Disappointing results with my setup. What's wrong?


After many days failing miserably I give up, need help.

Here is my setup

  1. Raspberry Pi 3 B
  2. FlightAware Pro Stick
  3. 8 elements Coax Antenna (7 1/2 actually)
    RGC213 antenna (11,3mm segments)
    RG58 feed cable (1m long)
    50 Ohms BNC connectors
  4. 19V PoE injector + Splitter + 5V DC-DC Stepdown
  5. Everything installed on the roof of my 4 floor building

The results are simply ridiculous, looks like there is no antenna at all. A 7mm wire give better results than the antenna itself (< 10nm)

What I did so far

  1. Tested without the PVC pipe (thought I could have some RF shielding material). Same results
  2. Tried to remove the self-amalgamating tape > Same results

Electrical connection on the feed line is ok

I really don’t know where to go, this is the 3rd attempt from ground I making. Tried with RG6 75Ohms cables and the results are almost the same.

Is there anything I could use to diagnose what’s happening?

Some pictures of my antenna and setup

1 Like

Only been doing this a month or two, but I would first recommend you try your setup with a cantenna or a four wire spider. They seem much easier to make and infinitely easier to get right.

I am sure the experts will be along later with more in depth analysis and assessment, but there are volumes on this forum about COCOs and heart break. Good Luck.

That’s a nice looking setup with the weatherproof box and rooftop pole mount up above trees. I’ve enjoyed making my own antennas, including my first and only failed CoCo, but if I had your setup I’d consider buying a FlightAware antenna. It would probably do great up there.

Good point about the ground plane comparison.

Brought the Pro Stick down and did some tests on my window using my reference ground plane


Ground plane: ~60nm
Coco: ~20nm

I’m doing something really stupid with my coco designs, but I really can’t find what.

The coax cables I’ve used on my “production” are 82% VF, so 275/2*0,82 ~= 113mm

Interestingly, the more elements I put on it, worst are the results :imp:

About buying a ready to go antenna, the whole goal of this project is achieving something DIY. I already bought the Pro Stick because I tought the bad results I was getting could be releated to my RTL dongle/sat amplifier. And it would be extremelly expensive to ship it overseas.

On the COCO, did you check each segment to make sure you were not shorting out? That is all I can offer. Sorry. Oh, your outfit looks very very good. Well done! Hope you can get it resolved,

Just a couple of things that may help the “team” help you. What results are you getting? What results are you expecting? Where on the planet are you located?

(Your stats page could answer most of these questions.)

You have a very nice installation with the advantage that the tip of the antenna element is within meter lead reach of the input to the dongle. Do you have continuity there? As another asked, have you insured that the center element is not shorted to the shield?

Good luck!



I get almost no result at all. Using the antenna in the first post (8 elements) or a 1/4 whip is the same (< 40km).

I managed to get some decent results some weeks ago (excelent at < 80km, good at < 150km and poor > 150km) with a very rudimentary antenna. Then I decided to make an “improved” permanent one. Now I even can’t reproduce those good results anymore.

flightaware.com/adsb/stats/user/ … tats-32697

Continuity is checked from the SMA connector to the tip of the antenna, no shorts.

The only thing that I haven’t fully explored was the options for the last element. Currently I’m using half element + 1/4 wave (5,7cm + 6,9cm). Might test another options (shorted, full element + 1/4 whip)

I’m from Curitiba/Brazil

Hi aldoir,
Can you experiment by joining the antenna directly to the short link cable at the pro-stick ?
(test without the long coaxial lead)
Do you have another short lead for prostick ? If so , I would check this as it seems to have tight bend and may be shorted.
Try simple spider using this method whilst watching the map screen/message rate.
You do not have to shut down each time you swap antenna or leads, you can have map screen running while you make some tests.
Observe message rate and map for activity when making changes.

CoCo antenna is very difficult to make correctly and is frustrating to obtain perfect results.

Cantenna or spider are extremely good and its very easy to get good station results .

