Review and Measurements of Purifi 1ET400A Amplifier

[audioBliss]

We are just hungry leechers over here so don't take it personally. You're doing great work for the whole audiophile community!

 

[rajapruk]

We are just hungry leechers over here so don't take it personally. You're doing great work for the whole audiophile community!

Word!

 

[somebodyelse]

CMRR at 1 kHz is only > 60 dB. I would like above, > 90 dB.

Are these CMRR figures achieved with real world sources where components are rearely better than 0.1% tolerance in the output?

 

[maty]

Real world? Maybe a very good PA from Germany or Austria. Only € 579 at home, Spain.

PAS 2002PCA, with input CMRR > 90 dB

https://www.thomann.de/intl/pas_2002pca.htm

PS: I have been researching for years to buy or build an amplifier for my second audio equipment, so I have saved a lot of information. Meanwhile I have been experimenting with the one I have until I get an excellent sound. Problem: I think I have reached the maximum possible improvement. The sound is still 2D (less than the original) but I would like it to be 3D and to perceive sound planes plans with certain very good recordings.

 

[rajapruk]

I am a bit surprised that it runs cool with the Hypex SMPS1200 in it.
I had this powersupply with bridged dual nc400, and then it was running pretty hot.

 

[miero]

@maty CMRR is not outstanding for Purify module, but you are comparing an amplifier module with a complete amplifier (with one exception). an input buffer can improve this parameter if needed.

 

[MZKM]

It's time to pick a number as a line in the sand as measure 'clipping' as being that figure. Everyone else has done it. Hypex do it. Stereophile does it. Audioholics do it. Magazines from the past did it. The power cube does it. Amir should do it.

I vote on -60dB (0.1%) THD, above 0.5W.

Human trials have shown that -40dB (1%) is about the audibility threshold in the treble (at 20Hz, even 100% THD can be tolerated), and that’s what Stereophile has chosen. With 4 devices in the chain (speakers, power amp, preamp, DAC), all having -60dB would sum to -54dB, so still inaudible.

Now, speakers distort a lot more, but if the speakers are -40dB and the other three are -60dB, then that’s just -39.9dB.

 

[maty]

@maty CMRR is not outstanding for Purify module, but you are comparing an amplifier module with a complete amplifier (with one exception). an input buffer can improve this parameter if needed.

Of course. The key is a something better input-buffer than of the kit. PURIFI is still on time, because the kit will not be for sale until October 2019. Or maybe the CMRR of the kit is > 90 dB at 1 kHz.

Amirm can measure the CMRR of the kit and so we will leave doubts.

Image optimized and rotated

https://www.audiosciencereview.com/...ng-on-an-eigentakt-amplifier.7671/post-191827

 

[audioBliss]

Real world? Maybe a very good PA from Germany or Austria. Only € 579 at home, Spain.

PAS 2002PCA, with input CMRR > 90 dB

https://www.thomann.de/intl/pas_2002pca.htm

View attachment 28917

View attachment 28918

View attachment 28919

PS: I have been researching for years to buy or build an amplifier for my second audio equipment, so I have saved a lot of information. Meanwhile I have been experimenting with the one I have until I get an excellent sound. Problem: I think I have reached the maximum possible improvement. The sound is still 2D (less than the original) but I would like it to be 3D and to perceive sound planes with certain very good recordings.

Very interesting amp! How do you buy one of the bigger models? I can't find them anywhere? I need around 2 x 450W 8Ohm for my subwoofers. Right now I'm running a Behringer EP4000 but it would be interesting to test something else

 

[maty]

A problem: blue color. Your wife can kick you out of the house. Years ago I wrote to the constructor asking about some values/specs but he did not answer my mail. Now I search.

http://www.brainaudio.com/products.htm

[PDF] https://images.thomann.de/pics/prod/131811_specs.pdf

Contact: http://www.brainaudio.com/feedback.htm


BTW: Notice that the ventilation grilles are just above the capacitor bank. The most delicate audio circuitry is well protected, even from the BIG toroidal transformer.

 

[Dj7675]

Is it possible to bridge this amp and run in mono like the benchmark? If so possible to test it in this configuration?

 

[somebodyelse]

Real world? Maybe a very good PA from Germany or Austria. Only € 579 at home, Spain.

PAS 2002PCA, with input CMRR > 90 dB

By 'real world' I don't mean what's on the spec sheet. I was reading a paper that criticised standard CMRR measurement procedure for audio devices for giving high numbers that were often unattainable when connected to actual upstream sources because the impedance mismatch in the upstream sources degraded the CMRR. Unfortunately I can no longer find the paper to reference. It may be that it's referring to a testing method that's no longer in use, that the author was wrong, or that I failed to understand it. The claim was that the first step of the standard involved adjustment so that the drive to the device under test had perfectly matched source impedances, providing ideal conditions for CMR. Real world sources are rarely perfectly matched, and this may significantly degrade the common mode rejection when these are connected to the device depending on the design of the input buffer. IIRC the classic opamp differential buffer was an example that may be significantly degraded, while a design licensed to THAT Corp. was significantly better. I'm looking to learn, so links to standards, schematics etc. would be much appreciated.

