Integrated Amp discussion...

[PENG]

Interesting that even Doug Self was brought into that discussion.

Are you experiencing negative inflation with that 0.002 cent?

I do understand the down side of negative feedback if used without discretion, but I think blanket statements about the side effects are not fair and not accurate. At least the author included tons of formula to support his claims, not just by hearsay, or parroting gurus such as Nelson Pass and others. Someone with a good background in feedback control theory should be able to follow his math and challenge him accordingly.

 

[PENG]

...and in fairness to Irv (for those that don't know the history on this), he does not attempt to argue that there is a problem with Class D amps, only that it is a possible concern which he chooses top avoid given there a fine A/B amps available at prices he is willing to pay to avoid living in Class D anxiety!

Agreed, he knows his stuff, and limits apparently, that's why I wanted to fix it for him.

 

[PENG]

Adding a lot of negative feedback causes higher order harmonics to rise in level at the expense of lower harmonics.

That's part of it, but I hope you read the article I linked. Like a lot of things, if done well we have have the benefits while minimizing the ill effects for good overall gain. By the way, the author happened to be working for Hypex, since you mentioned the name.

 

[Matthew J Poes]

That's part of it, but I hope you read the article I linked. Like a lot of things, if done well we have have the benefits while minimizing the ill effects for good overall gain. By the way, the author happened to be working for Hypex, since you mentioned the name.

There is nothing wrong with negative feedback. I’m not against it. I think I was misunderstood. My only point is that you can lower THD without improving sound. It can even make it sound worse. Putzy absolutely figured out how to do it right and lower overall distortion levels to such a low level that there is no chance any of the harmonics are audible. I think that’s the key. However you can’t compare every amp by THD and expect that to be a metric of sound quality. It’s simply not and I cited my source. Thd is not correlated with perceptions of distortion and it is not true that low THD guarantees no audible distortion, as Geddes showed in his experiment. Unfortunately there is no composite distortion metric in wide use that is correlated with perceptions of distortion so we are left having to look at the raw harmonic components.

 

[PENG]

There is nothing wrong with negative feedback. I’m not against it. I think I was misunderstood. My only point is that you can lower THD without improving sound.

I agree with you on that to a point, but not to the point when you are talking about numbers as low as 0.002% to 0.01% right up to and past the rated output, such as that of the Denon AVR-4308CI. You can also look at the harmonic spectrum (I cannot post them due to copy right..),the higher order harmonics are as good as amps that you probably respect, including NAD, Anthem, Arcam, among others. The individual components cannot be greater than the sum!! My 4308 does sound as good as my separate components, and it does use negative feedback.

https://www.soundandvision.com/content/denon-avr-4308ci-av-receiver-measurements

 

[Matthew J Poes]

I agree with you on that to a point, but not to the point when you are talking about numbers as low as 0.002% to 0.01% right up to and past the rated output, such as that of the Denon AVR-4308CI. You can also look at the harmonic spectrum (I cannot post them due to copy right..),the higher order harmonics are as good as amps that you probably respect, including NAD, Anthem, Arcam, among others. The individual components cannot be greater than the sum!! My 4308 does sound as good as my separate components, and it does use negative feedback.

https://www.soundandvision.com/content/denon-avr-4308ci-av-receiver-measurements

View attachment 27598

I think you did not read the article I posted. If you did you would see that claim is incorrect.

 

[M Code]

Bob Carver used to be regarded as a 'lord and savior' of the objective crowd many years ago. However, many from that side now believe he has since gone 'rogue' especially when he's peddling stuff like this from his site.

https://www.bobcarvercorp.com/silver-seven

That picture says it all. A horse-drawn wagon going from town to town selling an audiophile tube cure-all designed for any chronic solid-state conponent ailments.

Yup...
Bob Carver is a very unique guy...
And knows very well how to design something special and add a kool marketing twist to it..
Several years back, I was offered a senior, executive position in product development/marketing @ Carver but turned it down as his VP of Sales @ the time had a different viewpoint where to steer the Carver brand & products... I told Carver I didn't agree with the marketing direction....
Funny about a year later Carver Corp.ran into some difficult times due to some sales/marketing missteps..
Talking to Carver later on, he conceded to me that my original apprehensions had proven correct.. And over the years I continued maintaining contact with him.

Just my $0.02...

 

[PENG]

I think you did not read the article I posted. If you did you would see that claim is incorrect.

What claim? I did read the articles you posted, though not in detail and you posted a few so I might have missed one.

