ADS-B DIY Antenna

[ab cd]

Good evening!

No thanks needed - I'm new to EZNEC and this was a great opportunity to learn some of its features. Up until buying EZNEC, I had been evaluating my ADS-B antennas in the same fashion as you have been doing - maximum range evaluation. Here at my house, plane traffic varies greatly from day to day, so to get a good comparison between antennas, I've got to set up an antenna and let it run for a week or so before being confident I can compare the results. It appears that your location has a *lot* more airplane traffic, so you may not have to wait as long to have repeatable, comparable results. With EZNEC, I plan to model the designs, tweak the model, then do the maximum range comparison as the final test.

I'll be happy to share my tweaks, but I want to do two things before I do that. First, I need to double-check my model, as I think I may have an error in how I created the model of the stub itself. In my model, instead of keeping the stub length constant and moving the connection of the coax on the stub, I connected the coax at the end of the stub and varied the length of the stub. While the end result is still good, it may or may not represent the design you have presented. Second, I want to create a drawing of the antenna with the tweaks. While I think I can describe the tweaks in words without an issue, a picture will really make things clear. I would also like to show the antenna patterns predicted by EZNEC, as this antenna has a very directional response in elevation that I believe is very favorable for ADS-B reception. I may not be able to do this until this weekend, unless things are work calm down a little bit.

I'll be happy to model the shorted stub Franklin with 10mm spacing. If I'm looking at the diagrams correctly, the dimensions between the 10mm open stub version and the 10mm closed stub version are the same, with the only difference being the closed end to the stub. I will double-check this before I finish the models.

I am indeed an amateur radio operator, callsign KD5BYB.

thanks much and 73,
ben, KD5BYB

Thanks Ben. Waiting for completion of your work and posting of results. 73, abcd.

 

[kd5byb]

Good evening all,

So I got a little bit of time this evening and figured I'd share some of my findings modeling the Franklin Collinear Dipole antennas presented earlier in this thread.

The first one will be the 10mm, open-stub design as shown in this drawing:

This design is a very good design as-is. At 1090 MHz, it has a 1.5 SWR referenced to 75 ohms. The USB dongles that are generally recommended for ADS-B reception use the R820T tuner, which is a 75 ohm input device. A 1.5 SWR is very respectable. The antenna has a pattern that I consider very favorable for ADS-B reception - much of the gain is at a very low elevation. Planes at maximum distance will be present at a very low elevation.

Before I present plots of the antenna, let me make a few notes.

Notes:
1) Modelling software, when used properly, can give results that closely match the real world. While EZNEC itself is well known and respected, it is only as good as the operator. I am fairly new to EZNEC, so it is quite likely that I may have made some errors in the model or the application of the program.
2) It is a best practice to compare modeled results to real-world results. If this was antenna for HF, I have the tools to do that. However, at 1090 MHz, I do not have the tools to compare the model to a real-world specimen.
3) My models place antenna 2 meters above a ground. The ground model used is "real/high accuracy" and simulates a heavy clay soil which is considered an average quality ground. Antenna patterns are influenced by height above ground and the quality of the ground. My soil here is heavy clay, so that is what I used. The 2 meter elevation was nothing more than a nice, round number that I felt represented a typical ADS-B antenna mounting height.

Plots.

* First is the model of the antenna and the coordinates of the antenna wires.

Now for the SWR plot:

Discussion: as I noted above, this is referenced to a 75-ohm receiver such as the R820T used in common USB RTL-SDR dongles. This is a very nice SWR. While not a perfect 1.0, it is extremely close. A 1.5:1 SWR would induce a loss of about 0.18 dB. This is extremely low.

Pattern:

Discussion: Note that the max gain of 9.14 dBi occurs at an elevation angle of 1.9 degrees. For ADS-B, this is very favorable as you need your maximum gain at low angle to get that plane that is very far away and low on the horizon.

I did make some simulations where I moved the feed point from the 18mm position with very little benefit in SWR.

Bottom line: as drawn, this is a very nice antenna.

 

[kd5byb]

I was able to make some minor improvements to the above antenna. Specifically, the model showed favorable results if the little "stubs" at the end of the feedpoint were removed. I can show this better with pictures from the model.

Here is the original antenna feedpoint:

Here is the modified feedpoint:

The little "stubs" represented by wires 13 and 15 have been removed.

This results in improved SWR as shown below and slightly more gain:

Discussion:

* The difference between an SWR of 1.5 and 1.3 is going to be very negligible. Earlier I noted that the 1.5 SWR resulted in a 0.18 dB signal loss due to mismatch. A 1.3 SWR results in a 0.08 dB signal loss.
* However, the reduced loss *plus* the 0.7 dB gain increase may result in a range increase that while small, is noticeable.

