Stereophile | January 2017 Issue

[Mike Lavigne]

Seems to be quite a match between PH.D in Analytical Chemistry vs F1 Engineer and Nuclear physicist.

That said, the best sound is with a car salesman.

 

[microstrip]

IMHO Ayon Crossfire III with Ayon's recti is one of the most underestimated SET' s available.
This recti makes a difference.
My Ayon can drive AG Duo Mezzo XD louder without audible distortion than any other tube amp I have auditioned at home. And give much more organic and lifelike sound than any other SS amp I tried.

Can you confirm me if the Ayon Crossfire tubes are specific custom tubes or generic tubes?

 

[marslo]

Can you confirm me if the Ayon Crossfire tubes are specific custom tubes or generic tubes?

As far as I know Ayon has its own tube factory in Czech Republic and the tubes like 52 B and 62 B are only made for their amps.
This recti from my picture is probably made also there but there are people who think that this is a private label made by EMS Labs for Ayon.
Will ask Gerhard Hird one day to confirm.

 

[morricab]

That's ironic as I feel like I'm swimming in distortion with my past few SET auditions. Many people prefer 3rd order over 2nd order distortion, so that is probably why. I have not been happy with SET bass practically ever. No, I have not heard a KR in my system

By perusing Cheever, appears he did a Masters thesis in 2001 for his MSEE but isn't even in the industry - amplifiers have changed quite a bit since then and he never did any followup, nor was this reviewed or presented. He opens with the infamous Bryston vs Cary Stereophile examples from the early 90s (and yes, that Cary 300SEI is a crap amp as is the Bryston). He used a Hafler and his own custom built 1.5 watt SET, not exactly the creme de la creme. It also wasn't a scientific study (5 listeners, single blinded) that utilized 16 ohm Radio Shack speakers from the 60s (lol).

http://www.dancheever.com/

So "citing" Cheever isn't a big deal to me, although he makes some interesting points that you clearly agree with. If you have any more recent studies done on distortion, I'd enjoy reading them sometime.

Not quite sure Cheever, "not being in the industry" has a tiny bit of bearing on the quality of his thesis. Amplifiers have not changed substantially since 2001, circuit topologies very similar to the Hafler are being made right now and by some very high end companies. I agree the Brystons crap but I haven't heard that Cary; however, the Cary CAD-572SE monos I once had didn't sound crap at all. You have no idea how good or not good his own amp sounds...it measured pretty good for a 1 watt 45 amp. Single blind can still be valid if precautions are taken. Clearly the speaker only needs to be good enough to allow differntiation...it doesn't have to be SOTA even though that might make the demo even more compelling. Is it perfect? no. Is the research in the intro and the formulation of his metric pretty good? I think so, yes. Did he finish the Thesis? I assume so, but you are right this is not a signed off version from his mentor and the faculty head.

Not sure what is not recent about 2001. The previous work was from like the 1970s or 80s. There was some stuff from Otala but he never really tried to build a hearing model to correlate distortions with hearing. Again, there is little new under the sun in amplifiers. The same basic circuit elements have been there for a long time now and most combinations have been tried. Forget the high end hype.

The real innovation in SET is not so much circuitry (although Aries Cerat has come up with some new ideas actually) but new tube ideas (KR, EML) and better materials to lower noise and distortion and some hybrid concepts.

There are also the Geddes papers from a few years later. They approach things quite differently but the same basic conclusion is there that THD and IMD are very poor predictors of sound quality and that the content of the distortion is more important.

Please tell, me what means "swimming in distortion" for you? What speakers? How loud? When was the last time you heard, live, unamplified instruments up close? I tell you I had two nice experiences recently and I think in both case you would be telling me you are "swimming in distortion" if I took you into a room blindfolded to hear that and told you it was from playback and not live.

 

[andromedaaudio]

halcro by andromeda61, on Flickr

 

[morricab]

Seems to be quite a match between PH.D in Analytical Chemistry vs F1 Engineer and Nuclear physicist.

That said, the best sound is with a car salesman.