Did one more test now. Took my ground plane to the roof togheter with a 6 elements RG6 coco.

The ground plane found 10 aircrafts within 70nm. Using the same cable, just replaced the antena by the RG6 coco (post #4). Only 2 planes remained.

Definitely there is something pretty wrong in the way I’m making my CoCos

What are results if you join your spider 1/4 wave directly to prostick at pi enclosue ?

I just did this test… got 10 planes within 70nm range (I think I could get more if I leave it there for a while).

I think this is normal range for a 1/4 ground plane

Spider, easy to achieve 100+ miles (no amplifier).
On a roof 150+ mls, perhaps more with amplified prostick.

Cantenna is more efficient for longer range (200+mls sometimes) but not so good @ <50mls

It also depends very much on where you are. I was getting reports over 200nm on my spider indoors but there’s a lot of traffic in this area. Although at the moment (it’s 23:17 here) I only have 39 aircraft on my map. During peak times, that’s 200+.

They both are bad. Either one should be reaching out a couple hundred nautical miles.

I would suspect a connector problem

Tomorrow I will give this design a try.

f5ann.pagesperso-orange.fr/Anten … index.html

Not sure where the 0,66 factor comes from, but at least it’s using the same RG-213

Francois (F5ANN) is an expert class adsb antenna constructor.
To try to copy this design looks easy, but it is not.
I would suspect you have station problem with antenna connectors or cables not the antenna itself.
Perhaps go back to basics and use 1/4 wave spider (first checking for shorts on the antenna itself)
1 make a new or swap the short pigtail cable near the prostick
this has crimp/compression bnc connectors i think.
2 check sma pin connector male /female arrangement for faults or non-contact points, specifically with the centre pin, check this with care.The prostick has different sma connector than you may think.
You may have RP-SMA male (Reverse Pin-SMA male connector)

See this post for workaround


3 test long coaxial/swap out long coaxial

DIY CoCo’s can be a fickle beast, especially at 1090 MHz!

I also did a DIY one and the results were less than pleasing. Having worked in the RF industry (broadcast & telecoms) I know to get the phasing lengths spot on at 1090 Mhz you really need some test equipment.

I gave up in the end and bought a Jetvision A3 antenna (http://shop.jetvision.de/A3-ADS-B-Antenna) :smiley: … at the time Flightaware hadn’t released their antenna.

Very happy with the A3’s performance.



I have built many CoCos and my success rate is about 10% :open_mouth: Your build looks great.

The design I ended up with was just several lengths of coax cut to the appropriate length and pushed together like you have done. My VF is 0.66 so the lengths were 91mm. Double check the Velocity factor of the coax you have. How did you work out 113mm?

I don’t have the half wave length top end and I found that shorting the top made it worse. The bottom segment by the BNC connector is also 91mm.

I check continuity after the addition of each segment as it is too easy to get a stray piece of shield shorting somewhere. Once the antenna is complete I also give the whole antenna a bit of a shake and check continuity again just in case there is a bad connection somewhere. I also have about a 1mm gap between each segment.

Keep trying as the COCo is a great antenna :laughing:

If you do try the shorted top, make sure you put a DC blocker in place or you will burn out the stick.

Thanks for your tips

The RG6 antenna shown in post#4 was just a test. I’m working on the RG213 of the first post.

The first element is inverted, ie, the center of the cable is connected to the shield of the bnc, and the grass to center pin, is this correct or the first element shouldn’t be inverted?

About the VF, I never found a .66 cable, very strange

I’ll open the pro stick to check continuity directly in the circuit. I’ve soldered the connector myself and it’s ok

I played with CB radio in the past, have fair good knowledge about rf, SWR, but unfortunately have no access to test equipment

1 Like

Specsheet for RG-213/U (Belden 8267)
0.66 vf


The properties of the inner dielectric material of the coaxial is the important thing, see table on this page

radio-electronics.com/info/a … factor.php

You can calculate velocity factor (aka vg in %) of all sorts of cables and their losses per length @ ? freq here