FlightAware Discussions

High rise antenna?


How high is the tree?
Lets’s say 30’. if you can get 20’ up with a tree stand (I have done that), then you just need a 10’ extension (PVC) pipe to reach the tree canopy.


I posted above pics. It’s tall… 20-25’ is to the lowest branch.


Maybe you could use the tree stand to get yourself up higher with the bow and arrow, then shoot the arrow in such a way that it wraps the line around your ankle and sends you even higher in the tree. Then, you can place the antenna easily and only have to worry about how to get down- but gravity will be working in your favor in that case. :joy:


The fence and shed provides for a good scaling factor. I would say closer to 25’

Yep, you could easily place a tree stand anywhere in this area. Even if the bottom of the stand is at 30’ then, assuming you are 5’8” that puts your working height at 35’. Put a 10’ 1 ¼” PVC pipe at that height and you are at 45’ with much less leaves. A 20’ fence pipe might even put you above the tree. Now you are in lightning strike territory.

Above all, be safe, that is a long way to the bottom.


Long time since I didn’t work on this project, but today I finished it.

  1. Bought a slingshoot with 3 rubber tubes. With that I was able to launch a big nut from which I have tied fishing monofilament. That flew fairly up, around a branch, and down the ground.
  2. With the monofilament I have pulled a long parachute cord nylon rope. Thin, light and strong.
  3. With that rope I have pulled up a piece of PVC pipe (10’ long), hooking it from center of gravity. When the connection point is up at the branch, half of the pipe is sticking above that branch. This is about 70’ above ground.
  4. On that PVC pipe I have mounted at one end the FA antenna and at the other a sealed box with the preamp. The preamp is modified to accept power over coax, with a 5V stabilizer. I am feeding 9V on the other end, tried on the ground with a 200’ coax reel.
    Coax cable to house (overhead) and #10 copper cable for grounding down to ground next to the tree.
    This is the cable that I have bought: https://www.amazon.com/gp/product/B015UPRIU4/
    Those are the adapters: https://www.ebay.com/itm/2-Pcs-F-Type-Female-Jack-To-SMA-Male-Plug-Coax-Coaxial-Connector-RF-Adapter/272674830869
    And the power inserter (at the house end of the cable): https://www.ebay.com/itm/Commscope-power-inserter-for-your-cable-TV-RF-amplifier/263967080098
    And yes, I have a “shadow” where the tree trunk is, but I couldn’t make it higher…

Antenna cable advice requested

Results? The number of messages/sec more than doubled, finally I am above 1000! Probably the airpsy mini helped too, it did increased the message rate on old antenna location (3 ft above roof line).


The wider dynamic range probably helps a lot too. It allows the gain to be up and still hear most of the traffic. You probably get a lot more traffic at night, when it is quieter.

The newer code also works much better. When I upgraded recently the CPU dropped by almost 50%.


I am surrounded by trees (at some distance), so probably going higher helped a lot. There is still some foliage in the way to horizon, but I did go above a lot of it.
The Pi3 CPU is maxed out now, at 1000 msg/s. Older location I was getting 5-600 msg/s max.


You also switched to a different preamp with two filters didn’t you?
Anyway nice range :wink:

Cpu would be maxed out at 400 percent as there are 4 cores.

If running in dx mode the decoder loses quite a bit more cpu i believe.
That would probably really max out the pi.


Ah, that’s how CPU usage it’s counted in top? I’m good then, that’s with DX mode on (I am not sure what it does). I’ll post the htop screens below.
And yes that’s preamp, shown in the pic. Is the one with two stages and two SAW filters. Followed by 150’ of RG6 coax running into the house and a splitter (I am feeding the FlightFeeder too).

My rc.local looks like this:

echo performance > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu1/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu2/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
sudo sysctl -w net.ipv6.conf.all.disable_ipv6=1
sudo /home/pi/airspy_adsb -c -l 30005:BEAST -g 14 -p -w 4 -x &

I have to use the -l 30005 in the airspy because the SD card install, when used with “receiver type” other or relay, deletes the output to port 30005 from dump1090-fa. Ask Obj why :slight_smile:

Generated automatically by fa_config_generator

This file will be overwritten on reboot.

