FlightAware Discussions

Thoughts on optimizing gain

Not generally.

You quoted a discussion on someone using an amplifier and feeding the signal into a second amplifier.
Even for that discussion you will note that the range wasn’t reduced but rather the message rate.

Also without any amplifiers it’s basically impossible to have too much gain.

I use the Airspy in combination with RTL-SDR’s one stage 17dB (7x) LNA.
With the Airspy’s gain @ 21 there’s no clipping even with a helicopter hovering about my house so I guess that means there’s still some headroom available to crank up the gain by using a multistage LNA and play with the Airspy’s gain to turn the overall gain down to a point just before clipping?

In my experience turning up the weakest signal past -20 isn’t usually useful.

You can have the weakest signal at -30 as well and i don’t think it hurts reception.
Wouldn’t go lower than -30 or so, i think below -36 the signals aren’t detected as well.

The signals do clip in the ADC, but they are still demodulated.
In the rtl-sdr dongles you get overload issues in the analog stages, in the airspy those have more headroom.

The 2 staged LNA you have on the picture is the filtered one i assume?
(are the image texts swapped by chance?)

The image texts are correct, I only have the one stage LNA (below), the 2 stage LNA is from someone else called Wiedehopf I believe :smiley:
The 10dB gain difference between both LNA’s is clearly visible but this can be a coincidence because I don’t know at what value you have set your Airspy’s gain.

Yeah i’m running mine at lower gain.

Take a look at the graph below showing the signal level & aircraft seen in relation to the Airspy gain setting. I started with the Airspy gain at 21, then 20 minutes @ 20, 20 minutes @19 and 20 minutes @ 17. I didn’t try at 16 because 5x3dB equals about the gain of the LNA meaning that I could as well remove the LNA to obtain the same result.

That’s not quite true as those 5x3 dB are not completely in the first gain stage of the Airspy (which is also an LNA).
Also running 2 LNA stages with lower gain can be better than running 1 stage with maximum gain.

Anyway i don’t see any reason to try a gain of 16 either with the weakest signal at gain 17 already approaching -30.
That’s where i draw a very arbitrary line.

Well, you at least have to change back and forth a number of times and consistently see that difference in numbers.
Otherwise it can just as well be a random fluctuation.

I might agree that the difference from 19 to 17 is significant and that you lost a couple of airplanes in that transition.

But the difference in aircraft count from 21 to 20 and 20 to 19 looks like normal fluctuation and not significant of any performance difference to me.

How close to an airport are we talking?

It really mostly depends how many aircraft are really close.

Also depends what you want, sometimes you only get maximum range with higher gain.
That can mean sacrificing some update rate on the planes within 30 nmi and completely losing reception for planes below 5000 ft and closer than 5 nmi.

Anyway it shouldn’t get you much range anyway.
But if you feel like experimenting, go ahead.

If planes lower than 10 000 ft are flying closer than 15 nmi, and it’s a busy approach maybe even 2 runways or constant single runway operation, then your percentage of strong messages will probably be higher at the same gain as if you wouldn’t have so many low flying planes nearby.

I’m about 10 miles from a small airport that has 1 runway and probably 6 commercial flights a day. There’s probably more private single engine planes using it then anything else.

It will still increase the percentage somewhat.

Should be easy for you to look at the map and see if it seems you lose some of those commercial flights where you wouldn’t expect to lose them.

But unless your terrain allows 200 nmi range and more, setting gain for 5% strong messages usually doesn’t reduce range.
If your terrain and noise/interference levels allow 250 nmi range, then there is often some tradeoff between receiving close in aircraft and the ones really far away.

I can confirm it.

I had changed the gain from -10 (maximum) to the two lower levels (49.6 and 42.1).
Before it i had 10-12% messages > -3db, with the new values the amount went down to below 1%

There was no change in the max range, but trails of really close aircrafts ( less than 3 NM) were a flat line without any dotted parts in between.

For testing purposes i’ve switched back to “-10” after i installed my new antenna on Saturday.

Result is a lot higher number (30%) of messages > -3db
This also causes a lower number of messages in total and no change in aircraft seen.
The range is also slightly higher, but for getting this confirmed, i would need to run it for a longer period:

This is the moment i’ve changed the value:

And here’s an aircraft flying by pretty close, the line is broken around my location:

Hi, I have my value "Messages> -3dBFS (0.7% of messages) and an ADS-B Signal Median Level: -22.7
Are these values good, namely that the gain of Dump1090 is at 48? thank you in advance

From the 2nd post in this topic:

Thank you, what is the Dump1090 command line to use?

Same command line.

But you don’t need it, you already posted the percentage.

Yeah, I’m losing some planes below 7,500 to the north. The heywhatsthat site confirms it’s a terrain issue. To the south I can track out to 150+ nm but nothing near 250nm.

Terrain /houses/trees or is it just reception?

Seems something you did on Saturday improved your stats.

I would say near is mostly trees. My FA antenna is outside about 16’ above ground level.

On Saturday I changed out a NooElec to a FA Blue Pro Stick Plus.