I was looking for a scientific reviews of NAD d3020 v2 and some tips regarding its usage with Elacs DBR-62 and PC audio.
Unfortunately there are not many scientific reviews regarding d3020v2, and I did not had experience with audio before, so i looked for and measured as much data possible and here is the result.
About the amplifier:
It consists of modified Class D amplifier Hypex UcD102, which should have lower THD (between 0,005%-0.01%). SNR should be at -98dB.
Internal DAC is Cirrus Logic CS42528 (According to https://www.soundstagesimplifi.com/...d-d3020-v2-bluetooth-integrated-amplifier-dac), which should have -114dB SNR.
And these two components I primarily measured.
1. Software
Since I am not a pro when it comes to audio I used shareware called Audio Tester V3.0.
https://www.audiotester.de
In my usage, it had a lot of bugs, but it allowed me to:
- generate 1KHz sinewave (and others)
- to see FFT graph
- it has oscilloscope
- to perform frequency sweep test
- to measure frequency/distortion ratio
The software allows more tests, but I do not have necessary tools and skills to perform them (I am a software engineer, and I got a lot of help from people who do electrical engineering).
2. Hardware
Soundcards used for measuring:
ALC 1220 onboard of AsRock x399M
https://www.techpowerup.com/review/asrock-x399m-taichi/8.html
Creative LiveSB 1140, USB (for verification) on notebook
3. Things I have learned on this forums before performing the actual measuring
My only previous experience with audio quality was by year 2000 when I decided to remaster tapes which contain games for an old 8bit computer on audioCDs. So here are few findings or confirmations
a) I confirmed that Toslink (optical SPDIF) isolates amplifier from noise generated inside PC.
Unfortunately line in, used for the tests might still pick up noise.
b) -0.14dB gain in Equalizer APO prevents to trigger CAudioLimiter and avoids unnecessary distortion, which fixed an issue with "Flat" sound when using Cirrus DAC in the amplifier.
c) How AudioTester works, about its calibration and features.
So now for the results:
FFT Graph
All measuring is done over Toslink on -0.14dB gain in WASAPI Shared mode. Windows volume is on 100% on Speaker Out, Digital Out and Line In. Most measurements were done on 44,1KHz 24bits output and input set in Windows, while the software may indicate Float mode.
Amplifier Headphones output was calibrated by AudioTester to absolute maximum volume by 1KHz Sine, before LineIn on ALC1220 gets overwhelmed by the input, so the signal is on its possible maximum. AudioTester in calibration displayed 87% as highest readable value - whatever it means.
As you can see there are harmonics at 2, 3, 4, 5k and higher as its expected and in most cases they hit up to -98dB.
However I do not know how much power was sent to each channel. When the Amp is using headphones output, its power consumption "from the wall" was indicated to 11.6watts, 15,6watst when headphones are no longer plugged in. So my rough guess its 5 watts per channel.
As you can see, the measuring is picking up some noise, about that later.
Frequency sweep
For some reason I saved only test at 192KHz 24bit. Its highest value for CS42528, and as you can see the lines are nice flat, mostly within 0.3dB range.
Frequency Distortion
Distortion is mostly between -80 to -83dB and thus results are according to specs of 0.005% - 0.01%.
Again without necessary skills and tools, my best estimate of power for the speakers in this scenario is between 5-7 watts per channel as the wattmeter of whole amplifier indicated 11.6-15.6 watts depending if headphone jack was plugged or not. For more details check the link to V1 version of this amplifier.
Noise control
Even when there is no signal being sent to the amplifier, there are significant bumps in 80Hz, then at 2, 3, 4, 5 and more KHz even without 1KHz tone.
Unfortunately onboard ALC1220 LineIn picks up a lot of noise from the computer. So as a control I used small netbook and USB soundcard SB Live 1140.
Region at 80Hz is now clear, however harmonics above 2KHz are probably just below visible threshold of -110dB. Thats where I had to improvise.
Hearing Noise test
Previously when I measured GPU generated noise, i confirmed that when amp was on maximum volume, the sound could be heard from the speakers.
Instead I sent 2,3 and 4 KHz sinewave at -96, -102, -108, 110 and -114dBs directly to the speakers with amplifier on maximum.
Only 4KHz was slightly audible at -114dB (5cm away from the tweeter), which is interesting, because -114 is SNR of the internal DAC, and Equalizer APO already decreased all volume by 0.14dB. It should not be audible, not measureable.
When I turned off 4KHz sinewave, I could also tell its no longer there, and no other noise is coming out of speaker. I guess that this kind of noise is in fact only in Line-In, but not in amplifier, or its actually so low, its inaudible.
Edit: I did a verification measurement for this. I noted that noise spikes at 2-5K were at roughly -95dB. Then i sent 1KHz tone at -90dB and it wasnt visible on FFT. Then i maxed the amplifier and 1KHz showed up at roughly -63dB, but the 2-5K noise was still at -95dB. That means it wasnt amplified and its source is in LineIn or PC. For the same reason i can ignore spikes in 2nd verification measurement at 200, 400 and 600Hz. Unfortunately, reading 1KHz harmonics is due this unreliable.
Scientific reviews of similar amplifier:
https://www.stereophile.com/content/nad-d-3020-integrated-amplifier-measurements
Review for V1 should provide data I am unable to measure, as the modified UcD102 is used in both V1 and V2.
Conclusion:
1. Measuring of frequency sweep and frequency/distortion seems to show correct values.
2. Unfortunately tools I have available failed me at SNR.
Its much harder for me to tell SNR of this device, because harmonics of 1KHz could be higher due noise which is probably only in Line-In, but not in the amplifier. SNR can be between 90 and 103dB, depending which values I choose to trust. SINAD is probably between 79 to 83.
