THD+N Plugin Error?


When I run a plugin THD+N test on a repaired and refurbished NAD 3020e amplifier which has a rated output of 20W <0.05% THD I get the chart below. Even though the spec is without noise, obviously I was a little disappointed and assumed something wasn’t quite right with the amp.

However, on the main screen, if I do a direct test the amp does indeed manage to output over 20W with 0.002 THD / 0.005% THD+N.

Am I doing something wrong?

I’m using the latest stable S/W with a QA401 and QA450 as the dummy load.



Hi SABristol, in your spectrum we can see 22.39 dBV = 13.16Vrms and into 8 ohms that is indeed 21.6W as calculated. And the THDN at that point is -87 dB.

In the sweep, it looks like a similar number was hit at 9W = 8.5Vrms. So, this is a gap of 3.8 dB. You have -6 dB specified for QA450 in the sweep. Is your ExtGain also specified also as -6 dB in the spectrum plot?

Hi Matt,

Thanks for replying. As below, I have -6dB in the input gain to account for the attenuator in the QA450.

Capturing on a scope the Vrms across the 8 Ohm load it reads 13.53V which is 22.88W so ok.

Got it, thanks! Can you capture a plot of the THD(N) vs. Watts and the final spectrum display? From the THD(N) vs. Watts plot we can see the final measurement point, and from the spectrum display we can see the final measurement (once the graph is displayed, don’t run the analyzer again–just grab it’s display). That should indicate where the disconnect might be.


Hi Matt,

As you can imagine it was different this time and within manufacturers specification as both channels driven are showing as <0.005% THDN @ >20W :slight_smile:

I also ran the IMD/PWR test and the results didn’t look healthy as things get ugly @10W. Could be the same problem but I need to check whether this is in manufacturers spec and a realistic measurement.


Thanks for you help.


Ooops, I meant to insert this rather than the plugin chart.

OK, thanks and makes sense. I guess the point is that if you see something that doesn’t make sense, then adjust the stop point for the sweep at a value that you can manually inspect on the spectrum. For example, in your last plot, stop the sweep at 20W and verify the spectrum is as you expect (very clean). You can run a few extra cycles after you’ve stopped the sweep manually and poke around a bit to convince yourself what you are seeing is legit.

1 Like