PDA

View Full Version : What's the CORRECT Way to Build a CAROLINA WINDOM Antenna ?



WA3MOJ
12-04-2011, 04:24 PM
.

I'd like to build a Carolina Windom Antenna, but am limited to the total overall length.
Therefore, I'd like to design it around the 66' length cut for 40 Meters and all the related
bands on which that size antenna will operate.

There are a lot of articles out there of how to build a Carolina Windom antenna.
There is also a lot of seemingly conflicting information, too.

- Some have a vertical radiator coming down from the matching transformer.
- Some do not.

- Some use the 37.8% - 62.2% ratio for the location of the feedpoint; 25' / 41' .
- Some use the 80% on the long side and 20% on the short side ratio.

- Some use a 4:1 "current" balun for matching at the feedpoint.
- Some say you should use a 4:1 "voltage" balun in order for the vertical radiator to be
functional; then insert a 1:1 Isolation Choke at the end of the vertical radiator.

Then, there is the issue of "saturation" and getting a good balun. Just because a balun
is rated for 1.5KW doesn't mean it won't have problems with saturations using just 100 watts.

I'd like to get the straight answers on the correct way to build this antenna. I would like
to have a design "with" a Vertical Radiator section and based on the 40 Meter length.

Also, which other bands will this 66' antenna operate?

Thanks.

.

KC8VWM
12-04-2011, 04:52 PM
I found a lot of designs floating around either lack any technical information or they're simply cut and pasted from incorrect information from old designs. I also found many antenna designs only exist in antenna models and were never actually constructed by anyone too.

Some claim their designs have magical properties such as a vertical radiator. This is nothing more than common mode currents flowing on the feedline. I suppose you could technically call that a "radiator" since the feedline is in fact radiating, but I prefer to call it a lossy nuisance myself.

I like to construct antennas that are proven, provide well thought out detailed technical information about thier design and ones which can be contructed with repeated results.

With that said, all the OCF antennas are a compromise. They all exhibit nulls in the pattern and they generate common mode currents. Using the correct type of balun is important with this design and some use the wrong one.

However, if you want to achieve the best compromise design, I would recommend using this Windom off center fed design:

http://www.w8ji.com/windom_off_center_fed.htm

NI7I
12-04-2011, 07:18 PM
Would that be North or South Carolina?

KE4YOG
12-04-2011, 07:36 PM
If it works it is a North Carolina windom. I might be a little biased though...

KF5FEI
12-04-2011, 07:38 PM
Would that be North or South Carolina?

New, actually.

The plans noted by the OP do work -- I have one installed at the casa as an inverted V about 28 feet up, with a Wireman 4:1 balun, using #16 THHN and about 5 chokes 10' down on a piece of RG-8X.

I usually run QRP, but I've run 100 watts through it with no issues. IIRC, it is supposed to work on 40, 20, 15 and 10 -- mine is a bit long so it works best on the low end of 40 and 20 and not as well on the other bands. In any case, it is relatively inexpensive to build, and if it doesn't work right you can still use the bits and pieces to build something else.

AE4B
12-04-2011, 09:30 PM
i have used a version of this antenna for several years and i am very impressed with it. i used 23' for the short end and 46' for the long end,and with 10' for the "vertical radiator". a 4:1 VOLTAGE balun and 1:1 CURRENT balun (line isolator) works well in this application.
i consider it an excellent antenna for multiband coverage- 15m requires using a tuner but 40-20-10m has low swr. it is mounted at about 38' in the inverted v configuration. everyone seems to have an opinion of this antenna (vertical radiator in particular) -but it has worked very well for me-i have made contacts world wide with 100 watts.
as for me , it is a keeper. just saying...

G0GQK
12-04-2011, 09:35 PM
He does state it has to be 66 ft long because he's short of space

AE4B
12-04-2011, 09:54 PM
if short of space just let the short and long ends of the antenna dangle down about 8' or so on each side. the inverted v configuration uses less available space than a flat top version would use.
just keep the leg angles more than 90 degrees for best results.

KC8VWM
12-04-2011, 10:10 PM
Yup, they can be scaled smaller in size... 80/20 is still a practical feedpoint objective even when using 66' of wire.

However, there are far more efficient, yet smaller sized antennas which would take up far less space and provide the desired band coverage and yet exhibit cleaner patterns and more omni directional coverage performance.

