Canadian TV, Computing and Home Theatre Forums banner

Modeling Flower Pot Antenna for ham radio with 4nec2?

11K views 17 replies 2 participants last post by  holl_ands 
#1 ·
hi people,
i'm trying to simulate a halfwave dipole a.k.a flowerpot antenna for ham frequency (146 mhz nominal). it using rg58 coaxial. but i seem lost as it keep error on me
aprreciate if anyone can share if anyone do simulate a rg58 coax antenna

tq de 9w2xne
 
#2 ·
VK2ZOI's "Flower Pot" Antenna matches impedances using certain lengths of Sleeved and Unsleeved Coax:
VK2ZOI

I don't think that 4nec2 (with NEC2 Engine) is capable of simulating this sort of Feedline, with different currents flowing on the inside and outside of the Shield Wires (similar problem as Coax Balun).
 
#5 · (Edited)
It appears that you are trying to model ACTUAL RG-59 Coax by wrapping a central wire with a wire model of the shield.....but the Diameter of RG-59 would be 0.25-in vice your 1.6-in. BTW: An equivalent AIR DIELECTRIC Coax would provide 75-ohm Impedance with a MUCH SMALLER Diameter...and the LENGTH would increase since Air Velocity Factor = 100%:
http://www.pasternack.com/t-calculator-coax-cutoff.aspx

Looking at the fol. article re VK2ZOI's Half-Wave Flower Pot Antenna, note that there is a SELF-RESONANT CHOKE COIL at the base of the Antenna, where he provides a Table of Number of Loops vs PVC Tube Diameter. So it appears important to include this important element in the model.....somehow....
http://vk2zoi.com/articles/half-wave-flower-pot/

Note that the Flower Pot Antenna and the J-Pole BOTH use a Choke Coil at the base of the Antenna....however in the case of the J-Pole, it is NOT necessarily a RESONANT part of the structure, instead acting as a true CHOKE COIL, to suppress Common Mode Currents from descending down the Feedline. Some people claim the Choke Coil is Optional in the J-Pole....but it operates very differently in the Flower Pot Antenna and is probably NOT Optional:
Does my J-Pole antenna need the choke balun? | KB9VBR J-Pole Antennas

The Colinear J-Pole Antenna is similar to the Super-J-Pole, except using a PHASING COIL (NOT a Common Mode CHOKE) between the upper and lower resonant elements:
Improving the Super J-Pole Antenna
Jpole vs. Coaxial Dipole EZNEC Shootout
Note that the (presumably 1/4-wavelength) PHASING COIL can be replaced by unwinding it into a simple three wire structure (fig 1a)....a 1/4- Wavelength SHORTED STUB.
Hopefully, when wrapped around the PVC Tube, the proximity of additional Dielectric Material doesn't "detune" the structure very much....which is one more reason why most NARROW BAND Antennas need final tweaking with an SWR Meter....
 
#6 ·
First of all, it RG58 coaxial cable. I already homebrew couple flowerpot antenna and tested all using swr meter/analyzer. performance was good which under 1.5 SWR with decent RX/TX (depends on propagation and location) as mention by designer and other homebrewer.

Some how it quite Broadband(wide range) 138MHZ-156MHZ which below 1.5 SWR.
After awhile testing on 4nec2, it seem constant to have same radiation/gain depending on distance of 3 Director element. It do need choke coil but for now, i can add it using stub match options.

Still got more to learn as still a newbie to this software not very familiar at to use it and to improved it

thanks frenz




latest screenshot with added stub match
 
#7 · (Edited)
Yup, I've been working with TV Antennas (hence 75-ohm RG-59) way too much....
RG-58 is 50-ohm, VP=66% with 0.195-in PVC Jacket O.D., so SHIELD O.D. is a bit LESS than that. I didn't see any specific specs...my Calipers measured Shield O.D.=0.135-in for RG-58 marked as being REAL MIL-C-17/28 type....

I think the 1/4-Wavelength Stub should be located at the BASE of the Antenna....rather than simulated via an Output Matching Network on the Transmitter end of the Feedline (stub match option in 4nec2).
PS: It helps if you post the 4nec2 File for each version....

