Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: NEC for modeling OCF?  (Read 6670 times)
NZ5E
Member

Posts: 75




Ignore
« on: August 04, 2010, 05:45:49 AM »

I need to add 30, 60, 80, and 160 meters to my antenna farm and I am leaning towards the OCF dipole. I already have rotatable antennas that cover 6 through 20 and 40 meters.

The OCF dipole designs I have seen are compromised for multiband operation in an attempt to cover as many bands as possible. To fit my particular situation, I want to design an OCF dipole favoring 160 and 80 meters to be oriented with the wires running NE/SW mounted on a standoff at 94 feet on my tower. I want to design another OCF dipole favoring 60 and 30 meters with the wires running NW/SE mounted on a standoff at 97 feet on my tower.  Both antennas would be in an inverted-v configuration with a 140 degree included angle.

I have been utilizing the search engine to research different spinoffs of the NEC software and I am overwhelmed by all the choices. Please advise me as to which cut down version I should use to evaluate fairly straightforward antennas like the OCF dipole arrays described above. Free software would be great but I also don't mind paying for the software.
Logged
K0BG
Member

Posts: 9867


WWW

Ignore
« Reply #1 on: August 04, 2010, 05:56:45 AM »

Why OCF? You can model one, but when you do, don't forget to include the feed line. And, almost no matter what you do, you'll end up with a fair amount of common mode on the feed line, especially so if you use a voltage balun at the feed point. While some folks get them to play well, most don't!

One more thing. You might want to visit http://www.w8ji.com, and do a search for OFC.
Logged

K5LXP
Member

Posts: 4475


WWW

Ignore
« Reply #2 on: August 04, 2010, 07:37:58 AM »

I have EZNEC 5+ and have been playing lately with modeling OCF's.  My goal was to understand better feedpoint placement and Z over frequency of OCF's.

However, I haven't had much luck getting my models to match purported real world results.  I started off with a QST article by VE2CV on OCF multielement dipoles and no matter how I tweak the model, it won't match the published results.  There is likely a factor I'm not accomodating that's omitted from the article but it would be nice to reconcile the difference before I go much further.  Could be feedline/common mode but I wonder too if the published data was location specific (near field influences).  Even simpler single wire OCF designs (Carolina Windom, et al) I can't seem to tweak to meet published claims.  So I have to wonder how much of my problem modeling is the model, or the marketing.

I'm almost at a point where I'll put my own up and characterize that, and go from there.  Perhaps the vagaries of baluns, location, feedlines and other intangibles might explain what K0BG mentioned in terms of the success and failure of these antennas.  Would be nice to come up with a good "formula" to improve the chances of success.  It seems so simple in concept.

Anyway, if you want to play more with the modeling we could work on it offline if you like.  Any other input welcome.


Mark K5LXP
Albuquerque, NM


Logged
WX7G
Member

Posts: 5985




Ignore
« Reply #3 on: August 04, 2010, 09:24:00 AM »

The free 20-segment version of EZNEC does not have enough current segments to model this antenna above the 2nd band. You want to use at least 10 current segments per half wavelength.

EZNEC works well as does NEC-WIN PLUS. Both are used in the ARRL antenna modeling course. NEC-WIN PLUS does not run (at least for me) on the VISTA operating system so I have switched to EZNEC, NEC-2.
Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #4 on: August 04, 2010, 11:12:57 AM »

Why OCF?...

The following link shows an overhead view of the tower location with the yellow lines representing the guy wires and the red lines representing the best directions for tying off wire antennas to trees:

http://i581.photobucket.com/albums/ss252/N1048D/Amateur%20Radio/Shop.jpg

The longer span for tie off points, because of relatively flat slopes and not requiring any tree trimming, is the NE/SW axis which would put that antenna broadside to the NW/SE.  But, since the 80/160 meter antenna would be much less than one half wavelength above the ground, it would be relatively omnidirectional and it would not have good low angle radiation for DX anyway.  The ground elevation slopes off steeply in the NW/SE direction and there are numerous trees and tree limbs to deal with making the NW/SE direction more suited for the shorter 30/60 antenna and probably more desirable for DX being broadside to the NE/SW.