WB2WIK
12-04-2011, 10:37 PM
I'd just use a 66' center fed doublet fed with window line, a 1:1 current balun in the tuner and a good tuner.

The miraculous properties of a 10' vertical radiator added to the mix don't mean much to me. If there's lossy ground beneath the antenna, it will still be lossy.

WA3MOJ
12-04-2011, 11:03 PM
Yup, they can be scaled smaller in size... 80/20 is still a practical feedpoint objective even when using 66' of wire.

However, there are far more efficient, yet smaller sized antennas which would take up far less space and provide the desired band coverage and yet exhibit cleaner patterns and more omni directional coverage performance.


VWM-
To which antenna's are you referring that are more efficient, smaller, cleaner, etc.

Wire antenna's, or other styles?

.

KC8VWM
12-04-2011, 11:23 PM
...Seen one of these before?

80168

80169

6x6 foot span (foot print)
5 bands - 96% efficiency. Full size antenna elements.
No antenna tuner needed.
Folded dipole, DC grounded design, low, low noise levels

Omni directional pattern works better than a non rotatable dipole. No nulls in patern. Better coverage, cleaner pattern etc..

http://www.karinya.net/g3txq/temp/cobwebb_dipole_real.png



OCF Antenna Pattern:

http://www.karinya.net/g3txq/temp/ocf_low_angle.png

WA3MOJ
12-04-2011, 11:28 PM
i have used a version of this antenna for several years and i am very impressed with it. i used 23' for the short end and 46' for the long end,and with 10' for the "vertical radiator". a 4:1 VOLTAGE balun and 1:1 CURRENT balun (line isolator) works well in this application.
i consider it an excellent antenna for multiband coverage- 15m requires using a tuner but 40-20-10m has low swr. it is mounted at about 38' in the inverted v configuration. everyone seems to have an opinion of this antenna (vertical radiator in particular) -but it has worked very well for me-i have made contacts world wide with 100 watts.
as for me , it is a keeper. just saying...



AE4B -
Thanks for the info about using the VOLTAGE Balun at the feedbpoint. That's what I thought it should be and not a Current balun.
How did you come up with the 1/3 - 2/3 ratio instead of the 37.8% - 62.2% formula?
And, at an overall length for your antenna at 69' , wouldn't that put your resonant frequency at around 6.78 Mhz?
Or, am I doing something wrong?

.

WA3MOJ
12-04-2011, 11:35 PM
...Seen one of these before?

6x6 foot span (foot print)
5 bands
96% efficiency. Full size antenna elements.
No antenna tuner needed.
Folded dipole, DC grounded design, low, low noise levels

Omni directional pattern works better than a non rotatable dipole. No nulls in patern. Better coverage, cleaner pattern etc..




I have seen these before, but forget what it's called.
I don't think it's a Spider Beam...
Do these need a rotor? And, what is minimum height off the ground?

.

VE3EKJ
12-04-2011, 11:46 PM
First off, we have to nail down some definitions. The correct term is "OCF - Off Centre Feed Dipole". It is a design derived from the original Windom of decades ago, which had a single wire feeder.

Today it seems everyone calls an OCF dipole a Windom, which confuses things a bit but not that much, I guess.

The OCF is at heart just a dipole, no more or less. The only difference is that it is NOT fed at the centre! The offset is chosen to have a feedpoint where the impedance is in a useful range for more than one band. When up about 35-40 feet it shows a range around 288 ohms, give or take a few tens of ohms on 80, 40, 20, 10 and 6 metres. If you use the shorter 66 foot total length you lose 80 metres, although like with any shortened dipole you can get that back with some loading coils, at the price of some 80 metre bandwidth.

There is no 'perfect' offset feedpoint. Things vary in the real world, so the percentage of length on the long side and the short is subject to a bit of experiment. That's why you have seen different figures. Use what works for you.

That 288 ohms will match nicely if you feed it with a 4:1 balun, ending up between 50 and 75 ohms, depending on which band.

If you feed it with a voltage balun you will likely have a LOT of RF on the outside of the feedline, due to the off-set feedpoint! This can give problems with "RF in the shack" and is usually solved with "choke" baluns on the outside of the feedline before it enters the shack.

If you use a current balun, it will force equal currents into both the short and long elements of the OCF, greatly reducing the feedline radiation.

