Review and Measurements of Purifi 1ET400A Amplifier

[Julf]

i’m amazed, compared to my double ncore, it sounds clearer, more detailed, with a less round, firmer bass

Sighted listening? How did you match levels?

 

[somebodyelse]

Btw audiophonics is using a single stage lm317 / 337 voltage regulator. Does not sound that much high end.

And if their purifi amplifier use cheapest regulator like LM317 and not discrete regulator on this high end amplifier does’nt put the chances to reach a high level of quality

It could just be good engineering - why use a more expensive regulator if it makes no difference at the output? The same regs are used in the Neurochrome HP-1 for example.

 

[SIY]

with a sound level meter on a pink noise

You'd do well to get even 1 dB of repeatability. 2 or 3dB is more likely. The far, far better way to match levels is electrically, using a voltmeter at the speaker terminals and a sine wave. Even then, touch the volume knob while listening (unless you're using a precision attenuator as a separate box), and the level match is gone.

The difference you describe is very close to what a small level change can cause in an instantaneous comparison. My guess is that yours was at least a few minutes between listening to manually change amps, and your seating position changed as well from getting up and then back down.

 

[SIY]

I did this test by...

...peeking and not using any basic controls. I'm sorry, that makes your "observation" meaningless. And it's vastly more likely that the sound of the two is the same. This is the problem with not doing a good ears-only test.

 

[DonH56]

I did this test by always listening to the same song, by listening, listening again and again and again by changing the amplifier, it is clear that the ncore is less detailed, you must clearly turn up the volume of the ncore to hear the details easily audible with the purifi...

for the regulator, purifi don’t have put lm317 on their eval kit, they made it with discrete régulator, i don’t think they do this for the pleasure...

You sure it is not just because the gain is different between the two amps? I do not know, but your statement implies the detail is there, just at a lower level.

 

[SIY]

I come here to share my point of view, I did very very many tests

But without controls. That invalidates them straight up from any scientific viewpoint. The largest variables aren't considered at all.

I'm sorry that this is not something you want to hear, and I'm a bit blunt, but nonetheless, you really really should try this with real controls. And I say that in a very sincere spirit- putting your own hypotheses to the test is the way you get to truth. Science works. Use it.

 

[JimB]

I come here to share my point of view, I did very very many tests
I found this forum more serious than the others, I wanted to share my experience with you
I would not have allowed myself to post here a simple feeling and not a fact of which I was sure
you are skeptical with my method without trying to understand, it’s not the first time i do that, the report which I had made to you are not a one shoot listening, but the result of several sessions with different songs which I know very good in every détails.
I wanted to help, not be criticized at all ends
I did not come to put on a pedestal this amplifier what I have just finished and flatter my eguo compared to another amplifier which is for me an excellent reference
never mind ...
If someone want to come to my home to listen the difference with a coffee and or a beer it’s with pleasure that î’ll accept

I hear you. It’s a tough crowd. It’s just that many of us have experienced preferences which turned out to be quite small differences in something other than what we thought we were “testing”, volume levels being a PRIME example.

So, many of us doubt there are detectable differences in high quality components due to infinitesimal distortion specs (for example). I’ll also say that at least some of us are quite interested in any well controlled cases where differences ARE heard!

 

[Julf]

"Examination of published blind listening tests reveal that listeners are strongly disposed to report differences in sound quality when given two alternatives that are identical."

 

[BDWoody]

I found this forum more serious than the others, I wanted to share my experience with you
I would not have allowed myself to post here a simple feeling and not a fact of which I was sure

The thing is, you are right about the forum. This is a more scientifically oriented place than many of the other audio sites...which is why 'we' might expect or encourage more of a scientifically valid listening test protocol.

It wouldn't be worth much to science if everyone could claim whatever they wanted to, without having to back up those claims.

No one is thinking that you are lying, any more than when they do drug testing against placebo and the placebo takers get better...no one thinks they are lying, but that doesn't make the placebo a good treatment plan...

Psychoacoustics are complicated. It's more than just listening closely until you are 'sure.' I know I have heard things that disappeared once controls were put in place...many here have as well.

 

[BDWoody]

I would ask Madame to cross me the amplifiers without telling me which one I listen to while keeping my eyes closed ...

For a proper test...it should be Double blind...meaning she can't know what she is switching either...gotta watch out for clever Hans.

