A Very Strange Experience!
Today I tried the method suggested by PeterHR on my 8-element CoCo with element length 114 mm. I slid apart all the 8 elements in steps of 1 mm, and after every step checked message rate, maximum range, and plane count. I continued upto 4 mm and found best results of 30% increase, when the elements were pulled apart by 2 mm. This gave me optimum element length of 114 mm + 2 mm = 116 mm.
I had to slide open all the 8 joints precisely same amount 4 times. This was a bit tedious. I decided to repeat this exercise on a 4-element CoCo to find optimum element length for the coax I am using. This required adjusting only 4 gaps instead of 8, and was much easier. Once optimum element length found, I can apply it to 4, 8, 12 or 16 element CoCo.
I removed the top 4 elements of my 8-element CoCo, converting it to a 4 element CoCo, and repeated the procedure. This time the results were completely different. I ended up with an optimum gap of 1mm instead of 2mm, and to my surprise, the optimized 4-element CoCo 's performance reached 200% of what I got with optimized 8 element CoCo!!!
Unbelievable that a 4 element CoCo performed 200% of the 8 element CoCo made of same coax, same element length, both with optimized gap, at same location, during same hour of the day with steady air traffic, and connected to same system (CoCo >> 12 ft/4m RG6 Coax >> F-female to MCX-male RG316 pigtail 1 ft/30 cm to DVB-T Dongle >> DVB-T Dongle inserted directly into RPi >> 15 ft / 5 m Network cable from RPi to Network Switch/Hub >> 15 ft / 5 m Network cable from Network Switch/Hub to Router). No Amplifier.
Hi ab cd,
I’m doing the same thing with a 10 element CoCo. My element length is 91mm (VF = 0.66) When I had the antenna on the bench I was checking continuity as I moved each element. I found that although I was getting continuity, the resistance from end to end was varying. I made sure that end to end was consistently 1.3 ohms before taping the whole thing back together. 1.3 ohms was the lowest resistance I could get. Don’t forget that there are 2 continuity checks to be made with a CoCo
I’ve now run with gaps at 1mm, 3mm and 5mm and NOT found a difference in any of the usual parameters.
Hi Dave,
Whenever I disturb a CoCo, I always make a end-to-end continuity check (core to shield & shield to core for even number of elements, core to core & shield to shield for odd number of elements). I also make a short-circuit test between core & shield at each end. However I never measured resistance, just the continuity beep & light.
In order to be sure of my findings, I am now planning to make a brand new 8 element CoCo with same coil of coax, 114 mm element length, and repeat my experiment, hopefully next weekend. This time I will check the resistance also as suggested by you.
Your results are also strange: no difference in performance with element gaps 1, 3 & 5 mm
Hope you are fine. Did you used this LNA? Hows this performance ? If you recommend, i will buy this. Seems a good item to work. It have AMP+filter. I got it in a Chinese site. cart100.com/Product/521057659108
Looks interesting - looks like there is a low noise amp followed by a SAW filter followed by another amp - each amplifier has it’s own power regulator.
Would be better still if it was in a waterproof case and could be powered from the downlead … then it would be ideal for a masthead.
but by using a power injector in reverse or just bey putting an inductor between the output port and the power connector - powering up the wire would be possible.
Regarding the OP’s and other’s original DIY “ground” reflector antennas, if ADS-B is 1090Mhz then 1/4 wave would be about 6.9 cm or 2.71".
There were are a lot of posts to indicate that 2.575" was the ideal length of the elements but that would put it at 1/4 wavelength of 1147 MHz.
It won’t be a huge difference in reception but it is a 15% difference in center frequency. Am I missing something here?
Sorry if this has been answered before.
TIA
PS: There is a difference in velocity factor between bare copper wire and the speed of the propagation of the charge inside the coax cable which does result in shorter 1/4 wavelengths. The propagation through the coax can be a very significant difference to the speed of light down to 50%, but with the exposed copper wire conductor it should only be a 1-5% shorter wavelength in copper, not 15%. Is there a further tuning in the element lengths to match the impedance of the coax with the atmosphere? BTW, we are only talking about less than 3/10" but the target length should be exact.
The best way to determine the correct length of a dipole antenna using any given materials is to set up a calibrated RF transmitter at 1090 MHz and trim the center element of the antenna until you find the peak received signal. All the adjustments on wavelength would take into account the coax you are using, the connectors and everything else.
So far the online calculators I’ve seen don’t appear to be very credible with no explanation on how the calculations are generated.
This amplifier is available as a board on ebay for £10 (search for low noise RF amplifier 435 )- but the SAW would need to be changed fro 435Mhz to 1090MHx.
For bias power (up the wire) it just needs a 1/4" coil of insulated copper wire, maybe 12 turns fro the centre pin on the RF out connection to the +ve supply connector.
That’s an interesting pic, does it come with a cover? BTW, that image on the analyzer really does not indicate anything that I can discern. I would expect to see two traces. One for input and one that shows the output gain. That picture just shows a fancy analyzer with single trace noise.
