best antenna

[abcd567]

@gregy:
@Marsmaensch:

4-ELEMENT COCO
After correcting the Dielectric Constant error in my .nec input files. I ran simulation of revised file for 4-ELEMENT CoCo, and results are astonishing:
SWR is 5.03 at 114 mm length calculated by standard formula 1/2 Wavelength x VF.
Minimum SWR (2.04) is at 102mm element length.




Rev-SWR - Variable element length CoCo (4 elements) from RG6 VF 83 percent.PNG . Rev-GAIN - Variable element length CoCo (4 elements) from RG6 VF 83 percent.PNG . Rev-Gain Swr Pattern - 114mm element length CoCo (4 elements) from RG6 VF 83 percent.PNG

[abcd567]

@gregy:
@Marsmaensch:

8-ELEMENT COCO
After correcting the Dielectric Constant error in my .nec input files. I ran simulation of revised file for 8-ELEMENT CoCo, and results are astonishing:
SWR is 3.4 at 114 mm length calculated by standard formula 1/2 Wavelength x VF.
Minimum SWR (1.18) is at 108mm element length.




Rev-SWR - Variable element length CoCo (8 elements) from RG6 VF 83 percent.PNG . Rev-GAIN - Variable element length CoCo (8 elements) from RG6 VF 83 percent.PNG . Rev-Gain Swr Pattern - 114mm element length CoCo (8 elements) from RG6 VF 83 percent.PNG

[abcd567]

......I will VNA sweep it and see if i can find clues .... .

........and drop by a friend with an antenna analyzer to validate the calculations.

I am going to purchase This one when I get this one

[gregy]

I am going to purchase This one when I get this one

yes a friend with a VNA is a perfect friend!! partic if its a agilent, R&S or better!
mine is very modest ... VNWA3 .... £400 ... but does the job for what i need

[abcd567]

yes a friend with a VNA is a perfect friend!! partic if its a agilent, R&S or better!
mine is very modest ... VNWA3 .... £400 ... but does the job for what i need

Yes, good VNA is on my wish list, but not the professional one for $ 150,000. Yours £400 seems reasonably priced for a (rich ) hobbyist.
The one I linked, its name was originally HP (Hewlett Packard), then changed to Agilent, and now again recently changed to Keysight.

[abcd567]

@Trigger
Hi Dave!
Did you solve rain water/radome problem for your Franklin?
See photos below how I solved it

BEFORE I ATE THE RUSKS




AFTER I ATE THEM ALL

[gregy]

but did you attempt to determine the dielectric constant of bread rusks!
"rusk tuning"?

[abcd567]

...........i found that the shape and gain vary significantly with change in vf. is it possible for you to do a sweep in vf on your coco model?............
Benjamin

Yes it is possible to do a sweep with VF in my model.
However, if you want to use coax other than RG6, then changing only VF in the RG6 model is not enough, as dia of insulation & dia of shield also change when type of cable is changed. This means re-modelling of the shield. The 4nec2 does not have provision for metallic sheet, and sheets are simulated by mesh of wires. changing the dia of shield requires complete re-modelling of the wire mesh cylinder used as shield.

Please let me know specifically which cable/s you are interested in, so that I can re-model the nec input file accordingly, one separate file for each type of cable.

[abcd567]

@gregy:
@Marsmaensch:

4-ELEMENT COCO RG58 PE 0.66 VF
I ran simulation of new file for 4-ELEMENT COCO RG58 PE 0.66 VF CoCo, and results are:
SWR is 5.13 at 91 mm length calculated by standard formula 1/2 Wavelength x VF.
Minimum SWR (3.12) is at 98mm element length.

Download .nec file for this simulation:Rev-variable length CoCo 4 element RG58 PE insulated.zip

Update Oct 15, 2014: Core insulation set to "Polyethylene" and .nec file above replaced by new one




Rev-SWR - Variable element length CoCo (4 elements) from RG58 VF 66 percent.PNG . Rev-GAIN - Variable element length CoCo (4 elements) from RG58 VF 66 percent.PNG . Gain Swr Pattern - 118mm element length CoCo (4 elements) from RG58 VF 66 percent.PNG

[gregy]

smooth shield vs heliax shield
noting your earlier post "formula for smooth outer shield", and
- with heliax cable, the electricsl length of inner and outer are now different
- the fact that coco uses transposed cable sections
... this raises in my mind the dilemma of
how to determine optimum element physical length when using heliax cable sections ... indeed it seems like it may be a compromise ?
the outer sheild is used in each case as the "radiating" element, but the core length also plays equally important part and needs to also be correct length (the core and shield each have different VF .. which is corrected by your revised formula,
but not sure how to adjust for the effective different electrical length of core vs shield )

thinking aloud here ......
the physical length of shield and inner core inside dielectric are of course identical by definition, however the core protrudes beyond end of shield into free space
between the elements (the length of which is determined by the "gap" between elements)
hence its possible? to equalise to some extent the electrical length of shield vs core (given the heliax shield adds extra E length)
this additional core length across the gap is in free space (VF 1)
.. so perhaps its possible to compensate to some extent for the heliax longer E length
by adjusting the gap between elements? .. effectively lengthening only the core?
(in my current coco i minimised the gap to less than 1mm .... but perhaps this is
actually working against optimum outcome)

thoughts?

[abcd567]

@gregy:
@1090 Mhz:
Hi gregy!
Couple of months ago forum member "1090 Mhz" made a very successful coco from Heliax. He has posted some details & photos. He may give you some useful advise & tips.

[abcd567]

Making a coco is like treating cancer: some lucky ones survive, while many more die. Both are still dark art, not fully understood, and not fully under control.

[gregy]

@gregy:
@1090 Mhz:
Hi gregy!
Couple of months ago forum member "1090 Mhz" made a very successful coco from Heliax. He has posted some details & photos. He may give you some useful advise & tips.

yes i saw his details and "results" ... and it appeared he used the "standard formula" (his cable was smaller size
and diff VF) but otherwise mine was comparable

[gregy]

yes agree, there are too many variables, and by the nature of multi element,
the errors multiply fast!

[abcd567]

Like "Cancer Research Foundation" , we should form a "CoCo Research Foundation"