If you say so...without full experience from all parties a bold statement...

 

[marslo]

Marslo, which other SET amps have you compared to?

Not so many SET's ,mainly integrated Entropy by Amare Musica , also their pre De Forest with monoblocks
But I had on audition ARC ref 150 , Einstein The amp Ultimate , Leben 1000 among those which I remember as valve designs.
Have heard at Geoffrey Audiopax twice .
ASR Emitter II, AG XA integrated and XA combo pre/power, SPEC D class amp, Vitus SIA 025, Accu E450 and E460 as SS designs.
More then 10 altogether.
No SOTA amps though.

 

[bonzo75]

Not so many SET's ,mainly integrated Entropy by Amare Musica , also their pre De Forest with monoblocks
But I had on audition ARC ref 150 , Einstein The amp Ultimate , Leben 1000 among those which I remember as valve designs.
Have heard at Geoffrey Audiopax twice .
ASR Emitter II, AG XA integrated and XA combo pre/power, SPEC D class amp, Vitus SIA 025 as SS designs .
More then 10 alltogether.
No SOTA amps though.

Wow that's a good set

 

[morricab]

Lots of people love the sort of equipment like SET amplifiers and record players which have audible levels of euphonic distortion, often lots. Since most modern recordings are mixed from multiple microphones placed much nearer to the instruments than any listener (other than the player themselves) ever listens I think the concert hall experience can perhaps be more realistically imagined if these sorts of harmonic additions are made.
SET amps are very simple circuits, which would be a good thing IMO if they were also linear, but they are not, usually nowhere near.
I personally believe that these harmonic additions are audible and that is why some people like this sort of equipment (and speakers which have the same type of additions).

"Natural sound" may have to be simulated in recordings where microphone choice and position makes the raw recording somewhat unreal, as is often the case.

In my own amateur recording efforts I have found that microphone choice and position makes the most difference to sound quality, with the recorder less important. If that is not well done, and IME in many recordings it is not, a linear low distortion replay system will not sound as nice as one with loads of built in euphonic harmonic additions.

0.5% 2nd harmonic distortion definitely changes the timbre of a musical instrument.

Frequently I see good reviews of equipment which has major non-linearities in Stereophile. There is the age old question of "does it sound nice because of the distortion or despite it". I am firmly in the camp of those who believe the former.

See my two posts from a few weeks ago where I explain very clearly that I was very up close at two classical concerts and neither sounded like what you are describing...both sounded full and rich and present...I lived with a profi practicing her violin daily for 4 years (and it was always in the $1M + category of old violin) and it never had that exaggerated, wiry lacking body sound that people will blame on the close miking. Never.

Natural sound doesn't need to be simulated...most recordings actually have more in there than most people think but exaggerated high order harmonics give too much "leading edge" and make these high frequency effects too loud (high order harmonics is perceived as both "metallic" in nature and louder). It also impacts image dimensionality and soundstage perception (again exaggerating loudness makes spatial cues harder to process).

I personally believe it is the LACK of this kind of distortion is the big reason why SET has recaught fire in the audiophile world. Without high order harmonics you get more believable dimensionality and space and more palpbable imaging and dynamics breathe more easily...something I think has to do with lack of negative feedback and is based on the observation in the late 1950s by Norman Crowhurst that feedback amplifiers effectively create an artificial noise floor that is modulated with the signal so it is not true noise. That damages low level resolution and microdynamics.

BTW, Keith Howard effectively debunked the euphonic distortion non-sense in his Stereophile article "Naughty but Nice?". Go look it up. He wrote some code that allowed him to add distortion to digital recordings and he tried different patterns and different levels. His finding? That no distortion is the best (hardly revelatory but needed to be done I guess); however, the next best was a monotonic pattern of even in odd harmonics in an exponential decay. The worst was all odd harmonics, typical of a push-pull feedback amplifier. Higher levels sounded worse than lower ones. Nothing added was more "euphonic" was his conclusion...now it is just one man's opinion, as KeithR is fond of saying that maybe someone prefers the disonance...I can't dispute that because I guess there is always someone who likes it that way.

