Posted by
Stewart A. on
URL: https://www.es-forum.com/Intermodul-harmonics-mixing-signals-at-GSM-base-stations-Aaronia-Spectran-tp1543519p1543540.html
Hi Joris,
Here is the list of do's and do nots I did not get from the manual:
Use Only a 3MHz filter/bandwidth above 3GHz in combination with Peak or
Pulse mode.
You MUST use Pulse-mode to catch the higher digital bursts, like for WLan or mobiles. Celltowers are ok WITHOUT pulse-mode.
3.1GHz is the "switchpoint" between two generators on the Spectran, so a
false signal can show up at that point.
> I measure several ranges separately (for example the 2099-3011 MHz,
> 5470-5725 MHz, 5725-5825MHz ranges) and found that the high signals
> are very clear and also hearable (audio analysis), in many cases at
> around 200 meters from the base ststions ca. 2-6 V/m, in contrast to
> the for example 0,2 - 1,2 V/m values in the GSM / UMTS bands !
>
>
I have seen spikes in those areas too.
The 3GHz spike disappeared when I used correct settings.
The 5.8Ghz region is to be blamed on wireless internet and gadgets here.
> I think these high signals can be real, possibly 'intermodulation
> interference signals' (and out of band). I read a lot on the
> internet
> about measured signals and that all spectrum analysers have the
> problem
> of second and third (...) "harmonic frequencies", "intermodulation
> interference" from several different base station antennas, "mixing
> products from harmonics", etc, and that the problem is becoming more
> important, as more different antennas are being installed at each
> base station.
>
>
The harmonics to be expected are mentioned in the manual.
However I also do find harmonics that are not in 20db reductions, and I
think are real in some cases, like with separate frequency channels
within a band.
> See for example
>
http://www.tek.com/Measurement/App_Notes/2G_14758/eng/2GW_14758_0.pdf>
>
Right. The signal to noise ratio is troublesome for those wanting the
signal to get through. But such interference should not be ssen as
Higher db readings.
> The 900 (gsm) and 2100MHz (umts) signals from base stations could
> indeed also mix and give something around 3000MHz, and the readings
> between 5-6 GHz could also be explained by (third order) harmonics
> and mixing ?
>
doubtful at that level.
Does the first recommendation above fix any of this?
> The not-real 'mirror' signals ("in" the measured range) are normally
> weaker, but can the real (really in the air) signals
> like "intermodulation/mixing/harmonics" signals be more stronger
> than
> the original signals separately ?
>
>
Good question. The manuals say no, but 2 light bulbs are brighter than
one. The issue seems to be regarding harmonics...
> I also noticed that when I put my GSM on (in speaking mode), high
> signals
> in the 3000 MHz and 5000-6000 MHz range are also found (and in audio
> analysis also clearly the sound of a GSM). This can also be
> explainedby "intermodulation/harmonics/mixing" signals ?
>
>
So with a known transmitter nearby at 900 or 1900MHz, and the Spectran
set at 5-6GHz with a bandwidth of 3MHz, you are hearing signals that can
be definitely linked to the GSM?
Is this in Peak/Pulse mode?
Stewart