me to need a PP amplifier when I get enough power from SE? Oh ya, bragging rights over noobs who are fresh from the "moor whats is better" SS camp.

I really hate these kinds of cat fights, but this just isn't right at all.

X-over: This type of distortion is most common with Class AB transistors and under biasing. It does sound quite nasty. It's also the reason I favor BJT finals over complimentary MOSFETs -- NPN/PNP complimentary pairs are a good deal more complimentary than N-channel/P-channel pairs. Much less x-over, and less nasty x-over.

It's not such a big problem with Class AB hollow state since the onset is much gentler. (Attached) Here is the residual from subtracting the input from the output of a Class AB design that uses 6BQ6GA finals. There is no x-over seen here, and the only defect is very small glitches at the x-over points since there is no such thing as perfect balance in OPTs.

Harmonics: SE and PP amps make some odd order harmonics. It's unavoidable, but you can at least design to minimize the problem, and correct with NFB.

Matching tubes: This is highly over rated, and mainly an excuse to drive up the $$$$ of VTs. The Radiotron Designers Handbook describes an experiment done with a PP final where they put a 2A3 in one hole and a 45 in the other. The resulting distortion was a good deal less than you might otherwise think. It's better to be able to match Q-point plate currents to minimize core magnetization.

Mythical Class A PP Amps: Nothing "mythical" about it. Check out some spec sheets: many audio finals give example designs that keep the PP finals in Class A. All that's necessary for Class A PP operation is that the plate current of the finals never cuts off completely.

"Oh ya, bragging rights over noobs who are fresh from the "moor whats is better" SS camp".

I started out designing SS amps. I have some really good ones as well. SS doesn't have to sound as hideous as it all too frequently does. It's a question of picking the right types and topologies. Also planning a real speaker-buster with screen driven 36LW6s just because I can.

 

[Shoog]

My last amp build was SE and my current amp build is SE. The last one didn't match my PP amps so I thought I would go more pure and build an SE with gapped OT. I don't expect it to better my PP, but I am open minded.

I have come at it both ways and just found the compromises implicit in SE not to my liking so far.

Shoog

 

[M Gregg]

But what about the PSU..

What about it..well its a SE...so what!

That's half a push pull isn't it..

How many ways can you make a PSU..so what..DC is DC..oh sorry..I thought it was more than that.
You will have me tuning the PSU next...this audiophile nonsense... LOL
But what about the resonant choke....its resonant isn't it...

Regards
M. Gregg

 

[DF96]

I don't believe you would accept this.

I would accept it as a matter of freedom, but would argue about the truth of it. My point relates to silly remarks about 'underdogs' and 'tyranny'. SE lovers are wrong, but they are not harmful so I don't object to their existence. Strangely, some of them seem to object to being told they are wrong so in a sense they are objecting to my existence.

As I said earlier on, any circuit technique which genuinely improves SE (in the sense of making the output signal more like a simply-amplified version of the input signal) can be applied to PP. Thus PP has all the genuine advantages of SE, but in addition has some true distortion cancellation. For SE to be 'better' you thus have to accept that 'better' equates to more distortion. There are two plausible explanations:
1. SE lovers' ears tolerate odd-order distortion (even at very low levels) only when it is masked by significantly higher even-order distortion.
2. SE lovers' ears prefer some even-order distortion.

 

[Shoog]

But what about the PSU..

What about it..well its a SE...so what!

That's half a push pull isn't it..

How many ways can you make a PSU..so what..DC is DC..oh sorry..I thought it was more than that.
You will have me tuning the PSU next...this audiophile nonsense... LOL
But what about the resonant choke....its resonant isn't it...

Regards
M. Gregg

Power supplies are going to have some effect but if they are having a big effect then you are doing something fundamentally wrong.
The fashion for all film filter caps would seem to me to be doing something wrong since its nearly impossible to provide adequate filtering using just film caps. Even Theorston admitted that the only place for a film cap was in the last stage of the power supply - where it could be heard.
Chokes I like because they are great hash suppressors because their impedance rises with frequency - something a plain old resistor cannot match. Other than that, there not magic in any other way - unless they start ringing, as you say, and again that plain old bad design which can be easily avoided using PSU2.

Shoog

 

[M Gregg]

Just for interest,

I posted it some time ago (again for interest)

Resonant-choke Power Supply

Of course if you hear a difference something is wrong...is it..each to his own I guess...
As we know DC is DC...regulate it until it submits that will tame it...

Regards
M. Gregg

 

[Shoog]

Just for interest,

I posted it some time ago (again for interest)

Resonant-choke Power Supply

Regards
M. Gregg

Seen it done, but messing with PS resonances seems risky to me. The only reason to do so (for me at least) is for a cheap PS or a slightly more compact one. Can't see the point overall.


Shoog

 

[M Gregg]

Nothing like never trying it is there..

It avoids all the whats happening..nothing like finding a discharge resistor removed from the amp side of the choke making a difference and fitting it to the rectifier side..

then again SE is not to your liking so...

Try throwing the book out of the window...walk on the wild side...just for fun of course.

Use your engineering skill then use the suck it and see approach (Obviously within safety margins).

Regards
M. Gregg

 

[Shoog]

Nothing like never trying it is there..

It avoids all the whats happening..nothing like finding a discharge resistor removed from the amp side of the choke making a difference and fitting it to the rectifier side..

then again SE is not to your liking so...

Try throwing the book out of the window...walk on the wild side...just for fun of course.

