Atma-Sphere Class D Mono blocks

[Gregadd]

Hm. Maybe that's the issue I've been seeing. On the one hand, the 'only tube' people have trouble understanding that the class D sounds the way it does, OTOH the measurement guys don't like to see what we measured. FWIW this was never a problem with the tube amps!

I have heard class D. Maybe yours is better than the rest.

 

[Atmasphere]

Well Ralph since you have gone live with this experiment, we will see how it plays out in real time.
I think the snakes is taking a huge gamble. It is trading short term nutrition for long term survival. Certain animals are able to regenerate parts of their anatomy.

OK- that isn't the Auroboros that I was acquainted with...

With tube procurement having become much harder in the last 2 years, this really came in the nick of time. From our perspective though there was/is also an integrity issue; we did not want to offer something that we felt was not up to our standards. So the amp got a lot of scrutiny before it went out the door.

There is one issue that I've been aware of for decades. In moving to a solid state amp, we are doing what a good number of other OTL manufacturers have done. Every single one of them went out of business. I think the difference here is that we've been around for nearly 50 years whereas those others I mentioned weren't even around for 10, with the exception of Counterpoint, who went out on account of a divorce.

 

[morricab]

I've stated a good number of times (I think even on this thread) that the class D sounds a lot like our OTLs (if casual listening at work I have to look to see which is hooked up). While it has the same liquidity and smoothness in the mids and highs, in all of our listening tests and confirmed by customer feedback, the big tell that its the class D is that nice midrange is accompanied by greater transparency. The bass is different simply because the amp behaves as a proper voltage source so deals with the high impedance peak in the bass that is common with 99% of speakers made differently. Because our OTLs have a high output impedance on account of no feedback (similar to SETs using no feedback), on speakers that work with SETs the class D will sound likely sound a bit leaner by comparison. This is the case on my CARs, but they have level controls on the back (which is common with many speakers from the Power Paradigm era, like vintage EVs, JBLs or Altecs) so I was able to correct the midrange and tweeter levels so as to restore the proper tonality.

Because of the voltage source aspect, I did not expect the class D to work so well with ESLs, but we now have reliable reports that it works fine with old Quads and also with Sound Labs. The Quad owner felt that we were displacing the M-60 as being one of the better amps for use on that speaker, but that was the point in general, although we didn't expect that particular result I'll take it.

FWIW, my speakers at home employ beryllium diaphragms in the mids and highs (Classic Audio Loudspeakers T-3). I like the pure beryllium because the first breakup in the speaker is at about 35KHz, so they are very smooth and detailed.

Bob would have gained more traction at the time if he had simply pointed out that the distortion spectra of any amplifier is also its 'sonic signature'; the main difference we hear between amps (the minor difference being FR).

We are actually trying to get the distortion down as much as we can since its always barrier to transparency/detail. So zero intentional distortion. As I've said before (I can't speak for other designs), it worked out that the non-linearities in the amp (the encoding scheme and the deadtime in the output section; the latter of which dominates the distortion generation) tend to generate lower ordered harmonics rather than higher ordered. So it has a prominent 2nd and 3rd harmonic, its distortion spectra being somewhere between that of an SET and our OTLs in that regard (IOW, the 2nd is dominant but not as much as seen in an SET, whereas in our OTLs the 3rd dominates), but considerably lower overall; about 1/100th that of a and about 1/10th that of our OTLs.

Since the amp drives a much wider range of speakers than our OTLs and also since it sounds better on 99% of speakers made (and is more transparent on all of them), I don't see it as a lateral move at all. Any time you can get greater detail without sacrificing a smooth presentation its a step forward.

I have heard no correlation whatsoever between an amps measured distortion and the listeners perception of transparency. Yet, that seems to be what you are saying here. I have heard many of the best measuring amps in existence and they don't sound more transparent...apparent detail perhaps (but this seems more artifact than real resolution). Your amps are not the lowest distortion out there, nor are they the only Push/pull amps with some residual 2nd order distortion that is actually higher than the 3rd or subsequent higher orders...they don't sound remotely like your or other OTLs or a tube amp in general.

Audionet Max monoblock power amplifier Measurements | Stereophile.com

Please note especially the 1Khz FFT. 2nd is at -110db and everything else is lower. Audible? According to you no.


