ADS-B DIY Antenna
- Thread starter Mukundan
- Start date
- Status
I would also switch away from the old Windows client to the new one, or just shut it down if the data is coming from a new client anyway. That feeder has quite a few bugs in it!
ab cd
Senior Member
Following your advice, I just now tried to install new windows feeder version 2.0.2080 on my old faithful, Windows XP
Ah yes, we don't support XP, and neither do Microsoft ;-)Following your advice, I just now tried to install new windows feeder version 2.0.2080 on my old faithful, Windows XP
View attachment 2255
I think we are getting some Pi data from you and actually you should probably just turn the old feeder off. We won't be missing anything.
ab cd
Senior Member
The Desktop, and hence windows client is not always running. That is the reason I purchased and installed two RPis (each with an independent systems of antenna, Coax, Amplifier, DVB-T dongle).
These RPis run continuously 24 hrs a day/ 7 days a week / 52 weeks a year. Windows client runs intermittently, using its own antenna, coax & DVB-T dongle, only when the Desktop is running.
I would remove the Windows version then, could cause issues.
ab cd
Senior Member
Actually I have stopped using it since I got my RPi.
Only two days ago (on July 03) I started it just to see if it is still alive. Now I have again stopped using it.
ab cd
Senior Member
WIRE COLLINEAR
I have noted some contradiction in coil dimensions in this drawing of wire collinear by G-7RGQ.
It says the coil consists of 1½ turns of 10 mm dia. The length of wire in coil therefore works out to π x d x 1.5 = 3.14 x 10 x 1.5 = 47mm. Adding 4 mm wire to provide gap between turns, the total wire length becomes 51 mm. However the drawing mentions "total wire length in coil = 65.5 mm (¼ wavelength)".
SIMULATIONS
I have run 2 simulations of wire collinear by G-7RGQ:
Simulation 1: Using 10 mm dia, 1½ turns coil, as given in drawing.
Simulation 2: Using 14 mm dia, 1½ turns coil, calculated by wire length in coil = 66 mm, i.e. ¼ λ.
The simulation results show a marked improvement in Radiation Pattern as well as in Horizontal Gain when 14 mm dia coil (i.e. wire length = 66 mm) is used. SWR in both cases are more or less same and substantially high (i.e. SWR > 6)
10 mm dia coil
14 mm dia coil
NOTE:
Diameter of the coils in simulations is measured from center of the wire.
Hence the inner dia of coil (i.e. the dia of rod or drill bit on which the coil is wound) = dia of coil - dia of wire.
If a 2 mm dia wire is used:
The 14 mm dia coil should be wound over a rod/drill bit of dia = 14 mm - 2mm = 12 mm
The 10 mm dia coil should be wound over a rod/drill bit of dia = 10 mm - 2mm = 8 mm
AN EASIER WAY to make the coil accurately is to measure & cut 66 mm length of wire and then wind it over a rod or drill bit. In this case exact dia & number of turns is not important as the wire length is pre-cut to 66 mm (i.e. ¼ λ). A 1/2" (12.5 mm) or 3/8" (9.5mm) drill bit can be used to wind the coil. The coils and vertical sections can then be joined by soldering/brazing/welding.
I have noted some contradiction in coil dimensions in this drawing of wire collinear by G-7RGQ.
It says the coil consists of 1½ turns of 10 mm dia. The length of wire in coil therefore works out to π x d x 1.5 = 3.14 x 10 x 1.5 = 47mm. Adding 4 mm wire to provide gap between turns, the total wire length becomes 51 mm. However the drawing mentions "total wire length in coil = 65.5 mm (¼ wavelength)".
SIMULATIONS
I have run 2 simulations of wire collinear by G-7RGQ:
Simulation 1: Using 10 mm dia, 1½ turns coil, as given in drawing.
Simulation 2: Using 14 mm dia, 1½ turns coil, calculated by wire length in coil = 66 mm, i.e. ¼ λ.
The simulation results show a marked improvement in Radiation Pattern as well as in Horizontal Gain when 14 mm dia coil (i.e. wire length = 66 mm) is used. SWR in both cases are more or less same and substantially high (i.e. SWR > 6)
10 mm dia coil
14 mm dia coil
NOTE:
Diameter of the coils in simulations is measured from center of the wire.
Hence the inner dia of coil (i.e. the dia of rod or drill bit on which the coil is wound) = dia of coil - dia of wire.
If a 2 mm dia wire is used:
The 14 mm dia coil should be wound over a rod/drill bit of dia = 14 mm - 2mm = 12 mm
The 10 mm dia coil should be wound over a rod/drill bit of dia = 10 mm - 2mm = 8 mm
AN EASIER WAY to make the coil accurately is to measure & cut 66 mm length of wire and then wind it over a rod or drill bit. In this case exact dia & number of turns is not important as the wire length is pre-cut to 66 mm (i.e. ¼ λ). A 1/2" (12.5 mm) or 3/8" (9.5mm) drill bit can be used to wind the coil. The coils and vertical sections can then be joined by soldering/brazing/welding.