DECODER_OPTIONS="–max-range 360"
NET_OPTIONS="–net --net-heartbeat 60 --net-ro-size 1000 --net-ro-interval 1 --net-http-port 0 --net-ri-port 0 --net-ro-port 30002 --net-sbs-port 30003 --net-bi-port 30004,30104"
JSON_OPTIONS="–json-location-accuracy 2"
RECEIVER_OPTIONS="–net-only --net-bo-port 0 --fix"

This is htop now, in the moring (less traffic). The load on each core varies wildly on each sampling period (second).


For outdoor use, use Black cable ties.
I just noticed one (of two) of mine (white/natural) holding up an outdoor AP had failed from UV in only six months. These are 10mm wide ties found outside a power sub-station, not some ebay special.


Not sure why you would touch the governors but hey i’m sure there’s a reason.
You don’t need to worry about CPU you have plenty room i would say.

And as said you could use -l 29888 with airspy, set in piaware-config

receiver-type relay
receiver-port 29888

Then the system would actually work as intended and you could drop the -c 30004 on the airspy.
Beastplitter would feed dump1090-fa via connect 30004 and listen on 30005.
piaware and fr24 and other feeders could then get their data on 30005 :wink:
(Right now beast splitter just fails miserably at doing what it’s supposed to do and i’m surprised it’s not spamming your log with errors)
Oh nevermind you can just turn beast-splitter off by setting receiver-type to other right?
(Leaving the airspy options as they are)

Where did you buy that preamp, looks interesting?

Also i first thought you had 200 nmi range but you have 200 mi which is still very nice!


Yeah, that’s statute miles. Right now I am using the “other” as receiver and that turns off the beast-splitter.
The governor settings are recommended by Airspy.
The preamp is from eBay, China origin. I have added the power regulator and coil.


Murphy’s law. I have waited and waited for the lightning protector to arrive from China and I have decided to stop waiting and put up the antenna without it…

Next day what I get in the mail?

Oh, well, when life gives you lemons… make lemonade :slight_smile:
I have installed it on the end of the cable that’s in the house. At least will protect the receiver dongles, even if the cable might get compromised.


I have a question regarding use of multiple antennas.

In obstacle cases such as this, could you couple two antennas together and put one on each side of the tree trunk? To basically see through the tree by feeding signals from both sides?


Probably. But with my method of lifting it would be harder because I would hev to have a great aim on a branch on the other side. Climbing up there would definitely solve many problems.
However, until now it does not look like it is a very strong shielding effect.
Maybe I will run mode2deco on a different Orange Pi or VRS on PC and get the exact coverage anvelope.


Not to be a heretic but you could use the planefinder client, it gives a nice coverage display.
Also includes a somewhat minimal local 3D view.

Connecting multiple antennas to one dongle is really complicated, you are basically designing one big composite antenna. Without RF equipment like a network analyzer i wouldn’t start to be honest.
If you really want to do something like this using 2 dongles and combining the data is the FAR easier approach.


I didn’t think it’s free.


It is.

Thinking about your setup it came to mind that you could combine both receivers in one display and increase the gain on one of them to get more planes overall.

Piaware is receiving all data directly from the airspy mini so you can feed anything you want into the dump1090-fa without screwing mlat or anything else, it’s only for display.

I assume flightfeeder is running the default piaware config and offering MLAT results on port 30105 and normal beast data on 30005.
In the following bash script you will need to replace FF-IP with the ip of your flightfeeder.

while true
        socat -u TCP:FF-IP:30005 TCP:localhost:30004
        sleep 15
done &

while true
        socat -u TCP:FF-IP:30105 TCP:localhost:30004
        sleep 15                                                                                                                       
done &

You can run this bash script from your rc.local and then you will see all data combined on map of the RPi.

It will show a stupidly high message rate because many messages will be duplicates but hey i’m sure it will be nice especially for MLAT planes because normally only one of your 2 receivers will be participating and getting the position sent back.

Then turn the gain on airspy up by a few dB and the flightfeeder down a few dB. (The airspy is probably better at detecting the far away weak signals)
Any planes close in that would normally vanish will then be fed to the map via the flightfeeder.


I might try that, but on 30104 (30004 is closed on SD card setup that I have now) and I think that duplicated messages will be discarded, because they will have the same time stamp?
As for the level of airspy… It’s crazy! Even if it has the same front end like other receives, the actual DAC has soo much overhead! I am with that gain at 20, most of the planes around me (like miles away) have RSSI at 0dB and they still track a continuous line! And yes, I have messages with levels down to -10dB and a few with -49dB (that I guess are errors).
The other receivers would just stop getting messages at this level.

LE: Nope, socat the 30004 or 30104 doesn’t increase the count.