FlightAware Discussions

Coaxial Collinear (CoCo) Antenna - Hyps & Facts.

abcd567 2016-01-25 13:14:33 UTC #51

@Nigel:
Have you tried a non-conventional wire collinear, the Franklin-Spider?

It is actually a 1/4 wavelength groundplane (Spider) antenna, modified by adding an additional element at top of 1/4 wavelength whip. The additional element is a 1/4 wavelength hairpin (stub), on top of which is a 1/2 wavelength vertical element. Please see this post for simulation results & drawing:

Click here: My post #2601 dated Apr 7, 2015, in Planefinder Forum (In this post, click on thumbnails of Simulation screenshots to see full size images)

It had mixed results. It gave better results than 1/4 wave Spider to caius & xforce30164, both indoor installation, but gave poorer results than 1/4 wavelength Spider to jepolch (outdoor installation) and abcd567 (me) (indoor installation).

Later, I managed to improve it by adding an impedance matching network - I cut the feed coax at (0.85 x 1/2 wavelength x VF) from feed point, inserted a 3pF capacitor in series with core wire, and joined the cut shields so that it remained continuous.

abcd

LitterBug 2016-01-27 05:27:49 UTC #52

The key factor in my COCO success is the Disco ball at the top of my antenna as seen in my avatar…

Cheers!
Litterbug

omiez 2016-02-01 18:09:15 UTC #53

Hello Friends

Could someone tell me how long must each element bee if i use RG6 to do the elements and how much elements i have to do so that the antenna works fine and if i have to use a resistor at the top or is it no necessary?

Link me a ready project if you have some build with RG6.

Thanks

Devonian 2016-02-01 19:11:15 UTC #54

If you read these and many other articles on the internet, most have failed to make an adequately performing CoCo antenna with proven advantages over a standard 1/4 wave ground plane (spider).

Build yourself a quarter wave ground plane first and record the results you get from it over at least one week.
Without a ‘reference’ antenna, you won’t know if the next antenna you build that claims high gain is really working better or not.

Nigel

elpipila 2016-02-01 21:01:52 UTC #55

This is a good place to start: balarad.net/

And this thread has excellent suggestions: ads-b-flight-tracking-f21/built-my-first-antenna-and-doubled-my-coverage-t19517.html?hilit=doubled%20my

The velocity factor varies a little between manufacturers and types . It will likely be around 0.82. Experimenting with small adjustments in segment length may improve efficiency.

The graph at this site illustrates the effect of additional elements. w8ji.com/stacking_broadside_collinear.htm

I would suggest you leave the antenna “open” at the top.

As you assemble the elements, check the continuity between elements to be sure you do not have any “short circuits” caused by stray shielding wire touching the inner conductor.

A little patience building and you will have an excellent coaxial collinear antenna.

abcd567 2016-02-02 02:32:50 UTC #56

(1) I fully agree with views of Devonian (Nigel). You should fist make a 1/4 wave groundplane antenna and establish your system by running it for at least few days (or better few weeks), before attempting to make a Coco.

Please see this thread for how to make a 1/4 wavelength groundplane antenna (Spider or Cantenna):

Three Easy DIY Antennas for Beginners

(2) One of the most important thing in making a Coco is to accurately know the Velocity Factor (VF) of the coax used for making the Coco. This can be achieved by using a Coax of known brand & model whose Specs or Data Sheet is avaiilable on internet, from which you can find exact value of VF.

Once you know VF, the length of element is simply 138 mm x VF. For example if your RG6 coax has a VF of 0.85, then its element length will be 138 x 0.85 = 117 mm.

omiez 2016-02-02 10:51:31 UTC #57

Hello and first of all thank you for your suggestions.

I am already testing an antenna, the vertical antenna that came with the FlightFeeder.

Building an antenna will be no problem and i like to try out different types, i am Ham Radio Operator thats why i like to experiment with this things.

I will order some connectors so i can try out both antennas, first i will try the collinear and then a GP.

Thanks for your help.

LitterBug 2016-02-02 14:17:27 UTC #58

I started with 116mm segments for a proof of concept following the instructions at https://www.balarad.net/ After having great results compared to a simple ground plane, I built a second antenna and feeder and started tweaking the length 1mm at a time. Having two identical feeders from the same location made it easy to tell if the results were better or worse almost instantly by comparing plane counts, message rates, and range. If the results got better, I would shorten the other antenna to match and make sure the results were comparable. Eventually the results got worse (111mm) and I knew I was in the ballpark of optimal length for my setup. Using that method got me dialed in very close to the optimal length based on the VF for my cable. (112.4mm) Since then, I have played around with amplifiers and filters and have a setup that has been repeatable to build. Currently running two feeders 20 miles apart with nearly identical configuration. Tree cover, height above ground, feed line length, and geographical differences being the largest difference in config.