 

[maty]

About PSRR and CMRR, days ago I asked to the designer/owner of the Maraschino and MEGAschino class-D amps about the PSRR. https://www.cherryamp.com/

https://www.audiocircle.com/index.php?topic=164405.msg1749050#msg1749050

[ Don’t know exact PSRR off the top of my head, but likely 60-80dB mid-band. The Maraschino also runs from a regulated supply. The ripple is way out of band. It’s ultra quiet even at 20kHz. For the MEGA, the ripple is where PSRR is maxed out, so THD at high power is good (low) on both sides of the spectrum.

Somewhere, I have a photo of the MEGA MK2’s noise floor FFT. It’s amazing!! Pretty level at -140dB with -60dB stim. SNR is the ultimate “proof in the pudding” at about 120dB, so you know the power supply noise isn’t getting through! It’s a solid design. This also doesn’t involve the usual chokingly high feedback levels and low phase margins of other Class-D amps... ]

http://www.jeremykipnis.com/MEGAschino_Review_-_03_18.html

https://hometheaterreview.com/digital-amplifier-company-megaschino-power-amplifier-reviewed/

I admit, I have always been very inquisitive, and I sat in the front rows

 

[Ron Texas]

It's a nice trend, low distortion class D. Perhaps they can find an OEM to build these assembled. It's also awkward from a commercial standpoint to use a power supply from a direct competitor.

 

[maty]

...IIRC the classic opamp differential buffer was an example that may be significantly degraded, while a design licensed to THAT Corp. was significantly better. I'm looking to learn, so links to standards, schematics etc. would be much appreciated.

https://www.neurochrome.com/modulus-86/

[ The input to the composite amplifier is provided by a THAT1200 differential receiver. The differential input is ideally suited for differential connections, such as the XLR connections used in professional audio, but can also be configured to accept single-ended connections, such as the commonly available RCA connectors. Differential signaling provides many advantages, the most prominent advantage being the rejection of hum. This results in complete quiet during quiet passages of the music and results in a vast, wide, and open sound stage. ]

[ The Modulus-86 is a turnkey analog subsystem. It requires no adjustments and it achieves state of the art performance using readily available off-the-shelf parts. The Modulus-86 uses one channel of an LME49720 precision audio op-amp to perform error control on an LM3886 power amplifier. An OPA2277 precision op-amp is used as a high-end DC servo with third order filtering for an ultra-low output DC offset and fast settling time. The DC offset is less than 200 µV and is reached within ten seconds of power-up. A THAT1200 differential receiver provides common-mode rejection rivaling that of a line transformer, thus, minimizing hum injection by eliminating ground loops in the signal path. ]

[PDF] http://www.thatcorp.com/datashts/THAT_1200-Series_Datasheet.pdf

 

[Matias]

It's a nice trend, low distortion class D. Perhaps they can find an OEM to build these assembled. It's also awkward from a commercial standpoint to use a power supply from a direct competitor.

There will probably be a dozen vendors when these modules become available. From more traditional companies like NAD, Bel Canto and Marantz etc to smaller companies like Nord, Apollon, March etc.
I expect Purifi to release their power supplies soon too, but meanwhile we can use Hypex's SMPS.

 

[audioBliss]

A problem: blue color. Your wife can kick you out of the house. Years ago I wrote to the constructor asking about some values/specs but he did not answer my mail. Now I search.

Sounds inconvenient to have a wife like that

 

[Joachim Herbert]

A good friend of mine who designed loudspeakers for the German Canton brand always was full of praise for the B&O ICE modules for the professional audio projects he did thereafter.

But he insisted to use conventional power supplies instead of switched ones, claiming a tighter bass performance. Looking at noise figures this might be the way to go for ultimate performance, though a conventional supply will add to weight and price considerably.

But then, who needs ultimate performance when excellent performance is available for a fraction of the cost?

 

[amirm]

OK, fixed the faulty speaker cable. Here is the comparison against Hypex NC400 in stereo mode:

Note: the Hypex NC400 was measured using my older dummy load which created more distortion than my current one. So the gap between the two may not be as much as shown.

 

[amirm]

I second the idle power consumption request.

Here you go:

It would be nice to have a hard on/off switch to avoid the high idle current of the switching power supply. But otherwise, efficiency at power is 90% or so and gets there in a hurry.