I was looking into getting a filter which would help with blocking the very strong GSM signals in my region. I had done the “Do I need a filter” test described in this thread and noticed that I would probably benefit from a filter since the EGSM-900 signals were really strong. It seems that this frequency band is used almost everywhere around the world. I thought that the FlightAware filter would be a good choice since many posts online had recommended it and the description says that it “passes range 980 MHz - 1150 MHz”. Well, I assumed it would mostly block everything outside that range but in reality it passes a whole lot more than that. Here are my before and after pics.
As you can see the filter barely touches the 925-960 MHz range which is my main problem. Many of you pros probably knew this already but I would have appreciated a warning like this post ahead of time since the product description is quite misleading.
Be aware that you can’t directly compare spectrum power scans unless you’re careful to set a fixed scale & gain - the defaults will scale the brightness of the spectrum image based on the strongest signal seen, so all the “with filter” graph is really telling you here is that the GSM signals are still the strongest thing you hear, not that they’re the same strength as without the filter.
I am more than happy to re-run the tests with proper configuration. What parameters should I use? To generate the above scans I followed @abcd567’s guide.
The pro stick plus has an internal filter that cuts quite some of that 900 MHz band but it is still stronger than the 1090MHz band. (Which is expected since the transmitters of aircraft are less powerful and further away than the mobile towers)
Have a look at this post if you are looking at getting another filter:
Please not that while the rlt-sdr lna improved my reception the reception was quite good with the pro stick and filter.
The signal does not need to be eliminated just weakened sufficiently that it interferes less
There is a difference between “is not working” as in “is not magical” and “working within the designed limitations”.
Any pass-band filter will have slopes on the sides, it will not be a sharp cut-off. Some are “sharper” than others and most of the time you will need to combine multiple filters to get a sharper cut-off.
Anyway, even a slow roll-off filter is better than none, it will attenuate some degree of the offending signals and that might be just enough for some locations. It will cut even deeper the FM and TV stations that might be in a certain areas. But yes, some locations will require multiple filters.
The ProStick Blue from above has one SAW filter inside, and look at the curve. Because of different resonant modes, the curve is not perfect at all.
I am using a preamp with two SAW, but truthfully, adding multiple SAW filters one after another will not “cut” much deeper the lower sides where it “bleeds” signal.
If truly the antenna is close to a GSM transmitter, a more expensive approach must be taken, ceramic filters.
Something like this, used in airport TCAS systems:
Lower rejection of stoped band:
Thanks abcd567 for the graphs. They show very well what is happening. My situation is pretty much the same as in the orange Pro Stick example. I guess I should switch to the blue Pro Stick Plus.
It seems that the 925-960 MHz band (or parts of it) is used for cellular network traffic in Europe, Africa, Asia, Australia and parts of South America according to this map so it is not just a European problem.
The rtl-sdr lna is quite a bit better than the pro stick plus at eliminating the 940 MHz band.
But getting a rtl v3 dongle and LNA will run you around 50$ so it’s a bit more expensive than a pro stick plus.
Also for someone from Europe it will take a week or two to arrive from china.