Current config at both sites:
8x112mm coco --> short feed line --> Paladin 20db Satellite amp --> Feedline (20’ on one 60’ on other) --> Ground block --> 10’ feed line --> DirectTV power inserter (21V) --> 3’ feedline --> R820T2 Dongle --> RPi2

I also had to tweak the gain setting on the dongle one step at a time to get the “optimal” setting for each site. Too high of a gain setting will provide horrible results. One tick in gain level can have a drastic impact on the results.

Feeder 1 (Would like to get 20’ higher to clear tree line.)
38.6db gain, under tree cover, 15’ above ground
http://flightaware.com/adsb/stats/user/LitterBug#stats-5696

Feeder 2 (need to test with antenna mounted on opposite side of tower)
25.4db gain, on Ham tower, above trees, 50’ in air
http://flightaware.com/adsb/stats/user/LitterBug#stats-17370

Cheers!
LitterBug

The only filtering I am using is the 950mhz high pass filter built into the 20db Paladin Satellite amp

omiez 2016-02-02 15:00:06 UTC #59

Thank You LitterBug

I was doing a research on the cable I am using and it seems that the Velocity Factor of this one is 0.82.

So to have a good antenna the segments should be 1/2 wave length x 0.82 this would give the optimal element length.

In my case that would be 112.75mm.

EDIT:

Can someody tell me how i can fix the antenna to the mast ? Think about that the antenna is closed in a plastic tube, i will post some pictures soon.

backi 2016-02-22 08:52:42 UTC #60

I am just about to build my first CoCo. I am not sure which coax-cable I should buy. I have the choice of coax cable (RG6, 75Ω) with 65dB, 90dB or 120dB. Does it matter how many db the cable has? Which one should I prefer?

caius 2016-02-22 13:55:01 UTC #61

Those numbers don’t make any sense without more information. dB are a comparative measure, so unless you know what the comparison is, they don’t tell you anything - larger numbers could be better or worse.

LitterBug 2016-02-22 14:07:28 UTC #62

Do some research on the cable before buying it. You want a cable that you can find the manufacturer’s spec sheet for that indicates the “Velocity Factor”. Having that information will allow you to dial in the optimal segment length without having to do a bunch of trial and error. Personally, I have been using CCI 92041 RG6-Quad Shield with great success. This cable has a 81% velocity factor. Based on my testing, the ideal length of 112.4mm holds true.

Cheers!
LitterBug

triggers 2016-02-22 14:34:42 UTC #63

I didn’t bother with any of that 8) I had a few dozen metres of satellite cable in the garage just labelled SAT100 so probably something generic.

I started with a 6 segment one with each segment about 112mm. Open top, no baluns and put it on test in the loft. I got some idea of the range I was getting, took it down, chopped a couple of mms off and started again. Keep going until you find the “best” for your location. My element lenght was 91mm so it took a while. It looks like my cable had a Velocity Factor of 0.6. Most coax has Velocity factor of around 0.82 unless you choose to buy something exotic.

Blue ring is maximum distance calculated by Heywhatsthat.com at 40,000 feet, the red ring is 30,000 feet and the green ring is about 1 weeks worth of gathering data. I don’t think there is much more I can do – or is there

I’ll post some pics of my attempt at making a coco with copper tube & putting the inner + insulation inside it. I’m hoping it will be a bit more robust than just pushing the core between the outer and braid.

I lost my whole antenna a couple of weeks ago when it snapped at the bracket during some big winds. The antenna does tend to swing around a lot in a gentle breeze. The above is a 12/14/16 element one as I’m not sure how many of the segments came apart when it fell. I didn’t bother to check, just put it back up again.

backi 2016-02-22 16:37:26 UTC #64

So the screening attenuation does not have any influence at all? Its only all about the velocity in terms of the length of each element?

abcd567 2016-02-22 16:43:48 UTC #65

I have never seen a coax with such a high dB figures like 65dB, 90dB or 120dB. Such high dB figures are specified with Amplifiers.

The only dB figure specified with coax is Attenuation in dB/100 meter or dB/100 feet.

