Tweaking of Groundplane (Spider) Antenna

FlightAware ADS-B Flight Tracking
1

It is well known that in a Groundplane antenna, the slope of radials affects its impedance, and hence affects SWR.

  1. If slope = 0 degrees (i.e. radials are horizontal), Antenna Impedance is around 30 ohms
  2. If slope = 45 degrees down (i.e. radials are slanting), Antenna Impedance is around 50 ohms
  3. If slope = 90 degrees down (i.e. radials are vertical), Antenna Impedance is around 75 ohms

Before the advent of ADS-B, the Groundplane/Spider was mainly used by Hams. Since the defacto standard of ham transmitters & receivers is 50 ohms, the 45 degrees slanting Groundplane/Spider, which has an impedance of 50 ohms, well suited ham systems. Generally other antennas were also designed to have 50 ohm impedance, either inherently, or by use of impedance matching components. The coaxial cables used were also of 50 ohms rating.

When ADS-B receiving started, its equipment (receivers antennas and coax) followed the already established ham standard of 50 ohms impedance. The sites like Flightaware, Flightradar24, Radarbox24, Planefinder also followed the suite and offered 50 ohm equipment.

With introduction of DVB-T for ADS-B, the scenario changed. The DVB-T is designed for TV reception. The defacto standard for TV & Satellite is 75 ohms, hence the DVB-T and RG6 Coax both have impedance of 75 ohms.

Unfortunately, the 50 ohms Groundplane/Spider with 45 degree slanting radials was BLINDLY used for 75 ohms DVB-T & RG6. Logically a 75 ohms Groundplane/Spider (i.e. the one with vertical radials) should suite the DVB-T+RG6 system.

**I have done some simulations, which support above arguments. However Antenna making without proper test equipment & technical know how is a “Dark Art”. Hence the only way left to know if the simulation results are true, is to make prototypes, and put these on trial run. I intend to do this when I find time, as there are 4 models, and trial runs should be lengthy to be meaningful.
**

800Ă—1900 31.6 KB

2

Great work as always abcd567, I look forward to your future test results.
…Tom

3

Although I have done these simulations sometimes back, but did not post these results because I prefered to post my finding after I test these experimentally. I did not find time (or was just lazy) to do this lengthy exercise.

What triggered me to post my previous findings this morning (without first making and testing a prototype) is last night’s PM from forum member radiostationx (Mike), confirming that the Planefinder member Philips’ groundplane/spider with 90 degree bent radials was made by him, and that it has performed better than 45 degree radials.

A conclusion cannot be drawn on performance of only one prototype. When at least 3 prototypes, made by 3 different persons at 3 different locations are tested, then a conclusion can be drawn based on these 3 results.

Two snapshots of Spiders made & tested by radiostationx (Mike - M0RRQ)

4

Hi ab cd,
Strangely my actual spider antenna builds for Phillip did not in part agree with the theory ( nothing new there !)
I started in the conventional way by first cutting the radials and the centre driven element to 1/4 lambda (69mm) 69.5mm actual then put the unit in test.
Then I started trimming back until I got to a point where performance started to drop off .
Normally this is a situation akin to an over enthusiastic use of hair clippers, you have gone too far with the clippers and cannot put any hair back !
In this case though fine tuning is quite easy, I removed sleeving to wire & added a blob of solder to the cut wire element/radial and whilst still hot stretched it out so that an extra 4mm or so could be put back to the electrical length.
Fine tuning adjustments can then be made with a discarded fingernail emery board taking off a little at a time until optimum performance has returned.
Slide over the pvc sheath to cover the bare wire again, adjustments hidden.

I found the best results with radials @ length 69mm measured from the outer ring of the so-239 delrin insulator/ aka the start of the flange giving an electrical length of 69mm the same as the centre element. The flange being part of the groundplane (an integral part of the radial arrangement)
Radials are in fact actual length of 65mm.

I have a pictures attached to show what Phillips antennas looked like cut to correct length before bending so show what I mean.