If it's worth going through the trouble, it's worth getting it right...and you don't need to close your eyes...

 

[Julf]

For a proper test...it should be Double blind...meaning she can't know what she is switching either...gotta watch out for clever Hans.

The Clever Hans effect is especially strong with spouses. I guess it is a survival mechanism...

 

[PaulD]

Peeking is always a problem...
https://seanolive.blogspot.com/2009/04/dishonesty-of-sighted-audio-product.html

The ONLY valid listening test is one that is double blind, and level matched to 0.1db or better. This is difficult to achieve! Thus people (including myself and probably most) scrimp in various ways...

 

[Deleted member 65]

Peeking is always a problem...
https://seanolive.blogspot.com/2009/04/dishonesty-of-sighted-audio-product.html

The ONLY valid listening test is one that is double blind, and level matched to 0.1db or better. This is difficult to achieve! Thus people (including myself and probably most) scrimp in various ways...

Going all the way ...

Swedish AES has finalized build of their F/E device. To be used as a device supporting Swedish AES method of evaluating/testing amplifiers, DAC’s etc. Supports both A/B listening as well as the F/E (Before/After) method. Levelmatching within 0.05 dB.

Edit: F/E device is controlled by computer thru an Arduino box.

 

[somebodyelse]

Going all the way ...

Swedish AES has finalized build of their F/E device. To be used as a device supporting Swedish AES method of evaluating/testing amplifiers, DAC’s etc. Supports both A/B listening as well as the F/E (Before/After) method. Levelmatching within 0.05 dB.

Edit: F/E device is controlled by computer thru an Arduino box.

Is the design open so that others can repeat with nominally identical equipment?

 

[Deleted member 65]

Is the design open so that others can repeat with nominally identical equipment?

Don’t know being a passive member of Swe AES. The comparator consists of three parts, F/E device, computer software & Arduino box. Development has been a project with participants from KTH Royal Institute of Technology & some distinguished engineers.

@amirm intended to contact Swe AES on this, no idea if he had the time to do so. Other option is that I stop being lazy and do it myself ... ;-)

 

[TomJ]

if you trace the signal, the jumper bypass the buffer like the eval kit or simply put at 1 the gain ?
if they bypass, try it please...

Don't know. FYI the jumpers are adjacent to an LED (fault indicator?). Re gain options, this is what their website says. What's printed on their PCB is different, perhaps they did an ECO after the boards were made:

• Gain of + 7.2dB, for a total of 20dB (default) - Maximum power reached with a signal of ~ 4.2V RMS
• Gain of + 12.5dB, for a total of 25.3dB - Maximum power reached with a signal of ~ 2.2V RMS
• Bypass: in this case the signal passes directly from the inputs to the modules, allowing the lowest possible gain of 12.8dB. This mode can be interesting when using a powerful preamplifier, capable of driving a load of 2 to 4kOhm and reaching a voltage of ~ 10V RMS

 

[JimB]

Peeking is always a problem...
https://seanolive.blogspot.com/2009/04/dishonesty-of-sighted-audio-product.html

The ONLY valid listening test is one that is double blind, and level matched to 0.1db or better. This is difficult to achieve! Thus people (including myself and probably most) scrimp in various ways...

I remember when I discovered that I could distinguish tonal balance differences at about 0.1 dB. It was not a happy time, because I realized I had no chance of ever controlling everything to that degree. Careful level matching is very important.

 

[TimoJ]

Two things I have noticed, when comparing to my old (but rebuild) NAD216 amplifier:
Bass/drums seem to have more impact. Bass level is the same, but "something" is different.
At very low listening levels Purifi seems to be more "alive".

But I have not yet verified these. I will wire a test setup (with relays to switch speakers from my listening position) when I finish the build i.e. when speaker terminals arrive. Not sure how to mach levels without effecting signals.

 

[Julf]

Try to bypass the buffer stage and listen

But make sure you match levels...

 

[Vincentponcet]

Is someone aware of a vendor planning a 5 channels Purifi amp for HC usage ?
The current models are just 4 channels. So I would need one 4 channels + 1 stereo, which is 2 boxes instead of one, and more expensive than just 5 channels.
By putting amp board in height like in the NAD M27, you can put 7 channels.
One SMPS3kA400 would be enough for 5 channels.

Are you aware of a case vendor selling HC cases for 5 channels ?