For most RG6 75Ω cables, the attenuation at 1000 MHz is typically around 6.5 dB/100 feet or 21 dB/100 meters.
The Velocity Factor is typically around 83% for FPE (Foamed Polyethylene) insulation.

For the purpose of making a Coco, attenuation is unimportant as Coco’s length is less than 2 metes (6 feet), and hence attenuation is less than 0.42 dB (2 meter x 21 dB/100 meter), which is negligible.

backi 2016-02-22 16:59:13 UTC #66

right, Velocity Factor is almost the same, but attenuation varies more: (from 20dB/100m up to 40dB/100m), I assume the lower the better. Right?

produktinfo.conrad.com/daten … _de_en.pdf
produktinfo.conrad.com/daten … U_TERR.pdf
produktinfo.conrad.com/daten … 75_OHM.pdf

What about shielding? single, double, triple? Doesnt higher shielding make it harder for the ADS-B signal to get to the receiver?

obj 2016-02-22 17:12:01 UTC #67

Attenuation is not so important for the co-co itself, because attenuation is proportional to distance and the co-co isn’t very long. Lower attenuation is better, obviously, but probably your feedline is much longer than the antenna itself so worry about that more!

In a co-co you are using the coaxial sheath as one of the receiving elements anyway, it’s not shielding.

abcd567 2016-02-22 18:28:08 UTC #68

I fully agree.

In a co-co you are using the coaxial sheath as one of the receiving elements anyway, it’s not shielding.

This is partly correct.

The shield performs both the receiving as well as shielding functions.
The OUTER SURFACE of shield receives the RF signal when Radio wave from aeroplanes strikes it, but remaining thickness of the shield prevents the Radio waves to penetrate it and reach the core wire.
The core wire acts as phasing element, and to perform optimally, it should not intercept any Radio wave arriving from the aeroplanes.

Most dual shielded (1 foil+ 1 braid) coax have about 95% shielding and allow less than 5% RF signal to reach the core, hence perform well.
This is the most commonly used coax for Satellite/DBS and also used for coco making, and generally gives good performance both as coco & medium length TV/Dish cable.

The quad-shielded (2 foils + 2 braids) coax has shielding higher than 95%, and is used for transmission of RF signal where RF noise is high.
This is costlier compared to dual shield. If used in coco, it will somewhat improve the performance of coco, but not much.

https://farm4.staticflickr.com/3953/15519454692_46c08c7ab8_o.png

EVOLUTION OF A COCO / HOW A COCO WORKS
WIDE IMAGE - Scroll Right to see in full
https://farm1.staticflickr.com/400/18889859639_701b09b117_b.jpg

elpipila 2016-03-20 01:24:03 UTC #69

Some details of the direction we are taking for our new coaxial collinear installs and retrofits when needed.

We like coaxial collinears because they are cheap, easy to build, can have lots of gain, and can be protected in a sturdy radome.

Our goal, whenever possible, is to avoid the need to insert extra amplification and power inserters in the system so we work to maximize gain from the antenna.

Current designs have 20+ elements enclosed in a 2.8 meter pvc radome constructed in 3 approximately equal segments lengths of 1 inch, 3/4 inch, and 1/2 inch pvc pipe. The 1 inch segment is reinforced with a 3/4 inch liner, the 3/4 with a 1/2 inch liner and the 1/2 inch with a 3/8 cpvc liner. The 1 inch base includes an extra 30 cm reinforced segment for mounting.

The cable used in this data set is Perfect Vision RG6 solid copper core.
.
Tests were run to compare various combinations of antenna, cable, and 2 different amplifiers

They were; 1: At the antenna (including complete data set), 2: Antenna + 50 ft RG6 cable, 3: Antenna + Paladin amplifier + 50 ft cable, 4: Antenna + PCT amplifier + 50 ft cable.

http://i.imgur.com/uAeEpAk.jpg
http://i.imgur.com/hhtSYTv.jpg
http://i.imgur.com/XGSrGu0.jpg
http://i.imgur.com/CfjDv1R.jpg
http://i.imgur.com/nzj037D.jpg
http://i.imgur.com/a2R6A65.jpg
http://i.imgur.com/hU0DqY2.jpg
http://i.imgur.com/ay6AtdM.jpg

abcd567 2016-03-20 19:28:24 UTC #70

@elpipila:
This is great!
Your tests have proved it to be good.
Next step is trial run.
Waiting for trial run results.