Diversity Reception

[tjaekel]

Dear FlightAware team and community,

all works great, with FlightAware ProStickPlus (blue dongle). Cool stuff - amazing!
My ‘issue’ (question):
I see from my home (Orange County) so many airplanes (up to north of Bakersfield, entire LAX traffic - amazing!), but if a plane flies right over my home (I live right under a practice area with so many small planes here) - I do not see any plane (they disappear - sure).
OK, obvious, the antenna is vertically oriented and the ‘doughnut shaped receiving pattern’ cannot see anything vertically above me, just above the angled horizon from my location.

Here my idea/questions:
a) could I combine (like an inverse splitter) two antennas (one vertically, one horizontally orientated)?
(I tried, not a real improvement, makes it worse - sure due to impedance mismatch if not done
properly or a loss of 3dB, at least)
b) Could I setup two Rpi 3+ FlightAware receivers and combine the reception results into one map?
It means: Could I merge two channels into one display (based on the database, the final
receiver messages)?
c) OK, I am aware of this ‘MLAT’ feature (a friend having same receiver, many miles away, for
triangulation). But I want to use TWO receivers at my location and combine the results.

How to improve the reception of ADS-B right ‘above’ my home?
How to augment the results with a second, e.g. diversity receiver?

BTW:
If I let stream my receiver into the Internet (to your server) - would you combine the results, so
that my map is much more accurate (because somebody else sees the planes above my home)?

Is there any software API so that I could merge two receiver receptions into one map?
(extending the software running on Raspberry Pi)

Best regards
Torsten

[obj]

@tjaekel

a) could I combine (like an inverse splitter) two antennas (one vertically, one horizontally orientated)?
(I tried, not a real improvement, makes it worse - sure due to impedance mismatch if not done
properly or a loss of 3dB, at least)

With a single receiver this is unlikely to help. Your problem is likely to be too much gain more than anything.

Also, ADS-B signals are vertically polarized; mounting an antenna horizontally will produce very poor reception.

@tjaekel

b) Could I setup two Rpi 3+ FlightAware receivers and combine the reception results into one map?
It means: Could I merge two channels into one display (based on the database, the final
receiver messages)?

Yes, this is probably the most straightforward setup. Run one with a lower gain antenna (less directional) and/or lower gain settings in piaware. To combine the display, you can either run a separate dump1090/skyview instance (in --net-only mode) just for display, or use software like VRS that can take multiple data feeds.

The simplest thing to try to begin with is to decrease the gain on your existing install and see if that improves your overhead reception.

[wiedehopf]

Gain is most likely the problem, reduce it.

The inverse cone above your antenna where it doesn’t receive well is very small at low altitudes so the small plane should be visible except for 3 seconds directly over the house.

Some explanation in regards to gain:
Thoughts on optimizing gain - #2 by wiedehopf

As for combining two receivers with the --net-only method mentioned above, i’ve written some scripting to make that easier:
GitHub - wiedehopf/combine1090: Combine data from multiple ADS-B receivers into one readsb decoder / tar1090 webinterface

[geckoVN]

@wiedehopf

The inverse cone above your antenna where it doesn’t receive well is very small at low altitudes

It is ‘very small’ for a perfect antenna. I suspect it doesn’t exist at all for any physical antenna.

[wiedehopf]

@geckoVN

I suspect it doesn’t exist at all for any physical antenna.

I have a home-made quarter-wave combine with quite a sensitive setup made of an LNA and airspy-mini.

Still when aircraft at cruising altitude fly directly above there can be a short pause of signals.
I can also see the RSSI falling sometimes before i stop receiving the aircraft.
3 or 4 seconds later, reception resumes to normal good reception.

Maybe the FA antenna behaves differently with the cone being non-existent.
But quarter wave antennas seem to have this characteristic.

I assume it’s the combined alignment of the aircraft sending antenna being perfectly in line with my antenna.
But the effect is definitely observable.

[geckoVN]

@wiedehopf

But the effect is definitely observable.

I stand corrected.

(20chr)

[wiedehopf]

@geckoVN

I suspect it doesn’t exist at all for any physical antenna.

So i tried to give my argument some background, i’ve modified SkyView to draw dots for every position received:

Note: Circles at 2, 4 and 6 nmi, so these screenshots are fairly well zoomed in.

The coloring is according to RSSI but that is quite flaky, nonetheless the blue dots mean lower RSSI than the red dots.

So you can see the frequency of positions received reduces when the plane flies directly overhead.

A second example:

And another one:

The last screenshot also has somewhat more consistent RSSI.

[retman1222]

this is nice wiedehopf, a great way to explain

[wiedehopf]

Just to clarify, normally when people complain they are not receiving planes flying close-by, they just have the gain set too high.

This produces much larger gaps in coverage than this effect.

Too high gain is especially relevant with lower flying aircraft.
With aircraft flying higher up, the angle above the horizon reduces signal strengh enough to still receive them.

[obj]

You may also be seeing an effect where the body of the aircraft starts blocking one of the transponder antennas (typically, larger aircraft will have two antennas - top and bottom - and alternate between them when transmitting)

[modea]

BTW:
If I let stream my receiver into the Internet (to your server) - would you combine the results, so
that my map is much more accurate (because somebody else sees the planes above my home)?

I have a pretty simple setup using socat to pipe data from rx nodes into a central instance (similar to what wiedehopf has posted) you can view it here http://tbh.sdr.farm:8080 - it’s not beautiful but it works

I still need to work on a local setup that has hi/lo gain feeders