Tomorrow I'll put together a model of the same antenna but with the end of the stub shorted.

thanks much and 73,
ben, kd5byb

 

[ab cd]

I was able to make some minor improvements to the above antenna. Specifically, the model showed favorable results if the little "stubs" at the end of the feedpoint were removed. I can show this better with pictures from the model.

Here is the original antenna feedpoint:

Here is the modified feedpoint:

The little "stubs" represented by wires 13 and 15 have been removed.

This results in improved SWR as shown below and slightly more gain:

Discussion:

* The difference between an SWR of 1.5 and 1.3 is going to be very negligible. Earlier I noted that the 1.5 SWR resulted in a 0.18 dB signal loss due to mismatch. A 1.3 SWR results in a 0.08 dB signal loss.
* However, the reduced loss *plus* the 0.7 dB gain increase may result in a range increase that while small, is noticeable.

Tomorrow I'll put together a model of the same antenna but with the end of the stub shorted.

thanks much and 73,
ben, kd5byb

Good evening all,

So I got a little bit of time this evening and figured I'd share some of my findings modeling the Franklin Collinear Dipole antennas presented earlier in this thread.

The first one will be the 10mm, open-stub design as shown in this drawing:

This design is a very good design as-is. At 1090 MHz, it has a 1.5 SWR referenced to 75 ohms. The USB dongles that are generally recommended for ADS-B reception use the R820T tuner, which is a 75 ohm input device. A 1.5 SWR is very respectable. The antenna has a pattern that I consider very favorable for ADS-B reception - much of the gain is at a very low elevation. Planes at maximum distance will be present at a very low elevation.

Before I present plots of the antenna, let me make a few notes.

Notes:
1) Modelling software, when used properly, can give results that closely match the real world. While EZNEC itself is well known and respected, it is only as good as the operator. I am fairly new to EZNEC, so it is quite likely that I may have made some errors in the model or the application of the program.
2) It is a best practice to compare modeled results to real-world results. If this was antenna for HF, I have the tools to do that. However, at 1090 MHz, I do not have the tools to compare the model to a real-world specimen.
3) My models place antenna 2 meters above a ground. The ground model used is "real/high accuracy" and simulates a heavy clay soil which is considered an average quality ground. Antenna patterns are influenced by height above ground and the quality of the ground. My soil here is heavy clay, so that is what I used. The 2 meter elevation was nothing more than a nice, round number that I felt represented a typical ADS-B antenna mounting height.

Plots.

* First is the model of the antenna and the coordinates of the antenna wires.

Now for the SWR plot:

Discussion: as I noted above, this is referenced to a 75-ohm receiver such as the R820T used in common USB RTL-SDR dongles. This is a very nice SWR. While not a perfect 1.0, it is extremely close. A 1.5:1 SWR would induce a loss of about 0.18 dB. This is extremely low.

Pattern:

Discussion: Note that the max gain of 9.14 dBi occurs at an elevation angle of 1.9 degrees. For ADS-B, this is very favorable as you need your maximum gain at low angle to get that plane that is very far away and low on the horizon.

I did make some simulations where I moved the feed point from the 18mm position with very little benefit in SWR.

Bottom line: as drawn, this is a very nice antenna.

Thank you very much Ben. Excellent work. I am really impressed. Thanks for sharing your efforts and findings. Waiting for further progress in your work and more results. 73, abcd.

Later addition: Hi Ben, just noted that you have used #12 AWG wire for your Antenna Simulation. The antenna as well as the matching stub I built, have used the central wire of RG6U Coaxial Cable, which is #18 AWG (1.024 mm dia), while you have used size #12 AWG (2.053 mm dia). This is my fault as I have not mentioned on my antenna sketch what size of wire I have used to built the antenna & stubs. Since the calculations depend on dia of wire, using the wire size I have used ( #18 AWG) in your calculation by EZNEC will give somewhat different results. Best Regards, abcd

 

[kd5byb]

Good catch on the wire diameter. Yes, the wire diameter makes a difference. All things equal, a fatter wire will have a wider bandwidth. Something else I noticed - my models are for bare copper wire, while in the photos your antenna appears to be covered in insulation. I assume this was foam-dielectric RG6 cable? Is there non-foam dielectric RG6? I'll admit, while I'm familiar with 50 ohm coax, I knew very little about RG6 75 ohm coax. Most ham radio gear is 50 ohms, and up until now I've not had much need for a 75 ohm coaxial cable.

The foam insulation will also make a difference. If I remember correctly, the insulation slows down the velocity of propagation, making the wire appear to be longer than it is. I will rework the models with both the smaller wire and the insulation added either tomorrow evening or over the weekend.

Unfortunately, I created the model of the 10mm closed stub version before I saw this post. However, for comparison purposes to my past models, I've decided to present the results anyways even though the wire diameter is wrong and the insulation is missing from the model.

I think the results are very interesting. And they agree with what abcd has noted during maximum range testing.

So here is the the closed stub diagram that I modeled:

Here is the model:

And the wire table:

RESULTS:

The SWR is not quite as good as the previous version. I did move the feedpoint from the 18mm shown above to 20mm, as that resulted in the minimum SWR. It was slightly higher at the 18mm point.

The pattern is a slightly different, with the same gain as my modified design presented earlier with the stubs removed, but slightly more gain than the original open-stub version:

DISCUSSION:

An SWR of 2.05 is still very respectable, with a loss of 0.55 dB. While the SWR loss is higher than abcd's original open-stub version, this antenna has more gain which overwhelms the additional loss. (9.82 dBi gain here versus 9.14, a increase of 0.68) The model confirms that this closed-stub antenna will perform better than the original open-stub version. This agrees with the maximum range test results presented earlier by abcd.

As I've noted earlier, this antenna should work very well for ADS-B reception.

thanks much and 73,
ben, kd5byb

 

[ab cd]

Good catch on the wire diameter. Yes, the wire diameter makes a difference. All things equal, a fatter wire will have a wider bandwidth. Something else I noticed - my models are for bare copper wire, while in the photos your antenna appears to be covered in insulation. I assume this was foam-dielectric RG6 cable? Is there non-foam dielectric RG6? I'll admit, while I'm familiar with 50 ohm coax, I knew very little about RG6 75 ohm coax. Most ham radio gear is 50 ohms, and up until now I've not had much need for a 75 ohm coaxial cable.

The foam insulation will also make a difference. If I remember correctly, the insulation slows down the velocity of propagation, making the wire appear to be longer than it is. I will rework the models with both the smaller wire and the insulation added either tomorrow evening or over the weekend.

Unfortunately, I created the model of the 10mm closed stub version before I saw this post. However, for comparison purposes to my past models, I've decided to present the results anyways even though the wire diameter is wrong and the insulation is missing from the model.

I think the results are very interesting. And they agree with what abcd has noted during maximum range testing.

So here is the the closed stub diagram that I modeled:

Here is the model:

And the wire table:

RESULTS:

The SWR is not quite as good as the previous version. I did move the feedpoint from the 18mm shown above to 20mm, as that resulted in the minimum SWR. It was slightly higher at the 18mm point.

The pattern is a slightly different, with the same gain as my modified design presented earlier with the stubs removed, but slightly more gain than the original open-stub version:

DISCUSSION:

An SWR of 2.05 is still very respectable, with a loss of 0.55 dB. While the SWR loss is higher than abcd's original open-stub version, this antenna has more gain which overwhelms the additional loss. (9.82 dBi gain here versus 9.14, a increase of 0.68) The model confirms that this closed-stub antenna will perform better than the original open-stub version. This agrees with the maximum range test results presented earlier by abcd.

As I've noted earlier, this antenna should work very well for ADS-B reception.

thanks much and 73,
ben, kd5byb

Good catch on the wire diameter. Yes, the wire diameter makes a difference. All things equal, a fatter wire will have a wider bandwidth. Something else I noticed - my models are for bare copper wire, while in the photos your antenna appears to be covered in insulation. I assume this was foam-dielectric RG6 cable? Is there non-foam dielectric RG6? I'll admit, while I'm familiar with 50 ohm coax, I knew very little about RG6 75 ohm coax. Most ham radio gear is 50 ohms, and up until now I've not had much need for a 75 ohm coaxial cable.

The foam insulation will also make a difference. If I remember correctly, the insulation slows down the velocity of propagation, making the wire appear to be longer than it is. I will rework the models with both the smaller wire and the insulation added either tomorrow evening or over the weekend.

Unfortunately, I created the model of the 10mm closed stub version before I saw this post. However, for comparison purposes to my past models, I've decided to present the results anyways even though the wire diameter is wrong and the insulation is missing from the model.

I think the results are very interesting. And they agree with what abcd has noted during maximum range testing.

So here is the the closed stub diagram that I modeled:

Here is the model:

And the wire table:

RESULTS:

The SWR is not quite as good as the previous version. I did move the feedpoint from the 18mm shown above to 20mm, as that resulted in the minimum SWR. It was slightly higher at the 18mm point.

The pattern is a slightly different, with the same gain as my modified design presented earlier with the stubs removed, but slightly more gain than the original open-stub version:

DISCUSSION:

An SWR of 2.05 is still very respectable, with a loss of 0.55 dB. While the SWR loss is higher than abcd's original open-stub version, this antenna has more gain which overwhelms the additional loss. (9.82 dBi gain here versus 9.14, a increase of 0.68) The model confirms that this closed-stub antenna will perform better than the original open-stub version. This agrees with the maximum range test results presented earlier by abcd.

As I've noted earlier, this antenna should work very well for ADS-B reception.



thanks much and 73,
ben, kd5byb

Wire dia plays significant role in Stub impedance value and hence impedance matching/SWR.
Quarter wave Stub Impedance = 276 log (2S/d)
where d is stub wire dia and S is stub wire spacing.

For 10 mm wire spacing, a 2 mm dia wire will give an stub impedance of 276 log(2x10mm/2mm) = 276 x log 10 = 276 x 1.0 = 276 ohms.

For 10 mm wire spacing, a 1 mm dia wire will give 276 x log (2x10mm/1mm) = 276 x log 20 = 276 x 1.3 = 358 ohms.

For matching a 1600 ohm antenna to 75 ohm cable, required stub Impedance
Zstb = √(Zant x Zfdr) = √(1600 x 75) = 346 ohm

which is very closely provided by stub having 1 mm dia wire, 10 mm wire spacing (358 ohms).

 

[Nenad]

@ab cd
I made COAXIAL COLLINEAR: 3-AND-1/2 ELEMENTS (EACH ELEMENT 1/2 WAVELENGTH) WITH SHORTED TOP antenna from this thread. for now it's without amplifier. I also made jpole antenna for ATC listening.
Tell me is it possible and how to merge both so i have only one line on my rtl dongle.
Don't think its possible but I had to ask

 

[kd5byb]

@ab cd
Tell me is it possible and how to merge both so i have only one line on my rtl dongle.

Actually, I do believe it is possible with a diplexer. Some time ago when I was looking on eBay for satellite amplifiers like abcd has shown in his photo, I found some splitters / combiners that may work for this purpose. One can be seen here:

http://www.ebay.com/itm/380148691933

I bought a pair, but have not tested them yet. They are certainly inexpensive that if they don't work, you haven't spent a lot of money.

Looking at the ad copy, the UHF /VHF port will operate from 5 to 806 MHz, with the satellite port operating from 950-2400MHz. Hook the ADS-B antenna to the satellite port, the VHF ATC antenna to the VHF/UHF port, and the dongle to the IN/OUT port and i think you'll have what you want.

thanks much and 73,
ben, kd5byb

 

[Nenad]

It's even less expensive for me, i have couple of those in my toolbox..... will try it.

tnx ben...73

 

[kd5byb]

Good afternoon all,

I have edited the model on the 10mm closed-stub antenna. The results are not very good. As I suspected, the PE insulation over the conductor does cause the wires to look longer than they are, resulting in a high SWR at 1090 MHz with a lower SWR at a lower frequency. The far-field pattern plot is quite different.

In the plots below, I've left the feed-point at the 18mm distance (-51 mm in coordinates). I did vary the feedpoint from 2 mm to almost all the way in and while SWR did improve, it was not possible to get the minimum SWR "depression" to move up to 1090 MHz.

SWR:

Pattern:

Wires:

Discussion:

While the above seems pretty gloomy...let's put the results into perspective. The loss from the 10.9 SWR is 5.1 dB, and the lobes of the response plot come at angles I would think would not be optimal for ADS-B reception. I'm not saying that this antenna doesn't work, as clearly per abcd's testing, it works rather well.

I am going to try and tweak this design, staying with the 1.024mm conductor covered with foam PE insulation. It would be very interesting that once I've done this, to have abcd duplicate the design and try it out to see how well it works. This would give us a very interesting data point.

thanks much and 73,
ben, kd5byb

 

[kd5byb]

It's even less expensive for me, i have couple of those in my toolbox..... will try it.

Excellent! Please let us know how they work out.

thanks much and 73,
ben, kd5byb

 

[ab cd]

@ab cd
I made COAXIAL COLLINEAR: 3-AND-1/2 ELEMENTS (EACH ELEMENT 1/2 WAVELENGTH) WITH SHORTED TOP antenna from this thread. for now it's without amplifier. I also made jpole antenna for ATC listening.
Tell me is it possible and how to merge both so i have only one line on my rtl dongle.
Don't think its possible but I had to ask

Yes, it is possible with a diplexer, as KD5BYB (Ben) has mentioned.
I have done this with a home-brew diplexer (also known as demuxer). It is a combination DC-injector+Diplexer (Demuxer).
See schematic diagram & camera photo below:
NOTE: This is for a Dipole. For Shorted-top CoCo, you need to insert a capacitor between antenna & amplifier to prevent frying your DC power supply.
Schematic Diagram:

Camera Photo:

 

[Nenad]

still didn't install amplifire and dc power.
i'm going to wait on that for a while. ordered some capacitors and inductors.
so....
without power and amplifire i just need to connect as described?

 

[ab cd]

Good afternoon all,

I have edited the model on the 10mm closed-stub antenna. The results are not very good. As I suspected, the PE insulation over the conductor does cause the wires to look longer than they are, resulting in a high SWR at 1090 MHz with a lower SWR at a lower frequency. The far-field pattern plot is quite different.

In the plots below, I've left the feed-point at the 18mm distance (-51 mm in coordinates). I did vary the feedpoint from 2 mm to almost all the way in and while SWR did improve, it was not possible to get the minimum SWR "depression" to move up to 1090 MHz.

SWR:

Pattern:

Wires:

Discussion:

While the above seems pretty gloomy...let's put the results into perspective. The loss from the 10.9 SWR is 5.1 dB, and the lobes of the response plot come at angles I would think would not be optimal for ADS-B reception. I'm not saying that this antenna doesn't work, as clearly per abcd's testing, it works rather well.

I am going to try and tweak this design, staying with the 1.024mm conductor covered with foam PE insulation. It would be very interesting that once I've done this, to have abcd duplicate the design and try it out to see how well it works. This would give us a very interesting data point.

thanks much and 73,
ben, kd5byb

Thanks a lot for your diligent work & sharing the info.
If you simulate with reduced lengths of all elements by 85% (velocity factor) i.e. 69 mm element becomes 59 mm, and 138mm element becomes 117mm, we may get a respectable SWR. Don't reduce length of Matching stub, as it is bare conductor so as to make adjustable connection to coaxial cable. Thanks. 73, abcd

 

[ab cd]

still didn't install amplifire and dc power.
i'm going to wait on that for a while. ordered some capacitors and inductors.
so....
without power and amplifire i just need to connect as described?

You can use a simpler setup if you are not using Amplifier & power supply. What you need is a piece of wire to make a coil, and a housing of an ordinary TV cable splitter (cost $ 1.50). See Revised Schematic diagram below, and the Camera Photo of home-brew Demuxer/DiplexerI have posted earlier.

You can also buy it on eBay:
http://www.ebay.com/itm/Holland-Diplexer-Splitter-DPD2-Satellite-FTA-Dish-TV-/160922557948?pt=US_Splitters_Combiners&hash=item2577bb5dfc

http://www.ebay.com/itm/Holland-DPD2-2-Amp-Satellite-Diplexer-/130961366280?pt=US_Splitters_Combiners&hash=item1e7de7dd08

 

[Nenad]

did you test version with amplifier but without dc power?
i think there is no use of it without power but.....

 

[ab cd]

did you test version with amplifier but without dc power?
i think there is no use of it without power but.....

With DC power, the Amplifier Amplifies (increases signal strength).
Without DC power, the Amplifier Attenuates (reduces signal strength).
If you don't have DC power, remove the amplifier.

 

[Nenad]

I didn't install it yet.

 

[ab cd]

I have downloaded an antenna modelling and optimization software "4NEC2" from http://www.qsl.net/4nec2/ .
It is not as sophisticated as EZNEC, but it is very good considering that it is free of cost.
I then simulated my Franklin, and following 4 screenshots show the results (shows SWR = 2.49 for connection to 75 ohms cable/receiver, and maximum gain =6.95 dBi - see screenshot 3):

1- INPUT - WIRE GEOMETRY DATA

2 - INPUT - SOURCE/LOAD DATA

3 - OUTPUT - ELECTRICAL DATA, 3D GEOMETRY, RADIATION PATTERN

4 - OUTPUT - SWR vs Freq. Sweep

.

 

[Nenad]

what is good directional antenna to be made easy and quickly....
I have 2 receivers....one is for feeding "other site" with own antenna, and one is RTL dongle with DIY antenna.
now....i want rtl dongle to cover only one direction....