Use your engineering skill then use the suck it and see approach (Obviously within safety margins).

Regards
M. Gregg

That has always been my approach, but always based on sound engineering where possible. I refuse to reject what is proven to work and which has the evidence to support it. If a resonant power supply works - its because the physics/engineering defines it that way.

Shoog

 

[M Gregg]

What about,

Just try different PSUs with the same circuit,
even ones you just think will work ok and listen. (don't model it)
Use your engineering best guess..just for fun.

Resist the temptation to reach for the scope, see if you find something you like.

Then work out why you like it..

Regards
M. Gregg

 

[Shoog]

Seems like your deliberately introducing an artifact and then post hock justifying its existence.
The problem with this approach is we have all heard of people using underdamped input transformers and reporting it has a very "vibrant" sound which turns out later to be a high frequency resonance. It takes a while to spot that correct damping sounds more natural - but try convincing the builder, until he discovers it all by himself, that his underdamped transformer has a problem and he can't listen to it for more than half an hour.

The novelty factor can send us down many blind paths and many wasted hours.

Shoog

 

[H E Pennypacker]

Good golly miss molly.

Why would he deliberately introduce distortion just to try and defend it?

Two separate out of phase tubes are never going to be as coherent as one.

If what you say is true, we would never read the end of it here.

 

[Shoog]

Good golly miss molly.

Why would he deliberately introduce distortion just to try and defend it?

Two separate out of phase tubes are never going to be as coherent as one.

If what you say is true, we would never read the end of it here.

That is a statement which requires justification - not just faith.

-------

Lets face it SE was a Japanese response to the excesses of high gNFB Pentode Williamson style amplifiers.

Modern PP is a response to the excesses of japanese SE.

No one needs ever build another high feedback high power PP amp ever again - unless they specifically choose to.

People are still fighting last centuries war without realizing the enemy was vanquished about three decade's ago. The very fact that you can say "mythical Class A PP" and "parafeed isn't SE" with a straight face shows that you just don't get how things have moved on.

Shoog

 

[Wavebourn]

According to my own experience, properly designed and build SE and PP sound equally nice and clean. However, PP sounding equally nice consumes less power and generates less heat.

Disclamer: for nice sounding result SE requires less parts and knowledge, just copy somebody's design properly and it will sound well.

 

[devilsindetails]

The value of this thread to me is not in one view over another, but in learning much much more about the subject than I knew going in.

I commend those contributors who approached this discussion with an open mind and no real entrenched interest or position in one topology over the other. Your insights, observations and scientific arguments are useful where the provocateurs are simply argumentative.

Those who brought science, real-build experience--with both SE and PP designs--and some reasoned observations gave me a lot to chew on. contemplate and hopefully utilize as I design and tweak my own builds (SE and PP).

FWIW I own two near identical commercially-produced EL84 amps, one PP and one SE...the only major difference being OTs and the number of output tubes/ Used to drive identical loudspkrs: Dynaco A25s...they do sound different to my ears, with the SE providing wonderful detail to about half its rated output of 5 watts p/c and the PP not coming alive until beyond a quarter of its rated output of 12 watts p/c.

They are definitely beasts of a different feather for many of the reasons laid out here in this thread by others.

But the religious and mystic zealotry embraced by SE enthusiasts i've encountered over the past 20 years, who arbitrarily dismiss PP as an inferior approach leads me to believe they have no intention of challenging their own bias...and that's where the learning stops and a useless pis_ing contest ensues.

 

[DF96]

Two separate out of phase tubes are never going to be as coherent as one.

Fortunately, 'coherence' (whatever that means) is not an issue. Two similar valves fed two similar signals (in antiphase) will produce two similar outputs, but with odd order components in antiphase and even order components in phase. Combine these in antiphase in an OPT (or by some other means) and the odd order adds (incuding the wanted signal) and the even order partially cancels. That is how PP works. It really is that simple. In addition, for a given output you need drive the two valves less hard than you would need to drive one so you get much less high order distortion. It really is a win-win situation.

SE fans have to either accept lots of distortion, or try to partially cancel it using the curvature of the previous stage. This inevitably leads to more higher order distortion, as the lower order distortion from each stage gets multiplied together. (Note that in PP the distortion adds, not multiplies, so you can get cancellation without extra higher terms).

Then PP can add some NFB to reduce distortion still further. SE can't do this to the same extent because the forward path distorts too much. More win-win for PP.

 

[gsd]

Fortunately, 'coherence' (whatever that means) is not an issue. Two similar valves fed two similar signals (in antiphase) will produce two similar outputs, but with odd order components in antiphase and even order components in phase. Combine these in antiphase in an OPT (or by some other means) and the odd order adds (incuding the wanted signal) and the even order partially cancels. That is how PP works. It really is that simple. In addition, for a given output you need drive the two valves less hard than you would need to drive one so you get much less high order distortion. It really is a win-win situation.

SE fans have to either accept lots of distortion, or try to partially cancel it using the curvature of the previous stage. This inevitably leads to more higher order distortion, as the lower order distortion from each stage gets multiplied together. (Note that in PP the distortion adds, not multiplies, so you can get cancellation without extra higher terms).

Then PP can add some NFB to reduce distortion still further. SE can't do this to the same extent because the forward path distorts too much. More win-win for PP.

Thanks for the explanation.
That's where I've been going wrong.
So If by some means of measurement it measures 'better' then it sounds
'better'
I think I can hear the difference already!

Regards
gsd