Stereophile | January 2017 Issue

[Believe High Fidelity]

You make some very good points here. I have not heard this vaunted "authority" claimed for high powered amps compared to good quality lower powered ones...as long as both amps are running below clipping. This is an audio myth IMO. If a speaker is cruising at < 5 watts with peaks to 20 watts, then a good 50 or 60 watt SET (an expensive proposition to be sure) will have ample head room to deliver those peaks without sounding stressed.

I also agree that hte power supply is of key importance in making sure the amplifier is delivering as it should regardless of demands...as long as it does not exceed its absolute demands. I have seen this also with preamps where a highly regulated preamp will deliver more precise sound with more "balls" than other preamps...despite the argument that not much current is needed to be delivered.

I have had a number of experiences where the low powered amp sounded at least as powerful as the "beast" amp. The best example was a KR VA350i on an Acapella Violon speaker where it sounded far powerful and authoritative (especially in the bass surprisingly) than a 225 watt Electrocompaniet amp, which totally thinned out the bass...we were shocked. I have also heard a 30 watt PP Class A tube amp out muscle 130 watt Class AB tube monos on the Thiel CS3.7...another shock.

This could very well be mission statement behind the vast weight of the Aries Cerat Lineup in one area.

If it is one question I get more than anything is why the weight? 120lbs in a DAC is overkiill, 300lbs for 65W why?, etc

Somehow the mainstream design is that because you make a low powered tube amp means you have to power/input stage to match. Less complexity can attribute to better sound but I have yet to see a benchmark of what is enough for said Wattage output.

....and then of course he does add another 30lb to give plenty of reserve for peaks and transients

 

[microstrip]

That is true but I know of two notable exceptions: One is the Lamm ML2 18 watt SET monos. They use one 6C33C as an output tube and the other as a regulator. The other is the NAT Symbiosis SE that I had. I used a single large MOSFET per channel (each had its one huge heat sink!) for the output stage (it was single ended) and ANOTHER identical MOSFET, also with its own identical huge heatsink per channel, as a regulator. That meant 800 watts of continuous consumption from this monster. But what a cool concept.

Most other amps, as you have pointed out, simply don't regulate the output stage. Input and driver, increasingly common.

And some also have separate power supplies, including the power transformer, for the input and the driver.
One of best known power amplifiers with output regulation in Naim. Krell also used it in its Evolution One.

 

[853guy]

Whenever I have looked at the meters of Mike's Dartzeel, Justin's Accuphase on Apogees, and Luxman, they do way more than 1 - 3 watts.

The avant-garde engineer was asked to design the amp that would best run the speakers, he chose to do a class A 150w. Mike at AS has been selling a few demoing them with Vitus. One designer recommends his 80w valve SET for his horns. It all varies. No one agrees.

Yes, believe Mike at AS has said 300B Tri monoblocks are great but have nothing in reserve on his 104db sensitive Duos - in fact, I believe he prefers Vac 200iq P/P monoblocks on them.

I believe DDK on this forum has said Lamm 2.1s aren't enough power for AGs as well (18 watts). Ralph (Atmasphere) runs his 60 watt OTLs on 100db Classic Audio Reproductions speakers.

There clearly is variance of opinion.

Thanks for the compliments...just want to make sure that people realize that there is a lot more to good sound than "text book" engineering in particular because human perception is involved and it is not like designing car tires...

It does vary. There’s definitely variance of opinion. Human perception isn’t exactly like designing car tires…

And yet, it kinda is.

I think what I love about this thread is that aside from a lot of talk about horns and SETs, mostly what’s been discussed are real-world results. MikeL’s results are real, but there’s only one way to get them. We can pontificate from afar as to whether a bunch of ceramic drivers mated to a ribbon in a dual tower configuration can possibly do “X” or “Y”, but all that matters is “Z” - the result. You can design a tire in a computer and debate its theoretical parameters and then put it on a rolling road and measure every aspect possible, but it’s never going to give you the same result as this (check out the loading of Nico’s left-rear tire):



As someone who experienced first-hand the the Kassandra, Impera II and Concero 65 taking digits via a pathetic USB cable straight out of a tablet and sending current down some bits of wire into a speaker that would not make my personal 100, let alone my top 10 - while sitting on top of one another on the carpet in a tiny hotel room - elevate itself into one the most intensely ludicrous experiences I’ve ever had, I discovered that for all my theorising about amorphous core bifilar interstage-transformers and the “superiority” of resistor ladder network conversion, nothing comes close to laying down theory and opinion on the altar of being and event.

I love talk. It’s cheap - and it’s fun. But in the real world - in a non-linear and asymmetrical environment - all that matters is the result.