Benchmark Media Systems AHB2 power amplifier Measurements | Stereophile.com

Only a bit of 3rd is above -120db in 50Hz FFT but nothing is above -125db. Audible or is this amp functionally perfect?


Boulder Amplifiers 2150 monoblock power amplifier Measurements | Stereophile.com

Basically nothing above -120db.y

I guess you didn't actually read my post. I quote: "Just ordered a pair from USA Tube Audio, Scottsdale, AZ.. In a month or so, after maybe 300-or-so hours of burn-in...".. Since 'a month or so' has not transpired, I have no idea what differences I'll hear.. Or were you thinking/writing about some other Atma-Sphere poweramps?

Yes the OTLs

 

[Atmasphere]

I have heard no correlation whatsoever between an amps measured distortion and the listeners perception of transparency. Yet, that seems to be what you are saying here. I have heard many of the best measuring amps in existence and they don't sound more transparent...apparent detail perhaps (but this seems more artifact than real resolution). Your amps are not the lowest distortion out there, nor are they the only Push/pull amps with some residual 2nd order distortion that is actually higher than the 3rd or subsequent higher orders...they don't sound remotely like your or other OTLs or a tube amp in general.

Audionet Max monoblock power amplifier Measurements | Stereophile.com

Please note especially the 1Khz FFT. 2nd is at -110db and everything else is lower. Audible? According to you no.


Benchmark Media Systems AHB2 power amplifier Measurements | Stereophile.com

Only a bit of 3rd is above -120db in 50Hz FFT but nothing is above -125db. Audible or is this amp functionally perfect?


Boulder Amplifiers 2150 monoblock power amplifier Measurements | Stereophile.com

Basically nothing above -120db.y

OK- so we're doing this again. None of the amps you list conform to what I say needs to happen:

The Audionet does not seem like its higher orders are unmasked. We are looking for an exponential decay of succeeding orders and I don't see that. In addition distortion is rising with frequency. Distortion rising with frequency isn't going to be an amp that sounds like a tube amp with no feedback! FWIW dept.: in order to have the distortion vs frequency curve to be flat, you either run no feedback at all (such as in an SET) or you have enough Gain Bandwidth Product to support your feedback so that it does not rise with frequency. Apparently that is very hard to do with a class A or AB solid state design!

The Benchmark is interesting- look how it too has distortion rising with frequency. Honestly I didn't expect that since this amp otherwise has very low distortion.

We see this again happening in the Boulder with the 2nd and 3rd being insufficient to mask higher orders and the harmonics not falling off on an exponential curve.

So to my mind none of these amps would sound particularly musical, although the Benchmark has low enough distortion that would explain why the user reports are so varied. If you used a tube preamp with this you might like it just fine.

I've pointed all this out before WRT to distortion spectra, exponential decay (see Jean Hiraga) and distortion vs frequency; the latter is probably the easiest tell. So I agree- looking at the measurements I don't think these amps would sound like our gear. I withhold my comments about what I think they might sound like.

 

[Gregadd]

OK- that isn't the Auroboros that I was acquainted with...

With tube procurement having become much harder in the last 2 years, this really came in the nick of time. From our perspective though there was/is also an integrity issue; we did not want to offer something that we felt was not up to our standards. So the amp got a lot of scrutiny before it went out the door.

There is one issue that I've been aware of for decades. In moving to a solid state amp, we are doing what a good number of other OTL manufacturers have done.[emphasis supplied] Every single one of them went out of business. I think the difference here is that we've been around for nearly 50 years whereas those others I mentioned weren't even around for 10, with the exception of Counterpoint, who went out on account of a divorce.Veblen Effect

Ok so you hit me with two names with watch I am not familiar Auroboros and Veblen effect Google is helping me out.
1, I assume you have put forward your best effort.
2. I hope you dent go out of business.
3. I never knew you to be one that follows the crowd.

 

[Kingrex]

I thought I heard Ralph say the real pride of his company is the front end equipment. If I didn't really love my preamp, I would have gotten an Atmasphere to have the phono preamps. I bought a phono preamp last month. I asked Ralph if they made a stand alone unit but they do not. I could not find my way to part with the one I have.

I find it odd to say this, but I hope Atmasphere can soon raise the price of their class D amps. They are too good a bargain. I don't think I see enough profit margin in them. Not compared to what I see other companies getting for equipment. Then again, maybe the volume supports the product.