As I mentioned earlier, 4nec2 isn't particularly suited to analyzing the Flower Pot Antenna, so alternative Analysis Programs are advised....unfortunately, I don't have a feel for how BIG a difference there will be....for SURE the Whip in the Flower Pot Antenna would have a MUCH higher Unbalance than the typically analyzed direct connection of Coax to the (nearly "balanced") two Elements in a simple "Stick" Dipole forming the Active Element in a Yagi Antenna (see below).

NEC2 Engine (used in 4nec2, EZNEC, et.al.) is a simulation based on WIRE-MODELS, where currents flow THRU the wire segments. OTOH, Common Mode Currents flow on the inner and outer SURFACES of the Coax's Shield....which CAN NOT be simulated by the NEC2 Engine. When these Currents are Different (which WILL happen in the Flower Pot Antenna), there will be a Common Mode Current Unbalance problem...which can affect the Antenna Pattern of Directional Antennas, NOISE pickup in Receive Antennas due to imbalance or conversely unintended Radiation from the COAX (esp. INTO the Ham Shack) when Transmitting.....and also the Impedance Mismatch.

I have a strong suspicion that the Common Mode Currents flowing on the OUTSIDE of the 1/4-wavelength Shielded Coax work with the currents flowing on the 1/4-Wavelength UN-Shielded portion to approximate a 1/2-wavelength Vertical Dipole Antenna:
http://imageevent.com/holl_ands/omni/fmverticalstickdipole

More re effects of Common Mode Current Unbalance:
http://www.eznec.com/Amateur/Articles/Baluns.pdf
http://www.w8ji.com/common_mode_current.htm
Common Mode Curents in Antennas [Note VP is DIFFERENT for Inner & Outer Shield.]
Common-Mode Noise
http://www.audiosystemsgroup.com/RFI-Ham.pdf
http://www.yccc.org/Articles/W1HIS/CommonModeChokesW1HIS2006Apr06.pdf
 
#8 ·
yes sir, the flowerpot antenna or similar coax antenna is in fact a halfwave dipole which is 1/2 lambda. each element shielded and unshielded coax (center conductor) is about 1/4 lambda.

last screenshot is the same earlier but added stub match which i cannot save it to .nec format but .out

What i saw during testing with parasitic element a.k.a director, normally without it antenna will be around 2dbi. but it gain more dbi differently when distance of director from mast increase. tested with distance (32cm,51cm,84cm). i'm not sure how it perform on 3 way director. I inspired by lindenblad antenna.
 
#9 ·
#11 ·
I've been generating a fairly extensive 4nec2 Analysis of True OMNI and Quasi-Omni (usually Bi-Directional) Antennas for FM Band:
Omni & Quasi-Omni Directional Antennas - ImageEvent
Note that when analyzed in DETAIL, many of the so-called OMNI Antennas are actually Quadi-Omni's, with Bi-Directional performance when just Horizontal Gain is calculated and may only become OMNI when TOTAL (H+V) Gain is calculated.

I'm interested in doing the modeling to add the Original Lindenblad as well as the Parasitic Lindenblad (with traditional Vertical Dipole) to the list.....
This shouldn't take very long....found EZNEC version (lindnbld.ez) of the Original Lindenblad online:
Index of /EZNEC_FILES
Note that Cebik also did an analysis:
June 2008 Issue
 
#14 · (Edited)
2-m LINDENBLAD PARASITIC ARRAY + ACTIVE VERTICAL DIPOLE:

Analysis of 1/2-wavelength Vertical Dipole surrounded by a PASSIVE Lindenblad Parasitic Array...if you have RHCP signals, this is a VERY GOOD, TRUE OMNI Antenna...but it is only so-so for Horizontally Polarized signals:
2m Lindenblad Parasitic + Vert. Dipole
Recall that Raw Gain is about 2.15 dBi for a simple Stick Vertical Dipole, which is a TRUE OMNI for RHCP=LHCP as well as Vertically Polarized Signals (and nearly zero Response to Horizontally Polarized signals):
http://imageevent.com/holl_ands/amateurradioantennas/2mstickverticaldipole

The Lindenblad PARASITIC Array (with Active Vertical Dipole added in the Center) is SIMILAR to the Lindenblad Active Array (see analysis in next post). HOWEVER, note that the POSITIVE Angle of Diagonals in the Lindenblad Active Array must be changed to NEGATIVE Angles in the Parasitic Array to ensure that it primarily responds to RHCP signals.

