Interesting books....

Status
Not open for further replies.
Jump to Latest
#201
andy_c said:
Jan, just take the problem in reverse. Let's assume that we did manage to make a power amp that had, say, .01% third harmonic distortion at the high output level. What would it be at an output voltage that's 1/28 of this? It would be .00001276%. In other words, astonishingly low. Lower than we know how to make it.

And here's something else to consider. If these formulas are really correct under all circumstances (they aren't, they're only valid in the "weakly nonlinear" region) how do you explain that some power amps have more distortion at lower power levels than at higher?
Click to expand...


Check. I got confused when John mentioned the 28 squared and from that and the 0.01% input 3rd HD came to 7.5% at the output of the power amp.

Your second point, I don't know off hand. Are you talking about THD or about some specific harmonics? I know that on some curves it LOOKS that the HD is higher at low levels, but that often has to do with worse SNR at low levels; you're really measuring the noise instead of HD. But I guess that's not what you mean.

Jan Didden
 
#202
andy_c said:
Jan, just take the problem in reverse. Let's assume that we did manage to make a power amp that had, say, .01% third harmonic distortion at the high output level. What would it be at an output voltage that's 1/28 of this? It would be .00001276%. In other words, astonishingly low. Lower than we know how to make it.[snip]
Click to expand...


Andy,

Some more thoughts. I think we agree that we CAN built amps with less than 0.01% 3rd HD at, say, 28VRMS out for 1VRMS in. When you go down to 1VRMS out, the prediction is like you say that it would go down to 0.000012755%. I don't think we can measure that. But let's go down to a quarter, 7Vrms. The 3rd HD would then be predicted to be 1/16 which is 0.000625%, probably at the edge of measurement capabilities. Is that what we actually observe?

And what does it mean that it is only for 'weak non-linearity'. Does that exclude things like overloading or clipping only, or is there more to that??

Jan Didden
 
#204
Folks, I'm shocked! I show how a linear design engineer has to look at OPEN LOOP DESIGN, and mostly all that I get is 'static'. Bode doen't count for much in open loop design. OF COURSE, if we use GLOBAL NEGATIVE FEEDBACK, we can reduce our distortion to almost anything that we want. However, negative feedback has problems that even Bode can't fix. This is why many hi end designers strive to reduce or eliminate global negative feedback when possible.
 
#205
john curl said:
Folks, I'm shocked! [snip]
Click to expand...


John, you're ahead of me, I'm mostly confused!
All the discussions on that quadrupling of 3rd HD with doubling level, that must be valid for open loop as well as closed loop amplifiers, no?
The example I posted from Feucht was for an emitter-degenerated stage with at least some 12-15 dB of feedback I would guess, ok not astronomically so, but enough to make it a feedback amp.

BTW just out of curiosity, you must just be getting up. I just came back from work. Global economy??

Jan Didden
 
#207
Mikek, I can respect your opinion, but you have blinders on! I have been using negative feedback as a designer for more than 35 years, BUT it has serious problems in audio design. That is why Dr. Otala, and many others have written about its problems and limitations. You simply have not looked that far or listened to enough, if you think that negative feedback is the answer.
Basically, if I could make a high feedback circuit sound as good as a low feedback circuit, I would do it.
 
#208
#209
Well, Mikek, too bad you don't know more. In 1980, Matti Otala, Walt Jung, Marshal Leach and I wrote a complete critique of TIM and related distortions. It wasn't published, but it sure scared Bob Cordell, who was sniping at us at the time, as well as Dr. Cherry. ( Do these names sound familiar?) Instead, we published a cover letter 'Audio' July,1980 that saved Bob Cordell the embarassment of being corrected publicly in print.
Later, Matti wrote a seminal paper on PIM or PHASE INTERMODULATION DISTORTION, that was given, but NOT put into the AES Journal (for political reasons) Since then, Matti has not tried to play in the AES ballpark to any great extent, but went elsewhere.
Fairly recently, Barrie Gilbert has raised the question and PROVED what Otala had stated on PIM (about 20 years before) and even gave him grudging credit for TIM. I, personally, have make 100s of measurements, both in Dr. Otala's lab in 1976, or more recently with my own, virtually identical test equipment, PROVING the cause and effects of TIM. This is not an issue anymore. PIM is more challenging, BECAUSE we don't have an easy way to measure FM modulation with conventional test equipment. Why not make yourself useful, and show me a convenient way to measure PIM?
 
#211
john curl said:
Well, Mikek, too bad you don't know more.

In 1980, Matti Otala, Walt Jung, Marshal Leach and I wrote a complete critique of TIM and related distortions.

It wasn't published, but it sure scared Bob Cordell, who was sniping at us at the time, as well as Dr. Cherry. ( Do these names sound familiar?)

Instead, we published a cover letter 'Audio' July,1980 that saved Bob Cordell the embarassment of being corrected publicly in print.
Click to expand...

This is a new level of absurdity.....

What is the point of three people going to all the trouble of researching and writing what they consider to be a superior paper, only to shelve it to 'spare' the blushes of someone else??

:scratch:
 
#214
till said:
i would expect TIM and PIM
Click to expand...


TIM is not caused by feedback....and the near-zero phase intermodulation arising from using too little feedback is virtually eliminated by using sufficient feedback....

john curl said:
It was an editorial decision by 'Audio'........
Click to expand...

Good for them!!!

john curl said:
Get real....
Click to expand...

Indeed! Out with this subjectivist cult nonsense!

john curl said:
......have you ever worked in the real world?
Click to expand...

Certainly not in the one inhabited by quasi-scientific, neo-homeopathic subjectivist cultists. 😉
 
#217
This is not an elementary issue. It takes a real math background to derive or fully understand PIM. However, it is real, and probably the MAJOR problem with global negative feedback. Mikek is just denying its importance.
Still, this is what makes for 'different' audio designs. Those of you, who want to use lots of negative feedback, go for it.
 
#218
john curl said:
[snip]However, it is real, and probably the MAJOR problem with global negative feedback. Mikek is just denying its importance.
[snip]
Click to expand...


Depends on who you ask, really:

"Phase Intermodulation Distortion: Instrumentation and Measurement Results

Phase intermodulation (PIM) in audio amplifiers, particularly that which may arise from the use of negative feedback, is considered. Coherent SMPTE-IM instrumentation capable of resolving two nanoseconds of equivalent time dispersion and measurement results from several different real amplifiers are presented. The results show that PIM is not a problem in contemporary amplifiers and that negative feedback reduces total PIM in most cases. "

AES Preprint Number: 1842 Convention: 70 (September 1981)

I quote: "The mechanism for PIM is really straightforward [..] In essence, open loop amplitude inter-modulation [..] maps this into a signal dependent closed loop bandwidth, which in turn results in signal-dependent closed-loop phase variation or PIM". End of quote. There, didn't even hurt.


Jan Didden
 
#219
Andy_c, your thoughts?

janneman said:
All the discussions on that quadrupling of 3rd HD with doubling level, that must be valid for open loop as well as closed loop amplifiers, no?
The example I posted from Feucht was for an emitter-degenerated stage with at least some 12-15 dB of feedback I would guess, ok not astronomically so, but enough to make it a feedback amp.
Click to expand...


Andy, I am seeing this right or what?

Jan Didden
 
#220
Status
Not open for further replies.
Top Bottom