Problems with the spectrum in 700MHz frequency

We are working with LimeSDR over the 700MHz frequency with 15MHz rate and 40 gain.
When doing two spectrum measurements and changing the central frequency by 2MHz we see totally different results.
For example in image # 1 the central frequency is set to 708 MHz and the noise floor variation is from 702.5 MHz to 711.5 MHz.
In image # 2 the central frequency is changed to 706 MHz and we do not see the same noise in the frecuencies like in image #1.
We would like to understand why this is happening and whether there a way to correct it.

Central Frequencies upper 720MHz works fine.

Image #1


Image #2


pls post a screenshot when tuning to 2124 MHz


@joshblum, @IgnasJ, any thoughts on this?

After @c5e3 response we suspect we are seeing a signal from the 3rd harmonic of the higher frequency which is 2124 MHz.
Does it make sense? If so, is there any way to apply a digital filter to remove this from the 708 MHz frequency? Any recommended reading to do it? I am not an RF specialist.

since an IQ mixer uses a square wave as clock, you will see signal, that appear on the clock’s harmonics.
this is completly normal and cannot be filtered in software

1 Like

All of the low cost signal generators are just using square wave fractional N synthesizers as well. My signal hound VSG25A will show up at FS, FSx2, FSx3, etc… Totally forgot about this, so it threw me off for a bit. So I guess this is where RF filters come in handy :slight_smile:

Which Electronic filter you can recomended me?
I had tried with this

but did not filter by complete the signal