Oh i was mistaken then, thought you wrote that you got it somewhere along the way.
I guess the non-LNA stick will have to do then 
Be sure to try the PCB antenna and check if it makes a difference to the mag base.
Also as i said people have been running gains of 5 with still lots of >-3dB messages.
Still the reception was good, so you should also be able to use it with a ProStick if you so desire.
True, but the +10dB figure is a happy coincidence.
+10db into a 75 Ω load gives 0.86V
The 820T datasheet suggest a diode pair to protect the input - typically a BAV99 - that will start to clamp the input signal when it exceeds 0.7V
The RF path includes several filters and another variable gain amp on the output. This output amp is controlled by the AGC directly from the RTL chip itself.
Regardless of what is happening on the input, the AGC should be able to prevent the RTL being overloaded.
That might explain why ADS-B reception is working even with excessive gain in front of the 820T. (rtl-sdr LNA + pro stick for example)
Clamped signals are fine as long as they are âcleanâ signals before clamping.
You can in most cases still demodulate clamped ADS-B signals.
Might still explain this.
The demodulation might still work better with non clamped signals when two messages overlap.
When one message is letâs say 9 dB stronger than the other message, even with overlap reception should be possible.
But when you clamp both messages to the same level, reception might be impossible.
Still long range reception shouldnât be affected as those are weak signals anyway.
And overlaps of two such strong messages shouldnât occur too often.
When you say it works best with an attenuator, which other setups did you try in comparison using the rtl-sdr LNA + v3 combination?
Ah - good point.
Iâve got a set of 3, 5, 6, 7, 8 dB attenuators.
I started without any to get a baseline, then started adding attenuators either singularly or combinations.
By the time I got to 12dB, it was obvious I was on the downhill slide.
8dB seemed to be the sweet spot for this site.
How is the performance if you use no attenuator and adjust the gain so you have a similar percentage of >-3dB messages?
Itâs be an interesting test with two V3âs; one with physical attenuators and the other backing off the front end ampâŠ
I didnât try adjusting the gain on this site because it takes too long to edit the config and restart.
I tried adding the gain adjustment to skyview, but it conflicted with powering the bias-T.
It really shouldnât, you know what the problem was?
Or rather how/where is the rlt_biast called?
This should work, it edits the configuration and restarts dump1090-fa.
Automatic gain optimization for dump1090 fa · wiedehopf/adsb-wiki Wiki · GitHub
The problem is, the AGC has too long a response time to be able to back off the gain to prevent overloading. The AGC is designed to operate with carriers that are constant, however ADSB signals are just brief pulses and so do not raise the AGC voltage enough to reduce the gain.
Most people forgot, or never knew, these dongles were initially sold for DTV reception in Europe.
Thatâs absolutely true, but it will adjust the gain to suit the prevailing noise floor.
That may be true for those whoâve bought an ADS-B kit etc, but for those of us who started with a generic dongles, itâs hard to miss. On the case, (the one within reach) says âDVB-T+DAB+FM+SDRâ!!
My first (an E4000) didnât list SDR.
That in turn can mean it adjusts the gain so every aircraft in a 10 nmi radius is overloading the receiver and you canât receive them.
Also can mean that you get significantly less messages from the planes out to maybe 40 nmi.
If the noise floor is very good, which it should be with a well filtered LNA, itâs not good to adjust the noise floor as the wanted signal.
Generally speaking, AGC should be OFF for best digital signal decoding.
That makes sense, but what does âoffâ mean?
Itâll either be max, min or an arbitrary value in between.
AGC is the automatic gain control which is only enabled with gain set to -10.
Once you set another gain then itâs just a fixed gain and AGC is disabled.
That is better because you can check your signal by other means (RSSI) than the AGC controller in the dongle (reading continuous rx power).
This script does that for you: Automatic gain optimization for dump1090 fa · wiedehopf/adsb-wiki Wiki · GitHub
The script doesnât have anything to do with hardware AGC (-10) though.
It merely uses -10 as effectively a gain setting of 55.
This is from my software test rig, with an RTL STR dongle and an LNA-less antenna in a thatched roof space, so pretty bad installation.
My results would imply that I need some filtering, correct?
I have been given a 5 cavity filter by a friend who designs such things, and I will publish the results for this setup with the filter included.
Time and rain permitting, I will repeat the same two tests with a wideband LNA on an external pole.
Which command line did you use for that?
Seems like you have a little noise everywhere.
Iâd make sure the RPi with dongle is below the antenna and not at the same height.
Otherwise the dongle/RPi can feed noise into the antenna.
sudo rtl_power -f 800M:1200M:100k -i 30 -c 50% -e 30m -g 30 -F 9 >scan.csv
I will extend the cable to get the Pi away from the antenna after the filter run has finished.
With 10m cable.
There were about 10 aircraft without this cable, now 1 or 2 due to attenuation, but interestingly still strong interfering signals.
Which brand is the dongle?
It seems noisy.
And moving the RPi didnât help the noise so the other setup is clearly better 
You might just need a better antenna or a filtered LNA.
A better dongle might also help.
The filter might not even help much because your signals already seems very weak.
If the roof space has some kind of aluminium foiled isolation or other metal then you are out of luck anyway and wonât ever get acceptable reception.
Hello wiedehopf
It is an RTL-SDR.COM USB dongle
https://www.rtl-sdr.com/buy-rtl-sdr-dvb-t-dongles/