Last edited:
ab cd
Senior Member
Franklin Spider Failure Reason - Impedance Mismatch
The Spider has an impedance close to ideal 75+j0 ohms. The SWR is about 1.3
The Franklin Spider has an impedance 176+j0.14, far from ideal 75+j0 ohms. The SWR is about 2.35
As a result Franklin Spider gave inferior performance than Spider.
I now attempted a CRUDE impedance match (many approximations in calculations and factors & use of capacitor 10% lower than required) and still got a dramatic improvement. Please see data graphs below:
.
And this is how I achieved it:
Step 1 of 3: Starting point is antenna impedance by simulation
Step 2 of 3: Calculation of length of transmission line piece & value of series capacitor by Smith Chart.
Length of Transmission Line between Antenna & Capacitor (including length of Connectors) = 210 mm RG6, VF = 0.84 (assumed)
Capacitor = 2.2 pF.
I did not have a 2.2 pF capacitor, but had many 1 pF capacitors, so I put 2 capacitors of 1 pF in parallel, which gave me 2 pF, 10% less than required 2.2 pF. I will now purchase a 2 - 5 pF trimmer (adjustable capacitor) and replaced ceramic ones by trimmer and try to fine tune the response to optimum.
Step 3 of 3: Built & put on trial run
Later Addition: Later tweaking showed that using a 3 pF Capacitor (instead of 2 pF Capacitor) and 240 mm Transmission Line (instead of 210 mm transmission line) improved performance, and Franklin Spider surpassed Spider.
The Spider has an impedance close to ideal 75+j0 ohms. The SWR is about 1.3
The Franklin Spider has an impedance 176+j0.14, far from ideal 75+j0 ohms. The SWR is about 2.35
As a result Franklin Spider gave inferior performance than Spider.
I now attempted a CRUDE impedance match (many approximations in calculations and factors & use of capacitor 10% lower than required) and still got a dramatic improvement. Please see data graphs below:
.
And this is how I achieved it:
Step 1 of 3: Starting point is antenna impedance by simulation
Step 2 of 3: Calculation of length of transmission line piece & value of series capacitor by Smith Chart.
Length of Transmission Line between Antenna & Capacitor (including length of Connectors) = 210 mm RG6, VF = 0.84 (assumed)
Capacitor = 2.2 pF.
I did not have a 2.2 pF capacitor, but had many 1 pF capacitors, so I put 2 capacitors of 1 pF in parallel, which gave me 2 pF, 10% less than required 2.2 pF. I will now purchase a 2 - 5 pF trimmer (adjustable capacitor) and replaced ceramic ones by trimmer and try to fine tune the response to optimum.
Step 3 of 3: Built & put on trial run
Later Addition: Later tweaking showed that using a 3 pF Capacitor (instead of 2 pF Capacitor) and 240 mm Transmission Line (instead of 210 mm transmission line) improved performance, and Franklin Spider surpassed Spider.
Last edited:
ab cd
Senior Member
I have noted something wrong in the graph of collected data I posted above with impedance matching. When I changed ¼ wavelength whip with Franklin whip (i.e. converted Spider to Franklin Spider), all the graphs became nearly zero. I have previously swaped the Spider & Franklin whips several times, and each time the graphs dropped to about 40%, but not almost zero.
To clear the matter, tomorrow I will repeat the experiment about the same time and post graphs.
Inspite of above abnormal behaviour, the jump back of Franklin Spider graph to level of Spider after addition of impedance matching, clearly prove the effectiveness & success of impedance matching method I have used.
ab cd
Senior Member
In the above trial run of matching network for Franklin Spider, I had to make some assumptions & aproximations. As a result, the matching is aproximate, not optimum. I am now planning to optimize matching network by trial & error.
(a) I will replace fixed ceramic capacitor by adjustable (2pF to 5 pF) trimmer capacitor.
(b) I will make & try two more pieces of transmission line, one 5mm less (i.e.205 mm total length including connectors at both ends) other 5mm more (i.e. 215mm total length including connectors at both ends).
I hope I will be able to get still better results than the crude matching I have done now.
(a) I will replace fixed ceramic capacitor by adjustable (2pF to 5 pF) trimmer capacitor.
(b) I will make & try two more pieces of transmission line, one 5mm less (i.e.205 mm total length including connectors at both ends) other 5mm more (i.e. 215mm total length including connectors at both ends).
I hope I will be able to get still better results than the crude matching I have done now.
ab cd
Senior Member
Due to yesterday's abnormal graph, I checked my system and found a defective (dodgy/intermittent) connector. I replaced the connector and repeated yesterday's test, and results are now normal as follows.
1) When The ¼ λ whip was replaced by Franklin whip, there was a distinct drop in performance to about 40%.
2) When crude impedance matching was added to the Franklin Spider, there was a distinct increase in performance, jumping up from 40% to 85% of Spider.