There is only ONE version of the OCF design called a "Carolina Windom"! This design uses a voltage balun and puts a choke balun a quarter wave below the feedpoint for 15 metres. This encourages the quarter wave length of coax shield to act like a 15 metre 1/4 wave vertical, adding 15 metres to the list of bands for which the OCF will work. There is a lot of controversy as to how well this idea works. Some hams have found that if they make a second OCF cut for 30 metres, which of course will also work on 15 and attach it to the same feedpoint, drooping the wires under and away (preferably at right angles) from the main OCF it works much better, similar in having two ordinary dipoles fed from one feed line, which is a trick hams have been using as long as they've used dipoles! The reports are that this works very reliably and quite well! Especially with a current balun, as the vertical radiator/ Carolina Windom design often still has feedline RF problems.

Hopefully this will make things more clear.

WB2WIK
12-04-2011, 11:54 PM
AE4B -
Thanks for the info about using the VOLTAGE Balun at the feedbpoint. That's what I thought it should be and not a Current balun.
How did you come up with the 1/3 - 2/3 ratio instead of the 37.8% - 62.2% formula?
And, at an overall length for your antenna at 69' , wouldn't that put your resonant frequency at around 6.78 Mhz?
Or, am I doing something wrong?

.

The amount of offset a doublet is fed doesn't change its resonant frequency, only its feedpoint resistance.

A 1/2-wave wire fed at the center, at the end, or any other point is still resonant at the frequency where it's a 1/2-wave long.

AE4B
12-05-2011, 01:20 AM
the 23' and 46 'wire lengths were from an article in qst -may 1996 (page 65). - there are numerous wire lengths that can be used-radio works :25'=41 -carr:12'+57".-arrl uses:11'+58". google for many more.
the 69 ' length allows the antenna to be pruned to minimum swr by removing (the same amount) of wire from each end . i adjusted mine for 20m use and the swr was still ok on 40m +10m (tuner on 15m). remove 1/2 inch or so at a time while using swr /antenna analyzer.

WB2WIK
12-05-2011, 01:29 AM
Everything radiates if you can match it.:o

KC8VWM
12-05-2011, 04:29 AM
Do these need a rotor? .


http://www.karinya.net/g3txq/temp/cobwebb_dipole_real.png

Yes, clearly an omni directional antenna displaying the pattern above requires a rotor.

EI4GMB
12-05-2011, 12:30 PM
Personally, I would use a 4:1 Current Balun with a Carolina Windom. This acts as a matching unit and would make the windom easier to tune over a greater range of frequencies. It would also help reduce feedline radiation! :)

73,

Fred EI4GMB

G3TXQ
12-05-2011, 01:04 PM
It's important to understand that the feedpoint position on an OCF dipole is very much a compromise.

For example, feeding a half-wave dipole at the popular 33.3%/66.6% position produces an impedance which is just 1.3 times the impedance had it been centre-fed. In other words 65 Ohms if centre-fed it was 50 Ohms, or 94 Ohms if centre-fed it was 72 Ohms.

It's also a fact that a wire which is resonant at frequency F will not be exactly resonant at 2*F, 3*F, 4*F etc, because the "end correction" factor only applies at the ends of the wire and not at every current minimum.

Any departure from the "flat top" configuration will likely affect the feedpoint impedances differently on the different bands.

Add into the mix the effect on feedpoint impedance of any CM current in the feedline; then add in the fact that the balun is unlikely to maintain exactly the same impedance transformation ratio over a wide bandwidth, and you have a pretty complex situation.

Choosing a feedpoint position and an impedance transformation ratio amidst that complexity is an exercise in balancing compromises - there is no single "perfect" solution - which may account for the variations in design which we see.

On the plus(?) side, any losses in the balun and the feedline help to make the designer's compromise set of SWRs look better ;)

73,
Steve G3TXQ

K9SRV
12-05-2011, 02:07 PM
Regarding 66' vs. 137'vs. 260' OCFD antennas, does longer mean higher gain on ten? I know the
longer the OCFD the more the antenna radiates off the ends, but would a 260' OCFD, with an E-West
orientation have more gain to Eu. and Ja. than a 66' ft. OCFD?
Or, would the extra lenght be turned into usless heat?

