I am a new QA403 user. I cannot get the unit to auto-scale the Full-Scale Input. It stays at the value it comes up at regardless of input level. I have to change it by hand to set the best value. This feature is the most significant difference from the QA401 I have, and I am most disappointed it is not working.
Auto-scaling works with the supplied tests for THD vs. Frequency and THD vs. Level. I ran the self-test loopback tests from the QA403 Product Sheet Rev H, pages 15 – 17 (page 17 attached). As I watched the test run, the Full Scale Input changed level. The results match pages 15 – 17 of the product sheet.
Since I could match the product sheet, this shows that auto-scaling is working for these supplied tests, but it is not working in the most simple use case—applying an arbitrary level signal at the input of the QA403.
Specific to the attached THD vs. frequency loopback test, I have attached a loopback (page 15 of the Product Sheet Rev H with the Y axis max change to -80dB). The THD value drops to 0 at around 7kHz for all levels. This matches the Product Sheet, which stops at 7kHz.
I do not understand this. I get the same issue when I apply a single 6dBV tone and vary the frequency. Above 7kHz, the displayed value of THD is 0. I can see the second harmonic on the screen for signals between 7kHz and 10kHz, but the THD value on the display shows 0. Changing the sampling rate and expanding the x-axis above 24kHz does nothing.
Why is the THD calculation not working above 7kHz?
Hi @DavidR, none of our analyzer autoscale. The reason is that you will often set the input level based on what you are trying to optimize for (noise or distortion). So, if you are encountering an unknown signal, set the full input to max (42 dBV), connect the signal, note the level, and get the next highest input if you are seeking to minimize noise and perhaps bump up 12 or 18 dB if you are seeking to minimize distortion.
Bear in mind the measurement stop frequency is the upper limit for including harmonics (multiples) of the applied signal. At 24kHz, you will not be measuring enough harmonics to get a good picture of anything above 8kHz or so.
You want to sample at 192kHz, have an upper measurement limit around 90kHz and expand the horizontal axis on the plot out to 100kHz.
Note that 80kHz seems to be the standard measurement bandwidth for audio distortion plots. I presume this came from 256kSPS sample rate of many FFT analyzers, since an 80kHz anti-aliasing filter is fairly easy to design for Nyquist frequency of 128kHz, but its only a guess on my part.