To increase the AGT value, increase the wire radius of the excitation point, the wire referred to in the EX card (ie the balun connection point).
To decrease the AGT value, vice versa.

Man this is rough on an old geezers brain.

I never found the expert button mentioned in help

I don't see any wire size or radius for the EX card.

http://www.digitalhome.ca/forum/picture.php?albumid=604&pictureid=3112

Man this is rough on an old geezers brain.

Cheezebooger go open your design then in Geometry editor click or have the feed piont high lighted then look at the wire radius for that piont , making ajustments up or down in wire size for the little feed wire that repersents your balum 300ohm load you can ajust the antenna so it has an AGT of 1 you can get lucky and use the wire selections in the drop down box ,but sometimes you just have to make finer ajustments in radius sizes

I don't see any wire size or radius for the EX card.

Sure you do. The EX card refers to Tag 1. Thats wire number 1 tag 1 in this case. Tag numbers can differ from wire numbers.

The EX card also refers to 1 segment on this wire. Normally, good modeling practice says to use 3 or more segments on the EX card wire, but with VHF antennas and sharp angles like on the bowtie, that can produce other warning and errors that are difficult to correct, so 1 in this case is OK.
Also on the EX card wire, the number of segments must always be an odd number, otherwise there is no middle segment. No middle segment skews off the radiation patterns.

Man this is rough on an old geezers brain.
There is a learning curve, yes. But if you just pick up a few things on it each day, youll be a pro at it in no time. :p

The NEC modeling engine was developed by Lawrence Livermore Labs for use by the U.S Navy in the 70's to 90's. Lord knows how many millions of dollars was spent on this software. They made it public domain in the late 90's. Consider it like a tax refund, heh.

So the tiny diameter of the EX wire causes a resistance or impedance to the signal, a load as opposed to a short?

If so the diameter must be adjusted every time the gain changes to create the proper impedance.

NO....you're trying to outthink some mathematical trickery necessary to "normalize" the AGT.

If so the diameter must be adjusted every time the gain changes to create the proper impedance.
Well yeah, every time the antenna changes it will need to be readjusted.

If doing a single frequency calculation, you can add or subtract the db value that in the parenthesis.
But when doing a long frequency sweep, that would become very very tedious. :p

I can only get the AGT down to 1.15

I had to reduce the other wires to #14 .03205" R and made EX .0033 R. If I try to make it less it says:radius (3.e-3) differs more than 10 * radius (0.032) for wire 3
radius (3.e-3) differs more than 10 * radius (0.032) for wire 2

It still says there are segment errors, don't say what errors and when I run segment check it says it's OK.

Got segmentation on low. Fiddled with the length. No 1.0 AGT

Everybody around the world is getting our tax refund.

Can anyone help me figure out a conversion formula from
AGT (computed by the NEC engine) to gain correction in db?

the answer to that turned out easy...

the_gain_correction_in_db = 10*log(the_AGT_value)

so AGT=1.1 means gain correction of .4dB
AGT=.95 means correction of -.22dB and so on.

Does anyone know why 4nec2 removes the LD card from the nec file when calculating the AGT?

the engine definitely produces different AGT value with and without the LD line...

I think youll have to ask Arie Voors that technical program question on the 4nec2 forum. I know it removed the GN card, which makes sense, but I didnt know about it removing the LD card, which describes the wire material.

The "Nominal Impedance" of an antenna is a compromise across all operating frequencies.
The "Impedance" Chart depicts frequency dependence of REAL and IMAGINARY components.
To find "best" impedance for an antenna, change (JUST) "Char-Impedance" under the 4nec2
"SETTINGS" tab until you get the "best" SWR curve. [Click on SWR chart to see each change.]
Try 50, 75, 150, 300 & 600-ohms (etc) and "eye-ball" an answer....

Sometimes this is the impedance with smallest-maximum within the operating band,
but can also be a joint compromise with the Raw Gain curve so that the
Net Gain isn't adversely affected--loss due to higher SWR can be tolerated better
on those channels that have higher Raw Gain.....such as Hi-VHF 2-Bay below...