Now I have to optimize the matching network by:
1) Using an adjustable trimmer capacitor 2 - 5 pF instead of fixed ceramic capacitor of 2 pF.
AND
2) Try three lengths of transmission lines (i.e 205 mm, 210 mm & 215 mm) between antenna & capacitor.
I hope I will be able to achieve optimum matching, in which case the Franklin Spider's performance will (hopefully) exceed Spider's performance.
.
1) When The ¼ λ whip was replaced by Franklin whip, there was a distinct drop in performance to about 40%.
2) When crude impedance matching was added to the Franklin Spider, there was a distinct increase in performance, jumping up from 40% to 85% of Spider.
Now I have to optimize the matching network by:
1) Using an adjustable trimmer capacitor 2 - 5 pF instead of fixed ceramic capacitor of 2 pF.
AND
2) Try three lengths of transmission lines (i.e 205 mm, 210 mm & 215 mm) between antenna & capacitor.
I hope I will be able to achieve optimum matching, in which case the Franklin Spider's performance will (hopefully) exceed Spider's performance.
.
Last edited:
ab cd
Senior Member
The Franklin Spider is inferior to Spider if impedance matching is not used, but can be made superior to it by impedance matching, as shown in above post.
Today I have tried two tweaks in values of the two parameters of impedance matching as below:
(1) Soldered a third 1 pF Capacitor in parallel with existing two, increasing the total value of Capacitor from 2 pF to 3 pF
(2) Tried two different lengths of transmission line between the Antenna & the Capacitor. Original was 210 mm (including length of connectors at both ends). I made two more, one 225 mm (including length of connectors at both ends), other 240 mm (including length of connectors at both ends).
The combination of 3 pF Capacitor & 240 mm Transmission Line (including length of connectors at both ends) between Capacitor & Antenna resulted in Franklin Spider surpassing the performance of Spider.
Since each built of DIY antenna is slightly different, these values (240 mm / 3 pF) may not be optimum for every one, but sure will be close to optimum, provided RG6 Coax cable (with VF 0.83 to 0.85) is used as transmission line piece between Antenna & Capacitor, shown in photo in my post. #3630 above
Today I have tried two tweaks in values of the two parameters of impedance matching as below:
(1) Soldered a third 1 pF Capacitor in parallel with existing two, increasing the total value of Capacitor from 2 pF to 3 pF
(2) Tried two different lengths of transmission line between the Antenna & the Capacitor. Original was 210 mm (including length of connectors at both ends). I made two more, one 225 mm (including length of connectors at both ends), other 240 mm (including length of connectors at both ends).
The combination of 3 pF Capacitor & 240 mm Transmission Line (including length of connectors at both ends) between Capacitor & Antenna resulted in Franklin Spider surpassing the performance of Spider.
Since each built of DIY antenna is slightly different, these values (240 mm / 3 pF) may not be optimum for every one, but sure will be close to optimum, provided RG6 Coax cable (with VF 0.83 to 0.85) is used as transmission line piece between Antenna & Capacitor, shown in photo in my post. #3630 above
Last edited:
trigger
Member
ab cd
Senior Member
@ab cd, great piece of work as usual
You must be one of the only people in the world with a Smith chart, protractor and a pair of compasses readily to hand. I do have a copy of a Smith chart but have absolutely no idea what to do with it. I use it for standing my coffee mug on.
Thanks Dave!
If you want to learn impedance matching using Smith Chart, go through these pages in the sequence listed below:
(1) First: Load transformation.
(2) Next: Impedance Matching Basics - Series L and C.
(3) Next: Perfect Impedance Matching with Tx Lines, Series L, and Series C.
Remember all impedances on Smith Chart are normalized values, i.e. with reference to a base value chosen by user. For example if base value is chosen as 35 ohms, then on Smith chart, an impedance of 105 ohms becomes 105/base = 105/35 = 3.
The most convenient and practical value to choose for base impedance is system impedance. Since DVB-T dongle & common Coax have an impedance of 75 ohms, using 75 as base greatly simplifies calculations. Thus a antenna having an impedance of 135 ohm resistive and -90 ohm reactive impedance (135 - j90 ohms), will become 135/75 - j 90/75 or 1.8 - j1.2
Last edited:
ab cd
Senior Member
Exactly as I expected!Hmmmm, I read the first couple of sentences in the first link and I've decided to keep using the chart as a coffee coaster
This subject is is so imaginative & mathematical, it repels most, except very determined ones
Spider, Cantenna, and Franklin Spider are all inferior to my best CoCo (8 element, open top)
Last edited:
trigger
Member
Just as well some people are determined so they can enlighten us.
My eyes glaze over at the mention of imaginary numbers I've found a couple more videos on Youtube which I'll watch - assuming I stay awake.
In that case I won't make one as my CoCo also out-performs all the other antennas I've made.
My eyes glaze over at the mention of imaginary numbers
I've found a couple more videos on Youtube which I'll watch - assuming I stay awake.In that case I won't make one as my CoCo also out-performs all the other antennas I've made.
- Status
