[Carl V]

Most SE amps set the magnetic flux at half the saturation flux (north or south) and never go through zero flux, where PP amps are going from full north to full south through zero all the time.

SE output transformers do not suffer from the fact that most PP output transformers have un-equal coupling from each half of the primary to the secondary at high frequencies. SE transformers have a whole primary rather than two halves.  This is something I have worked hard to correct in my output transformers.

I am repeating what I think I heard the Designer of E.A.R. Tim de Paravicini mention in a show environment
that the transformer is the culprit in many amp desgins...or at least the weak link. 
And in PP pentodes or UL designs that the very low freq. can raise havoc.

 I had just purchased an EAR890 amp.  I was impressed with his design & to be honest I was impressed by his knowledge & ability to explain his design choices,....not unlike Roger's straight forward style.

[Roger A. Modjeski]

A good SET may have a lot more distortion than SS with a ton of feedback, but a good SET produces mainly 2nd order harmonic distortion which our brain isn't very sensitive to. SS amps (in general) produce more 3rd order and higher multiples which our brain is very sensitive to. So an amp that produces much more 2nd order harmonics with very little odd order or higher multiples of even order distortion may sound better than an amp that produces far less overall distortion, but of 3rd order with higher multiples. So distortion figures mean very little without seeing the FFT chart of what kind of distortion is being produced. I'm not sure the high voltage tubes run at makes too much of a difference, push-pull tube amps produce more 3rd order distortion and can be designed with feedback to lower THD to levels similar to an SS amp.

Keep in mind what I have said many times about large second harmonic distortion. That means large IM also and when pushed above several percent any complex (orchestral)  music will get muddy while simple (solo) music will be ok. I do agree that small amounts of high order, especially crossover distortion often found in SS amps is very annoying on all types of music and interestingly enough on the opposite types being most audible on solo music and less on complex music. In consulting for Seymour Duncan (guitar pickup maker) we found that a SS amp with moderate crossover distortion was just fine for lead electric guitar (where overdrive distortion is typical) however the same output stage used on a clean "acoustic jazz" type amplifier was most annoying and only on particular notes.  The result sounded a lot like "fret buzz". When we increased the bias to eliminate the crossover distortion the buzz went away. We did nothing else. Most SS amps have crossover distortion to some extent and it is most often seen at high frequencies and most often not talked about. Crossover distortion is the worst distortion of all and does have an almost infinite harmonic series. It should be discussed as a class of its own and not to be compared to 3rd and higher distortions. Frankly 3rd alone is not much different than second. The most important thing to look for is the oscilloscope trace that John Atkinson always provides (God bless him) right under the sine wave trace. If it is smooth without sharp edges that is benign. If there are sharp peaks, triangular tops or bottoms that's bad and will be annoying.

Here are the measurements along with some of Atkinson's comments. Note how the lower trace of the first graph is very smooth (other than the fuzz whose amount is exaggerated because the distortion is so low at 0.011%) This is a very quiet amp as JA points out elsewhere. If the distortion was 1% the trace would be smooth. The thing to note is that there are 3 cycles in the lower trace for each cycle in the upper trace which means that is almost all 3rd harmonic. The third cycle in the lower trace is the small amount of second. Because there are no sharp discontinuities in trace there is no evidence of high harmonics and very little 4th. I hope this lends more meaning to these traces for all.

For an example of crossover distortion (in a $55,000 amplifier) see http://www.stereophile.com/content/dan-dagostino-momentum-monoblock-power-amplifier-measurements. In figure 8 (oscilloscope traces of a sine wave and the residual distortion) JA points out that the spikes on the lower trace almost occur at the zero crossing of the sine wave (crossover distortion). In figure 9 you can see the long series of harmonics that go out to the 20th. at around -80 db. After the 3rd harmonic, the RM-200 is down 100-120 db. The SS amp also has a very rich HF IM spectrum also. Compare fig. 10 of Dan's amp to fig. 11 for the RM-200.

As to the thoughts about high voltage there is some merit especially in preamp design as the signal voltage is small compared to the large B+. However, in both power and preamps the simplicity of the circuit (at least mine are, many aren't) has more to do with the detail that is obtained. This is not to say all simple circuits are good. Even within a particular topology the choice of resistor and capacitor values, operating points, output transformer ratios tend to have a great influence. One parameter often missed in SE power amps is damping factor which if very low does not afford good control of the woofer and will also make speakers of varying impedance sound quite different from amp to amp.