Thanks for the info Hollands.

I see you have spent a lot of time modeling just about everything for UHF and VHF.

I have been experimenting with 4nec2 and coming up with designs with results that look good but the geometry varies so much from what has already been optimized that I am not confident that the actual performance will match my modeling results.

I just don't know enough about the raw gain vs net gain vs total gain, SWR, impedance and all those factors shown in your results so I am going to go by existing designs.

I do however believe that the best way to gang arrays together is with shielded 300 ohm twinlead. I bought a roll for that. All of the feed line designs say they are closest to twinlead so why not use twinlead?

The wavelength of the twinlead does not matter as long as the lengths are all the same and phased correctly as they enter the Balun or in my case, an amp with a 300 ohm input. Right?

I even bought dual 300 ohm input UHF VHF amps so I can have separate antennas.

I believe the bandwidth separation between VHF HI and UHF is too far to span with a single design without losing a lot of efficiency.

I just don't know enough about the raw gain vs net gain vs total gain, SWR, impedance and all those factors shown in your results so I am going to go by existing designs.
Since its not built directly into the program, Ill explain how to calculate Net Gain with 4Nec2. Youll need to use a spreadsheet program.

Here is a method to calculate Net Gain:

NetGain = RawGain+10*log(Feed-pointGain)
where Feed-pointGain = 4*Zr*Zo/((Zr+Zo)^2+Zi^2)

The [10*log(base10)]" converts to decibels, and Feed-point Gain is less than 1, since there is a loss.

Zo = characteristic Impedance for the transmission line connected to the antenna (For use with 300 ohm twin lead or a 4:1 balun, Zo = 300)
Zr = real part of antenna's complex impedance at a specific frequency
Zi = imaginary part of antenna's complex impedance at a specific frequency
RawGain = Gain output given by 4nec2 at the specific frequency


The Raw Gain, Real Impedance, and Imaginary Impedance data can be obtained when a Frequency Sweep is calculated. The 4nec2 program then outputs a Imp./SWR/Gain (F5) Line Chart. Under the "Plot" option is the option to plot Forw-gain, R-in (real), X-in (imag). It will give an error message, ie wGnuPlot.exe not found. Ignore it and it will output/display a text file with the needed data. Cut and paste the relevant data into a spreadsheet of your choice for your Net Gain calculations.

A quick way to look at the SWR figures to subtract from Raw Gain :

SWR Mismatch Loss dBi
----- -----------------
1.0 0
1.5 .18
2.0 .51
2.5 .88
3.0 1.25
3.5 1.60
4.0 1.94
4.5 2.25
5.0 2.55
I do however believe that the best way to gang arrays together is with shielded 300 ohm twinlead. I bought a roll for that. All of the feed line designs say they are closest to twinlead so why not use twinlead?

The best way to stack two identical antennas facing in the same direction is to use a properly designed and modeled phasing harness (ie custom formed wires)
I even bought dual 300 ohm input UHF VHF amps so I can have separate antennas.
Thats the best way to do it IMO.
I believe the bandwidth separation between VHF HI and UHF is too far to span with a single design without losing a lot of efficiency.
LPDA designs can do it, but they are generally 8 to 11 dbi antennas. With directors, maybe 13 to 14 dbi.

I'm trying to design an XGH8 (based on jed's) for a 497 MHZ frequency using the 4nec2 software. This is for a specific channel (Channel 18) that I presently can receive only at night time with my GH10 (jed's). Jed's XGH8 is optimized at 615 MHZ; hence the XGH8 parameters (i.e. lengths of the zigzags (a,b,c,d) and for the reflector elements (L1,L2,L3,L4)) are all optimized for this specific frequency. My question is how can I approximate these values for a 497 MHZ frequency to be used as initial conditions for the 4nec2 software? I tried to start with jeds XGH8 parameters for the 615 MHZ frequency, but the 4nec2 software is having hard time optimizing for the 497MHZ frequency using these initial parameters as initial guesses.

