Hi Nigel, Yeah, know the area well…I live off Lea Rd 
Let’s revive a little bit, this topic - trying to understand when and why, the bottom sleeve/stub on the feed line should be oriented to the end of the antenna or to the feed connector, depending clear of the endpoint of the antenna - some of them are 1/4 whip in air, some of the 1/4 whip in air but with shorted 1/4 coax, some of them just plain 1/2 coax and so on. The sleeve/stub sometimes is just a simple sleeve over the coax, soldered at one end, sometimes it’s 3-4 times in diameter than the coax and soldered either at 1/4 either at base of the feed line. On the internet are so many variations not just in ADS-b but also in HAM radio, and a lot of the models I saw are conflicting with the teachings of G8JNJ for instance. Can someone make a little light into this matter?
Many thanks!
Your query can best be answered by @G8JNJ (Martin).
Unfortunately I have not seen him visiting this forum for many years.
Might worth a PM to him
Thank you abcd567!!!
All the best!
1 Like
Hi,
It depends…
Normally you would wish the top end of the sleeve that is connected to the outer of teh coax to be at a low impedance point, so the open end of the sleeve (which is a high impedance point) will be positioned 1/4 wavelength away from the closed end.
In order for this to be effective the open end of the sleeve should be 5 to 10 times greater diameter than the coax, so that the coax doesn’t ‘load’ the end of the sleeve, due to it being too closely coupled to the outer insulation and braid of the coax.
When you are constructing a collinear, sometimes you want a high impedance node to appear at a certain point along the structure, so the sleeve may be the other way around.
The idea of the colinear is that all the radiating sections are fed in phase, so that the antenna presents itself as a number of dipoles stacked on top of each other with the correct phase relationship so that radiation is enhanced.
However there is a problem. The speed at which RF travels depends upon the medium, and RF on the outer of the coax will travel faster than RF passing along the inside of the coax, due to it’s velocity factor. So for a coaxial colinear to work properly it ideally needs to use air spaced coax, so that the inner and outer RF paths are equal.
It’s actually quite r=tricky to get a coaxial colinear to work properly, as any slight phase errors will cause the antenna pattern to tilt up or down rather than in the desired direction out towards the horizon, and measuring the impedance match at the feedpoint doesn’t really tell you that much about how well the antenna is actually working.
Adding more sections is often counter productive, as it makes it easy to introduce more phase errors and in addition gain is only achieved by reducing the vertical beamwidth. If you have too much gain you produce a very narrow pencil shaped pattern like a very thin doughnut or ‘lighthouse’ beam, and as a result may miss signals originating under or above the beam pattern.
Personally I always like to check any antennas I build against some sort of reference antenna. This is usually a dipole or 1/4 wave groundplane cut to be resonant at the required frequency. A simple A/B comparison will then give an indication as to the effectiveness of your new antenna.
Personally for simplicity, I’ve found it hard to beat a simple 1/4 wave length of wire fed against an inverted soft drinks can as a groundplane, cut to be 1/4 wave long from the centre of the can base to the open end of the can. I mount a pre-amp inside the can and use WF100 satellite TV coax that uses a copper foil, braid and inner. Although this is 75 ohm it doesn’t matter, as the pre-amp has sufficient gain to overcome any minor mismatch losses.
Regards,
Martin
3 Likes
Welcome back Martin.
Glad to see you after 8 years.
Thank you for your time & effort to provide valuable reply to @dallcompany’s queries.
Regards,
abcd567
Hi Martin!
Thank you for you time and your explanations. Still i’m in doubt so I will comment on your text and when you might have time and energy I would appreciate an advice:
“Normally you would wish the top end of the sleeve that is connected to the outer of teh coax to be at a low impedance point, so the open end of the sleeve (which is a high impedance point) will be positioned 1/4 wavelength away from the closed end.”
Which is the closed end in this case? The first λ/2 radiator ? Like in this picture? Does this picture works with λ/2 end element with no whip?
“In order for this to be effective the open end of the sleeve should be 5 to 10 times greater diameter than the coax, so that the coax doesn’t ‘load’ the end of the sleeve, due to it being too closely coupled to the outer insulation and braid of the coax.”
I see - in some HAM radio examples the sleeve was made from just a cooper pipe, slided over the last λ/2 element (which connects to the feed connector like below), usually soldered to he braid of the first λ/2 element, Is this a correct balun?
or like this
“When you are constructing a collinear, sometimes you want a high impedance node to appear at a certain point along the structure, so the sleeve may be the other way around.”