Given the details just described, an OCF seems to be the ideal antenna for my situation.  The OCF works well on even harmonics with 80/160 being ideal candidates and I would think 30/60 being close enough.  The transformation through a 6:1 balun should make the antenna more broadbanded than a center fed dipole.  Compared to multiple dipoles tied to the same feed, the OCF only requires one tie off point at each end and hopefully would not require tuning when mounted on the tower.

I am basing my information on what I read on the internet since I have not ever actually operated an OCF.  If I am missing facts that adversly effect the real world operation of the OCF, please advise.

One detail is that the guy wires above and below the level of the apex of these antennas will be non-conducting.  Another detail is that this is an internet remote station and I want to automate the operation as much as possible.


Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #5 on: August 04, 2010, 11:16:37 AM »

...You can model one, but when you do, don't forget to include the feed line. And, almost no matter what you do, you'll end up with a fair amount of common mode on the feed line, especially so if you use a voltage balun at the feed point. While some folks get them to play well, most don't!...

Why would it not work to place a choke balun immediately after the voltage transformation balun?
Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #6 on: August 04, 2010, 11:31:37 AM »

...One more thing. You might want to visit http://www.w8ji.com, and do a search for OFC.

W8JI states in the article: "Note: This feedpoint offset gives up 30 meters. This is the primary reason I avoided the 1/3-2/3 feedpoint position and used the 80% feedpoint position."

His design criteria was at least partially based on including as many bands as possible.  I am hoping to design a better performing antenna on my bands of interest (80/160 and 30/60) since I am only trying to optimize on two bands, not several.

Please let me state that I am not arguing with you, I am actually asking questions through my replies.  I have read several of your articles and have a lot of respect for your work!
Logged
K5LXP
Member

Posts: 4475


WWW

Ignore
« Reply #7 on: August 04, 2010, 11:41:38 AM »

> I am hoping to design a better performing antenna on my bands of interest

"Performance" implies optimizing your pattern, which is directly tied to the lengths of the elements (and any influence from the feeder).

So what "performance" are you looking for? 


Mark K5LXP
Albuquerque, NM


Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #8 on: August 04, 2010, 11:51:52 AM »

...My goal was to understand better feedpoint placement and Z over frequency of OCF's...

Exactly why I want to do the analysis myself.

...Perhaps the vagaries of baluns, location, feedlines and other intangibles might explain what K0BG mentioned in terms of the success and failure of these antennas...

I feel for the manufacturers that are trying to market antennas such as the OCF in an attempt to fit all the mounting situations of their customers.  Don't you think that the reason most yagis perform so much better than wire antennas is not that they have more gain, but that they are usually mounted up high and in the clear?

... Would be nice to come up with a good "formula" to improve the chances of success.  It seems so simple in concept...

Could not agree more.

...Anyway, if you want to play more with the modeling we could work on it offline if you like...

I might want to send you one of my designs to check.  I would like to have that warm and fuzzy feeling that I am doing things correctly.
Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #9 on: August 04, 2010, 12:14:55 PM »

..."Performance" implies optimizing your pattern, which is directly tied to the lengths of the elements (and any influence from the feeder).

So what "performance" are you looking for?...

I was referring to overall performance related to pattern, bandwidth, etc.  I want to design the antennas so that the resonance points fall within my bands of interest.  Since this is an internet remote station, it would be great if an antenna tuner was not required beyond possibly the internal tuner in the Kenwood TS-480SAT.  As long as a reasonable VSWR was obtained to keep the rig happy, I would also like to not use a tuner to reduce losses in the system.

I considered using 80 and 160 meter dipoles tied to the same feed with an SGC-235 automatic tuner at the feedpoint to allow full band operation on 80 and 160 and operation on 30 and 60 as odd multiples of 80 and 160.  This setup should result in minimum loss in the tuner.

