Do I Need A Filter?

[LawrenceHill]

That one is actually for GSM-R. Used by European railways. It could still cause problems though if you are near a railway line.
This is the normal GSM one

[abcd567]

@LawrenceHill

That one is actually for GSM-R. Used by European railways.

I dont know German language except “gutten tagen” and “auf wiedersehen” , thats why could not make out it is for railways. It caught my attention due to frequency mentioned and the word GSM.

[abcd567]

@LawrenceHill

This is the normal GSM one

@hdtvspace

Sorry, there are many mistakes in wikipedia. official gov lists are the law and not private organisations…

This official German Government one, quoted by LawrenceHill, also says 862-960 Mhz, same as in Wikipedia.

Wikipedia data is generally authentic as they mostly use data from government or reputable research organizations.

[pretoescuro]

I really need a filter to block the GSM, in a 500m radius I have 6 GSM antenas (3 in line-of-sight). Currently I’m using Pro Stick Plus with light blue filter and a PCB antenna.
As a future plan I will try to build a cavity filter and try to tune it using the “Nose” Generator and a regular SDR dongle, does any one here ever tried it?

[abcd567]

@pretoescuro

I’m using Pro Stick Plus with light blue filter

Cavity filter, if built & tuned properly will provide very good filtering.
However, an easier solution to remove GSM900 interference is to use Dark Blue “1090 MHz Only” filter.

Announcing the new 1090 MHz only bandpass filter!

FlightAware is excited to announce the availability of a 1090 MHz-only bandpass filter for the ADS-B community outside of North America. This is something many of you have asked for. The filter we’ve offered in the past wasn’t great at blocking GSM mobile phone interference that can be found outside of North America. This new filter is an economical bandpass filter option that provides better attenuation of 950 MHz GSM. Those of you building PiAware sites in urban areas or other areas with 950 M…

FlightAware 1090 MHz ADS-B Bandpass SMA Filter

[foxhunter]

Pretty sure it’s somewhere in this thread of almost 600 replies, but is this filter helpful if you’re using already the blue Pro Plus stick?

Has someone tested the differencies with or without it?

[LawrenceHill]

@foxhunter

Has someone tested the differencies with or without it?

Yes, I have. Well with the older light blue filter.
With just the ProStick plus I was receiving about 600,000 messages a day. Adding the filter increased the number of messages to around 900,000 a day.
Here is a screenshot from when I added the filter back in 2018. Filter was added on 3rd May 2018

[wiedehopf]

@foxhunter

Has someone tested the differencies with or without it?

It entirely depends on your location.
For some it’s a big difference, for some basically no difference.

[abcd567]

@foxhunter

but is this filter helpful if you’re using already the blue Pro Plus stick?

@LawrenceHill

Adding the filter increased the number of messages

I could not make out if “this filter” and “the filter” mean the FA dark blue filter or Cavity filter. Can you please elaborate? Thanks.

[LawrenceHill]

@abcd567

I could not make out if “this filter” and “the filter” mean the FA dark blue filter or Cavity filter.

Neither, the filter I added back in 2018 was the older FA light blue filter.

[Nitr0]

@rugomol

Two identical ADS-B stations sharing FlightAware antenna via splitter, one with Uputronics Wideband Preamp and this filter, other station with Uputronics 1090 MHz filtered preamp fitted with SAW filter. Nooelec SMArts as RTL-SDR dongles.
Two configurations: Antenna => LNA => Filter, then Antenna => Filter => LNA.

Results were surprising, I expected that an $80 combo will outperform a $45 filtered preamp. Lowered gain to 40 dB and repeated test with LNA first (which is the setup used in the 1090 Filtered preamp or the Pro Stick Plus). Range plots were almost identical as well, so did a control run to check the SMArts, as usual, flawless performance.

Those results are very surprising to me as well, especially the filter first tests since the insertion loss of these cavities are quite low in comparison with the SAW units. I’m wondering if it was slightly off tune…not that it could be far since it’s only a 8Mhz pass (and you’d know it for sure!), but 1090 could have been hanging on the edge of the curve. Did you happen to sweep the particular unit on a VNA by chance?

[foxhunter]

@wiedehopf

It entirely depends on your location.
For some it’s a big difference, for some basically no difference.

Thanks for all your replies. Maybe i give the dark blue one a try later as it’s not too expensive just for testing.

[LawrenceHill]

@hdtvspace

How do you test this?

I tested it a number of ways but the primary way at that time was by comparison with a nearby station that I used as a reference. The nearby station was consistently ahead of me in the rankings and adding the filter put me ahead. I could compare the hourly figures of both stations to see if my changes were having a beneficial effect. Since then I have set up my own reference site.

[bive]

I’m about to do some changes to my setup. Currently I have the ordinary FA-antenna on the roof and a RTL-SDR tripple-filtered LNA. Works ok but I’m suspecting that there are interference from the numerous cellphone towers around me. Living in a more rural part of Sweden so the cellphone transmitters are more powerful (and have lower bandwidth as a consequence) than the ones in the cities. There are 6 towers within 3km (2 miles for the US-people) and when I do a spectrum scan it looks like this:

I’m going to add a cavity filter (from Sysmocom) between the antenna and the LNA. Tried with the cavity filter indoors and after the LNA, but that didn’t have any obvious effect on things. So now I have put together a water proof box with the cavity filter and an RTL-SDR tripple filtered LNA and it will be put in place sometimes during the week.

