Review and Measurements of Purifi 1ET400A Amplifier

[CDMC]

How much power do I really need? Interesting video.

It is important to keep in mind that power needs are highly dependent on speaker sensitivity and listening distance. Here, the Harbeths sensitivity is a bit on a low side at 84.5db/2.83v/1m. They are a pretty good sized speaker, so will likely go in a reasonably large room. Lets compare how just changing listening distance can vastly change the power needed:

• At a 15 foot listening distance with 800 watts you get 106.3 db peak (loud but not super loud when playing dynamic music)
• Drop that listening distance to 9 feet and the power get 106.3 db drops to 320 watts.
• Go to a more efficient speaker like the Revel Salon 2 (86db/2.83v/1m), which is still on the lower side of efficiency and at 9 feet it takes 200 watts to get the same 106.3 db output.
• Go even more efficient, say a Revel F208 (89db/2.83v/1m), which is pretty average for floorstanding speakers, now your power requirement to reach 106.2 db at 9 feet is down to 100 watts.

The dynamics of the music will also effect how much power we need. Highly compressed music needs less power than dynamic music. Our ears perceive average volume. So if you have highly compressed piece of music you are playing at 86db average (which is pretty darn loud), its peaks are only 5 db and you never need more than about 10 watts, even with an inefficient speaker. Now you switch to a dynamic piece of music like played in the video with 20db peaks and the same 86db average level hits 106db on the peaks, while being no louder on average (in fact the compressed piece will generally sound louder at the same average level).

When looking at amplifier needs, it is always important to factor in the speaker efficiency, listening distance, and listening levels. When in doubt, go larger, you are less likely to damage a speaker from too much versus clipping the amplifier. I will say that many systems do end up with minor clipping due to not enough power and it often not noticeable until you have an amplifier that doesn't clip, the sound is effortless rather than getting strained or congested at high levels. (For reference my amp puts out 700 watts into the F208s for peaks).

 

[Matias]

The electronic music peaked at 750W, the strings quartet peaked at 28W, for the same volume levels, same speaker, same room, etc. So for me the music playing is by far the most determining factor of power needed.

 

[retro]

That's why I'm thinking nCore NC2K for Salon2..

Didn't Amir say/write somewhere he played his Levinson 53 monos into clipping on his Salon2's..?

 

[echopraxia]

That's why I'm thinking nCore NC2K for Salon2..

Didn't Amir say/write somewhere he played his Levinson 53 monos into clipping on his Salon2's..?

Wow, would be interesting to find that post and understand more of the conditions there.

That said, those ridiculously overpriced amps are only marginally more powerful than Hypex NC1200s: 500W vs 400W, which accounts for less than 1db difference I think. But yes, NC2K amps would be nice for peace of mind knowing that if there is every an SPL limit, it's most likely going to be your speakers and not the amplifier. However this also brings up the worry of blowing up the speakers themselves.

All the more reason why I really like active speakers and their self protection, but obviously that doesn't help those of us already with nice passive speakers we intend to keep for a long time

 

[echopraxia]

Well then, how about bi-amping with both NC1200 and Purifi..you get the best of both..?!?

Actually really interested in your comparsions, since I'm planning to get a pair of Salon2 in the near future..

I wonder how much additional power this would actually achieve in practice? I've been told in the past that bi-amping doesn't really increase effective amplifier power by 2x, but I have not done the work to understand why. I understand that it won't boost the power of any single sinewave tone, but nobody listens to single sinewave tones. For music with content distributed broadly across the spectrum, I don't see why the amplifiers wouldn't share the load to some extent.

And, I could assign the NC1200s to the lower frequencies and the Purifi 1ET400A to the higher frequencies where any distortion (if audible at all) would be most audible. But, honestly I think this is all purely theoretical. That said, theoretical is what we do over here in the electronics section (most of the time anyway)

 

[Matias]

Also Purifi is working on higher powered modules. Once they are released, upgrade is in order here.

 

[echopraxia]

Found this interesting discussion here about a 300W amp entering clipping when driving the Salon2's loudly (even when crossed over to subs!)

Interesting also because it brings up the point that at a certain level of power, you risk burning up the speaker's passive crossovers etc.

So I think the definition of an "endgame amp" (for passive crossover speakers) could be reasonably limited to how much power the crossovers and transducers can physically take for these low-sensitivity speakers.

Again, this is mostly theoretical because in my current room size, even the Purifi 1ET400A seems to be sufficient for me. But when we are discussing an "endgame" amplifier, it matters because we are discussing finding that amp that is fully capable of driving any passive speakers up to as much power as they can physically take. I may not drive my Salon2's that loudly right now, but maybe some time in the future I'll move them into a much larger room, etc.

