Review and Measurements of Hypex NC400 DIY Amp

[Julf]

* If the data are right *, SNR must be at max power and not at 1 watt -> Hypex NC250MP is the best option to listen to good/very good recordings with high DR at home.

I doubt your listening room has a SNR of > 100 dB.

 

[maty]

, the key is the SNR-A at 1 watt.

The difference is too big, hence ask to be verified.

 

[somebodyelse]

Audiophonics-PA-S250NC-stereo-NC250MP-specs

https://www.audiophonics.fr/en/powe...s-d-amplifier-ncore-2x250w-4-ohm-p-13545.html

View attachment 32143


Audiophonics-PA-S500NC-stereo-NC500MP-specs

https://www.audiophonics.fr/en/powe...s-d-amplifier-2x500w-4-ohm-ncore-p-13344.html

View attachment 32144


* If the data are right *, SNR must be at max power and not at 1 watt -> Hypex NC250MP is the best option to listen to good/very good recordings with high DR at home.

With modern commercial music, badly recorded and DR < 10 dB, I suppose it will be quite difficult to appreciate differences between the two modules beyond the power.

Can someone verify the data? The PDF manuals with all specs are not available on the Hypex website.
https://www.hypex.nl/oem-audio-amplifiers/

The manuals are available, but if you're using extensions like NoScript or uMatrix you'll need to allow a few things before you get to them. IIRC it needs script and cookie from google.com then clicking on the terms acceptance before they show up.

The figures you posted for the 250MP match the datasheet values. These are the same for the 252MP. The 122MP has the same figure for distortion and noise, but the SNR is 3dB down. The 500MP and 502MP match except for the SNR, which is 124dB on the datasheet.

 

[maty]

Even with MS Edge I get the message that I have to activate cookies. It is probably due to my HOSTS file, which almost weighs 461 Kb or... I avery short to Windows, which is very slob. I would have to log in with Manjaro.

The 500MP and 502MP match except for the SNR, which is 124dB on the datasheet.

Then 106 dB is wrong. Going from 30 uV to 40uV could not produce such a decrease in SNR.

Thank you very much!!!

 

[maty]

Since you do not specify it, it must be at 4 Ohms, 500 watts at 1%.

Then, 4 Ohms, 100 watts at 1% -> 117 dB.

At 4 Ohms, 1 watt -> 97 dB.

At 8 Ohms, 1 watt -> 94 dB.

SNR-A, 8 Ohms / 1 watt at 0.1% -> about 97 dBA, that it is very high!


If the 124 dB is at 8 Ohms, 250 watts at 1% -> SNR-A > 103 dBA. Too high, the 124 dB mut be at 4 Ohms/500watts.

 

[Julf]

SNR-A, 8 Ohms / 1 watt at 0.1% -> about 97 dBA, that it is very high!.

97 dBA at 1W is not very high, but why do you even care about what it is at 1 W? SNR is a measure of the range between the noise floor and the maximum signal.

 

[maty]

Really? You know why. Usually we need less of 10 watts continuous at home. 5 watts or less.

I have a lot of very good recordings with DR 15 dB or more. I need more watts to the peaks.

With my KEF Q100, 85 dB/W/m and minimum 3.9 Ohms, if I want 95 dB peak at 3 m I need 50 watts -> 100 watts at 4 Ohms.

PD: 80 dBSPL continuous at home is very high, among other reasons, because the walls are not so separated.

Modern music with DR < or << 10 dB requires much less peak watts.

 

[maty]

https://www.doctorproaudio.com/content.php?2273-calculators-proaudio-sound-dmx&langid=1#calc_spl

And in phase, we are in the center other vertice of the triangle: +3 dB -> 98.5 dB.

https://www.audiosciencereview.com/...er-one-can-find-for-2-500-or-less.6139/page-9

Spoiler: KEF Q100 and KEF LS50 measurements

KEF Q100

By the way,

KEF LS50, has true 83 dB/W/m.

 

[Julf]

Really? You know why. Usually we need less of 10 watts continuous at home. 5 watts or less.

Sure. What does that have to do with SNR? What you want to do is ensure the gear produces a noise floor below that of your listening room and your source material. If you want a 95 dB SPL peak, you only need to ensure your gear has a SNR (at full power) of at least 75 dB or (as I doubt your listening room has a noise floor below 20 dBSPL.

 

[somebodyelse]

Since you do not specify it, it must be at 4 Ohms, 500 watts at 1%.