 

[PENG]

If you meant the Gedlee article, I read that, but to understand most of how he develop the Gm metric I will have to brush up on Taylor and Fourier as well as partial derivatives, that I may or may not be willing to do. I am certainly interested in researching the topic further regardless. From my limited understand so far of what he has demonstrated is that THD and even IMD are not good predictors of sound quality. That does not contradict my own belief based on my understanding of the theories behind and years of experience in auditioning all kinds of mid to high end amps. In my previous posts I was simply saying that if THD is as low as 0.05% or less, there is no point worry about the individual harmonic components. I highly doubt Gedlee would be concerned with the harmonic contents for amps with specified and measured THD to be less than as high as 0.05% or even 0.1%, 20-20,000 Hz from fractional output to rated output. If I know him I would ask him myself but I don't so I am guessing.

I would also guess that if THD is higher than 1% (just an arbitrary number I can think of) then the make up of that 1%, i.e. the harmonic components, might be a factor significant enough for his metric. Again, I may have misunderstood a lot, but for now my impression is that for well design class AB amps nowadays, THD spec is practically irrelevant. The only reason I would still aim for those with very low THD specs is exactly to try and avoid excessive high order harmonics that get hidden by the THD number. To be clear, I agreed with you since your first post on the higher order harmonics resulting from negative feedback is not a good thing. I am simply of the opinion that since THD is the total, so in terms of Total = 2nd+3rd+4th+5th+........nth order harmonics, each component, whether it be 2nd, 3rd, 7th, 11th, will not be greater than Total. So if total is 0.05% I won't worry about any of the components regardless of the order. I have been saying this consistently, if you say it is not correct then we can certainly agree to disagree.

 

[mtrycrafts]

Bob Carver used to be regarded as a 'lord and savior' of the objective crowd many years ago. However, many from that side now believe he has since gone 'rogue' especially when he's peddling stuff like this from his site.
...

He is not the only one who went rogue and off the deep end. John Curl is in the same boat. You should see his posts at AA and past discussions with Jneutron.

 

[mtrycrafts]

If you meant the Gedlee article,...

While I have no standing, I did notice the paper to be only a convention paper, not a Journal paper that has rigorous peer review.
As well, each subjects had one pass through the test. No mention DBT even though paper mentions statistical analysis.
I wonder how well these same subject could replicate the same outcome if they did it 10 time, say hours apart, etc.

 

[JerryLove]

He is not the only one who went rogue and off the deep end. John Curl is in the same boat. You should see his posts at AA and past discussions with Jneutron.

It always worries me when I see someone I put stock in (and I don't have experience with this particular person, I'm talking in general) take some outlandish position on something else.

It makes me wonder "was he right here despite having jumped the shark there; or have I simply started drinking the coolaid".

And that does seem to happen a lot in audio.

 

[PENG]

While I have no standing, I did notice the paper to be only a convention paper, not a Journal paper that has rigorous peer review.
As well, each subjects had one pass through the test. No mention DBT even though paper mentions statistical analysis.
I wonder how well these same subject could replicate the same outcome if they did it 10 time, say hours apart, etc.

I read that paper again and no I am not going to re-visit the theory backgrounds even though I still have my books and notes from university. There is no point because he really did not share much of the needed information for me to agree, disagree or anything in between. He injected x number of stimuli without telling what they are, other than just THD, IMD, transfer functions etc.. Yet he seemed to have concluded THD and IMD should not be used to say one amp would sound better than the other based on THD/IMD graphs like a blanket statement.

I don't believe THD/IMD measurements can be used to say a Bryston amp should sound better than a Marantz AVR either, but I do feel, like Power output, it does matter as they in general (not always) may indicate the quality of the design. So yes, all else being equal I would stick go with an amp measured 0.001% distortions that is designed to be transparent, rather than one that measured 0.1% and also designed to be transparent. Just like I would take a 300 WPC amp even though a 100 WPC amp is already an over kill for my application, all else being equal.

It is disappoint to see that, invariably, people would take some of those views such as Dr. Geddes's to make their own blanket statements that I suspect Dr. Geddes himself may disagree.

Back on the feedback vs distortions thing, just because feedback would result in lower overall THD but increase higher order THD does not mean manufacturers such as D&M, Yamaha would go crazy with feedback in order to produce amps with low distortions. Why wouldn't their design team design their amps with minimum distortions and then apply feedback to further reduce distortions to inaudible levels? There are no shortages of publicly available schematics for amps with little feedback for DIYers, not rocket science.. We just can generalize and imply that amps with lower THD simply mean they use more feedback, and/or if that's true, that it should be ignored, without also questioning why the other amp has much higher distortion specs, just in case it may indicate something else that has nothing to do with how feedback was applied in the design.

