So, I'm working on some automated tools for optimizing GH antennas, and I was wondering what the relationship between SWR (Standing Wave Ratio) and raw gain were for 'real' gain. I saw on j3d's page with the GH 4 6 8 and 10, he had his fancy plots for 'adjusted' gain with respect to SWR. I already found, by looking through the source to xnecview, how to calculate the SWR from the text output of nec2, I just want to make sure I'm looking at the right numbers for optimizing antennas.

This is looking to be a java program that invokes nec++, a c++ port of nec2 for unix based operating systems. It will use a list of channels, directions, and tvfool's noise margin values to tune a custom antenna using brute computational force combined with randomized algorithm finesse.

So far I have the part that generates the nec files and random variations of same... the parsing of the output data is looking less daunting now. And yes, it will be open source. I guess GPLv3 since I'm basing the models off the GH!

firimari: I have no idea what you are talking about,but look here.The calculation for Net Gain was explained in post #47 in this thread.
http://www.digitalhome.ca/forum/show...8&postcount=47

This is looking to be a java program that invokes nec++, a c++ port of nec2 for unix based operating systems. It will use a list of channels, directions, and tvfool's noise margin values to tune a custom antenna using brute computational force combined with randomized algorithm finesse.

Im not quite sure what youre trying to accomphish in the end, and wouldnt want you to waste time reinventing the wheel. Arie Voors 4nec2 program already interfaces with HF Ionospheric Communications Propagation Analysis and Prediction. Of course, keep in mind that the value of that type of information is mostly if these antennas were used to transmit.

Im not quite sure what youre trying to accomphish in the end, and wouldnt want you to waste time reinventing the wheel. Arie Voors 4nec2 program already interfaces with HF Ionospheric Communications Propagation Analysis and Prediction.
As a computer scientist, hobby projects are often wasting time reeeinventing the wheel. I know they have a genetic algorithm optimizer in 4nec2, but I want to do it myself.

I have no idea what you are talking about,but look here.The calculation for Net Gain was explained in post #47 in this thread.
Thanks! I missed that. easy peasy.

Ericball,

I sent you a PM, but I didn't get a reply, so I guess you haven't visited the forum for a while.

Here is a method to calculate Net Gain:

NetGain = RawGain+10*log(Feed-pointGain)
where Feed-point Gain = 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 ( Zo=300 for the GH)
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

If you have any programming skills, I would dearly love to obtain an executable that can parse the 4nec2 F8 output file for the RawGain. and Antenna's complex impedance (real and imaginary parts) for each frequency that is stepped in the frequency sweep at the appropriate Phi and Theta values of 90 and 0 degrees.
Hah. I am working on this. Another great point is that (unless someone can point me to the configuration to change this) 4nec2 assumes the feed-point impedance is 50 ohms, so for instance it thinks j3d's gapless GH10 has a SWR of 6.5102 at 537.5 MHz (or at least the model I've recreated with 1/4" aluminum and 6 gauge copper, using the appropriate material conductances...)
Calculated using the
sqrt( ( (zr-r0)*(zr-r0) + zi*zi )/( (zr+r0)*(zr+r0) + zi*zi ) )
equation from xnecview (zr = real impedance, zi = imaginary impedance, r0 = feed point impedance (300 ohms here)) I get 2.131648

Now I just need a user interface for this Java code, and to maybe add on NARODs, and I'll have a NEC-file generator for Gray-Hovermanns.

Selection from my output parsing code:
Frequency 537.500000 MHz
Impedance 145.650000 + 49.261000 i ohms, SWR 2.131648
Theta 90.000000 Phi 0.000000 Raw dB 14.420000 Net dB 13.812296

Hah. I am working on this.
Hah, if you would have read further down the thread, you would have seen where I was working on a COBOL program to parse the data from the old 80 column punch card data for Autofils, when low and behold, I discovered the data was already there, ready to be input into any spreadsheet program, to be manipulated by the user as he sees fit.

Another great point is that (unless someone can point me to the configuration to change this) 4nec2 assumes the feed-point impedance is 50 ohms, so for instance it thinks j3d's gapless GH10 has a SWR of 6.5102 at 537.5 MHz (or at least the model I've recreated with 1/4" aluminum and 6 gauge copper, using the appropriate material conductances...)

Ahem, err, etc. As a computer scientist, whats the first thing you tell the users of your program to do ? READ THE D**N DOCUMENTATION ! The input impedance is selectable in 4nec2. From the main screen, Settings > Char-impedance.

firimari, Ill admit it, your life working for me would have been miserable, heh. But dont you dare let me discourage your creativity !

Dcoumentation? We don't need no steenkin documentation!

Yeah, the RTFM is well deserved. thanks. But COBOL? seriously?

You might as well be writing in intercal.

But COBOL? seriously?

Yep, seriously. COBOL is very good at parsing thru 80 column punch card data. The NEC2 program engine, as written by Lawrence Livermore Labs around in the late 70's, was written in Fortran IV. And for all practical purposes (unless you want to analyze what gain youll get with the antenna if its underwater or in your body, heh) is the engine to use. All the same laws of physics havent changed since then, heh.

