Review and Measurements of Purifi 1ET400A Amplifier

[maty]

@amirm
Hi, my first post, by any chance do you have a 10khz phase shift (degrees) measurement at the output, without the use of any A.P. aux hf filtering involved?

Bruno Putzeys said in the interview, PURIFI 1ET400A: 1º at 20 kHz. Presumably 0º to 10 kHz.

Hypex NC400, 16º at 20 kHz.

BOSC - HiFi Monoblock Class-D GaN Amp, flat up to 30 kHz.

https://www.diyaudio.com/forums/cla...ass-amplifiers-affects-sound-post5929443.html

 

[March Audio]

Excuse me, don't know why your answering, I asked @amirm , who's doing the measuring on this, not you.

What you posted up doesn't represent what I asked him for at all, just confuses it.

And I like to see what's coming out of the output of an amp, with and without the AP filter masking it.

I am answering because we have already been through all of this elsewhere. It would be wasting everyone's time to go through it again here.

I will ask again, what do expect to see or learn different from what I have already told you?

 

[March Audio]

Bruno Putzeys said in the interview, PURIFI 1ET400A: 1º at 20 kHz. Presumably 0º to 10 kHz.

Hypex NC400, 16º at 20 kHz.

BOSC - HiFi Monoblock Class-D GaN Amp, flat up to 30 kHz.

https://www.diyaudio.com/forums/cla...ass-amplifiers-affects-sound-post5929443.html

Correct.

 

[georgehifi]

Like I said please I'm asking @amirm, sorry not you, for the measurements and screen shots.
So we can see the amount of switching noise on the 10khz square wave vs the amount of phase shift at 10khz without any A.P. filtering involved.
IcePower1200as also have 50 kHz audio bandwidth, yet have 70 degrees phase shift at 10khz (attached), both have 500 to 600khz switching frequencies, so seems someones fibbing.

Bruno Putzeys said in the interview, PURIFI 1ET400A: 1º at 20 kHz. Presumably 0º to 10 kHz.

Sorry Maty if only 1' at 20khz then there's an extraordinary amount of switching noise going through to the speakers, that why I would like to see a screen shot of a 10khz square wave not using the AP's HF filter.

Cheers George

 

[March Audio]

Like I said please I'm asking @amirm, sorry not you, for the measurements and screen shots.
So we can see the amount of switching noise on the 10khz square wave vs the amount of phase shift at 10khz without any A.P. filtering involved.
IcePower1200as also have 50 kHz audio bandwidth, yet have 70 degrees phase shift at 10khz (attached), both have 500 to 600khz switching frequencies, so seems someones fibbing.


Sorry Maty if only 1' at 20khz then there's an extraordinary amount of switching noise going through to the speakers, that why I would like to see a screen shot of a 10khz square wave not using the AP's HF filter.

Cheers George

This is simply not correct as explained to you elsewhere.

Why do you think your speakers respond to 500kHz?

This is not an ice power amplifier.

No one is lying, it is simply that you do not have an understanding of the subject.

 

[georgehifi]

Why do you think your speakers respond to 500kHz?

never said they did. yet I do have plasma tweeters that probably could.
Now can you please let amirm respond when he wakes!, again he's the one doing the measuring not you.

 

[March Audio]

never said they did. yet I do have plasma tweeters that probably could.
Now can you please let amirm respond when he wakes! Agian he's the one doing the measuring not you.

So if you don't think your speakers respond to 500kHz then you will agree that the switching frequency is of no relevance. You will agree that it is both electrically (and mechanically) filtered out.

If so you will agree that looking at a plot with bandwidth higher than your speakers response has no relevance or importance. The "noise" as you put it is *not* in the audio band.

So again I will ask what are you going to conclude from seeing an unfiltered plot?

 

[restorer-john]

Now can you please let amirm respond when he wakes!, again he's the one doing the measuring not you

George, this is what is called a forum. It has members who contribute, add commentary, argue, present facts (or not) and generally (not always) have fun discussing high fidelity devices, systems and equipment.

Your claim to fame is an LDR based volume control isn't it? How about sending a "lightspeed attenuator" to Amir to have thoroughly tested. Then we can talk, or argue about non-linear distortions, channel balance and other non-linearities.

 

[georgehifi]

Let amirm speak please tomorrow, your not his mother so please I ask nicely.

 

[March Audio]

Let amirm speak please tomorrow, your not his mother so please I ask nicely.

George,

Am I correct in assuming you beleive the switching noise that you see on a wide bandwidth unfiltered plot is in the audio band? Ie it is presented to the speaker and interferes in the audio band?

If so you have a fundamental misunderstanding.

IIRC the AP filter Amir uses is pretty much flat to 100kHz which is way, way beyond the response of any speaker. So the 10kHz waveform above IS what's being presented.

 

[maty]

That high frequency noise is suffered by all the class D logically. But it is also obvious that low pass filters are used so that this does not happen!

Unlike excellent and fast class A and AB amplifiers, the bandwidth in class D is much smaller.