800Ă—600 164 KB

800Ă—600 95.3 KB

600Ă—800 136 KB

I think perhaps this could be further improved by using a 75 ohm flanged connector to achieve better matching for a 75 ohm unbalanced dongle input.
In theory, the swr would be lower and the antenna would be slightly more efficient especially on weaker adsb signals.
Again , in theory, as we are dealing with microwave signals very small changes and tweaks could mean a station seeing a distant aircraft or not seeing it.

I am glad I got nice performance from this pair built for Phillip, the tweaking took a while , but he is very happy with the results.

I wondered if this would also work for the cantenna in the same way making the radial section of the cylinder shorter.

5

In my setup, I use an inline 15dB LNA that is 50 ohms in and out. It’s fed with a 50 ohm antenna through a 50 ohm FA filter. The LNA output goes to the RTL dongle antenna port, and this is where things get sketchy…maybe… I have the 50 ohm SMA to MCX plugged right in and it appears to work famously.

It may be assumed that the RTL antenna port is 75 ohm since TV coax is 75 ohm, and this is after all a TV receiver by design…

However, if you examine the MCX connector plug on the end of the factory supplied RTL antenna, it has the dielectric sleeve, while a true 75 ohm MCX does not. So, by using 75 ohm inline amps and RG6, one creates a 75 ohm antenna system feeding a 50 ohm load. The RTL dongle antenna port is actually 50 ohm.

See the images and compare with your RTL antenna plug/port:

http://i63.tinypic.com/t83gyf.jpg http://i65.tinypic.com/wtxgyu.jpg

This is not a terrible thing. The important thing is to keep your system free of changeovers as much as possible to extract those last few tenths of a dB :slight_smile:

(actually, it doesn’t matter as much on receive, since a load mismatch between a pure tuned 75 ohm antenna/amp/coax system and a 50 ohm input impedance will simply very, very slightly reduce the power delivered to the receiver front end, with no effect on the actual received SNR. It won’t affect things unless it is so attenuated that it drops below the receiver noise floor.)

Why do we look to 75 ohm? IMO, it’s because of two things:

  1. RG6 coax is cheap
  2. inline satellite TV amp is cheap and can be powered through the coax, out of the box.

Problem with inline satellite TV amps is that they are designed for wideband performance, and are not necessarily low-noise amplifiers since they already have a good clean, amplified signal from the LNB feeding them.

Problem with 50 ohm LNA’s is the cheap ones usually are in hobby form, and need an enclosure, bias tee power injector and a power supply.

That makes it all fun to play with, just to see what really works vs. works on paper.

In the end, though, SWR is the ratio of load impedance to transmission line impedance. In receive, the load is the receiver or inline amp in front of the receiver.

6

Hi K5ted,
Thats interesting about the mcx connector variants, thanks for posting .

It is always satisfying to get a few squitters received in the 250mls + range column of your flight aware stats for the day to show that all is well with your station.

Its getting those signals from afar using off the shelf gear and a bit of homebrew that is the challenge I enjoy.
Those few 10ths of a db can sometimes mean receiving a distant aircraft or not.

Most sat TV amps here in Europe are 75 ohms and Belden Webro WF100 75 ohm coaxial cable is very reasonable quality and cost.
The aim is an impossible one… to have a balanced antenna feeding system using unbalanced cable and connectors.
But we do like experimenting !

Regards,
Mike

7

@K5TED
The use of a connector of 50 ohms characteristic impedance to a load of 75 ohms input impedance has negligible effect on the impedance seen by the source. This is due to very small length of the connector.

Please see calculation below which show that the input impedance seen by feed coax at the input of the DVB-T dongle (through MCX connector) is 74 ohms, negligible change due to 50 ohms MCX connector.

650Ă—750 17.5 KB

.

8

Please show documentation from the manufacturer stating the DVB-T dongle antenna port impedance is 75 ohms. I see a 50 ohm jack and a 50 ohm factory supplied antenna plug.

That indicates the receiver antenna port is indeed 50 ohms. A flurry of formulas doesn’t change that.