I work with enough engineers to know not to diss FORTRAN (or Matlab), but COBOL's relevance died out with y2k bugs. Anyway, back to UI design if I want anyone else to use my learning-exercise toy.

Linux users should take heart, I'm making sure the ability is there to use nec2++ instead of the exes in 4nec2.

but COBOL's relevance died out with y2k bugs.
Yeah, poor foresight (and ignorance of good techniques and style) on the part of many programmers. I didnt do that kind of crap, heh. I needed the full 4 digits of the year to sort by for one thing.

Also keep in mind, Arie Voors who wrote the 4nec2 Windows front end to the NEC2 engine (and gives it away and maintains it for free) is an extremely good programmer. Hats off to Arie Voors ! I think once you read thru his documentation, youll agree that he has documented the program much better than 99.9% of all the programmers out there.

I have a question on a antenna I'm playing with in the 4Nec2 program.

I have a total again for horizontal of -9<dbi< 8.23 at 575 mhz
S.W.R. 1.24
Impedence of 366-j25.7
Parallel form of 368//-j5251
Series comp 7.e-3 uH
Parallel comp 1.454 uH
My question is having the E [phi] [v/m] on the horizontal plane good
if shown like so in the pattern

lower right corner shows 1.95 < V/m < 14.1

The above figure on SWR and impedance look good.

The E [phi] [v/m] or 1.95 < V/m < 14.1, volts per meter figures deal with transmitting. Youre not planning on transmitting with the antenna are you ?

No i was not 300ohms was just wondering how i ended up past 0 hey but maybe i could transmit over to the buds house hehe thank you 300ohms and your NARODs play a big part in the antenna , just wish I had not closed to many windows at one time as i lost my one design that held the impedence at 300 ohms across the bands .

Now for another question about the front to back how does that come into play for gain and what would a good front to back be anyways .

That all depends on how much of a front to back ratio you want. A lot of times, I want less. Generally, higher gain and higher front to back ratio go hand in hand.

as i set here amazed at your tallent for antenna i cant figure out the 4nec program
as amazed that all vcr read 100 after power failure
what came first chicken or egg
how do you know wher to put tees for frame do you have the zzz part built first and how do you get it to fall ever so perfect take the build picture with cloud back ground how did he know where to put all those bends and at what didtance what is best sbgh build and correct build for 15235 zip code after digital switch to cover the aviable channelswhat are the dimensions for a bending jig for the zzz wire for that
confused
hutch67

I'm a 4NEC2 beginner, trying to model a log-periodic UHF-band antenna in 4NEC2. The log periodic design requires connecting monopoles to each side of a twinlead, in alternating directions. In the following example file, are wires 3 and 4 electrically connected to the twinlead? This is not a final file, and I'll have to swivel the monopoles. Right now, my main problem is connecting the monopoles separately.


CE
GW 1 5 -1 0 1 -1 0 -1 0.05
GW 2 5 17 0 1 17 0 -1 0.05
GW 3 15 0 0 7 0 0 0.25 0.05
GW 4 15 0 0 -0.25 0 0 -7 0.05
GS 0 0 0.0254 ' All in in.
GE 0
TL 1 3 2 3 50 0
GN -1
FR 0 1 0 0 470 0

In your 4nec2 folder under models > HFmultiband, you have a few log periodic antenna examples, ie logp.nec, logper.nec, lpda.nec, LPDA1.nec and maybe some others.



The log periodic design requires connecting monopoles to each side of a twinlead, in alternating directions.
Keep in mind, thats only one way log periodic antennas are made. There are many other ways.

In your 4nec2 folder under models > HFmultiband, you have a few log periodic antenna examples, ie logp.nec, logper.nec, lpda.nec, LPDA1.nec and maybe some others.
Thanks for the pointer. I'll start with one of these and rescale to my needs.

All these examples show a 180 degree twist between elements, which works OK. For future reference, this begs the question... how do I tell 4NEC2 when I want it to twist the twinlead 108 degrees? I apologize if this is in the "help" file. I'm running 4NEC2 under wine in linux, and I seem to have everything except the "help" functional.

For future reference, this begs the question... how do I tell 4NEC2 when I want it to twist the twinlead 108 degrees?
Youre going to have to elaborate on that question. I dont see where youre getting 180 degrees from, much less 108 degrees, heh.

I apologize if this is in the "help" file. I'm running 4NEC2 under wine in linux, and I seem to have everything except the "help" functional.
Well, the 4nec2.rtf file is the exact same as the windows help file and the .rtf file opens fine in OpenOffice 2.4. I would just print the whole thing out and bind it in notebook. (I did, for easier portability)

Youre going to have to elaborate on that question. I dont see where youre getting 180 degrees from, much less 108 degrees, heh.
If I could type, I'd be dangerous :( Anyhow, with the "transmission lines" icon in effect, select the transmission line segment and click on "Cross connection" to twist the transmission line 180 degrees

Well, the 4nec2.rtf file is the exact same as the windows help file and the .rtf file opens fine in OpenOffice 2.4.
Abi-Word works for me. Thanks for the pointer.

There will probably be more questions coming tomorrow about SYMbols.