1.920 THD is > THD+N

Hi @Gruesome, You can limit the harmonic count by setting the measurement stop frequency. For example, if you are sampling at 192k with a fund=10k, and you wanted to see just the first three harmonics (20, 30 and 40k), then you’d set Measurement Stop to 45 kHz. The harmonic count is more elegant, for sure.

In other tests one now often sees exponentially spaced (evenly spaced on a logarithmic scale) multi-tone intermodulation test signals. That’s not something the QA403 can generate, or could it?

The Multitone button in the GENERATOR section will let you specify a number of tones per octave and a desired RMS level. For example, these settings:

will generate this output:

Duh! Of course, the button is right there. @matt , how does one set the number of tones (or the number of octaves) to be generated? I see the field to set the number of tones per octave. But where do I set the range?

(As an aside, in general I find the various frequency range or bandwidth settings difficult/non-intuitive to find. What I mean is the distortion integration frequency window, the noise integration frequency window, the multi-tone generation window at issue here, and the display window. Wouldn’t it be better if there was one settings menu where these settings would all reside?)

Ideally, I’d like to specify frequencies and amplitudes myself. One example would be to generate Bob Cordell’s 1980’s three tone MIM signal (19kHz, 10.05kHz, 9kHz). A general way to edit, save and read in a frequency & amplitude (and starting phase?) list for the generator would be most helpful.

Similarly, it might be helpful to be able to specify a multi-notch filter, so one can have the software integrate the distortion power (or sqrt(power)). It would fill a gap between looking at the amplitudes of n-th order harmonics directly & separately when only a few signal lines are generated, and visually assessing the amplitude and density of the ‘sea’ of distortion lines generated by many signal lines. Related: in some of your posts, like post #7 above, from December 2021, you (@matt) show highlighted frequency bands, for example around signal lines excluded (notched) from a THD calculation. I have not found a way to turn this feature on.

Having just recently started to think about it, I believe looking at the peak level of the ‘sea’ of distortion lines will just evaluate the dominant harmonic, often the 2nd (or maybe third, if there is a cancellation mechanism for even orders). Having the software integrate the multi-line distortion signal might be better. With sufficient resolution it might even be possible to do this by order, i.e. separately sum all n-th order distortions that can arise from a multi-line signal spectrum.

Or maybe this already exists, and I just haven’t found it yet?

Added: I double checked that I am not overlooking menus or options because of my tablet interface.

One deficiency I noted is that while right-clicking works on a tablet (long touch = right click), the hovering tooltips that display for example definitions for the measurement boxes above a graph screen do not appear with a tablet, unless a keyboard with a mouse (equivalent, e.g. touch pad) is attached.

I double checked with a keyboard the right-click menus for the underlined buttons in the ‘hardware’ left side menu, and I do not see any additional options besides the ones that appear on long-touch with the tablet touch screen interface.

WRT spectral contamination or multitone source this paper A New Class of In-Band Multitone Test Signals AES Convention Papers Forum » Spectral Contamination Measurement explains the limitations of the usual multitone signals and suggests specific sets adjusted so that IM and harmonic products don’t land on source tones.

I have a spreadsheet but the forum software won’t let me upload it. Here is an image:

Thanks, @1audio , I had read about that issue (avoiding overlap of signal and product tones). That would be one reason to specify the exact tones manually, with a list. Are the frequencies in your list already adjusted so they fall into the center of frequency bins for customary FFT resolutions?

I don’t know. I suspect they may be but the list is 20+ years old and adc’s and soundcards have evolved.