[Carl V]

Here are the measurements along with some of Atkinson's comments. Note how the lower trace of the first graph is very smooth (other than the fuzz whose amount is exaggerated because the distortion is so low at 0.011%) This is a very quiet amp as JA points out elsewhere. If the distortion was 1% the trace would be smooth. The thing to note is that there are 3 cycles in the lower trace for each cycle in the upper trace which means that is almost all 3rd harmonic. The third cycle in the lower trace is the small amount of second. Because there are no sharp discontinuities in trace there is no evidence of high harmonics and very little 4th. I hope this lends more meaning to these traces for all.

What graph are we supposed to to look at the new $55K SS amp by Dan...or ?

[Roger A. Modjeski]

What graph are we supposed to to look at the new $55K SS amp by Dan...or ?

 

I added the figure numbers. Let me know if there are any questions.  When one is familiar with looking at scope traces the high order harmonics become very obvious. Those little peaks (JA calls them spikes) are evidence of high order harmonics and they are not small. I would have had a lot more to say about them if I measured an amplifier at that price point. Luckily there was a pair of amps because one died after 50 minutes at 1/3 power and JA tried but could not bring it back to life. It appears that Dan's very special, venturi shaped, solid copper heat sinks don't do their job very well as the amp overheated.  Read Dan's 3 line letter in MFGs comments (Feb 2013, p. 133). Why anyone would want this amplifier is a mystery to me.

I once again recommend Stereophile as great entertainment and Atkinson's measurements are priceless. This is the only American magazine that I trust and the only one that measures things. At one dollar a copy its a bargain. As an engineer and manufacturer I find it most amusing that these very expensive amplifiers perform poorly and are often unreliable. This is the only source I know for finding out the weakness of a product before buying it.

[mlee]

Does anyone know of a primer or handbook available for deciphering Atkinson's measurements and graphs ?  I tried the Stereophile website and could not find one; and as a subscriber, have never seen one.

Once I can recognize and understand what it is I am looking at, then the next step is to find another text or multimedia file which can tell how deviations from the norm translate into audible differences.

Jason/

[OzarkTom]

Just like the lower powered Class A SS and SET tube amps, audiophiles will learn that the lower powered Class D amps will be the future for listening. I have had 200. 300, and now a 500wpc Class D amp here, and the 15wpc Class D on 12v battery has the most transparency. On AC, it does not.

The higher powered Class D amps are definitely needed for dynamics and is the most fun when playing loud. But since that will just blow out my hearing if I play it loud everyday, the lower Class D on battery gets played the most. But I am happy to own both.

[Roger A. Modjeski]

Does anyone know of a primer or handbook available for deciphering Atkinson's measurements and graphs ?  I tried the Stereophile website and could not find one; and as a subscriber, have never seen one.

Once I can recognize and understand what it is I am looking at, then the next step is to find another text or multimedia file which can tell how deviations from the norm translate into audible differences.

Jason/

Welcome to Audiocircle. I see you are new and I approved you... I suppose that's part of the process now. Your suggestion is a good one and perhaps you could write John Akinson and suggest someone (perhaps me) write one. You can say I offered. While (not always) easy for me since I make the same kind of measurements, they can look like hieroglyphics to someone who does not know his way around a test bench. I doubt 10% of his readers get the whole thing. For now I will take them as they come to point out certain things. Did you understand how the residual distortion trace of Dan's amp has problems that are likely audible? They are evidence of grit that is often associated with transistor amps.

Although Atkinson is kind with his criticism of faults but if you read between the lines, or often lack there or, at the last paragraph he usually slips something in such as in the review of the CJ in December 2011 where he said "CJ's LP125M measures well for a traditional tube amplifier". Whereas in the same issue he said. As did the original version of the RM-200, the MK II edition offers superb performance for a tubes design, particularly in terms of (low) midband distortion and the ability to drive low impedances. And I remain impressed by the quality of its output transformers". As he was only testing one of a monoblock he did not encounter the hum problems which caused Robert Deutch considerable trouble which was only solved by floating the ground of one amp. For $9,500 a pair you would think that CJ could have solved this problem as there are several solutions better than having to float one amp which does not always work and is a bit dangerous. But then isn't is a bit dangerous for two Dr's of economics to be designing amplifiers.