Love and hugs,

853guy

 

[bonzo75]

Nice to know that vinyl is not the exclusive land of horns. My only excellent exposition to horns was also digital.

There are a lot of demos of trios, Acapella, Anima, on digital. I too heard Cessaro on digital. Sure, WE and David's system sound much better on vinyl than they do on digital. But so will a SS+low sensitivity Apogee. SPUs seem to work better with horns though, and it is said that started in Japan where people used to sit very close to large horns in small rooms, and the SPU helped with the very high high frequency energy (as opposed to having a Lyra on it at that distance).

 

[LL21]

Hello opus.100% correct about the 'correct' measurements,what they really mean and if are helpful.Measurements are what statistics are for politics.On official way to lie.

Now,the easiest thing when designing an amplifier,is to make one with low THD.Just add gain,and feedback.and there you have it.Designing a good sounding amplifier is another story.
So the question should be,is it easier to design a good sounding amplifier or a low THD amplifier?Answer is of course the latter.A 50 dollar chip amp measures great.So what?

The fact that low efficiency speakers sound more compressed than high efficiency speakers like horns,have nothing to do with what is driving them,but the physics behind the transducers themselves.

Example.A flea 2W amplifier with minuscule PSU capacitors,coupled to a 100Hz flare tractrix horn,will blow away a conventional cone driver driven by an SS amplifier with welding spec output stage,and there are many SS like that out there.

The 100Hz-flared horn,will present 0.2% H2 at 115db for the two octaves it covers,where as the conventional driver will be near meltdown at that levels,let alone measuring distortions.

It has nothing to do with what amplifier is used.The energy transfer efficiency for the horn is couple of magnitudes higher than a typical midbass cone driver.Imagine a compression driver's voice coil barely moving inside it's magnetic gap,producing much higher acoustic energy than a conventional cone driver moving at it's Xmax limits (gross distortions) and it's voice coil dissipating triple digits Watt(dynamic compression).

Best
Stavros

Hi Stavros...now THAT is interesting. Thank you for that, and I have very much appreciated the incredibly sophisticated level of discussion this thread is taking to share and teach many of us who love audio have and [in my case] have that "a little knowledge is a dangerous thing" problem which we are working on solving.

 

[microstrip]

There are a lot of demos of trios, Acapella, Anima, on digital. I too heard Cessaro on digital. Sure, WE and David's system sound much better on vinyl than they do on digital. But so will a SS+low sensitivity Apogee. SPUs seem to work better with horns though, and it is said that started in Japan where people used to sit very close to large horns in small rooms, and the SPU helped with the very high high frequency energy (as opposed to having a Lyra on it at that distance).

I was not asking about demos - sorry usually I do not care about them - but regular use and users opinion. I must say that at some moment I was tempted by horns and vintage, but when I saw the high correlation between them and vinyl, my enthusiasm diminished very fast. Along recent years I am getting the feeling that people opinion on this subject is mostly due to their equipment, and even musical preference. Surely, with vinyl the possibilities and scope of fine tuning in the source are much larger. In digital thinks are more difficult - unless you go the Lampizator way and get a lot of tubes!

 

[bonzo75]

I was not asking about demos - sorry usually I do not care about them - but regular use and users opinion. I must say that at some moment I was tempted by horns and vintage, but when I saw the high correlation between them and vinyl, my enthusiasm diminished very fast. Along recent years I am getting the feeling that people opinion on this subject is mostly due to their equipment, and even musical preference. Surely, with vinyl the possibilities and scope of fine tuning in the source are much larger. In digital thinks are more difficult - unless you go the Lampizator way and get a lot of tubes!

With Lampi we end up listening to 1 or 2, across SS+Cones and SET/high sensitivity .

I think the correlation might exist for other reasons.

I enjoyed Geoffrey Armstrong's Trios with an Oppo

 

[microstrip]

(...) I love talk. It’s cheap - and it’s fun. But in the real world - in a non-linear and asymmetrical environment - all that matters is the result.

Love and hugs,

853guy

You are my hero! (This forum should have an applause smile!)

 

[microstrip]

(...) I think the correlation might exist for other reasons. (...)

Which?

 

[bonzo75]

Which?

Different tastes. Maybe there is a large number of people who thinks vinyl better than digital who also thinks horns are better than other speakers - which does not necessarily mean horns and vinyl need to be matched with each other in a system. It just happens because of a taste. Maybe a more advanced stage of the hobby?

 

[Aries Cerat]

Hi Stavros...now THAT is interesting. Thank you for that, and I have very much appreciated the incredibly sophisticated level of discussion this thread is taking to share and teach many of us who love audio have and [in my case] have that "a little knowledge is a dangerous thing" problem which we are working on solving.

