Hi,
Persackly my point too. Glad to be in agreement once again.
Oh, if ever you have the spare time and courage to wade trough the thread I mentioned you'll soon realise how far ahead of times some folk really are.
Yes, I'm being sarcastic here but after thirty years wouldn't you be too?
Ciao Brett San,
Persackly my point too. Glad to be in agreement once again.
Oh, if ever you have the spare time and courage to wade trough the thread I mentioned you'll soon realise how far ahead of times some folk really are.
Yes, I'm being sarcastic here but after thirty years wouldn't you be too?
Ciao Brett San,
Hi,
Oops, forgot to paste in the link.
This is the one:
A very interesting thread, in fact it's one of the best I've ever encountered here.
What surprises me though is that apparently a lot of people don't see the distortion mechanisms in SE designs, complementary design and PP operation etc.
Moreover, people don't seem to have a clue about the influence of the spectral content of THD with regard to the sound of an amplifiier.
To me at least this is fundamental to audio design.
Don't kill yourself over the solid state design propositions, while the analysis is certainly interesting, it still hasn't come to a conclusive stage.
I would be interested to hear from other members how they view THD, pretty meaningless as such IME, and how it's spectral content correlates with the perceived sound of an amp.
The influence of global NFB on dynamic behaviour, it's influence on spatial imaging of the soundstage and so on...
There's a lot more I'ld like to discuss on the forum with regard to distortion but this would be a nice place to start.
Cheers,
Oops, forgot to paste in the link.
This is the one:
A very interesting thread, in fact it's one of the best I've ever encountered here.
What surprises me though is that apparently a lot of people don't see the distortion mechanisms in SE designs, complementary design and PP operation etc.
Moreover, people don't seem to have a clue about the influence of the spectral content of THD with regard to the sound of an amplifiier.
To me at least this is fundamental to audio design.
Don't kill yourself over the solid state design propositions, while the analysis is certainly interesting, it still hasn't come to a conclusive stage.
I would be interested to hear from other members how they view THD, pretty meaningless as such IME, and how it's spectral content correlates with the perceived sound of an amp.
The influence of global NFB on dynamic behaviour, it's influence on spatial imaging of the soundstage and so on...
There's a lot more I'ld like to discuss on the forum with regard to distortion but this would be a nice place to start.
Cheers,
How much feedback should you use?
1) At what frequency?
2) Depends entirely upon the open loop distortion characteristic.
There isn't a general answer. It depends upon the circuit. Hence the lengthy debates. It's like asking how much food should I eat.
Tubes and BJT and FETs all have different effects on distortion and so does topology so it is no surprise that there are widely differing prescriptions for feedback. Ironically, I maintain that the lower the open loop distortion the more feedback you can use without unwanted side-effects. But this is too sweeping a statement because while arguable in the limit, the type of distrotion really matters, not just the total.
1) At what frequency?
2) Depends entirely upon the open loop distortion characteristic.
There isn't a general answer. It depends upon the circuit. Hence the lengthy debates. It's like asking how much food should I eat.
Tubes and BJT and FETs all have different effects on distortion and so does topology so it is no surprise that there are widely differing prescriptions for feedback. Ironically, I maintain that the lower the open loop distortion the more feedback you can use without unwanted side-effects. But this is too sweeping a statement because while arguable in the limit, the type of distrotion really matters, not just the total.
Hi,
This is absolutely correct.
Which doesn't automatically imply that you have to apply all the FB it can handle of course.
Agreed.
I think the harmonic content of the distortion define how we percieve it.
An amplifier with a predominantly odd order harmonic distortion will be percieved as harsh and glary whereas a spectrum with predominatly even order harmonic distortion will be perceived as much less objectionable.
In fact the latter is what a lot of people like so much about SE amps...whether this is a goal to strive for is open to debate but in view of the current state of digital playback I can certainly see the attraction.
Cheers,
This is absolutely correct.
Which doesn't automatically imply that you have to apply all the FB it can handle of course.
Agreed.
I think the harmonic content of the distortion define how we percieve it.
An amplifier with a predominantly odd order harmonic distortion will be percieved as harsh and glary whereas a spectrum with predominatly even order harmonic distortion will be perceived as much less objectionable.
In fact the latter is what a lot of people like so much about SE amps...whether this is a goal to strive for is open to debate but in view of the current state of digital playback I can certainly see the attraction.
Cheers,
Um yes, traderbaum sure,
When it comes to tube amplifiers? For sure you can use 30 dB of global feedback without running into instability with some amps, provided the amp has sufficient open loop gain for this (which is another matter) and provided you take proper measures to handle the excessive phase shift of the transformer.
But can someone explain me why many, many, people have reported over the last 50 years that if you put more than 10 dB to 15 dB of global feedback to whatever tube amp with OTF’s it throws out all the music of the amp?
I have experienced this quite noticeable myself too and a well build tube amp without any global feedback stays sounding the best to my ears.
Needless to say, yes I am fully aware that tube amps are a totally different piece of food than transistor amps.
Cheers
When it comes to tube amplifiers? For sure you can use 30 dB of global feedback without running into instability with some amps, provided the amp has sufficient open loop gain for this (which is another matter) and provided you take proper measures to handle the excessive phase shift of the transformer.
But can someone explain me why many, many, people have reported over the last 50 years that if you put more than 10 dB to 15 dB of global feedback to whatever tube amp with OTF’s it throws out all the music of the amp?
I have experienced this quite noticeable myself too and a well build tube amp without any global feedback stays sounding the best to my ears.
Needless to say, yes I am fully aware that tube amps are a totally different piece of food than transistor amps.