 

[KEW]

We just can generalize and imply that amps with lower THD simply mean they use more feedback, and/or if that's true, that it should be ignored, without also questioning why the other amp has much higher distortion specs, just in case it may indicate something else that has nothing to do with how feedback was applied in the design.

Yeah, given we are talking major established companies with deep experience in amp and feedback design, I cannot get too concerned over the idea that Denon (or Marantz, Yamaha, etc) is using negative feedback in such a reckless manner. It seems more "theoretical possibility" than "likely reality". I am nowhere close to understanding the case about total THD. What you say makes sense to me - If THD is truly Total Harmonic Distortion, the distortion from reckless use of negative feedback (which I understand to be harmonic) should be included in that value. However, I cetainly am above my competence attempting to understand the counter-argument.

(Note: I know the below will further derail this thread, but this is relevant to the question of how much difference in sound is there between AVR's or Amps. I know we usually qualify that with "competently designed, in proper working order, and operated within its design capabilities", and this does violate the "competently designed" criteria - which is surprising for any major Amp/AVR manufacturer! So it does seem to happen!)

On the other hand, I can understand Matthew's skepticism (though I'm not sure how skeptical he is about Denon's use of feedback) given that he discovered an Onkyo AVR which starts to roll-off at 5kHz and is 10dB down at around 17kHz. I certainly would not expect that out of any modern AVR (it is an older model, but not that old)! The measurement was with the DSP engaged (but no processing applied). It is pretty shocking that Onkyo would not have tested this and corrected it! I wonder how many other models used the same circuitry for the DSP! Clearly this is an audible f$%k-up!*

https://www.avnirvana.com/threads/outlaw-audio-model-976-7-2-surround-processor-review.3778/

If the specific model of Onkyo AVR was identified, I did not see it.

* My OCD requires that I hold fast to Matthew's presentation. As can be expected, he was shocked to get this result. While I present it as fact, Matthew presents it as - (paraphrasing) he double checked all settings and if there is some sort of processing engaged, he did not find it (leaving open the possibility of human error). I appreciate the care he took in this presentation (which I casually violated)! I also can very much relate to his concern (and now reset the AVR to eliminate the likelihood of some "derelict" processing "hiding out" in my AVR prior to set-up)!

 

[PENG]

Yeah, given we are talking major established companies with deep experience in amp and feedback design, I cannot get too concerned over the idea that Denon (or Marantz, Yamaha, etc) is using negative feedback in such a reckless manner. It seems more "theoretical possibility" than "likely reality".

That was more or less my point.

On the other hand, I can understand Matthew's skepticism (though I'm not sure how skeptical he is about Denon's use of feedback) given that he discovered an Onkyo AVR which starts to roll-off at 5kHz and is 10dB down at around 17kHz. I certainly would not expect that out of any modern AVR (it is an older model, but not that old)! The measurement was with the DSP engaged (but no processing applied). It is pretty shocking that Onkyo would not have tested this and corrected it! I wonder how many other models used the same circuitry for the DSP! Clearly this is an audible f$%k-up!*

From his posts, I don't think he's too skeptical about that but I could be wrong.

If the specific model of Onkyo AVR was identified, I did not see it.

* My OCD requires that I hold fast to Matthew's presentation. As can be expected, he was shocked to get this result. While I present it as fact, Matthew presents it as - (paraphrasing) he double checked all settings and if there is some sort of processing engaged, he did not find it (leaving open the possibility of human error). I appreciate the care he took in this presentation (which I casually violated)! I also can very much relate to his concern (and now reset the AVR to eliminate the likelihood of some "derelict" processing "hiding out" in my AVR prior to set-up)!

When I first acquired my Marantz SC-7/SM-7 pair, I was shocked to see similar roll off with the SC-7 in the loop. I double, triple checked and could not find the reason other than there is something wrong with the >30 year old preamp. It was much later, probably days, when I suddenly noticed the high pass filter was turned on. I probably still have those plots, really similar to those Onkyo roll off.

 

[mtrycrafts]

It always worries me when I see someone I put stock in (and I don't have experience with this particular person, I'm talking in general) take some outlandish position on something else.

It makes me wonder "was he right here despite having jumped the shark there; or have I simply started drinking the coolaid".

And that does seem to happen a lot in audio.

Well, yes, one should know if a person who is used as an example jumped the shark about something they are well versed in. Yes, Curl did design good amps earlyer in his career but something changed later.
As we are talking about this, same happened with a respected Hawksford and his article titled Essex Echo. One has to evaluate carefully.

 

[M Code]

That was more or less my point.



From his posts, I don't think he's too skeptical about that but I could be wrong.