 

[Atmasphere]

I find it odd to say this, but I hope Atmasphere can soon raise the price of their class D amps. They are too good a bargain. I don't think I see enough profit margin in them. Not compared to what I see other companies getting for equipment. Then again, maybe the volume supports the product.

Thanks. We price to a formula. I might have mentioned this before; it usually means its priced lower in this market. I feel a bit weird pricing to what the market will bear. I see a lot of my competition doing exactly that, some even going so far as saying that's exactly what they are doing. But I want to be able to look inside and see what I spent my money on. When priced to what the market will bear (Veblen Effect) you can't see that.

 

[morricab]

OK- so we're doing this again. None of the amps you list conform to what I say needs to happen:

The Audionet does not seem like its higher orders are unmasked. We are looking for an exponential decay of succeeding orders and I don't see that. In addition distortion is rising with frequency. Distortion rising with frequency isn't going to be an amp that sounds like a tube amp with no feedback! FWIW dept.: in order to have the distortion vs frequency curve to be flat, you either run no feedback at all (such as in an SET) or you have enough Gain Bandwidth Product to support your feedback so that it does not rise with frequency. Apparently that is very hard to do with a class A or AB solid state design!

The Benchmark is interesting- look how it too has distortion rising with frequency. Honestly I didn't expect that since this amp otherwise has very low distortion.

We see this again happening in the Boulder with the 2nd and 3rd being insufficient to mask higher orders and the harmonics not falling off on an exponential curve.

So to my mind none of these amps would sound particularly musical, although the Benchmark has low enough distortion that would explain why the user reports are so varied. If you used a tube preamp with this you might like it just fine.

I've pointed all this out before WRT to distortion spectra, exponential decay (see Jean Hiraga) and distortion vs frequency; the latter is probably the easiest tell. So I agree- looking at the measurements I don't think these amps would sound like our gear. I withhold my comments about what I think they might sound like.

You said that below -110-120 db is inaudible so why would masking then matter? I simply found examples of standard Class AB amps that basically meet that criteria. Also, if your 2nd and 3rd are down around -100db they aren’t masking anything. Your claiming an exponential decay of harmonics? Then you are the only person to ever achieve this with Class D.

 

[Atmasphere]

Also, if your 2nd and 3rd are down around -100db they aren’t masking anything.

They aren't? How did you arrive at that?

 

[Gregadd]

I hope they raise the output power. I would think 200 watts rms into 8ohms minimum.

 

[Gregadd]

Getting back to the merits. Have you changed your position on having the amplifier double as the impedance decreases?

 

[microstrip]

I hope they raise the output power. I would think 200 watts rms into 8ohms minimum.

Considering the trend of modern speakers 100W is enough for most applications, if the amplifiers are able to supply the needed current along full bandwith of the system. 200W only increases the current by of 40% compared to 100W .

 

[morricab]

They aren't? How did you arrive at that?

You were throwing these levels around and saying how much lower the distortion was compared to your OTLs, so I guessed it to be around that level. If not well that might be better or worse. Again, if each subsequent harmonic isn’t about 20dB lower than the preceding one then according to Cheever it is likely audibly degrading the sound. Also, if your distortion is not increasing with frequency, you might again have the only Class D amp in the world that doesn’t do that.

 

[Atmasphere]

Getting back to the merits. Have you changed your position on having the amplifier double as the impedance decreases?

Sort of- that's a 'depends'.

I have to admit I like to be finally able to sell amps to people that have speakers meant to be driven by a voltage source- which is about 99% of speakers made.

The reason I didn't like that concept so much with tubes is that to do it, you had to add feedback to the design so that it would behave as a voltage source. You can do this with almost any tube amp; certainly our OTLs. But that isn't the same as saying the amp sounds its best because the simple fact is if you add feedback, if you want it to sound right you have to add a really lot of it! Like over 30dB and I mean over 30dB at all frequencies. The amps @morricab listed just above do not do that and its literally impossible with a tube amp.

Its really hard to do with solid state class A or AB designs too. For most of the history of solid state amps, the semiconductors needed simply didn't exist.

But you can do that with a class D design if it is self-oscillating and can switch fast enough. So now you can have the feedback you need so it no longer makes the amp sound brighter and harsher than real life- it can have the same relaxed presentation as a zero feedback amp, but of course much lower distortion so it can be more transparent.