Hello Lloyd
Glad to be among the forum company.though on strick time limitations

 

[Robh3606]

I must say that at some moment I was tempted by horns and vintage, but when I saw the high correlation between them and vinyl, my enthusiasm diminished very fast.

Hello Micro

Don't let that sway you! I have two horn set-ups my Array 1400 clones and my DIY set-up and they both use digital sources. The horns don't care what the source is although the owner may! Give them a shot they are lots of fun and definitely different from your run of the mil cone and dome speakers.

Rob

 

[Mike Lavigne]

For many reasons high powered amplifiers can give us "the authority and scale of the music ". But I have seen many high powered amplifiers that could sound anemic and small, irrespective of being very powerful in the specifications.

Mike, you have masterly tuned your system around the NH458. IMHO any amplifier will sound limited in your system, irrespective of being low or ultra high power!

I've been trying to figure out how I feel about your comment. there are a few different inferences I can guess at from what you wrote.

it seems you are saying that somehow only the dart 458's could work properly (at similar overall performance levels) in the context of my system. if that is your meaning, why do you think that?

is your opinion that this is because I have a darTZeel preamp?

if I choose a completely different brand preamp and amp would you then expect it could reach the same performance levels?

or is there another aspect (or aspects) of my system which cause you to say that about the dart 458's?

or....is it something(s) unique about the 458's themselves?

 

[opus112]

Hello opus.100% correct about the 'correct' measurements,what they really mean and if are helpful.Measurements are what statistics are for politics.On official way to lie.

Hi Stavros - perhaps we could re-purpose Mark Twain for audio - 'there are three kinds of lies.... and measurements'. But I feel that's a bit harsh, a measurement is just a number and numbers have no meaning.

Now,the easiest thing when designing an amplifier,is to make one with low THD.Just add gain,and feedback.and there you have it.Designing a good sounding amplifier is another story.
So the question should be,is it easier to design a good sounding amplifier or a low THD amplifier?Answer is of course the latter.A 50 dollar chip amp measures great.So what?

I agree here - it is easier to design an amp for lowest possible THD than design an amp for best SQ. Because for the former you've just got a single number to hit, you can do much of the work in a simulator. But design for best SQ requires constant build - listen - tweak loops, its time-consuming. And good sounding has to apply to the whole range of recordings, not just GWAG. And to back this up there's a principle from economics which I feel is relevant here - 'a measure which becomes a target ceases to be a good measure' (Goodhart's Law).

The fact that low efficiency speakers sound more compressed than high efficiency speakers like horns,have nothing to do with what is driving them,but the physics behind the transducers themselves.

This is news to me, I do hope I'm going to learn something.

Example.A flea 2W amplifier with minuscule PSU capacitors,coupled to a 100Hz flare tractrix horn,will blow away a conventional cone driver driven by an SS amplifier with welding spec output stage,and there are many SS like that out there.

The 100Hz-flared horn,will present 0.2% H2 at 115db for the two octaves it covers,where as the conventional driver will be near meltdown at that levels,let alone measuring distortions.

It has nothing to do with what amplifier is used.The energy transfer efficiency for the horn is couple of magnitudes higher than a typical midbass cone driver.Imagine a compression driver's voice coil barely moving inside it's magnetic gap,producing much higher acoustic energy than a conventional cone driver moving at it's Xmax limits (gross distortions) and it's voice coil dissipating triple digits Watt(dynamic compression).

Still can't see how this negates the amp factor. I accept that voice coil heating will tend to reduce dynamics - so you're saying that accounts for the _whole_ difference in dynamics? The cone driver doesn't need to reach its Xmax limit in order to notice the difference in dynamics.

 

[microstrip]

I've been trying to figure out how I feel about your comment. there are a few different inferences I can guess at from what you wrote.

it seems you are saying that somehow only the dart 458's could work properly (at similar overall performance levels) in the context of my system. if that is your meaning, why do you think that?

is your opinion that this is because I have a darTZeel preamp?

if I choose a completely different brand preamp and amp would you then expect it could reach the same performance levels?

or is there another aspect (or aspects) of my system which cause you to say that about the dart 458's?

or....is it something(s) unique about the 458's themselves?

A little bit of some of them ... Your system is so finely and precised tuned as an whole that no amplifier will have a fair opportunity against the NH458. Particularly IMHO it would be very difficult to change electronics keeping the balance you achieved between speakers, tape, vinyl and digital.

There are some aspects that also IMHO are unique to the NH458 - an equilibrium of continuousness, sweetness and authority, but keeping the texture of life music. Some systems will not show it, but I am sure you understand what I mean.

