Headphone amplifiers - do they make audible difference (output impedance) - test

[pma]

Just curious how much difference there would be between two different headphone amplifiers, both running below clipping, of course. One of them has output impedance of 2.7 ohm and the other one 50 ohm.

Test setup

Headphone amplifiers were connected to the analog output of Topping D10s DAC. JBL460NC headphones were used as headphone amplifiers load. Output from the headphone amplifiers was split to JBL460NC and to the input of E1DA Cosmos ADC and the sound samples were recorded at 96/24 format. Max. power during playing and recording music samples was about 20mW.

Music samples are uploaded to https://pmacura.cz/Promenade.zip

If you are interested, please download the file, unzip it and do the test. Foobar ABX report is preferred as a result of your effort. There is also a poll valid for 7 days.

Below is the measured impedance of JBL460NC headphones

 

[Bow_Wazoo]

Topping A90D vs. Schiit Lyr+

Do all amplifiers sound the same? Or not?! Decide for yourself. The components: Topping A90D Arichitecture: discrete NFCA modules. https://www.tpdz.net/productinfo/852442.html Schiit LYR+ Architecture: Hybrid. 6SN7 tube in the preamplifier, and discrete output stage in Class AB operation...

www.audiosciencereview.com

 

[pma]

Topping A90D vs. Schiit Lyr+

Do all amplifiers sound the same? Or not?! Decide for yourself. The components: Topping A90D Arichitecture: discrete NFCA modules. https://www.tpdz.net/productinfo/852442.html Schiit LYR+ Architecture: Hybrid. 6SN7 tube in the preamplifier, and discrete output stage in Class AB operation...

www.audiosciencereview.com

That's a very different comparison. Here the distortion profiles of both compared headphone amplifiers are much closer. The main difference is in output impedance, here.

 

[KSTR]

Output impedance difference will directly reflect in frequency response so this test is mainly about that difference*). Pavel is sure aware of this but maybe others are not.

The headphone used as test load has only little impedance variation vs frequency, therefore the impact will be small. But we know that even fractions of a dB can be audible when the cover a large frequency range.

A problem with comparisons of different frequency responses is that strict level matching is not possible for obvious reasons. Therefore best practice is to test at stepped level differences of 0.1dB over a range of +-1dB or so and see if a perceived difference in an AB (or ABX identification) is persistent.

*) As a very small secondary effect, there might be a difference in distortion, too.

 

[pma]

Some Deltawave and technical information:

1. File matching

Initial peak values Reference: -1.149dB Comparison: -1.119dB
Initial RMS values Reference: -20.633dB Comparison: -20.596dB

2. Deltawave plots

both spectra from 2 amps almost overlap

3. Harmonic distortion measured at 134mV voltage at headphone connector directly in the test setup

 

[pma]

So, not a single one positive ABX result yet, not mentioning it should have p < 0.05.