When I first acquired my Marantz SC-7/SM-7 pair, I was shocked to see similar roll off with the SC-7 in the loop. I double, triple checked and could not find the reason other than there is something wrong with the >30 year old preamp. It was much later, probably days, when I suddenly noticed the high pass filter was turned on. I probably still have those plots, really similar to those Onkyo roll off.

For ur information...
The Marantz 7 Series was actually designed in the USA but built in Japan. During the mid-late 70s, Marantz USA based in California hired some very capable design engineers mainly from JBL. The phono pre-amp circuit in the SC7 uses a full complimentary outputs(14 transistors including FETs) and can output up to 10V RMS with an MM cartridge. Also the Low & High Filters were 18dB/Octave compared to the normal 6 or 12dB, allowed the bandwidth to be wider. Their frequency response curves are included in the owners and service manual...

Just my $0.02...

 

[PENG]

For ur information...
The Marantz 7 Series was actually designed in the USA but built in Japan. During the mid-late 70s, Marantz USA based in California hired some very capable design engineers mainly from JBL. The phono pre-amp circuit in the SC7 uses a full complimentary outputs(14 transistors including FETs) and can output up to 10V RMS with an MM cartridge. Also the Low & High Filters were 18dB/Octave compared to the normal 6 or 12dB, allowed the bandwidth to be wider. Their frequency response curves are included in the owners and service manual...

Just my $0.02...

I know, I have the original manual, even invoices as it was a gift from my ex audiophile brother who turned to MP3 in recent years. He bought them brand new in 1979. I have also downloaded the service manuals last year when I wanted to recap them. It also has the flat amp circuit, but you know that already obviously.

 

[PENG]

I know we usually qualify that with "competently designed, in proper working order, and operated within its design capabilities"

True, we also typically mentioned in pure direct mode, or with no eq no dsp blabalba etc etc.. With DSP involved, all bets are off, especially if done sighted.

 

[PENG]

Interesting that even Doug Self was brought into that discussion.

Are you experiencing negative inflation with that 0.002 cent?

You got me interested enough that I decided to re-read one of his article about his demo.

http://www.silcom.com/~aludwig/Amplifier_distortion.htm

I found that article more relevant to the THD criteria for amps than Geddes Matthew linked, because Doug's findings were directly applicable to the topic whereas Geddes tended to focus on THD/IMD related to speakers. Neither one seemed to contradicted one of the common belief that all else being equal, an amp designed to be accurate/well designed, if used well below (I also included the word "well" to give myself some wiggle room..) it's output capabilities...........

Both emphasized THD is not a good, or poor indicator, Geddes's were more unclear as he talked about speakers mainly and at times seemed to say while non linear distortions in speakers weren't that bad as it is in electronics, but also that for amps, non linear distortions weren't a problem anyway (again it was unclear and convoluted so I might have misunderstood him). The linked (by Matthew) article talked about how he created the files with distortions using MathCad so he's not really testing the amps in that study.

In Doug's demo, he clearly pointed out the SS amp dut could do better if powerful enough as he state in the linked article:

"Effect of Amplifier Power

Suppose the SS amp puts out 4 times the power as the SE tube amp. If you make the output volumes equal, the alpha value for the SE amp is twice as large as for the SS amp. Compared to a SS with alpha = 0.4, the SE amp with alpha = 0.8 sounds much worse. A SS with alpha = 0.2 compared to a SE with alpha = 0.4 is also better, but the difference is less dramatic. So in this situation the advantage of higher SS power outweighs the advantage of less discordant SE distortion products.

So far I have mainly tested myself on one short music segment, the Vivaldi sample. Based on this, my detection threshold is close to a value of alpha = 0.3 for the SS case, where 1% of the samples are clipped, and I am 90% accurate in selecting the distorted file. This corresponds to an average power output equal to 9% of the maximum power output. Further testing can only result in a lower threshold. I am guessing that if I run my amps at 5% of their rated power that clipping will not be audible."

So based on what he found (I don't/won't),for those who want sound quality as good as a 50 W SE tubes, can do well in a good 300 W SS amp too.

Whether one believes him and other gurus who might have made similar claims about how TDH relates to perceived sound quality, the safe bet is to get more than enough voltage/current capability to ensure the amp never clips. Also, when they claimed an amp rated much higher THD than another one rated much lower or even 0%, I think we have to take it with a grain of salt, as the caveat is, more about clipping (e.g. Doug Self),and/or where the distortion is on the waveform (Geddes metric?). I put ? behind Geddes because I found his stuff much harder to read, as he seemed (just seemed..) to jump from one thing to another, really hard not to take him out of context. In fact, he himself had said "People always take my statements out of context.." I don't want to be one of those "people".