How do you add a high impedance node in a CoCO? As, for instance, the sleeve in this picture is oriented face down to the feeder line(connector). Is just about the whip or the entire element with or without closed whip?
or without closed whip like below
This one is identical like the four λ/2 collinear from above, but instead of the sleeve, they made the balun from coax itself - kinda of Pawsey balun resemblance…
And the last headache 
Which one from the below picture is correct as according to my “read” experience and not the scientific one, the stub should not be oriented upwards in the λ/4 top ended elements ones
Maybe a simple hand written schetch will do 1000 words
Many thanks Martin - Looking forward to learn new things - experimenting without learning is fun, but learning makes it double. People like, you having the knowledge and the time, or energy, to share the knowledge, are a exceptionally welcome. It’s always a pleasure to read your posts and you website.
Have a great day!
Yes just like that
You know that’s a DC pass-through amp right?
You don’t “add” a high impedance node, it’s a mathematical characteristic of the construction (if you’ve built it right 
)
1 Like
To be honest none of these really make much sense to me.
Let’s take a look at these four examples.
I have marked out what I think are the high (shown in Red) and low (shown in Blue) impedance points on the outside ‘radiating’ part of the structure.
The 1.4 wave sleeve balun should be low impedance at the bottom closed end and high impedance at the top open end.
If we then mark where the rest of the Hi and Low impedance points are along the structure, then where the coax transitions occur are at at low impedance points.
This seem to be OK, except when you look at the top of the final section. In theory looking at antenna A it should be a low impedance point, but it is low open circuit, so it is likely to appear as a high impedance point. Something seems wrong there.
Likewise in antenna B, the top section is now a 1/4 wave so the end should be a Hi impedance point, but it is short circuit, which also seems to be wrong.
Antenna C is the same as antenna B, but with a 1/4 wave top section, so this should be OK, as the very top will be Hi impedance and 1/4 wave down where it is connected to the short circuit section it will be low impedance. However as with antenna B the chort circuit seems to be at a high impedance point, which doesn’t seem to be correct.
Antenna D is pretty much the same as antenna C, even though it doesn’t have a short circuit at the top coax section, it still has the 1/4 wave stub, which will make it seem like a low impedance point.
If you work down from the top, but transposing the impedances it doesn’t seem to be correct either.
So none of these would seem to work as intended, or do they ?
By stacking the 1/2 wave sections in this way it is as if you are connecting several 1/2 wave dipoles in parallel with each other. If you do this without some form of impedance compensation, you will not have a 50 ohm feed point, but something much lower, depending upon the actual number of dipoles you are using.
In order to bring the feed point impedance back up to 50 ohms you have to apply some form of impedance transformation, and I think this is what the mismatched end section may somehow be achieving.
In addition to this as I said before the speed of RF flowing on the outer of the coax (jn free space) and forming the actual antenna sections, will be a lot faster than the RF travelling through the coax on the inside, due to the dielectric constant of the insulation, and it is this that messes up the phase relationship of the ‘dipole’ sections.
Owen Duffy does a much better job of explaining these sorts of problems than I can.
https://owenduffy.net/blog/?p=5065
https://owenduffy.net/blog/?p=157
https://owenduffy.net/blog/?p=16560
The bottom line is that I don’t trust any coaxial colinear designs that can be found on the internet, and even if the design could work, the construction tollerances need to be pretty good for it to have the correct phasing and vertical radiation pattern.
I have seen commercial designs that seem to work, properly but they don’t use standard coax. They either use sections etched on printed circuit board with specific dielectric properties, or use air spaced coax sections.
Basically for them to work, several things need to be correct.
-
The 1/4 wave decoupling sleeve need to be large enough to work properly.
-
The velocity factor of the coax needs to be 1 (air spaced).
-
The feed points need to be connected in parallel at high impedance nodes.
-
There need to be enough sections connected in parallel to bring down the feed point impedance to be something close to 50 ohms, or there needs to be an additional impedance matching section, to convert whatever feed point impedance exists back to 50 ohms.
-
Good quality low loss materials need to be used.
The first Owen Duffy page I linked to, has a photo of a commercial 3GHz printed circuit board version (figure 6) that seems to have been designed properly.
Many people think they have got a working coaxial colinear, but there is a lot that can go wrong, and as is the case with all home built antennas I’d always suggest making a comparison against a much simpler reference antenna if at all possible.
Regards,
Martin
1 Like
Yes, I am fully aware of that fact, but the whip is open ended, no DC short.
As an additional precaution, I have retained white color insulation of the whip wire. The small portion (few mm) visible at point where whip wire enters the F-female was intentionally kept visible for taking the photograph to make it clear to viewer that it is a wire with insulation. After taking photos, it was pushed to fully insert in the hole, leaving only the white insulation exposed.
1 Like
OFF-TOPIC
@G8JNJ
There was a new type of design I have developed & posted here in 2020, but as you have not visited here for last 8 years, most probably you have not seen it. I have named t “V-Stub Quick Spider” The word “Quick” in it’s name is to show that it can be made quickly as it does not requite any connectors or soldering, and is solely made of coax.
