Rohde & Schwartz FPC1000 Spectrum Analyzer 'gitch'

[JohnPi]

I have an R&S FPC1000 SA. Under some conditions I notice an artifact step in the display at just about 1 MHz. This is not a function of the signals I put in -- in fact it also occurs with no input signal and just a 50 Ω connected to the input. I have seen this step as large as 10 dB in some cases.

It seems this is somehow related to relative values of RBW and VBW settings, and/or the averaging mode, but I can't figure out what.

Here is an example of no input, and (in this case a small step) a step is visible at 1 MHz (not precisely, generally 989 kHz).



Any ideas on what is causing this ?

[TheUnnamedNewbie]

When I just looked at the picture I thought it was something we see all the time in spectrum analyzers, though usually the other way round (sudden increases in noise-floor) usually when the analyzer mixer switches to a different harmonic. This might be similar, but more a case of an additional block needing to be engaged to measure below 1 MHz? some kind of upconverter?

Is there a block diagram available of the FPC1000?

[JohnPi]

No block diagram available from R&S. I have contacted their tech. support; I'll see if they can explain what it is doing.

One thing, while the SA is described a 5 kHz - 1 GHz, DANL (noise level) is only specified about 1 MHz. I hope it isn't horrible below that as that's what I'm most interested in...