The Length of the Vertical Dipole was adjusted so that SWR (50-ohm) was minimum on the 145 MHz Design Frequency while the Lengths of the Diagonals in the Lindenblad Array were adjusted so that LHCP (Left Hand Circular Polarization) was minimum on 145 MHz (there is some interaction). When RHCP signals are Reflected off of more or less Flat Objects, most of the time the Rotation is Reversed, changing them into LHCP (Right Hand Circular Polarized) signals. So, when RHCP signals are Transmitted, it is desirable to minimize the response to LHCP signals from ALL Directions (Rear, Sides and Ground Bounce from the Front) to minimize Multipath Interference.

The Antenna is a TRUE OMNI for either RHCP signals, Vertically Polarized signals and (with less Gain), Horizontally Polarized signals (so the Diagonals are coupling the Horizontal Signals into the Vertical Dipole). There is moderate suppression of LHCP across all Azimuths and Frequencies. (Total) Gain is significantly improved on 145 MHz (but only on SOME Azimuths), losing less than 1 dB on either end of the 2m Band.

Note that for this "first look" analysis, I assumed that the Angle of the Diagonals was 45-deg and they were Separated from the Center of the Antenna by 1/4-wavelength (below Active Array looks closer together). So there may be room for some improvements and different tradeoffs.....for example, decreasing Separation improves suppression of LHCP, but at the expense of lower Gain and increased SWR....and vice versa....

BTW: I have yet to find a way to get nikiml's Python Optimization Scripts to Optimize an OMNI Antenna for equal Gain on all Azimuths...either with or without the --omni option....so I'm forced to use various "manual" methods.....











 
#15 · (Edited)
2-m LINDENBLAD OMNI ARRAY:

The Lindenblad Omni Array (from the 1940's) consists of four Diagonal Elements, where they are ALL Active. The fol. Analysis assumes a "PERFECT" Lossless Feedline System by assigning an Active SOURCE on each of the four Elements. A REAL Antenna would split a single Twin-Lead into four Paralleled Twin-Leads, one for each Diagonal Element, which would result in some degradation to performance....this configuration would require a separate analysis....that I may or may not do.....

The Antenna is a TRUE OMNI for either RHCP signals, Vertically Polarized signals and with less Gain, Horizontally Polarized signals. There is moderate suppression of LHCP across all Azimuths and Frequencies. (Total) Gain is about 1 dB lower than the preceding Lindenblad Parasitic Array + Vertical Dipole....which has a MUCH simpler Feedline.

Note that I searched for and downloaded an EZNEC File for the Lindenblad Array analyzed here. I did NOT change ANY of the Dimensions, including assuming the use of 46.91-deg Angles in the Diagonals. An OPTIMIZATION should be run to find the "best" Dimensions.











 
#18 ·
Yes, the above Lindenblad Parasitic Array is "better" than Original Lindenblad ACTIVE Array....wrt to OMNI-Directional performance for either RHCP (Right-Hand Circular Polarized) or Vertically Polarized signals....where "better" means somewhat higher Gain in SOME (not all) Directions and slightly higher suppression of undesired LHCP response to Multipath Reflections. And only requires a SINGLE connection to the Central Vertical Dipole...rather than a more complex 1-to-4-Way Interconnect to all four Active Diagonal Elements (which I did NOT model...incl. resultant INPUT SWR, but would probably cause even further degradation).

BTW: These are both fairly NARROWBAND Antennas, intended for the 2-m Amateur Radio Band. Note degradation of Parasitic Array's SWR at the band edges.....
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top