You were throwing these levels around and saying how much lower the distortion was compared to your OTLs, so I guessed it to be around that level. If not well that might be better or worse. Again, if each subsequent harmonic isn’t about 20dB lower than the preceding one then according to Cheever it is likely audibly degrading the sound. Also, if your distortion is not increasing with frequency, you might again have the only Class D amp in the world that doesn’t do that.

No- most self-oscillating class D amps should have that characteristic. This is because its easy to get GBP with a class D amp on account of how it makes gain, which is the relationship of amplitude between the incoming audio signal and that of the 'clock' signal in the encoding circuit.

 

[morricab]

Sort of- that's a 'depends'.

I have to admit I like to be finally able to sell amps to people that have speakers meant to be driven by a voltage source- which is about 99% of speakers made.

The reason I didn't like that concept so much with tubes is that to do it, you had to add feedback to the design so that it would behave as a voltage source. You can do this with almost any tube amp; certainly our OTLs. But that isn't the same as saying the amp sounds its best because the simple fact is if you add feedback, if you want it to sound right you have to add a really lot of it! Like over 30dB and I mean over 30dB at all frequencies. The amps @morricab listed just above do not do that and its literally impossible with a tube amp.

Its really hard to do with solid state class A or AB designs too. For most of the history of solid state amps, the semiconductors needed simply didn't exist.

But you can do that with a class D design if it is self-oscillating and can switch fast enough. So now you can have the feedback you need so it no longer makes the amp sound brighter and harsher than real life- it can have the same relaxed presentation as a zero feedback amp, but of course much lower distortion so it can be more transparent.

No- most self-oscillating class D amps should have that characteristic. This is because its easy to get GBP with a class D amp on account of how it makes gain, which is the relationship of amplitude between the incoming audio signal and that of the 'clock' signal in the encoding circuit.

Sorry, should have WHAT characteristic? Try to be clear...

 

[Kingrex]

Ralph,
What's your take on my Dartzeel NHB 108 Model 1. How are its distortion profiles. Sonically I hear the cleanest amp I have had in my system. But then again, its one of the only SS amps I have had. It does a good job of getting out of the way of the music. At times I am not fully sure this is what I want. Its confusing. I am very familiar with a tube sound as I have had quite a few models over a 7 year period. And there are aspects of it I like. But there are times they seem to push me away. The Dartzeel never pushes me away. There are times I might yearn for something a little different. But that is a lot different.

Is the Atmasphere Class D amp on paper more clean with less higher order harmonics?

 

[Tuckers]

Also, if your distortion is not increasing with frequency, you might again have the only Class D amp in the world that doesn’t do that.

Ding, ding, ding.

 

[Holmz]

Thanks. We price to a formula. I might have mentioned this before; it usually means its priced lower in this market. I feel a bit weird pricing to what the market will bear. I see a lot of my competition doing exactly that, some even going so far as saying that's exactly what they are doing. But I want to be able to look inside and see what I spent my money on. When priced to what the market will bear (Veblen Effect) you can't see that.

Ralph - I am already hand wringing at the current price.
(So please don‘t raise it on my account…)

How does your class Class-D differ from Purifi based ones?
(It looks like the distortion profile of the Purifi is not decreasing monotonically.)

 

[morricab]

Ralph,
What's your take on my Dartzeel NHB 108 Model 1. How are its distortion profiles. Sonically I hear the cleanest amp I have had in my system. But then again, its one of the only SS amps I have had. It does a good job of getting out of the way of the music. At times I am not fully sure this is what I want. Its confusing. I am very familiar with a tube sound as I have had quite a few models over a 7 year period. And there are aspects of it I like. But there are times they seem to push me away. The Dartzeel never pushes me away. There are times I might yearn for something a little different. But that is a lot different.

Is the Atmasphere Class D amp on paper more clean with less higher order harmonics?

So, are you saying the DarTZeel doesn’t push you away but also doesn’t engage you in the music?

Measurements are available from Stereophile. It is one of the poorer measuring SS amps of recent vintage. It doesn’t use any global feedback, I think, but does use some local feedback. Overall, a low active part count, low feedback Class AB design.

 

[Bjorn]

These are specifications, not measurements.

Correct. Specs like that says very little.