Cheers
fdegrove said:Hi,
Oops, forgot to paste in the link.
[snip]
[/snip]
There's a lot more I'ld like to discuss on the forum with regard to distortion but this would be a nice place to start.
Cheers,
Phew what a tread Frank,
But however everybody seems interested in odd versus even harmonic distortion, nobody took notice of this post:
phase_accurate said:
IMD is one of nastiest and “un-musical” kinds of distortion audio reproducing equipment produce IMO. It has a non-harmonic relation to the original signal. However nobody seems interested in.
Cheers
FOOD FOR THOUGHT...
Hi,
Tell me about it...it took me an entire afternoon to catch up.
And that was exactly what I was hinting at in a previous post.
There's IMD and also of course THD, and while SE and Charles came close, they failed to explain the importance of the relationship between the harmonic content of distortion and the perceived listening results.
Much of what I know stems from meetings with the french "L'Audiophile" team and of course a lot of reading and countless experiments.
Just like musical instruments have their own timbre allowing us to recognise them, so do amplifiers with a certain harmonic distortion spectrum have their typical sound, or colouration if you like.
This is most noticeable on SE amps where often little or no global NFB is used and to a lesser extent and with a totally different harmonic spectrum also in PP amps where the distortion cancelation technique in the PP stage pushes 2nd harmonic into the background making odd order harmonics more dominant.
All this gives these amps a typical sound and much to my surprise even amps with very little THD still show this typical character.
This is rather strange and considering the figures measured it seems to indicate that human hearing is far more perceptive of subleties than we tend to believe.
\rant off...I don't want to write 10 pages on this thorny topic.
Cheers,
Hi,
Tell me about it...it took me an entire afternoon to catch up.
And that was exactly what I was hinting at in a previous post.
There's IMD and also of course THD, and while SE and Charles came close, they failed to explain the importance of the relationship between the harmonic content of distortion and the perceived listening results.
Much of what I know stems from meetings with the french "L'Audiophile" team and of course a lot of reading and countless experiments.
Just like musical instruments have their own timbre allowing us to recognise them, so do amplifiers with a certain harmonic distortion spectrum have their typical sound, or colouration if you like.
This is most noticeable on SE amps where often little or no global NFB is used and to a lesser extent and with a totally different harmonic spectrum also in PP amps where the distortion cancelation technique in the PP stage pushes 2nd harmonic into the background making odd order harmonics more dominant.
All this gives these amps a typical sound and much to my surprise even amps with very little THD still show this typical character.
This is rather strange and considering the figures measured it seems to indicate that human hearing is far more perceptive of subleties than we tend to believe.
\rant off...I don't want to write 10 pages on this thorny topic.
Cheers,
Hi,
Regarding harmonic distortion:
I believe that it is not the total (THD) that matters at all, but the exact proportions of the elements. I'm not saying that even is good, and odd is bad, nor that high's are worse than lows. Just that when there is an un-natural distortion spectrum, we can sense it.
Oh, and distortion IS natural: check your ears.
Regarding IMD:
I agree with Pjotr. Also I do not believe that 2 tone tests tell anything like the whole story.
Back to topic:
NFB affects all of these profoundly.
This is just the start of this thread...I hope.
Cheers,
Regarding harmonic distortion:
I believe that it is not the total (THD) that matters at all, but the exact proportions of the elements. I'm not saying that even is good, and odd is bad, nor that high's are worse than lows. Just that when there is an un-natural distortion spectrum, we can sense it.
Oh, and distortion IS natural: check your ears.
Regarding IMD:
I agree with Pjotr. Also I do not believe that 2 tone tests tell anything like the whole story.
Back to topic:
NFB affects all of these profoundly.
This is just the start of this thread...I hope.
Cheers,
dhaen said:
I do not believe it either
but it is part of the story and too many times overlooked IMHO. Also I don’t blame NFB, it are engineers implementing it improperly. I am too much involved in control theory to get things working and into production for my daily work.Let’s put up the following hypothesis (and just as a hypothesis):
IMD is the interaction between several signals producing new signals due to non-linearity in the amplifier. Now when it comes to stereo imaging, spaciousness, depth and so on, these qualities are largely dependent on the phase relations between left and right signals (and their intensity of course). Now with IM, original signals that come from various places on the stage interact with each other producing new signals at places in the image not very related to the place of the original signals, thus apparently at random locations. IMHO these new signals “blur” or “fog” the original soundscape.
You can observe this also with some mid-speakers driven quite loud (with a blameless non-clipping amp). They start producing quite a lot IMD, turning the whole soundscape in a noisy mess.
Now the question arise: Do harmonic distortion signals affect the soundscape in the same way or are they less harmfull?
Cheers
Pjotr said:
Yep. I've spoken here before several times about noise-floor modulation in some topologies, and a large part of it is IMD. As signals get more complex, though not neccessarily loud, the IMD can be very dense, even if low, and looks just like the grass on the bottom of a spec-an disply is moving up and down with the complexity. The denser the spectrum in music, the worse it will sound.
This is the basis for a lot or the 'electronic' colourations of many systems. The easiest test is to use a good choral recording.
I agree. Insensitive speakers that have to move the cone a lot to get any level out of them always sound bad to my ears. And that's almost everything I've ever heard.
Horns and 'stats have much lower IMD than dynamics.
Add to this too, a real world amp that may be producing a lot of it's own IMD crap that the speaker is going to modulate......
IMO, HD seems to be less harmful and just makes everything a bit rounder and fatter sounding, and IMD predominantly sounds harsh.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.