I'm assuming that my ignorance of roughly sizing the zigzag sides for a specific frequency/frequencies is delaying arriving at reasonable initial guesses for the 4nec2 program. My general understanding of approximating the lengths of elements (length = wavelength/2) of any antenna comes from the relation of (wavelength = 300/frequency). My problem is that I know how to apply this relation to approximate the length of antennas using straight horizontal/vertical elements, but I'm not sure how to apply it to a zigzag with elements at a 45 degree angle.

I tried to start with jeds XGH8 parameters for the 615 MHZ frequency, but the 4nec2 software is having hard time optimizing for the 497MHZ frequency using these initial parameters as initial guesses.

OK, start with JEDs XGH8 model and make sure the FR card is set to 615 mhz. Then in Geometry Edit, select all wires so they are all red, then select Edit > Rescale wires to the new frequency of 497 mhz. Then run the calculations to make sure it peaks where you want it to. Its that easy. :p

Thanks 300ohm. I usually rely on the Notepad editor, and never thought the code may have rescaling capability to a given frequency. Thanks again.

The Geometry Editor in 4nec2 is fantastic. You can copy and paste single wires or groups of selected wires and rotate and move them very quickly. :p

It's simple to rescale any given 4nec2 design.
Simply edit the file to add a SYmbol statement "SY F=1.00", and then mulitply each
and every X, Y, Z dimension by the Rescaling Factor "*F". Small changes in "F" can
be used to recenter a given design or start with F=3.0 to repurpose a given UHF design
for the Hi-VHF Band and tweak to optimize. I have provided many examples......

Since this is impractical to do for the (hundreds? thousands?) of wire statements in a
Screen Grid Reflector, don't bother Rescaling a Screen Grid. If repurposing from one
band into another, re-Build the Screen Grid.

Presently, I'm designing/modeling with 4nec2 a very high gain antenna for low UHF (peak at 497 MHZ). Since all my antenna builds are destined to the attic, I'm planning to use reflectors made of aluminum sheets. My question is it possible to simulate an aluminum sheet reflector by using the "rectangular-grid" option in 4nec2 (and using a large number for vertical cells)? Or is it better to use the "Geometry Builder" routine to achieve that?

Since basically less than 1/10 wavelength will approximate a solid sheet for all practical purposes, I would go with the Geometry Edit and draw my own 2" grid. I feel it gives me more control over the final outcome.

Its very easy to do. Just draw a small section of the desired grid, copy it, and move it to attach to the original section. Then copy that whole section and attach and repeat.

There are two ways to build a Screen Grid using 4nec2:
1) Build a rectangular PATCH using GEOMETRY BUILDER (under RUN tab).
2) Use GEOMETRY EDIT to build a RECT-GRID (under CREATE tab...not active for 3D view).

FIRST, I build the Grid using an arbitrarily high GW number to avoid number overlaps,
and then SAVE IT for use with a variety of future projects. I usually resegment ALL wires
to just ONE segment so that the total number of wire segments is a reasonably (large) number,
knowing full well that it's going to take a minute (or many minutes) to run and will be
approaching the maximum number of segments allowed in my chosen NEC Engine (5000).

I also place the Screen Grid BEHIND the origin by the (odd-ball!!!!) 4.555-inches so I can readily
find/replace-all the resultant fixed numbers with a variable reflector SEParation SYmbol = Sep.
[The odd-ball number avoids unwanted replacement actions to similar Y and Z-axis values.]

THEN Copy/Paste the antenna design of interest into a newly named file containing the desired
Screen Grid. [Or "borrow" a part of a (HUGE) file someone else built by deleting the ACTIVE part.]
Since there are usually 100's-1000's of wire statements defining the Screen Grid, it is very
difficult to select portions of that file....so copy/paste INTO a copy of that file.