Back to the original question, what is the ideal 1/4 wave dipole length for ADS-B? And please explain why in detail with real numbers .
One mistake I have seen is that a bare copper wire protruding from a coax in air has the same impedance as the characteristic impedance of the coax. That is incorrect. The bare copper wire has a velocity factor of 1-5% less than the speed of light, not the 80-50% of the coax. But it does present an impedance mismatch between the exposed bare copper wire and the interface as it re-enters the coax. This could be optimized with very small adjustments in dipole length, but I still don’t see 15 percent off center frequency as suggested by other posters. More like 5 percent.
Feel free to fill in the blanks if you think I am missing something!
Would you mind running a simulation for a wire collinear antenna using the core wire from an RG-6 coax? I believe the thickness is 1 mm. I’m interested in building one, and I’m really curious as to what your modeling software recommends for the coil diameters. Thanks in advance!
Ok, this weekend I will design & simulate a wire collinear using 1 mm dia (18 AWG) core wire of RG6 Coax, and post the results.
Please note that ALL Collinears (Coaxial, Wire, Franklin etc) are easy to make but hard to get right. The reason is that collinears require high degree of accuracy in both the design as well as in construction. Their dimensional tolerance is much smaller than what can be achieved by a hobbyists DIY design & DIY construction. On top of that hobbyists lack the fancy test equipment required to test & tune/trim the antenna.
On the other hand a quarter wavelength (69 mm) whip connected to core of coax + some sort of ground-plane connected to shield of coax make a good DIY antenna, as it has (1) large dimensional tolerance and (2) has a natural impedance matching with 75 & 50 ohm system of coax, dvb-t receiver, and amplifier. Easy antennas in this category are Cantenna & Spider. The 1/4 wavelength antennas have disadvantage of low gain (2 dBi) compared to collinears (typically 5 to 10 dBi), but this can be overcome by adding a low cost satellite tv amplifier.
abcd567, Any news on this? Very interested to hear the results. You sound like you know what you are talking about.
BTW. What sim software are you using?
While amplification can compensate for weak signals, it can not amplify what does not exist. So, I would suggest one should look at the source of db gain when evaluating. I want max gain from the antenna (raw signal strength) and then top it off with amplification as needed.
I have done simulation of 3 designs of wire collinear whips with pepsi can as groundplane. One whip is franklin, the other two are coiled whips of different sizes of coils. Optimized by simulation software, these have gain of 3 to 4 dBi and SWR 2 to 1.1. My default antenna isCantenna. It’s simulation gives a gain of 1.5 dBi & SWR less than 1.5.
Based on simulation results, all 3 wire collinear Cantennas should perform better than standard 69mm whip Cantenna, but when I made proto type whips, and put to trial run (by replacing Cantenna’s 69mm whip with the collinear whip under test), all these collinear whips proved to be inferior to the cantenna’s original 1/4 wavelength (69mm) whip.
I am planning to try trimming these antennas this week end to see if i can improve their performance.
I have used two different simulation software, and both give closely matching results. The two software are “4nec2” and “manama-gal”. I have expierienced earlier that these software can give an error of 5% to 10%, which is too high for an accurate design of a collinear. That is why I am planning to try by tweaking & trimming the prototyps I have made.
This is a misconception that weaker signals are non existant in a low gain antenna.
If you place a high gain and a low gain antenna at same location, same radio signals, weak & strong, will hit both the antennas, and produce an identical output at output terminals of both the antennas. The only difference will be in strength of output signals. Not a single signal will be missing from the output of low gain antenna.
This is like a family photo having toddlers, under 10, teens, and adults. If the photo is reduced to 1/5 size, will the toddler, being small in size, become non existing in the smaller picture?
I like the toddler analogy but unfortunately that only works while the weaker signals are above the noise floor of the amplifier otherwise thy may as well not exist. Luckily there are plenty of nice low noise and easy to use amps around like the PGA103 based ones with hopefully a <1 db noise floor at 1ghz but there are plenty of others. Ive had good results with W1GHZ’s design, if you are handy with a soldering iron its worth a try.
You are right. I was about to add the sentence about this, but left out in order to make reply simple, and focused on main issue of “non existance”.
If an amplifier is used, the signals which are below the noise level of amplifier get mixedup in noise and are practically lost. This can be overcome by using low noise amplifiers. The common satellite tv amplifiers have a noise level around 2 to 5 dB, and most brands wont mention the noise figure. The special low noise amplifiers like the one you mentioned have noise level below 1 dB.
That said, we have to consider another fact: the curvature of earth prevents reception beyond 300 nm. The aircraft’s transponder emmits powerful enough signal (200 watts?) so tht even after travelling 300 nm, its strength is still above the noise floor of common satellite tv amplifiers. I did not make any measurements of signal strength, but the fact that I have been able to steadily recieve signals from planes as far as 250 nm+ using a satellite tv amplifier confirms this conclusion.
:mrgreen:
.