 

[LL21]

Hello Lloyd
Glad to be among the forum company.though on strick time limitations

Totally respect that...this is our goof off time when we are not working. Appreciate your taking the time here.

 

[amirm]

A little bit of some of them ... Your system is so finely and precised tuned as an whole that no amplifier will have a fair opportunity against the NH458. Particularly IMHO it would be very difficult to change electronics keeping the balance you achieved between speakers, tape, vinyl and digital.

There are some aspects that also IMHO are unique to the NH458 - an equilibrium of continuousness, sweetness and authority, but keeping the texture of life music. Some systems will not show it, but I am sure you understand what I mean.

You have not heard Mike's system. How can you tell all of this from afar?

 

[Aries Cerat]

Hi Stavros - perhaps we could re-purpose Mark Twain for audio - 'there are three kinds of lies.... and measurements'. But I feel that's a bit harsh, a measurement is just a number and numbers have no meaning.



I agree here - it is easier to design an amp for lowest possible THD than design an amp for best SQ. Because for the former you've just got a single number to hit, you can do much of the work in a simulator. But design for best SQ requires constant build - listen - tweak loops, its time-consuming. And good sounding has to apply to the whole range of recordings, not just GWAG. And to back this up there's a principle from economics which I feel is relevant here - 'a measure which becomes a target ceases to be a good measure' (Goodhart's Law).



This is news to me, I do hope I'm going to learn something.



Still can't see how this negates the amp factor. I accept that voice coil heating will tend to reduce dynamics - so you're saying that accounts for the _whole_ difference in dynamics? The cone driver doesn't need to reach its Xmax limit in order to notice the difference in dynamics.

Hi

It is not just the voice coil heat up or the Xmax limitations.These were only an example of how far apart can two transducers be in their effectiveness converting electrical energy to acoustic energy.
That is why i said even a tiny 2W would beat a welder-spec amplifier if the latter drove X and the former Y.

The example's horn will convert 50% of it's input energy into acoustic energy,where as the typical midwoofer would struggle at 0.4% (not a typo).

You cannot expect a driver to receive 125 times the energy,and behave as linear.Unfortunately there is no way around it.And i say unfortunately, because it would be nice to have an alternative to a half a tonne midbass horn.....but when you hear one,there is no going back

 

[marslo]

Let me congratulate all participants of this thread for the high level of discussion.
I learn a lot from it .
The post of Stavros is absolutely key for me and explains what I experience with my setup for last 6 years.
I hear less compression and more palpable music with highly effective speakers ( 109 db) driven by rellatively small power SET amp( 30 Watt per channel)
I would also point out that the very important factor is the gain of the source in the chain, for example my Lampi is very easy load for the amp and thanks to that the amp operates at very low power with fair SPL.
Selection of the right amp is also -or maybe mainly - taking in consideration the synergy of the setup.
From the other hand the Einstein The Amp Ultimate (OTL) sounded best for me from above 10 different amps I tried as far as the coherency, timbral accuracy, dynamics and other audiphile criteria are concerned but had to much gain for my setup and I couldn't listen to it more than 20 minuts.
I share also the experience that some of more powerful SS designs had less authority and slamm than weak SET's , and interistingly- it concerned mainly low and mid bass.

 

[morricab]

Hi

It is not just the voice coil heat up or the Xmax limitations.These were only an example of how far apart can two transducers be in their effectiveness converting electrical energy to acoustic energy.
That is why i said even a tiny 2W would beat a welder-spec amplifier if the latter drove X and the former Y.

The example's horn will convert 50% of it's input energy into acoustic energy,where as the typical midwoofer would struggle at 0.4% (not a typo).

You cannot expect a driver to receive 125 times the energy,and behave as linear.Unfortunately there is no way around it.And i say unfortunately, because it would be nice to have an alternative to a half a tonne midbass horn.....but when you hear one,there is no going back

Compression of low sensitivity drivers actually sets in well before they have reached their power handling capacity. You get both thermal compression and dynamic compression effects.

One thing a lot of people don't talk about though is at the other end of the loudness spectrum... the quiet end. It's down at that end that another kind of compression occurs. Specifically, soft sounds simply don't get efficiently transformed into cone motion. The signal comes through and the cone does not respond or responds inadequately to the signal magnitude.

This is due to electrical inefficiency and simple mechanical resistance to motion from relative rest. I heard this with a number of insensitive speakers where they really fall apart at low volume and coupled with a compressed high volume end means that their overall dynamic range where they perform well is narrow.

An exception I have heard to this is electrostats where they do soft very well and so are only limited at one end (loud) of the dynamic range spectrum. They sound less compressed than boxes of similar sensitivity but less than horns.