Page 1 of 1

dBr on peak - why level > 0dBr ?

PostPosted: Mon Nov 07, 2022 9:09 am
by RJA4000
On spectrum analyzer, when I select dBr Vr=Peak for the Y Scale, I happen to have values above 0dBr.

Is that normal ?

In below example, V1 level is at 7.24dBr, as may be read in the measurement summary above the plot.

Peak greater than 0dBr.png
Peak greater than 0dBr.png (92.6 KiB) Viewed 3175 times

Re: dBr on peak - why level > 0dBr ?

PostPosted: Mon Nov 07, 2022 5:11 pm
by VirtinsTech
It is normal. The Normalized Amplitude Spectrum uses the apparent peak value in FFT bins as the 0dBr reference. Therefore the spectrum curve will never exceed 0dBr. However, the true magnitude of those peaks in the spectrum curve will be equal to (for the cases of no spectral leakage) or higher than (for the cases of spectral leakage) the magnitude represented by a single FFT bin. Therefore, their values may be higher than 0dBr after spectral leakage compensation. We use the apparent but not the true peak value as the 0dBr reference for the Normalized Amplitude Spectrum because there is no guarantee that there exists a true peak in the spectrum (e.g. white noise).

Re: dBr on peak - why level > 0dBr ?

PostPosted: Sat Dec 31, 2022 6:12 pm
by RJA4000

OK, but it's very confusing, so that should be an option, IMO.

Also why not to allow "Mark peak" as in THD measurements ?
This does nothing in this case, while it would be very useful.