ABX test of NC252MP x A250W4R into 4ohm//2.2uF load, files for download

[pma]

Hello, following my latest thread where I compare behaviour of Hypex NC252MP vs. A250W4R class AB amplifier into 4ohm//2.2uF, I have prepared files for the listening test, that were recorded from the outputs of these amplifiers, loaded with 4ohm in parallel with 2.2uF capacitor. This load simulates difficult load impedance of some electrostatic speakers, as explained in that thread. For technical questions, please visit that thread.

The amplifier were driven to 20Vrms max. output voltage (100W/4ohm).

The test sample is from wonderful Beethoven piano concerto, played by Krystian Zimerman.

https://pmacura.cz/piano_test_abnc.zip

If you are interested in the test, please download the test file, unzip it and make an ABX listening test. Or just tell us, what you hear from the test files, differences between both. If you wish, please post your result. This is mine:

foo_abx 2.0.2 report
foobar2000 v1.4.8
2023-03-23 19:06:45

File A: ABD_piano.wav
SHA1: 4252a76844e07e4fff77c55c593fda323e953320
File B: DAB_piano.wav
SHA1: 513e04fd56e7ec4fbb744c4f99f2da1e6f74c546

Output:
ASIO : Topping USB Audio Device
Crossfading: NO

19:06:45 : Test started.
19:07:20 : 01/01
19:07:26 : 02/02
19:07:31 : 03/03
19:07:36 : 04/04
19:07:41 : 05/05
19:07:47 : 06/06
19:07:57 : 07/07
19:08:02 : 08/08
19:08:11 : 09/09
19:08:17 : 10/10
19:08:22 : 11/11
19:08:27 : 12/12
19:08:32 : 13/13
19:08:38 : 14/14
19:08:46 : 15/15
19:08:50 : 16/16
19:08:50 : Test finished.

 ----------
Total: 16/16
Probability that you were guessing: 0.0%

 -- signature --
bc8dc32ab3ed3397620815e7f08532911403bc88

It should be quite easy. I have also made a Deltawave comparison by Paul's @pkane nice software. Some screenshots below:



And finally, the link to the test files:

https://pmacura.cz/piano_test_abnc.zip

 

[Sokel]

That's me:

If a 13 yo interface paired with 40 yo headphones can tell with such certainty things are not very nice.

 

[pma]

If a 13 yo interface paired with 40 yo headphones can tell with such certainty things are not very nice.

And it was recorded at high level. As the background whistles do not change with level, they get worse at lower volume. Not only this, there is also a strange level change in DAB file with strong piano strokes near 32s.

 

[amirm]

You can't use an audio interface to capture for this test. You must capture the output of the speaker using a microphone. The oscillations will intermodulate with the switching frequencies to create in-band spurs. This would not happen with a speaker as it won't play the switching frequency to being with, or even the oscillation itself. So there is nothing to intermodulate.

 

[TNT]

@pma - care to comment?

//

 

[Rja4000]

You can't use an audio interface to capture for this test. You must capture the output of the speaker using a microphone.

And by plugging the interface in parallel with the speaker (with a proper attenuator), would that work ?

EDIT: OK, I understand the interaction of high frequency oscillation with the interface is the issue.
You'd need a LPF, as you have to use with the AP to measure a class D amp.

 

[pma]

You can't use an audio interface to capture for this test. You must capture the output of the speaker using a microphone. The oscillations will intermodulate with the switching frequencies to create in-band spurs. This would not happen with a speaker as it won't play the switching frequency to being with, or even the oscillation itself. So there is nothing to intermodulate.

@pma - care to comment?

//

And by plugging the interface in parallel with the speaker (with a proper attenuator), would that work ?

When recording at high level, I use voltage dividers with low pass filters to avoid HF intermodulations in the audio interface. The idea to use a microphone to prepare files for the ABX test is incorrect, for well known reasons.

BTW, it was shown earlier that there comes audible buzz like sound from the tweeter when Hypex is loaded with the 2uF capacitor.