Unfortunately there are not many scientific reviews regarding d3020v2, and I did not had experience with audio before, so i looked for and measured as much data possible and here is the result.
About the amplifier:
It consists of modified Class D amplifier Hypex UcD102, which should have lower THD (between 0,005%-0.01%). SNR should be at -98dB.
Internal DAC is Cirrus Logic CS42528 (According to https://www.soundstagesimplifi.com/...d-d3020-v2-bluetooth-integrated-amplifier-dac), which should have -114dB SNR.
And these two components I primarily measured.
1. Software
Since I am not a pro when it comes to audio I used shareware called Audio Tester V3.0.
https://www.audiotester.de
In my usage, it had a lot of bugs, but it allowed me to:
- generate 1KHz sinewave (and others)
- to see FFT graph
- it has oscilloscope
- to perform frequency sweep test
- to measure frequency/distortion ratio
The software allows more tests, but I do not have necessary tools and skills to perform them (I am a software engineer, and I got a lot of help from people who do electrical engineering).
2. Hardware
Soundcards used for measuring:
ALC 1220 onboard of AsRock x399M
https://www.techpowerup.com/review/asrock-x399m-taichi/8.html
Creative LiveSB 1140, USB (for verification) on notebook
3. Things I have learned on this forums before performing the actual measuring
My only previous experience with audio quality was by year 2000 when I decided to remaster tapes which contain games for an old 8bit computer on audioCDs. So here are few findings or confirmations
a) I confirmed that Toslink (optical SPDIF) isolates amplifier from noise generated inside PC.
Unfortunately line in, used for the tests might still pick up noise.
b) -0.14dB gain in Equalizer APO prevents to trigger CAudioLimiter and avoids unnecessary distortion, which fixed an issue with "Flat" sound when using Cirrus DAC in the amplifier.
c) How AudioTester works, about its calibration and features.
So now for the results:
FFT Graph
All measuring is done over Toslink on -0.14dB gain in WASAPI Shared mode. Windows volume is on 100% on Speaker Out, Digital Out and Line In. Most measurements were done on 44,1KHz 24bits output and input set in Windows, while the software may indicate Float mode.
Amplifier Headphones output was calibrated by AudioTester to absolute maximum volume by 1KHz Sine, before LineIn on ALC1220 gets overwhelmed by the input, so the signal is on its possible maximum. AudioTester in calibration displayed 87% as highest readable value - whatever it means.
As you can see there are harmonics at 2, 3, 4, 5k and higher as its expected and in most cases they hit up to -98dB.
However I do not know how much power was sent to each channel. When the Amp is using headphones output, its power consumption "from the wall" was indicated to 11.6watts, 15,6watst when headphones are no longer plugged in. So my rough guess its 5 watts per channel.
As you can see, the measuring is picking up some noise, about that later.
Frequency sweep
For some reason I saved only test at 192KHz 24bit. Its highest value for CS42528, and as you can see the lines are nice flat, mostly within 0.3dB range.
Frequency Distortion
Distortion is mostly between -80 to -83dB and thus results are according to specs of 0.005% - 0.01%.
Again without necessary skills and tools, my best estimate of power for the speakers in this scenario is between 5-7 watts per channel as the wattmeter of whole amplifier indicated 11.6-15.6 watts depending if headphone jack was plugged or not. For more details check the link to V1 version of this amplifier.
Noise control
Unfortunately onboard ALC1220 LineIn picks up a lot of noise from the computer. So as a control I used small netbook and USB soundcard SB Live 1140.
Region at 80Hz is now clear, however harmonics above 2KHz are probably just below visible threshold of -110dB. Thats where I had to improvise.
Hearing Noise test
Previously when I measured GPU generated noise, i confirmed that when amp was on maximum volume, the sound could be heard from the speakers.
Instead I sent 2,3 and 4 KHz sinewave at -96, -102, -108, 110 and -114dBs directly to the speakers with amplifier on maximum.
Only 4KHz was slightly audible at -114dB (5cm away from the tweeter), which is interesting, because -114 is SNR of the internal DAC, and Equalizer APO already decreased all volume by 0.14dB. It should not be audible, not measureable.
When I turned off 4KHz sinewave, I could also tell its no longer there, and no other noise is coming out of speaker. I guess that this kind of noise is in fact only in Line-In, but not in amplifier, or its actually so low, its inaudible.
Edit: I did a verification measurement for this. I noted that noise spikes at 2-5K were at roughly -95dB. Then i sent 1KHz tone at -90dB and it wasnt visible on FFT. Then i maxed the amplifier and 1KHz showed up at roughly -63dB, but the 2-5K noise was still at -95dB. That means it wasnt amplified and its source is in LineIn or PC. For the same reason i can ignore spikes in 2nd verification measurement at 200, 400 and 600Hz. Unfortunately, reading 1KHz harmonics is due this unreliable.
Scientific reviews of similar amplifier:
https://www.stereophile.com/content/nad-d-3020-integrated-amplifier-measurements
Review for V1 should provide data I am unable to measure, as the modified UcD102 is used in both V1 and V2.
Conclusion:
1. Measuring of frequency sweep and frequency/distortion seems to show correct values.
2. Unfortunately tools I have available failed me at SNR.
Its much harder for me to tell SNR of this device, because harmonics of 1KHz could be higher due noise which is probably only in Line-In, but not in the amplifier. SNR can be between 90 and 103dB, depending which values I choose to trust. SINAD is probably between 79 to 83.
Last edited: