ADS-B DIY Antenna

Status
Not open for further replies.
airbus340 said:
Thank you for the help!

Here is a picture of my piece of art... :D
Not looking very good compared to other antennas in this thread, but it perform great.
Click to expand...
Thanks a lot. This picture is a great help to all forum "experimenters".

In your photo, the connection of two coax cables (interconnector & feeder) are hidden behind the wooden stick, need a closeup photo of connection arrangement.

What is the length of interconnecting piece of coax? It forms a phasing harness and it's length is critical.

Thanks again for posting the photo. It will trigger another series of experiments.
 
Ok, here are some more pictures.

All four dipoles are 5/8 wavelength = 86 mm.
The dipoles are made of 1 mm copper wire.
Distance between center of top and bottom dipoles are 1 wavelength = 272 mm.
The cable start at the top dipoles and then connected to the bottom dipoles + the cable to the dongle.

Feeder cable is 50 cm and connected to a TV-amplifier and the dongle is hooked direct to the amplifier.

Ab cd, I didn´t know the cable length was important so we just cut it long enough to reach between the connectors.
 

Attachments

  • both.jpg
    both.jpg
    271.3 KB · Views: 125
  • bottom.jpg
    bottom.jpg
    253.8 KB · Views: 123
  • top.jpg
    top.jpg
    234 KB · Views: 116
giacomo1989 said:
@ab cd No infos about the bakooka'ed dipole? :p
Click to expand...

RPi's crashes last week destroyed all the bazooka experiments. After RPi crash problem was over, I ran it for couple of days, then on Friday trimmed it's limbs by 4mm, making these 65mm instead of 69 mm. The performance dropped, as is clear from the daily collection data below, which dropped from 48,000 to 41,000 after trimming on Friday.
bazooka 69mm - 65 mm.PNG



I will now make another 69 mm dipole, run few days with bazooka sleeve, then remove bazooka sleeve, and run again for few days.
 
Last edited:
airbus340 said:
Ok, here are some more pictures.

All four dipoles are 5/8 wavelength = 86 mm.
The dipoles are made of 1 mm copper wire.
Distance between center of top and bottom dipoles are 1 wavelength = 272 mm.
The cable start at the top dipoles and then connected to the bottom dipoles + the cable to the dongle.

Feeder cable is 50 cm and connected to a TV-amplifier and the dongle is hooked direct to the amplifier.

Ab cd, I didn´t know the cable length was important so we just cut it long enough to reach between the connectors.
Click to expand...
Thanks for posting additional pictures & additional information. I think with this much details, experimenters can start.
 
Hi Airbus/Ab cd,
Looks like the phasing harness is just over 1 wavelength.
As I remember too ab cd , the velocity factor of the phasing harness is also very important here.
I take it that you have "played around" with the dipole spacing and this is the best for you Airbus ?

Just a thought but I have had great results on HF using twin speaker lead or rig power lead as a balanced feeder , it has an typical impedance of around 72 ohms. I run my balanced feeder to a balanced antenna (like yours airbus..only mine is a lot lot bigger..for HF !) then I have a 1:1 balun near the tranciever/receiver with a short stub (around 1ft) of good coax to join to the rx/tx.
This way the only unbalanced section is the short coax stub before the antenna feeder.
This antenna feed method has superb performance and little in the way of interference/rfi/emc problems as all currents are equal throughout a large part of the antenna system.
I digress...Sorry Chaps.o_O

Anyway, I wondered if performance on the stacked dipole design could be enhanced by substituting a link of 1 wavelength twin speaker wire as a phasing harness.

Cant get twisted lighting chord cheap anymore...costs a fortune, but makes great feeder..
There are 75 ohm balanced to unbalanced devices (baluns) used in cctv ..neat little things they are for making receive antennas.

Thanks airbus for posting your design.

I envisage some members already have " mental blueprints" and the materials to hand for some home construction & they will like me conduct "sea trials" of a stacked dipole array clone..the progress of which will no doubt be listed here.

now im off to botch that early 4 element "experimental" franklin into a stacked array..:)

Like the cantenna,spider,plantenna etc these ideas and variations are great for experimentation, they dont cost much and are fun to construct. Best of all we learn lots by practical experimentation.
Thanks airbus again for your post.
 
Last edited:
ab cd said:
Congratulations for finally receiving the Amplifier!
Did the amplifier arrive by a sail boat or inside a bottle with a note? :D
Click to expand...

It occurs to me that my last reply may have sounded curt. If so, I apologize.

For background, I ordered this low noise amplifier (LNA) a couple of months ago:

http://lna4all.blogspot.com/

I was attracted to it because of high gain (18 db at 1090Mhz), small size, and 5v power. The plan was to put the LNA and dongle up near the antenna and run a USB cable from the Pi to it. I got a USB splitter, hacked one side to get at the power wires and soldered them to the LNA after adding the jumper discussed on the page.

ab cd's comment was based on a previous exchange where I complained about the LNA taking so long to cross the Atlantic. I was out for a walk when our postal person just came up and handed it to me.

I connected the LNA to the spider and the rest of the rig and dump1090-mutability immediately started crashing. Some Internet research suggested it was caused by power problems. So I eliminated a 15' USB extension cable and stuck the Pi up on the antenna mast with the dongle and LNA. Much better. Dump1090-mutability stabilized.

I started playing with the gain on the dongle. Based on earlier discussions here, I had turned the gain down to 30db to try to solve the crashing. With the crashing solved another way, I started there. Oh, Wow. I immediately saw more aircraft, at greater range, with more messages than I ever had. After an hour, I raised the dongle gain to 40db. Better yet. After another hour, I set it to "agc". I'm not so sure this one is better. The CPU load is nearly 50% and not so many messages, but that might be normal daily variation.

It is too early for an quantitative results, but qualitatively, I'm impressed. Before, when an aircraft moved out of range, its Virtual Radar signal strength would fade down to 4 or 5, then the plane would disappear. Now, the plane keeps a signal strength in the teens and disappears much further out. This is telling me that the plane is going behind one of the many obstructions around here instead of getting lost because of signal strength.

I'm going to run this set up for few days and maybe tweak the dongle gain. I'll keep y'all posted.

-- Doug
 
Last edited:
DougJohnson said:
It occurs to me that my last reply may have sounded curt. If so, I apologize.

For background, I ordered this low noise amplifier (LNA) a couple of months ago:

http://lna4all.blogspot.com/

I was attracted to it because of high gain (18 db at 1090Mhz), small size, and 5v power. The plan was to put the LNA and dongle up near the antenna and run a USB cable from the Pi to it. I got a USB splitter, hacked one side to get at the power wires and soldered them to the LNA after adding the jumper discussed on the page.

ab cd's comment was based on a previous exchange where I complained about the LNA taking so long to cross the Atlantic. I was out for a walk when our postal person just came up and handed it to me.
Click to expand...
No need to apologize, it did not sound curt, but it looked strange, as if you forgot the previous conversation on this issue.

It is too early for an quantitative results, but qualitatively, I'm impressed. Before, when an aircraft moved out of range, its Virtual Radar signal strength would fade down to 4 or 5, then the plane would disappear. Now, the plane keeps a signal strength in the teens and disappears much further out. This is telling me that the plane is going behind one of the many obstructions around here instead of getting lost because of signal strength.

I'm going to run this set up for few days and maybe tweak the dongle gain. I'll keep y'all posted.

-- Doug
Click to expand...
I discovered the "Magic" of Amplifier very early.
Initially I used 2 amplifiers in cascade, and could reach limit of line-of-sight (450 kms/250nm) with my following INDOOR antennas. Later I removed one out of two amplifiers, and this did not cause any appreciable reduction in range:
(1) Tiny Tot 1/2 wavelength Dipole (http://forum.planefinder.net/threads/ads-b-diy-antenna.23/#post-210)
(2) 4-Element Franklin WITHOUT matching Stub (http://forum.planefinder.net/threads/ads-b-diy-antenna.23/#post-221)
 
Last edited:
radiostationx said:
I take it you have completed some trials using various swarm patterns as a reflection technique ?:)
Click to expand...
ab cd said:
Hope the bee in your profile picture produces lot of honey and does not sting :D
Click to expand...
Lots of opportunity to experiment but dealing with a living thing you have to be careful :cool:
My bees do sting but only if they feel they are under threat. I did get a lot of honey last year. Beekeeping year starts now so antenna experiments will have to wait until autumn :(
 
just back from my local hardware store, got some "goodies" including 100 meters of 2.5mm solid coper wire for antenna-experimenting. (pretty cheap actually, only 30 euros for 100 meters, which is 30 eurocents per meter which is around 10 dollarcents (0.1$) per foot!) Let's see what kind of crazy antenna designs I can come up with :D

Also, After a power-outage on friday, the spider antenna (I swapped it on saturday) already showed a nice improvement at my parents place!
spider_home.png
 
Last edited:
7
xforce30164 said:
just back from my local hardware store, got some "goodies" including 100 meters of 2.5mm solid coper wire for antenna-experimenting. (pretty cheap actually, only 30 euros for 100 meters, which is 30 eurocents per meter which is around 10 dollarcents (0.1$) per foot!) Let's see what kind of crazy antenna designs I can come up with :D
Click to expand...
In metric systen, generally copper wire, which are mainly used for electrical wiring, are sized in crossection area, not diameter. Is your wire 2.5 mm dia or 2.5 mm² crossectional area? If it is 2.5 mm², then it's dia is 1.8 mm (1.784 mm).

Electrical copper wires in usa & canada are sized in wire gauge numbers (American Wire Gauge or AWG). To find the diameter, or crossectiona area, one has to refer to an AWG table. For examle AWG 18 has dia of 1 mm, and AWG 12 has dia of 2 mm

Also, After a power-outage on friday, the spider antenna (I swapped it on saturday) already showed a nice improvement at my parents place!
View attachment 1475
Click to expand...
Any ¼ wave whip with a ground plane - disc, horizontal radials, slanting radials, single vertical radial pointing down (= dipole), drink/food can - is better than stock antenna which comes with dvb-t dongle. Out of all these ¼ wavelength whippers, experimenters have found spider to be the best.
 
Last edited:
ab cd said:
In metric systen, generally copper wire, which are mainly used for electrical wiring, are sized in crossection area, not diameter. Is your wire 2.5 mm dia or 2.5 mm² crossectional area? If it is 2.5 mm², then it's dia is 1.8 mm (1.784 mm).

Electrical copper wires in usa & canada are sized in wire gauge numbers (American Wire Gauge or AWG). To find the diameter, or crossectiona area, one has to refer to an AWG table. For examle AWG 18 has dia of 1 mm, and AWG 12 has dia of 2 mm


Any ¼ wave whip with a ground plane - disc, horizontal radials, slanting radials, single vertical radial pointing down (= dipole), drink/food can - is better than stock antenna which comes with dvb-t dongle. Out of all these ¼ wavelength whippers, experimenters have found spider to be the best.
Click to expand...

Thats true, and you're right, the cable is specced in mm^2 (so it is a 2.5mm^2 wire and diameter is 1.8mm). =]
 
Last week's trial run conclusively proved that using 65mm (=2.56 inches or 2-9/16 inches) as ¼ wavelength gives poorer results than using 69mm (2.72 inches or 2-11/16 inches).

bazooka 69mm - 65 mm.png
 
ab cd said:
Any ¼ wave whip with a ground plane - disc, horizontal radials, slanting radials, single vertical radial pointing down (= dipole), drink/food can - is better than stock antenna which comes with dvb-t dongle. Out of all these ¼ wavelength whippers, experimenters have found spider to be the best.
Click to expand...

One of the appeals of the Franklin for me was that it was DC grounded. I assume that has fewer static build up problems. Not so with the spider. I have seen a number of suggestions including a reasonably high value resistor (say 1K to 10K), an inductor, or back to back diodes from the 1/4 wave to the ground plane. Any thoughts?
Thanks,
Doug
 
Hi Trigger,
No, not shielded as the shield will make another "ghost"/"phantom" antenna . The shield may cause excessive interference pickup.
We dont need that .
Any parallel twin (aka shotgun) lead will suffice, bell wire,speaker wire, red & black dc/ rig power lead etc they all have around 72 ohms impedance.
The theory being that the cores have same electrical properties on each lead, they are seperated at a constant distance along the entire length of around 1 single conductor diameter as a spacing done by the wall of the insulations all the way along the cable. So, its near enough 1:1 feeder @ 75 ohms.
A better way is to use twisted lighting cord, sometimes cloth covered, popular in the 1960s (if you can get it cheap), the chord is as above but twists are also of a constant distance, this lowers the possibility of bleed through from unwanted freqs . A bit like a magic invisible screening if you will, before unbalanced foil screening as we know today. The small electrical currents from a received signal flow equally along the twists of the feeder.
A bodge workaround for this would be to strip down some cat5 or installer/fixed wiring grade network cable or BT telephone outdoor dropwire (these have solid cores and are twisted) and use 1 pair.
Another alternative, stripped down cat 5 patch lead, again twisted and again you will only use 1 pair, fiddly to work with stranded cores though.
On the plus side..A 2m patch lead is cheap !
Back to the speaker cable,
You can unzip 86mm from a length of twin, thats one feeder and 5/8 dipole done in 30 secs
Now same again and support & position the 2 dipoles sets as per Airbus photos using his spacing, join at interconnect you should trim so that 272mm of speaker wire exists as a phasing harness joining the two to start.
The feed back to rtl dongle keep as short as practicable, then terminate the speaker wire feeder with a 75 ohm blaun, (balanced to unbalanced) transformer such as this

http://www.ebay.co.uk/itm/2x-4x-8x-...t=LH_DefaultDomain_3&var=&hash=item566ddd4429

The rtl dongle has an unbalanced 75 ohm mcx feedpoint, your antenna is in (theory anyway) totally balanced.
By adding a stock rtl mcx pigtail and a sma-bnc converter for the balun join, you should have only 6 inches or so of unbalanced feeder, the rest being 75 or so ohms totally balanced throughout the whole array.



Thats the theory anyway ! :confused::confused:
 
Status
Not open for further replies.
Back
Top