How come there is so much noise from 1080MHz to 1100Mhz? What setup do you have?
This is my test unit. Pi3 B+ with a Pro stick plus, onto a bias tee power injector into a RTL-SDR Blog ADS-B Triple Filtered LNA (Bias Tee Powered) but the noisy part, into the coax line, about 12 metres, up to a small 3db whip. As mentioned, my test unit with the PI3 on the desk, in the office. I will in due course move this unit to the pole, utilising POE and getting rid of the noisy coax. My main unit is up the pole, poe powered a lot higher with a DBP high gain stick. I am waiting for Raspberry to release their POE unit for the new B+'s before setting the test unit skywards and replacing the current POE with a raspberry one too (much better designed). Now that this script has worked, I will tonight install it on my main unit and see what results I get.
What gain setting are you using? Sorry if you already mentioned it, I cannot recall.
Potentially, this is the total gain available to you: 53 dB (if using -10 for gain setting), or 49.6 if not using -10. Add to that around 20 dB from the ProStick Plus pre-amp. Another 27 dB from the RTL-SDR Blog pre-amp. If you add an FA antenna, an additional 5 dB.
You cannot keep adding gain and expect performance improvements, even with extra filtering. These dongles/sticks are little marvels, but they are only 8 bits.
Well based on the standard gain tests, the best figures for those results is 25.4. Now using these visuals, what we you suggest?? Now re the other set up and your comments on this one, what would you get rid of to reduce the gain or swap to a pro stick?? I would think the RTL sdr??
On my main unit, I am running a 7.7 and with this unit my longest detect is 512 kilometers, (that was a freak unique day)average though between 430 - 460 k’s.
Ok…good. I think you are pretty much at the limit. Granted that every installation is different, I have mine currently set to 36.4, but my stick does not have a pre-amp. I use an RTL-SDR Blog pre-amp/filter combo as well, and an FA antenna. So in my case, my total gain is 36.4+27+5=68.4.
Your total gain is 25.4+20+27=72.4.
I think you are already in the shrinking SNR region.
My suggestion is to reduce the gain to 20.7, or try without the RTL-SDR Blog pre-amp, or replace the ProStick Plus with a regular dongle/stick, one that does not include a pre-amp.
In my limited experience, once you go under 16.6 gain in the dongle/stick, no pre-amps, performance seems to degrade.
It’s difficult to see any signal at all in all that noise – it looks like static throughout the entirety of the filtered in-band part, as if that noisy coax you mention isn’t connected properly and so acting as a 12 metre antenna, or the power injector is throwing all kinds of noise into the mix.
It could be interesting to try again with bare minimum – just the Pro Stick Plus and antenna and see how that compares to these scans. Then add in parts and see where all that noise comes back in. You can test a single gain that you know to be currently noisy by stopping dump1090-mutability and running the rtl_power command directly using the relevant lines from the script.
I don’t think it’s your gain setting because the noise and the signal are about equal all the way up through the run from the very start, whereas what you’d expect to see is that the noise disappears and leaves good signal once the gain gets low enough.
I do have a 4G mobile phone tower about 300 metres from home. With ABCD’s filter test, there was some noise coming from that but I found the pro stick plus did reduce that 850 900 noise. I having been, with that tower near, thnking filtering is what I need, subsequently using the pro stick plus and the RTL _ SDR both of which supposed to have good filter features.
I think filtering is essential in your case, but that still does not give you unlimited room for increasing gain. The losses introduced by the filters can be compensated, of course, but the law of diminishing returns (a.k.a. input saturation) will eventually kick in.
If you had left the gain on AGC with the FA antenna, that’s why probably you got less numbers. The signal is too strong for AGC and will clip lots of signals.
Personally, in my location, I had to lower the ProStick Plus to 36 dB to maximize the number of connections. It’s very simple - keep the web page with the flights open and follow the number of messages.
I assume you use dump1090-fa. Then, in the SSL connection to the Pi, type this (copy/ right-click):
sudo nano /etc/default/dump1090-fa
Change the number after the --gain to something else. Then save and close. And:
sudo systemctl restart dump1090-fa
Watch the webpage for the Message numbers. Wait 20 seconds till the numbers stabilize a little. Then repeat until you get the highest numbers.
Until I try this on my main unit, this is all experimental. I know it works now and that option of reducing the tests is in the plan. Is there a general gain figure you aim for, given a “perfect” physical set up. I think I have that with the main unit, 1 M from the antennae, cat 6 network cable to the WWW. A 9bb antennae. Th variables are what gain you are adding. The only negatives would be “external” noise etc.
Unfortunately, no. The goal is best performance, and that has no correlation to gain sometimes. Every setup is different, and if that is not enough, every RF ambient is different.
Experimentation is key, but you cannot go around the fact that gain cannot be added non stop, and that these dongles are not top of the line receivers to begin with. They are great for what they cost, but have limitations.
Try the ADS-B Receiver Project graphs, they are an eye opener.
If he is using Piaware, it’s easier to use this command:
sudo piaware-config rtlsdr-gain 36.4
Where 36.4 is the gain I have my receiver setup, just change that to the gain you want.
Then do the:
sudo systemctl restart dump1090-fa
Less chances of accidentally corrupting a text file.
Yes i use those graphs. My SNR at the moment for both units is good, 12 or less.
Now here is a good question. What is better, physical gain or gain induced via the script control of the gain of the dongle???
LOL I nearly know these scripts off by heart … 
To use your term, I think physical gain is better. By physical gain you mean antenna gain, hopefully. Gain from pre-amps can be the same or worse than the dongle’s script control gain. The noise figure of the pre-amp is an important specification.
Running the dongle’s gain too high, or too low, is best avoided in my opinion.
That’s good! Try a setting of 20.7 then, as I suggested earlier. It may increase the SNR.
Echoing what @Dxista said but also to add that even on a single setup the best gain can change from hour to hour depending on traffic density, interference, temperature and so on. In fact my highest readings so far have been using AGC and I’m tempted to put mine back to that and leave it alone. It seems to be able to keep me at max performance throughout a 24 hour period.