 

[microstrip]

That's ironic as I feel like I'm swimming in distortion with my past few SET auditions. Many people prefer 3rd order over 2nd order distortion, so that is probably why. I have not been happy with SET bass practically ever. No, I have not heard a KR in my system

By perusing Cheever, appears he did a Masters thesis in 2001 for his MSEE but isn't even in the industry - amplifiers have changed quite a bit since then and he never did any followup, nor was this reviewed or presented. He opens with the infamous Bryston vs Cary Stereophile examples from the early 90s (and yes, that Cary 300SEI is a crap amp as is the Bryston). He used a Hafler and his own custom built 1.5 watt SET, not exactly the creme de la creme. It also wasn't a scientific study (5 listeners, single blinded) that utilized 16 ohm Radio Shack speakers from the 60s (lol).

http://www.dancheever.com/

So "citing" Cheever isn't a big deal to me, although he makes some interesting points that you clearly agree with. If you have any more recent studies done on distortion, I'd enjoy reading them sometime.

As you point Cheever work is a Master Thesis, nothing else. Anyone reading if fully can see it was supposed to be a starting point, proving he has the competence and capacity to go on more in depth studies. It is outdated, and as we can expect most of the time from a master thesis, superficial in fundamental aspects - no one has time to carry significant work in such a reduced period.

It would be great if someone decided to take a PhD level thesis on this subject, particularly now we have access to much better analyzers and software tools but I have little hope of seeing such thing. People look for subjects that are not a lifetime battle, and particularly, the resources needed to carry the subjective part with success are enormous. Besides, imagine what the poor guy would feel if the thesis examiner is a member of the "all the decently designed amplifiers sound the same" club?

 

[morricab]

Can you confirm me if the Ayon Crossfire tubes are specific custom tubes or generic tubes?

As I have one I will comment. The output tubes are custom Ayon designed 300B derivatives (called 62B by Ayon). They are large, with plate structures the size of a 211 or 845 and similar dissipation capabilities in terms of watts. They have lower output impedances though than big transmitter tubes, more like 300Bs, so constructionwise they are more 300B like.

The rest of the tubes in the amp commercially available types, NOS 6SJ7s and new or NOS 6N30s from Russia. The 5U4G can be either new or NOS. I currently run NOS Svetlana "winged C" but I have tried the Ayon branded 5U4G and it is better (it is also twice as big with much bigger plates and I would guess capacity as well). I think the Ayon one is a rebranded Emission Labs mesh plate rectifier tube. No matter, it is better than my NOS rectifiers.

 

[morricab]

Wow that's a good set

Agree, that is a good set...quite diverse as a comparison and some nice products in there as well. He thinks the Crossfire beats all those? (well I think it is darn good too)

 

[morricab]

As you point Cheever work is a Master Thesis, nothing else. Anyone reading if fully can see it was supposed to be a starting point, proving he has the competence and capacity to go on more in depth studies. It is outdated, and as we can expect most of the time from a master thesis, superficial in fundamental aspects - no one has time to carry significant work in such a reduced period.

It would be great if someone decided to take a PhD level thesis on this subject, particularly now we have access to much better analyzers and software tools but I have little hope of seeing such thing. People look for subjects that are not a lifetime battle, and particularly, the resources needed to carry the subjective part with success are enormous. Besides, imagine what the poor guy would feel if the thesis examiner is a member of the "all the decently designed amplifiers sound the same" club?

I have a Ph.D student working for me now on his thesis in pharmaceutical research...should I change the course of his research from laser Mass spectrometry to audibility of distortion studies??...I sure would like to...

 

[microstrip]

As I have one I will comment. The output tubes are custom Ayon designed 300B derivatives (called 62B by Ayon). They are large, with plate structures the size of a 211 or 845 and similar dissipation capabilities in terms of watts. They have lower output impedances though than big transmitter tubes, more like 300Bs, so constructionwise they are more 300B like.

