[Objective Benchmark Part 1] Op Amps comparison in voltage gain stage.

The different Op Amps were measured after sufficient stabilization / burn-in time, using identical conditions. Those measurements are my first trial at doing some Audio Precision measurement, so any feedback will be appreciated.
Note: Please do not consider those measurement as an absolute reference, nor that they accurately depict how the amplifier will sound. Those are worth sharing with the community to see how different Op Amps behave and try to draw correlation between Subjective and Objective.
See my Subjective sighted impressions here (blind test coming soon).
Let's start with the overview and conclusion on the performances, and keep the separate and more detailed measurement further down this article:
_________________________________
Performance Overview
_________________________________
THD+N in function of the generator level in Volts, tested at 1kHz.

We can see that v6 Vivid has the most singular behavior, with distortion slowly rising well before clipping level. In the opposite direction, it also seems to be the cleanest at low levels. Purely looking at the best THD+N figure of the lot is not very relevant, but at this game, LME49720, OPA1642 and LM833N are the winners. Looking at individual measurements, LME49720 pulls ahead at highest output voltages.
Now, testing the components to its limits is good, but we must also focus on the conditions in which we are actually listening to the components. With loud listening levels using HD58X and high dynamic range musics, I reach around 1Vpeak, = 0.707Vrms, which corresponds to ~0.25Vrms position on xAxis, as the amp volume knob is calibrated to +9.3dB. This basically means we should only be looking at the curve below 0.25Vrms if we wanted to correlate objective and subjective impressions.

Next figure shows THD in function of the frequency, with a fixed 2Vrms output (corresponding to xAxis point of previous figure at 0.7Vrms):
THD (without Noise) differs quite a lot between Op Amps. Some of them manage not to have any THD increase at high frequencies, which gets closer to an open loop system.
OPA2107 seems to have an odd discrepancy between its two channels below 600Hz. 
From here on, there will be a dedicated section for each Op Amps tested, which gives much deeper view over it's actual impact on sound quality. 

How to easily compare measurements: open this page two times in separate windows (not just tabs), click on the first picture of the desired Op Amp, and use "alt+tab" to alternate between windows. You can then alternate between 2 Op Amps measurements while keeping the full screen picture, since the order of measurement will be kept consistent.

_________________________________
OPA1642
_________________________________

Single tone and Sweep tests:

  • 1kHz - 7.3Vrms output (+9.3dB gain):
Apart from the 0.4dB gain gap between the L and R (inherent to my whammy, will be compensated in the future), we can see that my measurements still suffer from high frequency (4k+) and 50/150/250Hz+ harmonics noise. This is the best I could do for the moment. Better measurements will have to wait until the lab construction is finished.
  • 100Hz - 7.3Vrms output:
  • 10kHz - 7.3Vrms output:

THD+N = f(frequency, Hz) (2Vrms fixed):


THD+N = f(input Level, V): (+9.3dB gain):

Inter Modulation Distortion:
  • 60Hz+7kHz 4:1 (-2dB gain)

  • 19kHz+20kHz:

  • 1kHz+5.5kHz:

 Noise measurements:
Performed at 0% volume, APX generator OFF.


_________________________________
OPA2107AU
_________________________________

 Single tone and Sweep tests:
  • 1kHz - 2Vrms output (-2dB gain):

  • 1kHz - 7.3Vrms output (+9.3dB gain):

With its residual (+80dB):

  • 100Hz - 2Vrms output:
  • 100Hz - 7.3Vrms output:

  • 10kHz - 2Vrms output:

  • 10kHz - 7.3Vrms output:

THD+N = f(frequency,Hz):

THD = f(input Level, V): (+9.3dB gain):




Inter Modulation Distortion:


  • 60Hz+7kHz 4:1 (-2dB gain)

  • 19kHz+20kHz:

  • 1kHz+5.5kHz:

Noise measurements:
Performed at 0% volume, APX generator OFF.

Performed at 100% volume, APX generator OFF.



_________________________________
LM833N
_________________________________
Single tone and Sweep tests:
  • 1kHz - 7.3Vrms output (+9.3dB gain):

  • 100Hz - 7.3Vrms output:

  • 10kHz - 7.3Vrms output:

THD+N = f(frequency,Hz):

THD = f(input Level, V): (+9.3dB gain):




Inter Modulation Distortion:

  • 60Hz+7kHz 4:1 (-2dB gain)

  • 19kHz+20kHz:

  • 1kHz+5.5kHz:
_________________________________
LME49720NA
_________________________________
Single tone and Sweep tests:
  • 1kHz - 7.3Vrms output (+9.3dB gain):
Looks very clean to me!
  • 100Hz - 7.3Vrms output:

  • 10kHz - 7.3Vrms output:

THD+N = f(frequency, Hz) (2Vrms fixed):

THD+N = f(input Level, V): (+9.3dB gain):




Inter Modulation Distortion:

  • 60Hz+7kHz 4:1 (-2dB gain)

  • 19kHz+20kHz:
    • 1kHz+5.5kHz:

    Noise measurements:

    Performed at 0% volume, APX generator OFF.

    _________________________________
    Burson V6 vivid
    _________________________________

    Single tone and Sweep tests:
    • 1kHz - 2Vrms output (-2dB gain):

    • 1kHz - 7.3Vrms output (+9.3dB gain):

    With its residual (+80dB):
    • 100Hz - 2Vrms output:
    • 100Hz - 7.3Vrms output:
     
    • 10kHz - 2Vrms output:
    • 10kHz - 7.3Vrms output:

    THD+N = f(frequency,Hz):

    THD = f(input Level, V): (+9.3dB gain):


    Inter Modulation Distortion:

    • 60Hz+7kHz 4:1 (-2dB gain)
    • 19kHz+20kHz:
    • 1kHz+5.5kHz:
    Noise measurements:
    Performed at 0% volume, APX generator OFF.

    Performed at 100% volume, APX generator OFF.


     Due to some noise (50,100,150,200Hz..) and some ~5.1kHz polluting and impacting the measurements. See below what I measure when probe is connected to the APX555, but not yet to the amplifier:



    I shall revisit those measurements once I learned more about APX ; have a clean environment ; have a second round of Op Amps to test.

    !!!!!!! PAGE IN CONSTRUCTION !!!!!!!

    Comments

    Post a Comment