It would be difficult for me to tie off a full size 160 meter dipole running in the NW/SE direction, but two antennas at right angles to each other would be neat since I could switch antennas to fill in the nulls of the patterns.  That would be the next best thing to being able to rotate that antennas.
Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #10 on: August 04, 2010, 12:26:51 PM »

The free 20-segment version of EZNEC does not have enough current segments to model this antenna above the 2nd band. You want to use at least 10 current segments per half wavelength.

EZNEC works well as does NEC-WIN PLUS. Both are used in the ARRL antenna modeling course. NEC-WIN PLUS does not run (at least for me) on the VISTA operating system so I have switched to EZNEC, NEC-2.

I have noticed that EZNEC seems to be very popular and I have downloaded the free version to play with.  I played with a DOS version of miniNEC over 20 years ago.  I was thinking that 20 segments seemed very limited.  I was always assuming that I would probably purchase a workable copy, just wanted to make sure which one to buy.

I am still scratching out a living and have limited time to play with this.  I was thinking that I could model the antennas at my given height with average soil conductivity and not bother with modeling the tower, the other metallic guy wires beyond the non-metallic ones, and the feedline.  Would this not probably get me close enough to have a reasonable assurance of resonance and performance when mounted on the tower?
Logged
WX7G
Member

Posts: 5985




Ignore
« Reply #11 on: August 04, 2010, 12:29:18 PM »

For 60 and 30 meters (5.3 and 10.1 MHz) the 20-segment free version of EZNEC works.

92' OCF, fed 27.5% from one end:
5.3 MHz, 200 ohm VSWR = 1.5:1
10.1 MHz, 200 ohm VSWR = 2.1:1

Fed with a 4:1 current balun and 50 ohm coax the internal tuner should tune this.

A 260' OCF fed 27.5% from one end with a 2:1 current balun covers 160 and 80 meters.

These are free space models. You will want to model them at appropriate height above Sommerfield-Norton ground. The balun/feedline effects can be modeled by using a wire (the coax shield) and a series L-R to represent the balun common-mode impedance. The RF source is placed to one side of the OCF/feedline wire junction.

I ran the inverted-vee (140 degree) configuration over average GND. The dimensions are correct here too. 5.3 MHz VSWR = 1.5:1, 10.1 MHz VSWR = 2.3:1.

« Last Edit: August 04, 2010, 12:44:33 PM by DAVE CUTHBERT » Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #12 on: August 04, 2010, 12:49:51 PM »

For 60 and 30 meters (5.3 and 10.1 MHz) the 20-segment free version of EZNEC works.

92' OCF, fed 27.5% from one end:
5.3 MHz, 200 ohm VSWR = 1.5:1
10.1 MHz, 200 ohm VSWR = 2.1:1

Fed with a 4:1 current balun and 50 ohm coax the internal tuner should tune this.

A 260' OCF fed 27.5% from one end with a 2:1 current balun covers 160 and 80 meters.

These are free space models. You will want to model them at appropriate height above Sommerfield-Norton ground. The balun/feedline effects can be modeled by using a wire (the coax shield) and a series L-R to represent the balun common-mode impedance. The RF source is placed to one side of the OCF/feedline wire junction.

I ran the inverted-vee (140 degree) configuration over average GND. The dimensions are correct here too. 5.3 MHz VSWR = 1.5:1, 10.1 MHz VSWR = 2.3:1.

Thank you very much for running this for me!  This will give me something to compare to in my initial modeling with unfamiliar software.


Logged
WX7G
Member

Posts: 5985




Ignore
« Reply #13 on: August 04, 2010, 01:57:22 PM »

I can send you screen shots of the EZNEC setup along with a working modeling file that will work with the free version.
Logged
NZ5E
Member

Posts: 75




Ignore
« Reply #14 on: August 04, 2010, 07:16:08 PM »

I can send you screen shots of the EZNEC setup along with a working modeling file that will work with the free version.

I just sent you a personal message with my email address.

Thanks in advance,

Terry
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!