Will make a new scan when it’s up and running.

[bive]

@hdtvspace

Perhaps try this tips which I’ve tested and erase the most RFI and mobile and FM signals at the dongle:

FA antenna with no coax or only a very short (cm) good coax pigtail direct to the dongle with an 1090 bandpass filter! And then use a long usb 2.0 cable up to 5m is here perfect working from outside in the house.
Often seen and posted here, but a big mistake, dont plug the usb dongle (Fa Pro Stick Plus eg) direct into the raspi / PC! PC RF noise interference eg is here strong.
Only with a usb extension with ferrites and not in the near of the raspi/pc and never in the near of the power supply. Often seen here on photos in little boxes outside, but very ugly and wrong for perfect signals.

Put the dongle in a shielded metall housing / extra box with shielding, but beware of the heat and cooling air. Do not use PoE near the dongle. It’s a source of noise and bad for signal qualitly if you plug the dongle near that.

If you need for long coax feed lines an LNA, it must be installed BEFORE the long coax cable feed near the antenna. And not for short lines direct to to stick → voodoo triple filter and quad filter.

An cavity filter is the only thing which I would prefer in your case, but is very expensive. Like this one I use on one of my now three antenna setups… because I have a lot mobile provider towers and FM around of my house here in the city: https://shop.jetvision.de/epages/64807909.sf/en_GB/?ObjectPath=/Shops/64807909/Products/71010

The antenna is approx 7m from where I have my dongle and RPi. A bit too long to run a USB-cable even if it’s an active one. I have a low-loss cable (0.3dB loss per meter) after the LNA so the loss is not a factor. The cavity filter (7 pole) will be connected with a 0.5m long LMR400 cable to the antenna and with a 15cm cfl200 cable to the LNA). Should be good enough.

I have a good quality USB cable with ferrites between the airspy mini (metal housing and probably good enough shielding, @prog usually knows how to make good stuff).
I also have taken measures to keep every RF-producing part separated as much as possible from the airspy.

I feel that the only thing left if this doesn’t work is to build a couple of faraday cages for the airspy and the RPi, but I suspect that the benefits from that are so small (pre covid-19 I had 6-700 aircrafts on a really good day and an average coverage in all directions of 220 nautical miles’) compared to the cost of building proper faraday cages. Probably more to be gained by installing a tall mast and removing a couple of trees that block in some directions.

Thanks for the tips though. They are sound and may help others to squeeze out some more from their systems.

[bive]

Today the weather was good enough to make a trip up on the roof. A new box with the 7-pole cavity filter from Sysmocom and a tripple-filtered LNA from RTL-SDR replaced the old box with only a triple-filtered LNA.

Triple-filtered LNA only (1m CFL200 cable between the antenna and the LNA):

7-pole cavity filter with triple-filtered LNA (0.5m LMR400 cable between the antenna and filter):

The scan is made with Spectrum Spy and with a gain of 17 both of the times.
Looks a bit cleaner on the lower frequencies (terrestrial TV transmissions) and perhaps a little less interference from the 4G downlink (960 MHz).

I haven’t done any testing yet, but I see an increase compared to yesterday’s numbers. I also the that I perform better compared to some neighbouring sites (they are all below yesterdays numbers) but it’s too early to draw any conclusions from it since the traffic is what it is in these strange times.

Now I will start fiddling with gain and preamble settings for some time to see if things are different than before.

[randys77]

@hdtvspace

Often seen and posted here, but a big mistake, dont plug the usb dongle (Fa Pro Stick Plus eg) direct into the raspi / PC! PC RF noise interference eg is here strong.
Only with a usb extension with ferrites and not in the near of the raspi/pc and never in the near of the power supply. Often seen here on photos in little boxes outside, but very ugly and wrong for perfect signals.

My godness!

I’m planning to buy the Fightaware pro stick plus + ADS-B1090 antenna (0.6m, 5.5dBi) to upgrade my current setup (just from 1.0v to 1.2v ). So do you recommend to plug-in the FA stick into a USB hub instead direct to the Raspberry right? what kind of USB hub?

Thanks in advance!

[wiedehopf]

@randys77

I’m planning to buy the Fightaware pro stick plus + ADS-B1090 antenna (0.6m, 5.5dBi) to upgrade my

@hdtvspace

Often seen and posted here, but a big mistake, dont plug the usb dongle (Fa Pro Stick Plus eg) direct into the raspi / PC! PC RF noise interference eg is here strong.

I’ll just say that probably quite a few of long time posters would disagree about not connecting the SDR directly to the RPi.
Connecting the SDR directly to the Pi is how it’s usually done and provides very good results for many people.

@hdtvspace seems to have jinxed hardware as it seems everything is behaving different for him.

I always get the impression that using a USB hub only improves the results because they usually have their own power supply.
On the other hand, trying a USB hub doesn’t hurt.
But the the USB connection of the FA stick is designed with RFI in mind.
So i’m not sure it’s valid to say that this is always an improvement.

[abcd567]

@wiedehopf

I’ll just say that probably quite a few of long time posters would disagree about not connecting the SDR directly to the RPi.
Connecting the SDR directly to the Pi is how it’s usually done and provides very good results for many people.

Fully agree.
Using a USB extender cable often results in problems (even if it is a short length), if quality of USB cable is not good.

[Nitr0]

If external noise is of concern, do not use a USB 3 hub. That said, a Pi4 already has the noisey USB 3 chipset anyhow, so…