Thanks,
John

W8JI
12-05-2011, 02:50 PM
Regarding 66' vs. 137'vs. 260' OCFD antennas, does longer mean higher gain on ten? I know the
longer the OCFD the more the antenna radiates off the ends, but would a 260' OCFD, with an E-West
orientation have more gain to Eu. and Ja. than a 66' ft. OCFD?
Or, would the extra lenght be turned into usless heat?

Thanks,
John

Assuming heat losses are minimal, which they usually are, gain only comes from removing radiation in some other direction or directions.

When you cannot rotate the antenna, or plan the exact position of lobes to useful directions without adding holes in other useful directions, gain becomes a bad thing to have. It is a bad thing because the nulls or holes in coverage are always deeper than the gain increases other directions.

Also, as a wire is longer and longer, heat losses go up and efficiency goes down. Rhombic antennas have poor gain for the beamwidth, because they have very long wires that have to convey current from end-to-end. More than half the power is lost as heat in really long antennas.

The last antenna parameter that is important, unless you plan lobe and null placement, is gain. If I had the choice of any general coverage non-rotatable antenna for higher bands, it would be a simple dipole or vertical.

73 Tom

K9SRV
12-05-2011, 03:47 PM
Thanks Tom! I specifically ask because I am going to mount 1 {Short} end to a fixed spot on the house, up 30'.
The long side I can move around, as I have many 100' trees in my yard. I thought it would be fun to follow some DXpeditions by simply relocating the 1 movable side. Today I might shoot for ST0 if they are stiil in some daylight. Is this plan stupid, or just something
Bill Murry might concoct? :)

John

EI4GMB
12-05-2011, 05:18 PM
There is only ONE version of the OCF design called a "Carolina Windom"! This design uses a voltage balun and puts a choke balun a quarter wave below the feedpoint for 15 metres.

The Radioworks Carolina Windom uses a 4:1 Current Balun as a matching unit and employs an RF isolator 10 feet below the feedpoint which allows the vertical section to radiate.
In my experience, it does not significantly improve the DX performance of the antenna. Instead it makes it noisier and mostly fills in nulls useful for high angle local work. That is not to say the Carolina Windom is not an excellent DX antenna, for I use one, and I can tell you it surely is. However, the vertical radiator has less to do with this. Hope this helps. :)

Kind Regards

Fred EI4GMB

VE3EKJ
12-05-2011, 08:19 PM
Personally, I would use a 4:1 Current Balun with a Carolina Windom. This acts as a matching unit and would make the windom easier to tune over a greater range of frequencies. It would also help reduce feedline radiation! :)

73,

Fred EI4GMB

I could be wrong Fred but I don't believe that a Carolina Windom should use a current balun! You see, it seems to WANT to have lots of RF on the feedline! That's what radiates from the vertical radiator portion!

After the vertical radiator is where they put a choke balun, to stop the RF from traveling any further on the outside of the shield.

So using a current balun with a Carolina design strikes me as reducing the very feedline radiation you're intending to radiate!

Or is this old guy missing something?

G3TXQ
12-05-2011, 10:12 PM
I could be wrong Fred but I don't believe that a Carolina Windom should use a current balun! You see, it seems to WANT to have lots of RF on the feedline! That's what radiates from the vertical radiator portion!

I think that's probably correct - a voltage balun will tend to drive CM current onto the feedline.


After the vertical radiator is where they put a choke balun, to stop the RF from traveling any further on the outside of the shield.

I often hear that said, but it's a rather misleading description. If the choke is physically small and can be treated as a lumped component, there can't be current flowing one side of it and not the other! What it can do is drive a current minimum in the CM current standing-wave pattern at the point it is located.

73,
Steve G3TXQ

EI4GMB
12-05-2011, 10:24 PM
I could be wrong Fred but I don't believe that a Carolina Windom should use a current balun! You see, it seems to WANT to have lots of RF on the feedline! That's what radiates from the vertical radiator portion!

After the vertical radiator is where they put a choke balun, to stop the RF from traveling any further on the outside of the shield.

So using a current balun with a Carolina design strikes me as reducing the very feedline radiation you're intending to radiate!

Or is this old guy missing something?

My apologies. You are indeed correct. The Radioworks literature states that the Carolina Windom uses a 4:1 Balun but doesn't state whether this is a current balun or a voltage balun. However your explanation about the use of a voltage balun in this application makes perfect sense. I stand corrected.

Kind Regards

Fred EI4GMB

ad: giga-rw