Yes, really.
Innovation comes from challenging old truths with new thinking and new technology.
Just look around and you will find many examples of what we do today that was seen as impossible to improve just a few generations ago. The same applies to radio technology.
The wireless cellphone industry developed the outstanding LNA chips of today that can work in harsher conditions (on the cell tower, in close proximity of emitter antennas) than what we have in our houses.
Without actual measurements tools to prove that the current LNA design is not working, IMO this is just a solution looking for a problem.
That wasnāt the issue. Rather do you need the filter in front of first stage or it can be after.
Also, those scans donāt prove anything about the actual recovery of ADS-B signals. Yeah, we all live in areas with RF noise. We donāt need to add dedicated filters in front of our TVās or Radios to work.
Are you aware you are arguing with the very guys who made your setup work. We only exist at our scale because we do something better than the rest of the industry. I am more than happy to discuss the technical details, but just saying that our industry solved some problem ages ago and we should look elsewhere because of magic, thatās not how engineering works.
What we are discussing here is way beyond the filtering solutions you mentioned and might be subject to a patent if we see an interest in doing so.
In any case, I shared the schematics with RTL-SDR Blog and we could see something in the upcoming weeks.
I think that sounds very positive if it can be achieved. Thatās going to bring up the wanted signal and reduce the unwanted signal. It sounds like it can really help the signal to noise ratio. Iām watching with great interest.
Me too
hehe me coooool
Still playing around with the gain settings and the -e option
Cannot decide using 18 or 19
In this chart you can see the time where i used 18 (the part where weakest was lower).
Do i stress that question that too much and at the end it does not matter?
And the corresponding message rate and range
Gain is far less relevant with the Airspy than with an RTL dongle. If you have a shufti at my graphs on the three month view, you can see where I dropped the gain down a notch and my weakest signals dropped from around -36 to -42 and itās made very little difference to the overall performance.
Pick either one. The difference (if any) may be some more messages from aircrafts close to the horizon and perhaps a few more messages from aircrafts blocked by trees close to the antenna (these are my own experiences, may not apply to all setups and so on).
But the number of messages extra you may get will probably not make a big difference on the FA stats.
Thanks for the input.
Iāve set it down to 18 from 19 and monitoring it. So far no impact, the values are similar to the yesterday
I was discussing only whatās available and public knowledge now.
IMO if indeed is something new and improved, the cellular industry has deeper pockets and would be more interested in better front-end filtering than the ADS-B hobby industry.
Good luck!
I usually am one of those that is lucky enough to have disposable income to spend to try new things.
I have bought the Airspy Mini and R2 (donāt use the Mini anymore), the RTL-Blog LNA (used it as is, then modded the power supply and now not in use anymore).
So most likely I will buy a new product just to compare with what I have now.
We employ guys from the same ācellular industryā who used to build 4G base stations. Our main customers are not hobbyists. But yeah, getting that ADSB algo work on a commodity SBC and beat FPGA based implementations was a funny personal challenge. Solving the RF problem is just another challenge.
I grew up in days of Silicon NPN transistors (like BFY89) when we only dreamed of Gallium Arsenic components of today. So I have a reference point of how far the RF electronics advanced because of cellular industry, that pushed the development and miniaturization of RF blocks - both active and passive (filters).
Example: My LNA has in the front end a SPF5189Z (GaAs pHEMT, soldered by me, as an upgrade to the TQP3M9008 E-pHEMT the Chinese LNA had in there). Both those devices would had been pure magic to the younger me.
MMIC are good for lazy engineers and students, but impose severe design limitations in more complex systems. Discrete RF is not only more flexible, but also much cheaper.
Check this, this and this.
Hereās an example. A single transistor preamp for HF/VHF with negative feedback loop vs. a very good MMIC: https://twitter.com/LondonShortwave/status/1304113277648150532
VHF can be done by transistors with a hand tied at back 
Hi all, noob here and just got the Airspy Mini a few days back, Iāve adjusted nothing as yet but struggle with the whole gain and sample rate thing, and is there anything else to change? Here is my 24 hour graph for signal, would you change the gain at all? I do have a 1090mhz Uputronics preamp on the way so just trying to learn this now in anticipation!
And my 24 hour message rate
Here is also my 7 day view which shows the change in signal when I attached the Airspy, previously for gain I was using the script that someone wrote on here to handle it but that doesnāt work for Airpsy
Luckily with the airspy itās easy to get right.
sample rate ā¦ just leave it be itās not as stable on 20 MHz and often not much to be gained.
Just put the light blue line in the graphs around -30 and youāre usually fine.
Having the Airspy now installed on a different site i am thinking which " -e " setting might be the best one.
Itās my spare Raspberry 4B and iāve set it to 11.5 where the CPU usages touches the 90%
I remember reading it somewhere here that something between 80 and 90 is ok?
With the current gain setting the weakest signal level is around 26, ok as well?
I have the feeling that with a lower gain which brings it to 30 the average message rate is lower.
Just check the graph for the max CPU used ā¦ check the journal that you donāt lose too many samples.
If you donāt lose samples ā¦ no issues.
Message rate feelings are unreliable, i wouldnāt trust them 
But yeah itās not an issue.
Also itās not 26 and 30 ā¦ itās -26 and -30.