The rest of the tubes in the amp commercially available types, NOS 6SJ7s and new or NOS 6N30s from Russia. The 5U4G can be either new or NOS. I currently run NOS Svetlana "winged C" but I have tried the Ayon branded 5U4G and it is better (it is also twice as big with much bigger plates and I would guess capacity as well). I think the Ayon one is a rebranded Emission Labs mesh plate rectifier tube. No matter, it is better than my NOS rectifiers.

Thanks. Is there any second source for them? How expensive are the 62B's?

 

[morricab]

Thanks. Is there any second source for them? How expensive are the 62B's?

I am not even sure that they can be bought privately unless you own an Ayon amp. You would have to contact Ayon directly as they are not sold separately on retail like KR tubes to the best of my knowledge. I think they are about 1000 euro a pair though when obtainable.

 

[marslo]

As I have one I will comment. The output tubes are custom Ayon designed 300B derivatives (called 62B by Ayon). They are large, with plate structures the size of a 211 or 845 and similar dissipation capabilities in terms of watts. They have lower output impedances though than big transmitter tubes, more like 300Bs, so constructionwise they are more 300B like.

The rest of the tubes in the amp commercially available types, NOS 6SJ7s and new or NOS 6N30s from Russia. The 5U4G can be either new or NOS. I currently run NOS Svetlana "winged C" but I have tried the Ayon branded 5U4G and it is better (it is also twice as big with much bigger plates and I would guess capacity as well). I think the Ayon one is a rebranded Emission Labs mesh plate rectifier tube. No matter, it is better than my NOS rectifiers.

This is also what I heard but the shape of EML 5U4G is different so I am not 100 % sure.

 

[LL21]

Following from Bonzo's point about the complexity of the debate going on here, speaking for myself, I am enjoying the tennis match. While i cannot say i fully appreciate the technicalities of what is being stated, i admit it is fun to watch and try to pick up a few things along the way.

No doubt, this debate could go on even among sophisticated equipment designers who no doubt each have their own philosophy (Aries Cerat, Ayon, Atmasphere, Gryphon, CJ, Zanden, Kondo, Vitus, Nelson Pass, et al) which is based on a mix of technical fact, overview of acoustical science, personal opinion, personal priorities and I suspect no small amount of personal hypothesis/theory.

 

[morricab]

Following from Bonzo's point about the complexity of the debate going on here, speaking for myself, I am enjoying the tennis match. While i cannot say i fully appreciate the technicalities of what is being stated, i admit it is fun to watch and try to pick up a few things along the way.

No doubt, this debate could go on even among sophisticated equipment designers who no doubt each have their own philosophy (Aries Cerat, Ayon, Atmasphere, Gryphon, CJ, Zanden, Kondo, Vitus, Nelson Pass, et al) which is based on a mix of technical fact, overview of acoustical science, personal opinion, personal priorities and I suspect no small amount of personal hypothesis/theory.

For sure...don't forget Lamm, who claimed (not sure if it is still true) that he designed his gear based on his own hearing models and didn't do listening "tuning" at all. He felt that once the design hit the right measurement target it will sound "right"...at least that was supposedly for the original ML2, which was universally praised for its sound and is still a damn good amp today. If you don't need more than 18 watts then it is possible that if you bought one...you were done.

 

[microstrip]

I am not even sure that they can be bought privately unless you own an Ayon amp. You would have to contact Ayon directly as they are not sold separately on retail like KR tubes to the best of my knowledge. I think they are about 1000 euro a pair though when obtainable.

Do they have lifetime estimations? I must say I dislike being completely dependent on a single manufacturer.

 

[bonzo75]

So, not liking 300bs, I once tried the Ayon 32B in my Big 7 in a system near to me in London. The 32bs are steroidal 300bs, and then it was by far the best sound I had heard from a Lampi. We were swapping valves, and this just jumped up 20 levels. There was some sort of gain match. So I got all excited, but then when we tried in Bill's system, it distorted, and Lukasz said it could potentially be dangerous for the dac, as it required excess voltage - don't exactly remember.