They specify the following conditions:
Signal-to-Noise Ratio: <10Hz-20khz AES17
Output Noise: Unweighted, <10Hz-20kHz AES17, 0R termination

I'm not familiar with AES17 so someone else will have to explain what that means for the reference level.

 

[Julf]

I'm not familiar with AES17 so someone else will have to explain what that means for the reference level.

AES 17

 

[somebodyelse]

AES 17

I found that one, saw that the reference level was defined in terms of digital full scale (section 9.3 for SNR, following back to 3.3 for full scale definition), and assumed I was missing something and needed an explanation from someone familiar with its application to equipment without a digital input.

 

[Julf]

I found that one, saw that the reference level was defined in terms of digital full scale (section 9.3 for SNR, following back to 3.3 for full scale definition), and assumed I was missing something and needed an explanation from someone familiar with its application to equipment without a digital input.

AES 17 is for measuring digital gear (even their analog outputs). In that situation "full scale" has a pretty clear definition.

 

[somebodyelse]

AES 17 is for measuring digital gear (even their analog outputs). In that situation "full scale" has a pretty clear definition.

So why are Hypex using it for analog gear? What's the equivalent standard for power amps that they should be using instead?

 

[Julf]

So why are Hypex using it for analog gear? What's the equivalent standard for power amps that they should be using instead?

It is quite relevant for analog gear too, as long as a suitable point of "maximum power" is selected to represent "full scale".

 

[maty]

The human being begins to perceive sounds from 19 dBA - 21 dBA (source at 1 m). It is something to keep in mind when choosing the fans, the SMPS power supply of a custom PC and with the air conditioners.

I have a good-sized window with double glazing and modern Spanish blinds (i.e. with built-in insulation). Thanks to that old Persian invention, in addition to adjusting the light in the room and reducing energy loss, the noise is lowered.

When a recording excites me, I darken the room, under the blind and I notice the (noise) difference.

Suppose it is not summer and the window is closed. And at night, blinds down. I guess I should have 25 dBA. I want SS amplifier wiht SNR-A near 90 dBA. I need 65 dB.

I usually listen to very good vinyl rip 24/96 or 24/192 WAV/FLAC recordings. Vinyls have 56-65 dB of dynamic range. The vast majority are rips without noise, clicks and other discomforts that are usually attributed to vinyl, but I rarely listen to them with my loved speakers. They are usually very good editions or pressed -> 65 dB of dynamic range.

Finally, 25 dBA + 65 dBA = 90 dBA.

 

[maty]

How loud is too loud?

https://www.osha.gov/SLTC/noisehearingconservation/loud.html

PD: in a closed room you get 80 dBSPL with fewer watts than in the open field, without walls.

 

[Julf]

Suppose it is not summer and the window is closed. And at night, blinds down. I guess I should have 25 dBA. I want SS amplifier wiht SNR-A near 90 dBA. I need 65 dB.

I usually listen to very good vinyl rip 24/96 or 24/192 WAV/FLAC recordings. Vinyls have 56-65 dB of dynamic range. The vast majority are rips without noise, clicks and other discomforts that are usually attributed to vinyl, but I rarely listen to them with my loved speakers. They are usually very good editions or pressed -> 65 dB of dynamic range.

Finally, 25 dBA + 65 dBA = 90 dBA.

Not sure I understand your logic. You don't need a much greater SNR from your gear than from your source material. If your gear is capable of 90 dBSPL in volume, a 70 dB SNR will keep the noise floor below 20 dB SPL.

 

[maty]

What I want to highlight is that in a room of a relatively modern Spanish house, that is, well insulated, the ambient noise at night is much lower than in other countries. More near of 25 dBA than 45-50 dBA.

The other day I read that the use of the Spanish blind is spreading in other countries, especially in Mexico. But in that country it is more for security reasons, with metal shutters.

 

[somebodyelse]

It is quite relevant for analog gear too, as long as a suitable point of "maximum power" is selected to represent "full scale".

I understand the relevance - it's one of the tests "substantially identical to those used when testing analog equipment" as mentioned in the AES17 abstract. I had hoped there would be a standard for identifying that "suitable point" given the number of different approaches to "maximum power" currently in use, or at least a generally accepted method that those familiar with it would be able to point to. I also know from testing in other fields that the standards you may expect to exist often don't, or have multiple competing forms. If that's the case here then please say so.