 

[Kal Rubinson]

Stereophile has been doing it longer too. But they seem to publish some of the measurements which support their opinions and not the other way around.

I do not understand your point. The measurement suites that accompany each review is quite uniform. Specific tests are not eliminated to fit the product. In many cases, the published results conflict with the the accompanying subjective comments.

 

[Matias]

I do not understand your point. The measurement suites that accompany each review is quite uniform. Specific tests are not eliminated to fit the product. In many cases, the published results conflict with the the accompanying subjective comments.

Agreed. The problem with Stereophile measurements is not cherry picking which ones to show and which ones to hide. They are very consistent with that throughout the years.

The problem is a lack of reference, having benchmarks of what is good and what is bad, and when it is bad, not pointing it out loud and clear. This for example.
https://www.audiosciencereview.com/...perfectwave-directstream-dac.9100/post-232837

 

[Labjr]

I do not understand your point. The measurement suites that accompany each review is quite uniform. Specific tests are not eliminated to fit the product. In many cases, the published results conflict with the the accompanying subjective comments.

In Stereophile, there are never more than a few basic measurements compared to many that are done here. And generally they support the opinion of what the reviewer claims to hear. I rarely see criticism. Everything seems to be "superbly engineered" as long as it's not a total disaster. IMO, ASR has turned the reviewing business on it's head.

 

[echopraxia]

In Stereophile, there are never more than a few basic measurements compared to many that are done here. And generally they support the opinion of what the reviewer claims to hear. I rarely see criticism. Everything seems to be "superbly engineered" as long as it's not a total disaster. IMO. ASR has turned the reviewing business on it's head.

I agree that ASR is disrupting the speaker review business, in a very good and healthy way (for consumers).

So this leads to the question: Will Stereophile (and/or other reviewers/magazines that wish to avoid becoming obsolete) be purchasing a Klippel NFS for future reviews? If not, why? Are they unable to afford it? Or some other reason?

 

[Deleted member 19195]

After this test I did with 2x500W at 4 ohms into my 2.7 ohms current hungry speakers, playing loud and bass heavy music, it really had me rethinking how much power I really need. The 1ET400A could be all I really need. But how much is it of an audible improvement over my current NC500OEM based main amplifier? That is the question.

The answer, most likely, is none at all.

 

[firedog]

I think it’s hard to define how much power in an amp is truly “endgame”. ~200 watts at 8 ohms and ~400 watts at 4 ohms seems to be enough for most people and speakers, but this is relative, and I think more headroom is always appreciated.

I’m actually quite curious as to whether the improved SINAD and IMD of Purifi 1ET400A is actually audible vs the Hypex NC1200 (which is twice as powerful, at ~400 watts into 8 ohms and ~700 watts into 4 ohms).

While it’s nice to know your amp has ultra-low distortion/noise/etc, once you’re already at the level of performance as the Hypex NC1200 (for example), it’s hard for me to imagine how any further improvements would be audible. When you look at speaker distortion measurements, it’s clear that the amount of distortion from either of these amplifiers is going to be completely overpowered by the distortion caused by the speakers themselves.

Don’t get me wrong — the improved quality performance from the Purifi is very welcome. It’s really beautiful engineering. But if the difference isn’t audible, then the Hypex NC1200’s greater power capability (twice as powerful as the Purifi 1ET400A) would make it the better “endgame” amplifier.

It turns out I own both Hypex NC1200 amps, and Purifi 1ET400A amps... almost as if I planned to try a blind ABX test between them at some point

I'd certainly be interested in hearing your comparison. Especially blind, but also sighted.

 

[mk1classic]

Testers of the new Vera Audio amplifiers indicate a difference in the sound to the better when comparing the latest Purifi-Audio based Vera Audio P150/600 RS vs the more powerful Hypex based Vera Audio P400/1000.
Minute differences, but enough to go for the 1ET400A amp with less power.

 

[VintageFlanker]

Testers of the new Vera Audio amplifiers indicate a difference in the sound to the better when comparing the latest Purifi-Audio based Vera Audio P150/600 RS vs the more powerful Hypex based Vera Audio P400/1000.
Minute differences, but enough to go for the 1ET400A amp with less power.

Hum... Would be glad to know how such "tests" have been conducted exactly.
Such claims from a manufacturer (who will obviously tell you their more expensive gear will sound better) have to be taken with a big grain of salt...

Minute differences, but enough to go for the 1ET400A amp with less power.

That's how marketing works. "Minute difference, but enough to go for the higher model"... And so on! That way, the all ranges are perfectly cohesive each other.
Still waiting for blind listening tests of Purifi amps, once we know they measure stellar.