Below are some screenshots of V-Stub Quick Spider. I would love to have your comments when you have some time to spare.
NOTE-1: The open mouth of V in the diagram below is shown 45 mm wide. However, subsequently I discovered that increasing it from 45 mm to 50mm or 55mm improves performance.
NOTE-2: A simple A/B comparison with Flightaware’s 1090MHz 5.5 dBi antenna, to give an indication as to the effectiveness of V-Stub antenna, was done and posted here:
Comparison of V-Stub QuickSpider with Bench Mark Antenna (FA Antenna)
Geometry
Radiation Pattern in Vertical Plane
3D Pattern
Current+Phase
Built With RG6 Coax Core Wires
On top, the small sliding piece of wire (with core insulation to hold the sliding wire in place) is for conveniently adjusting the length of top wire for best MESSAGE RATE (if no swr meter is available) or minimum swr if swr meter is available.
I have just taken a look at an old six section commercial UHF colinear, that is basically the same as the ‘Franklin’ design you have just illustrated.
The 1/2 wave phasing section is folded back on itself and rolled around a Tufnol Rod, forming a non inductive transmission line. By rolling it like this, you achieve a more compact section, radiation from the phasing line is reduced and as a result you obtain a more symmetrical horizontal radiation pattern.
The ends of the phasing section are 1/4 wave apart, and the radiating sections are 1/2 wave long. The top section is slightly less than 1/2 wave and has a sliding outer aluminium tube that can be used to extend or shorten the top section in order to ‘tune’ it for the best match at the required frequency.
By using 10mm diameter aluminium tube for the radiating sections, the bandwidth is slightly wider than would be achieved with a thinner conductor.
You could also use a Coke can type decoupling sleeve in place of the 1/4 wave radials, as this would help improve the feed line isolation and may also produce a more symmetrical / predictable horizontal radiation pattern.
Just some thoughts that may help to further optimise designs.
Regards,
Martin
2 Likes
Thank you Martin for your valuable comments.
I will try to redesign the V-stub to incorporate your suggestions.
By folding back the two 1/4 wave sections (the two arms of the V) on themselves and rolling these around a Tufnol, or a non-metallic tube, will enable antenna to be housed in a PVC pipe of say 1/2 inch or 3/4 inch dia.
I will next try your suggestion to use a Coke can type decoupling sleeve in place of the 1/4 wave radials to improve the feed line isolation.
Thanks again for your analysis of my V design, and suggestions to improve it further.
Regards,
abcd567
Just be careful about the material you use, as some dielectrics could ‘load’ the transmission line phasing section requiring it to be slightly shorter.
If you used stiff wire you maybe be able to wind it on a former and then remove it.
The decoupling sleeve could have a smaller diameter than a Coke can, maybe Red Bull, to give it wings , but that was what I had to hand. As long as it is 5 to 10 times the diameter of the coax it should be fine.
Let us know what you find.
Regards,
Martin
Thank you Martin for the useful tip. This is what I had in mind.
In current design, the entire whip and V is made of single piece of RG6 core wire, which is copper-coated steel, which is stiff, so it is easy to follow your tip.
I have planned to wrap the V around a pencil, then slip out the pencil. This will give me a one-piece whip whithout need for jointing pieces by screws or soldering, and without former.
Yes, you are right. I will try a smaller dia can such as Red Bull.
Below are 3 Cantennas I made many years ago with different can diameters. The 68mm dia regular can gave the best performance, the 54mm dia can performed somewhat less than the regular 68mm can, and the 20mm dia water pipe performed poor.
Click on image below to see larger size
Click on image below to see larger size
There is a problem with the Cantenna shown in the previous post.
The decoupling sleeve needs to be an odd multiple of 1/4 wave along the outer edge from the feed point to the rim of the can.
In the example shown it has a 1/4 wave radius top section (shown as ground plane) and a further 1/4 wave length of can, making it 1/2 wave in total.
1/4 waves are high impedance at one end (usually the bit that s not connected to anything) and low at the other (usually the point required to be a low impedance).
1/2 wave are the same impedance at each end, so if one end is in the air (high impedance) the other end will seem to be the same.
For the Cantenna to work as shown in the picture, the can length would need to be 1/2 wave, making the total radius + length = 3/4 wave.
Regards,
Martin
Thank you Martin for pointing out the error in length of the can.
However I would like to point out that the top section’s diameter (and not the radius) is 1/4 wave, so it’s radius is 1/8 wave.
In this case to make (radius + length = odd multiple of 1/4 wave), the can length should EITHER be 1/8 wave (making total 1/8+1/8=1/4 wave) OR 5/8 wave (making total 1/8+5/8=6/8=3/4 wave).
I was busy so could not make a new antenna with wrapped V.
I will make it in a day or two and perform comparison with unwrapped V antenna, and post the results.
Regards,
abcd567