Cool, but on the other hand, you could simply slow down the timebase on the 'scope to get more waveforms on the screen to increase the resolution. I can easily and completely separate two tones at 1000Hz and 1005Hz (with levels shown in dBV) on my cheapish Siglent 'scope's FFT.
True, you can have better tones definition changing the timebase, my bad. However, the dynamic range is about 30 dB better using the spectrum analyzer (at least with my instruments) and it's easier to measure. Thanks for the input.
Impressive! I must admit I was skeptical at first, because a good sound card can achieve close to 100dB dynamic range, with the frequency resolution basically only depending on capture length. However, it seems your project can easily match that performance, and on second thought that’s not too surprising with the high quality, low noise components you are probably using. One question though, why the attenuated followed by amplifier stages? Wouldn’t that just introduce more noise?
Thanks for the comment. The noise contribution of one attenuator pad is about 0.0203 uVrms (-140dBm) for a BW of 500 Hz (span chosen), so doesn't impact much on the measurement, also because there's also the noise contributon of the generator and amplifiers. The timebase phase noise is about -150dBm@100Hz. The attenuators were introduced mainly to adapt the levels, but the second attenuator has also the function of providing a consistent 50 Ohm resistive load to the mixer, minimizing reflections.