I see a 50 ohm connector, so that means 50 ohms input impedance.

9

Please download the data sheet (pdf) for the DVB-T’s Front End Tuner chip R820T from the link below.

http://superkuh.com/gnuradio/R820T_datasheet-Non_R-20111130_unlocked.pdf

Please see diagram on page 18, left side where RF_IN is written (in red), below which impedance is given 75 ohms (in green):

1281Ă—791 56.4 KB

10

I see where you’re coming from, but these are application notes for the tuner chip, not dongle manufacturer specs. Fact remains, the factory antenna and the dongle antenna port are 50 ohms, at least on my Nooelec dongles. The RTL SDR dongle has an SMA connector, also 50 ohm.

11

@k5ted

Yeah that’s what he’s saying. Yes, they’re using a 50ohm connector but the chip itself has a 75ohm input. The formula’s he posted are showing that, even with the 50ohm connector and the 75ohm chip the effective impedance of the receiver is 74 and change

12

Perhaps… Guess that’s what we get for $15

13

LOL! :mrgreen:

14

Or you can get one built with a F connector and not worry about mismatch…


Comparison between these ThumbSat R820t2 and the the NooElec R820t2 dongle with SMA to type F cable show negligible difference. The simplicity of having type-F connectors throughout my whole setup without adapters… PRICELESS!

Cheers!
LitterBug

15

https://farm2.staticflickr.com/1656/25760597306_ce168ea609_m.jpg.https://farm2.staticflickr.com/1464/25160068463_c02d18e5cd_m.jpg
DVB-T facts:

(1) Manufacturer has fitted the dongle with 50 ohm mcx connector.
(2) The input impedance of front end tuner chip (R820T) is 75 ohms
(3) Manufacturer’s use of a 50 ohm connector at the input of a 75 ohms chip has a very insignificant effect on input impedance of dongle i.e. the input impedance changes from 75 ohms to 74 ohms (please see calculations in my previous post).
(4) Manufacturer supplies a whip antenna with the dongle.

 (a) The whip with groundplane has an impedance of 50 ohms. The coax of whip is also 50 ohms (RG174)
 (b) No attempt has been made for impedance matching of 50 ohm antenna with 74 ohm input impedance of dongle.

.

16

@Radiostationx:

Hi Mike!
Here is simulation result for your variant of the Spider.

800Ă—516 8.72 KB

.

17

I crossed over from the dark side (coaxial collinear) to check out your simulations.

Use high quality F connectors with a couple of 50 mm plaster washers along with some RG6 copper core for 2 builds.

One build had fixed spiders extending 69 mm from the edge of the 50 mm ground plate (94 mm total distance to axis).
The antenna length was trimmed (65 mm above ground plane) and the spider angle was adjusted (15 degrees down) to minimize SWR.

The second build had a fixed antenna length of 69 mm extending above the ground plane.
The spider legs were trimmed (45 mm from the edge of ground plane - 69 mm distance to axis) and down angle adjusted (75 degrees) to minimize SWR.

In both cases the as measured SWR was very acceptable with the 75 degree shape form being excellent and about as good as you can get with a SWR of 1.07 and 79 impedance .

(http://i.imgur.com/M3ThcGW.jpg)
(http://i.imgur.com/a7igTQQ.jpg)

18

@elpipila:
Thanks for making & testing prototpes based on simulations.
I want to buy low cost equipment for testing antennas. Can you please tell me which equipment and software you used for these measurements? Thanks.

19

I have the 1400. The net says you can use the lower cost ones if you jump through some hoops using a harmonic multiplier.
rigexpert.com/index?s=aa1000

1 Like
20

@elpipila:
Thanks for the information. I checked the site, the 1400Mhz one is CAN$1500. Even 600 Mhz model is CAN$885. This is way up my budget. :frowning: Will wait till I win a lottery. :smiley:

For now I think I will choose one from here:
http://discussions.flightaware.com/ads-b-flight-tracking-f21/low-cost-signal-generator-35mhz-4-4ghz-for-antenna-testing-t36877.html