FlightAware Discussions

Cantenna versus Coaxial Collinear Antenna


I am new to all this. I was testing few different antenna. From what i managed to gather, most of the people who tried to do collinear antenna ends up with not so satisfactory result. Well, that is not the case for me, I was surprise with the result if compared to cantenna

Antenna > DC blocker > generic LNB In-Line Amplifier -> 10 meter RG6 -> Bias-T -> TA1090EC -> Generic dongle

The generic in line amplifier are very noisy, I am currently working on a BFP420 based amplifier, just waiting for PCB from JLPCB. I am using toroidal transformer with LM317 to avoid unnecessary noise from SMPS. If i use SMPS, I am not able to get any plane/message. I am still working on a design that uses TPS7A4700


You are using no filtering and an amplifier.

Those two facts favors colinear antennas as they act as a filter but often don’t always have the best gain (unless you are quite precise and hit exactly the correct segment length)


I did use filter before the dongle, TA1090EC. I was quite precise when measuring the segment length, spend around 3 hours to make sure all segment is at the same length.

With cantenna i was only able to receive a maximum of 100msg/s but with collinear i can easily hit 300msg/s


Congrats. Yes, the key to co-linear antenna is to measure twice and cut once. Probably most people don’t have that patience or eye-hand coordination.

I am using a preamp with two SAW filters at 5 meters from antenna, because I have a very long cable following that. In your case, with short cable, you might want to adjust the dongle gain.


That’s some RF porn right there.



I see your amp/filter pcb you got it from aliexpress?
Yes, i did adjust the gain after changing the antenna


I did get it from eBay, but they have it on Aliexpress too.
I had installed the 7805 source and the blocking coil because I had to feed it via coax (150ft = 46 meters). I am pushing 9V on my end, at stabilizer arrives some 7.5V, enough to make the 5V.


You are comparing two antennas. The essential conditions of comparision of two antennas are:
(1) Both antennas are at same location and height
(2) Both have same type and length of coax between antenna & receiver
(3) Both have identical hardware i.e. both should have identical LNA, Filter, and Receiver (DVB-T or Pro Stick etc).

Was your stats comparision chart done with above conditions met?


I see… do you by any chance know what transistor were on the board? 20USD for a board is quite a steep price for me, i would rather create my own pcb and with parts should cost less than 20USD for 10 complete board.


@abcd567 yes, all the said condition was met… :slight_smile:


Great… cleared the doubt.


Unfortunately the seller erased the markings on them, SOT89-3 packages. I suspect, for the first one, this:

Note that that amp has a compression point, to deal with high level signals:

Noise is decent at 1GHz:


Which brand/type of coax did you use for the antenna and which length did you use for the pieces?
How did you terminate it if at all?

Also cantennas are not quite as good as spiders in my personal experience.
Maybe you have the time some day to compare it to a spider :slight_smile:

Are you gonna use SAW filters with your amplifier?
Good luck making it, RF design is not easy (and i don’t have enough knowledge so i trust the rtl-sdr-blog LNA version)

I have a colinear with 8 elements in my attic which i just hacked together.
It’s not quite as good as the spider i have in regards to gain. But i’m pretty sure the noise is less.
Probably next spring i’m gonna put a 12 element on the roof and check how it compares.
(Using CFD200 coax)


Simulation of 8-element Coco

Coax V.F = 0.86
Element length = 118 mm

Screenshot 1 of 4 - Radiation Pattern, Gain


Screenshot 2 of 4 - SWR, Impedance


Screenshot 3 of 4 - SWR, Frequency Sweep 500Mhz ~ 2000Mhz


Screenshot 4 of 4 - GAIN, Frequency Sweep 500Mhz ~ 2000Mhz



(4) Run the two systems in parallel for a test to confirm similar performance.

This will give you confidence that differences in performance are solely due to changes made.