OK, so Please ‘Splain’ and thanks for your patience.
If the Vf is based on the properties of A) flawless known braid and B) known center conductor, and the C) dielectric properties of the insulator to come up with a velocity factor that is used to determine the length of the elements of a CoCo, what happens when one of the above are not true ?
If I take the copper off of the steel, it “works” but only taking the fact that it is steel into the calcs, which I have to believe would change something in the conductance, impedance etc. Copper-steel-copper multiple times over has to make a difference. We are talking about less than 2 microns of copper on coax center conductor. A knife edge cut around the diameter of the wire is many times that and given the skin effect I think it is a possible defect in a given element.
If the braid is corroded in a section it isnt going to work the same, no ?
Sorry for the mindless shot-gun questions, but I have a very hard time accepting that two antennas, made of the same basic materials using the same measurements ( within the limits of a utility knife blade-width) one is junk and the other is stellar.
“Forensics-wise something is different between them, and it may not always be “length” of the elements.”, he said, preparing for the rotten tomatoes…
[At 1090Mhz the depth of current/ signal on copper is ~1.97 microns or .0075 thousands of an inch]
There is a reason that copper clad steel is used in coax…it works better.
The velocity factor almost solely depends on the insulation material.
PE (polyethylene) 0.66
FPE (Foamed Polyethylene) depends on degree of foaming used by manufacturer. Most FPE insulated coax have VF in the range 0.80 ~ 0.86
The manufacturer measures the velocity factor by tests on several manufactured lots and documents it in his specs. The user does not have to, and cannot work it out. He has to use the Vf specified by manufacturer in cable specs.
Despite designing my own antennas I’m not an antenna expert, nobody is, it’s a black art, guided by experience and a few rule of thumb calculations. I reckon for every successful antenna there’s a few dozen in the bin!
Many ‘house brand’ cables come from a number of suppliers, so detailed datasheets are not available.
Using a noise source like the cheap BG7TBL Noise Source. (a few people here have them) and a generic RTL dongle to make a Spectrum Analyer, you can analyse the performance of a stub filter to determine the Vf.
Where I was headed in my thinking is that datasheets ( if available ) are fine, but they are the “as-designed” specs for the coax, not the “as-built” whats in your hand after its been coiled, bent, cut, crushed, had the center conductor damaged, shield splayed out etc. particularly in regard to why some Cocos work and some do not.
It doesn’t seem plausible to me that so many intelligent people try and fail based solely on applying the Vf of the coax to the element measurements yet it seems 99% of the discussions are focused on exactly that and the other 1% is metering to make sure the elements aren’t accidentally shorted.
What I am suggesting is that perhaps there are other factors that are impacting the builds that are just as important and less well documented ?
Perhaps just me talking out loud here, but the WHY of it bothers me. I built 4, one is Da Bomb, 3 work no better than a 1/4 spider.
Related to not having test equipment, yesterday I did a little experiment to dig a little more at whether a damaged coax center conductor could be a factor. This was a quick and dirty effort involving a 1/4 spider so really doesn’t prove much in respect to a Coco, but all I had time for.
I scored a replacement 18 gauge copper coated top element to simulate the cuts made in constructing a Coco and compared that to the existing unblemished one, making sure I went through the thin copper cladding, perhaps 6 or 8 heavy cuts in a 68.8mm element. As you might guess, it made little to no difference I could observe. ( Even rather severe bends and distortions of the element had little impact on reception)
And that is the key: that I could observe. What I think I learned learned was Using the dump1090 and VRS or Piaware software to do a test like that masks a lot of things because they will still do their best to bring in planes and positions until a signal is too weak or corrupted to decode. The software is too good to be easily defeated by minor reception issues. SO rather than planes, positions and messages, if I wanted to directly see antenna A vs Antenna B I should be looking at comparable signal strengths for the same aircraft or be running RFscan/ HDSDR etc. to directly observe the incoming signal.
So @abcd567 and @geckoVN Have either of you made a Coco that worked to your satisfaction ?
Edit: Another point where my brain cells are slowing remembering things from long ago prior learning: Digital signals, not analog. Works perfectly- until quite suddenly, it doesn’t.
. No, not a single one out of over one dozen Cocos I have made. All performed worst than a Spider.
.
Using following method I could somewhat improve one of these failed ones:
Below are the simulation results for a CoCo (8-Elements, made of coax VF=0.86).
These results show how a 1 or 2 mm error in element length can grossly affect SWR and Radiation pattern.
Gain is relatively less affected by 1 or 2 mm error.
The error in element length can either due to using wrong VF, or wrong cutting or a combination of both.
In this simulation the a VF of 0.86 was used, giving element length 118 mm.
So +6mm or -6mm element length on the model results in SWR ranging from 1.22 to about 6 and gains ranging from a high of 11.39 to a low of about 5.7 ? (that’s quite a range)
I see the pattern changes- but in a simplistic sense it appears the worst case high Swr/ malformed/tuned gain of 5.7 still looks to be much higher than the spider would achieve. And yet the realized signal reception is that of the spider or worse…
Does the signal to noise ratio go crazy or something and add another aspect to consider ? So far all I have is endless possible points of failure- sorta glad I didnt go into RF engineering.
Speaking realistically here, I think that maintaining a ± 2mm measurement is easily achievable for a careful DIYer. So…a little more than 3.2 - 3.3 on the SWR and, say 11.2 to 9.5 on Dbi gain ( looking strictly at the model charts.
)