It is more important, whatever BP says, that the phase does not oscillate much in the audio band. In class A and AB it is not a problem, hence it is not usually specified. In any case, it must be less than 6º at 20 kHz, something that did not comply with all known Class D until the appearance of the latter two. At least those who deign to explain it. In the case of PURIFI because I wrote the question in the other forum, with links.

I do not know of any reason why this restriction should not also apply to class D.

 

[March Audio]

That high frequency noise is suffered by all the class D logically. But it is also obvious that low pass filters are used so that this does not happen!

Unlike excellent and fast class A and AB amplifiers, the bandwidth in class D is much smaller.

It is more important, whatever BP says, that the phase does not oscillate much in the audio band. In class A and AB it is not a problem, hence it is not usually specified. In any case, it must be less than 6º at 20 kHz, something that did not comply with all known Class D until the appearance of the latter two. At least those who deign to explain it. In the case of PURIFI because I wrote the question in the other forum, with links.

I do not know of any reason why this restriction should not also apply to class D.

You need to take another read of Brunos commentary on the phase shift. If the phase shift is linear it just becomes a fixed time delay across the audio band, which is of no consequence.

 

[maty]

I know what he said from the first moment, because I asked the question. But I want as in the other technologies, with a minimum phase shift in the audible band. That is why I became so interested in the new design, the PURIFI 1ET400A.

The second question (H2 dominant) was not answered or asked by the interviewer.

With the kind of music I listen to, well known, it is what I want.

 

[March Audio]

I know what he said from the first moment, because I asked the question. But I want as in the other technologies, with a minimum phase shift in the audible band. That is why I became so interested in the new design, the PURIFI 1ET400A.

The second question (H2 dominant) was not answered or asked by the interviewer.

With the kind of music I listen to, well known, it is what I want.

So what is your concern?

If you know what he said then you will understand that a linear phase shift is of no consequence because it is just a time delay across the audio band. If you understood Brunos statements you would know that his previous Hypex designs have no issue with phase shift.

 

[ahofer]

While he’s here, any way to test GeorgeHifi’s insistence that no class D Amplifier can drive a low impedance load like a Thiel CS3.7 (which dips periodically below 1 ohm) and remain stable. My money is on the Class D (particularly the NC1200 w/49 amps of current on tap) although it would be interesting to see if I’m wrong.

 

[maty]

I have never been interested in buying a class D except when the IcePower 1200AS2. As soon as I discovered his lousy phase graph I gave up!

For me it is important that the elements of my audio system behave like live music, with acoustic or electrical instrumentation, without synthesized instrumentation. I have repeated a few times that I will not be among the first to build the PURIFI (assuming a reasonable price). Since this thread started I have managed to improve the sound of my second system even more, so I am in no hurry.

 

[Joachim Herbert]

@georgehifi Five posts and already insulting people that provide helpful information. Can't help to jump in here.

 

[Matias]

George is a well known "anti-class-D crusader" on Audiogon forum... Straight to the ignore list.

 

[ahofer]

I wish we could get to some blind testing of Class D amps. They seem like a huge step forward in efficiency. I’d be interested to see if they are as indistinguishable in controlled tests as solid Class A/AB designs. Put up a new Bryston, a McIntosh, a Krell, and one of these.

My own experience was that my old Pass-designed Adcom 5802 (a backbreaker purchased to drive my own Thiel CS3.6) and March Audio’s p252 were audibly indistinguishable once I got my cable switch time down. Everything I thought I was noticing on one was audible in the other. However, my listening was with an easier load (Harbeth SHL5+).

 

[georgehifi]

It is more important, whatever BP says

I'm more interested in what @amirm has measured that's why I asked if he's done the ones and saved them so they can be posted up.
1: Which are a phase response graph from 50khz down to 5khz without A.P. filter inline.
2: A screen shot of a 10khz square wave without the A.P. filter inline.
3: Now that it's been mentioned rms wattage all just before clipping at 8ohms, 4ohms, and 2ohms without any understating of the 8 or 4 to make the 2 look good, that seems to be getting more popular to do with manufacturers own specs.

All that I have asked is specifically directed at @amirm because he is the one doing these independent measurements no one else, and what I ask seem to get other backs up around here, There seems to be a lot of product protection going on around here.

Five posts and already insulting people that provide helpful information. Can't help to jump in here.

I specifically ask @amirm for these measurements if he has them on file one one else. They are the ones with their backs up over this.

George is a well known "anti-class-D crusader" on Audiogon forum

No sorry, if you care to search I praised a few times the BelCanto modded NC400 sound with a very nice pair of of $$K Rahl ribbon 2 ways. And I'm a big promoter of GaN Class-D as it helps fix the 2 Achilles Heel's Class-D has, dead time and switching frequency

I have never been interested in buying a class D except when the IcePower 1200AS2. As soon as I discovered his lousy phase graph I gave up!

This is why I would like to see from amirm the same phase graph and 10k square wave on the Purifi without any external A.P. HF filtering.

Cheers George

.