 

[Willem]

I admire technical excellence but in the real world of audible differences more power is likely to beat lower distortion etc at a level that should alŕeady be below accepted levels of human hearing ability.

 

[echopraxia]

I admire technical excellence but in the real world of audible differences more power is likely to beat lower distortion etc at a level that should alŕeady be below accepted levels of human hearing ability.

I agree; however, my reason for trying the Purifi (beyond admiration of the technical specs) is that Bruno Putzeys said in an interview that the audible difference he heard was more than he expected. I understand that despite my/our great respect for his engineering, this could just have been his own expectation bias. But I must say it made me curious.

So I plan to blind ABX test the Purifi 1ET400 vs Hypex NC1200 at some point, to see for myself if there’s an audible difference at this exceptional level of performance other than differences in power.

Personally I 100% agree and I do not think there is a difference we can hear, but I’m curious to test this because there is always some small possibility of audible issues we haven’t thought of getting past the standard set of measurements. Unless we somehow test amplifiers with complex audio signals driving complex loads, there will always be some lingering question here that only an ABX test can truly set to rest.

 

[Deleted member 14468]

SoundStage! Hi-Fi also measured this amplifier with an APx555. They are starting to do measurements in house too (before Bascom's BHK Labs did it for them). Nice, @Doug Schneider , congrats!

https://www.soundstagehifi.com/inde...asured-performance-of-the-eigentakt-amplifier

Hello,

It's been a couple weeks since I visited here, but now that I'm back, it's nice to see this post.

Yes, we did purchase an APx555 and, of course, you're going to see many more measurements in the very near future. As you mentioned above -- BHK Labs was doing our electronics measurements for many years. We couldn't get everything to Bascom King's lab, but we tried to get as much as we could. But it's time to bring it all in-house now -- so you'll be seeing much, much more from us!

Thanks,
Doug

 

[mk1classic]

Hum... Would be glad to know how such "tests" have been conducted exactly.
Such claims from a manufacturer (who will obviously tell you their more expensive gear will sound better) have to be taken with a big grain of salt...

That's how marketing works. "Minute difference, but enough to go for the higher model"... And so on! That way, the all ranges are perfectly cohesive each other.
Still waiting for blind listening tests of Purifi amps, once we know they measure stellar.

I see in hindsight I shouldn't have written the response above, as this is a highly flammable topic on a forum as ASR. I was not talking about supplier marketing texts or scientific tests. This was based on responses from a few lucky users in Norway testing the different samples in their homes with their audio setups.

Since these tests were highly non-scientific tests, I reported my general understanding on their feedback to contain "minute differences", differences that may be due to peoples opinion and "imagination" at the time of listening. i.e. there might not be any differences and you buy what you want in the end.

On the topic of exact tests. How can we measure the "minute differences" we "see or hear" between low-distortion units, like the different versions of 1ET400A and Hypex NCORE based amplifiers? There must be more to measure, as the measurements we see performed here on ASR doesn't differentiate them good enough for "us analytic measurement geeks" to ward off attacks from the angry "listening mob".

 

[CDMC]

I see in hindsight I shouldn't have written the response above, as this is a highly flammable topic on a forum as ASR. I was not talking about supplier marketing texts or scientific tests. This was based on responses from a few lucky users in Norway testing the different samples in their homes with their audio setups.

Since these tests were highly non-scientific tests, I reported my general understanding on their feedback to contain "minute differences", differences that may be due to peoples opinion and "imagination" at the time of listening. i.e. there might not be any differences and you buy what you want in the end.

On the topic of exact tests. How can we measure the "minute differences" we "see or hear" between low-distortion units, like the different versions of 1ET400A and Hypex NCORE based amplifiers? There must be more to measure, as the measurements we see performed here on ASR doesn't differentiate them good enough for "us analytic measurement geeks" to ward off attaches from the angry "listening mob".

The difficulty is that you first have to establish they are in fact hearing minute differences. To do so requires blind testing. If the blind tests show an audible difference, and the measurements show they perform the same, or within differences that are inaudible, then you go searching for what is not being measured that makes them sound different. In reality, if you do a blind test between the hypex and purifi it is extremely unlikely anyone could hear a difference between the hypex and purifi amps.

Bruno is in a tough position. His Purifi amps are an improvement over his hypex designs. They have lower distortion, a wider bandwidth, and other objectively better measurements. The problem is those improvements are below the level of audibility. For an objective consumer that wants state of the art, they are a no brainer upgrade. For the larger group of subjective consumers, there has to be a story that they sound better, because for that group, their ears are a far more accurate measuring device. He would sell very few amps if he said they measure better, but sound exactly the same while costing more. That doesn’t sell to the subjective crowd, they need a story.