[tull skull]

"But then isn't is a bit dangerous for two Dr's of economics to be designing amplifiers.".................................

Love it!!!

[steve f]

Unfortunately there are far too many amps designed by everyone from Economic Phd's to carpenters. There are too many audiophiles who don't approve of designers with EE or physics credentials. I don't get it. No, you don't need an engineering degree to design a working amp, but how in heck can you learn all of the foundation knowledge without an education in the area? Why do a lot of audiophiles assume that someone with an engineering degree can't have the listening skills of a music loving audiophile? I've read many comments over the years that insult EE grads. I've never replied to those posts because frankly they are ludicrous statements. The very best amps I've auditioned over the years were designed by guys like Pass, Modjeski, and Rozenblit, not ARC or CJ. Then there is the audiophile fascination with boutique parts, biwiring, cryogenics, power cords, etc. Laws of physics need not apply.

I am aware of only two people who take measurements of high end equipment, Atkinson in stereophile, and Aczel in the audio critic. I suppose we audiophiles are to blame, because we don't ask anyone to verify listening tests. There is a great need for a how to understand basic test primer. For most of us, it's the window dressing of the review, after we turn to the how does it sound part.

[G Georgopoulos]

But then isn't is a bit dangerous for two Dr's of economics to be designing amplifiers.

Hi Roger,long time no see.I agree with you that two Drs of economics designing amplifiers is a bit gangerous
especialy with high voltage tubes...

kind regards 

[mlee]

... perhaps you could write John Akinson and suggest someone (perhaps me) write one...

Thank you.

My connection to John Atkinson is marked by one degree of separation.  The person whom I know is also adamant about measurements anchoring a reviewer’s opinion and he was also in the industry for quite a while, but he has no easy primer that he could refer to me.  As for an audible  translation, he said that although he agrees that it would be instructive, he could not generate enough interest and present sound demonstrations to a group.

I believe a starting point is needed, and once introduced, may lead to greater involvement and reliance in the community upon faithful and reproducible measurements, at least as a counterpoint if not a grounding to what we believe we hear.  Mr Atkinson would be a good choice, as well as you, Mr Modjeski, as you have the interest in the subject and since I can only provide intermittent involvement, not being in the industry.  But the service to the community is there, at least to have a number of subscribers start and recognize a common “scientific” vocabulary other than the worded opinions we are so familiar with.

Btw, I also understand the necessity of the tightrope balancing act of Mr Atkinson in his measured reviews.  I will consult with my degree of separation and will follow up.  Regards.

[Ericus Rex]

Roger,

I can't imagine JA wouldn't take you up on that offer to write up an explanation on measurements given your expertise and his respect for your work.

I have emailed JA about this.  I hope it works out!

[Ericus Rex]

Looks like a Stereophile article is a no go.  JA replied that allowing a manufacturer to write the article for the mag would put him in hot water with the other manufacturers.  He did agree that a primer was needed so maybe he'll write it himself.

Roger, maybe your article could 'published' here on AC?  It could be made a 'sticky' in the Lab section for all to see?

[BobRex]

Looks like a Stereophile article is a no go.  JA replied that allowing a manufacturer to write the article for the mag would put him in hot water with the other manufacturers.  He did agree that a primer was needed so maybe he'll write it himself.

Roger, maybe your article could 'published' here on AC?  It could be made a 'sticky' in the Lab section for all to see?

JA has done it before - Scott Frankland wrote an article on single ended topology a few years ago.  Don't know if he got a load of crap for that one, but I'm betting he did.

[Roger A. Modjeski]

Looks like a Stereophile article is a no go.  JA replied that allowing a manufacturer to write the article for the mag would put him in hot water with the other manufacturers.  He did agree that a primer was needed so maybe he'll write it himself.

Roger, maybe your article could 'published' here on AC?  It could be made a 'sticky' in the Lab section for all to see?

 

I do see JAs prediciment. I will see what I can do here perhaps bit by bit or when I get a chunk of time, which is rare with all the projects going on.

[harley52]

The people building equiptment need all the info they can get. They are the exception to my logic. You folks are making the joy of listening to music, way too complicated. Keep it Simple Silly.