I’m using piaware with an RTL-SDR v.4. I’ve been picking up messages at a rate of 350-400/second and I’m usually tracking around 50 aircraft up to 150-200km away. Max RSSI in view1090-fa hovers around -3 and min around -27.
I picked up a second hand flight aware 1090 bandpass filter (dark blue) and with that in-line, the max signal drops to about -10/12 and I pick up about a quarter of the aircraft and message rate. I was expecting some drop in signal strength but that seems a bit much.
I’ve tried fiddling with the gain but that gets me no where. Is the filter simply not all that useful to me in my location? Have I bought a faulty filter? What steps can I take to debug this?
If RSSI has dropped 9 dB, it sounds like it’s a faulty filter. However, do you have an LNA after the filter, or is it just a filter ahead of the RTL-SDR dongle? If you do have an LNA, that could be another factor to consider.
If you don’t have test equipment for making actual measurements, there’s not much more that you can do since you have already fiddled with the RTL-SDR gain to no avail.
If you don’t have strong interference that filter is unlikley to improve your reception at all.
Especially if you have an SDR without builtin LNA or an external LNA.
Hard to say if the filter is faulty.
max RSSI will change around quite a bit.
graphs1090 comparison will help there.
I assumed @fettle was looking at graphs1090 to see the 9 dB reduction in RSSI. However, if you’re not using it you should install it using wiedehopf’s script. It’s very straightforward.
The insertion loss of the filter at 1090 MHz should be around 1 dB, however I wasn’t able to find the spec for this at FlightAware store. Anyway, that’s why I think it’s a faulty filter, or maybe connections to the filter.
With your setup, it won’t really matter much whether the filter is near the antenna or the RTL-SDR. However, it is unusual to just connect a filter without an LNA. How long is the coax from antenna to the filter/RTL-SDR, and what type of coax is it?
About 5m of coax - which is longer than I ended up needing so I ought to cut it down at some point. I’m not sure of the type, it came attached to the antenna.
That SDR has input → LNA → SAW filter → rtl-sdr
(very similar to the adsbexchange blue SDR / FA blue SDR)
It’s not impossible but unlikely that you need a prefilter for that.
Also frees up the v4 for other endeavors.
If you’re not too much into the software, adsb.im makes it a bit easier to give acars / vldm or sonde decoding a try.
Limited options but easier to set up.
That’s interesting, it looks like the signal level dropped about 2 dB. Just what you would expect. Did you try to increase the gain in the RTL-SDR dongle by a few dB and that didn’t help the overall performance?
rtl-sdr without LNA, gain is at 49 anyhow for ADS-B in pretty much all cases.
And going to 58 / -10 / tuner agc is not really that big of a gain either.
The rtl-sdr internal noise on the input stage is just too high.
That’s why people use LNAs after all.
Also without an LNA, rtl-sdrs usually don’t need filters.
You’d need to be stupid close to an interference source.
Testing if a filter improves things is a much better test than using rtl_power to create spectra anyhow, that’s mostly guessing.
Ah - I think producing the graph is probably better than me just glancing at the figures in view1090-fa. You’re right the loss is a lot less than I described (if I look at the peak figures before/after for example).
I have rtlsdr-gain set to 60 both before and after.
