yah, funny how I can always remember just about any phone number after seeing it just one time, but i have to get that 698 backwards!

Interesting tidbit from here: http://www.nec2.org/part_3/cards/ex.html
Useful to know when playing around with stacking distances.

When there is more than one source on the structure, only the impedance of the last Source specified will be collected .

I can always remember just about any phone number after seeing it just one time
I cant. Damn that Tommy Tutones !

or this is as clear as mud....

http://en.wikipedia.org/wiki/Normalization_(statistics)

Thank you, all excellent, helpful posts here!

a few more questions:


1. what exactly is the "normalization" setting in the expert doing, i see some members use it, others dont, i didnt see much in the documentation re. this

2. when initially drawing/designing an antenna in geo edit, say for the UHF broad band, what is the frequency that should be entered in that windows frequency box, I assume since the auto segment uses this, it should be set at the highest frequency, say 689 Mhz

3. I believe Holl_Ands, you have said that in a manual frequency sweep, the program uses the azimuth with the maximum gain for the lowest (start)frequency in the sweep for the gain azimuth of all subsequent frequencies in the sweep, so does this mean that setting the PHi/Theta manually in the sweep window box will not give me the frequency sweep gains at those set angles

thanks for all the assistance ;)
Normalization means to plot Gain RELATIVE to the Max value.

For example, in the PATTERN plot, default appears to be "No Normalization"
(under Far Field menu), resulting in a plot of the actual Gain values.
If "Normalize Overall" is selected, azimuthal plots at ALL frequencies will be
RELATIVE to the Max Gain (defined to be 0 dB) across ALL frequencies.
If "Normalize Each" is selected, azimuthal plots at each frequency will be
RELATIVE to the Max Gain FOR that frequency. (0 dB is single freq Max).
The Max Gain used for normalization is found elsewhere in the box.

I would design at the center frequency (e.g. 198 or 584 MHz) and "ass.u.me"
the segmentation is good across the entire band...and BTW it may or may not
be adequate to achieve AGT=1.0 for both Hi-VHF and UHF....even with
different SOURCE wire radius choices.

Manual Freq Sweep (under F7 Generate Menu) is way too easy to make
mistakes....and even when I know I've entered data correctly, I've also seen
too many incorrect results, so I don't use it....so I won't comment on using it.
Instead, I ALWAYS copy/paste my "standard" set of FR/RP Cards and comment
out everything except the ones I want....consistent results nearly every time....

Thanks Holl_Ands, im trying to absorb AMAP, to really be able to use the program with some degree of confidence, ill review your posts, and practice :D



it may or may not
be adequate to achieve AGT=1.0 for both Hi-VHF and UHF....even with
different SOURCE wire radius choices.

im not sure I understand this...

Frequently, autosegmentation for 584 MHz will also work at 198 MHz, although it is likely that
the SOURCE radius for AGT=1.0 is different for UHF and Hi-VHF bands.

And sometimes you'll have to use separate 4nec2 files, with separate autosegmentations
and probably different SOURCE radii.

if im concerned with designing for ONLY the UHF OR ONLY the VHF, does AGT have to = 1 for BOTH bands

AGT is different for every frequence.
So it should be 1.0 for those you are interested in.
I am assuming though, that having it close to 1.0 across both bands, would mean
that your model is good.

Normalization means to plot Gain RELATIVE to the Max value.

Ahh, for PLOTTING. Basically, Ive always used normalization because the opposite is ab-normalization, heh. (and because oldsparks used it in his RP card)
So using option 5: tot-gain normalized, will plot the correct beamwidths.

if im concerned with designing for ONLY the UHF OR ONLY the VHF, does AGT have to = 1 for BOTH bands
No, AGT has to be 1 (0db) only for the frequencies the antenna is designed for. VHF and UHF bands are separated by a wide amount of frequencies, so thats why AGT can vary a whole lot between them sometimes.

nikiml,

AGT is different for every frequence.
So it should be 1.0 for those you are interested in.
I am assuming though, that having it close to 1.0 across both bands, would mean
that your model is good.

that is the point of my question....


300ohm,

thanks!

300ohm,


when you say to run the AGT (for adjusting) on narrower bandwidth, what exactly do you mean by that,



what im doing is I run calculate (AGT) on say middle UHF=585, check the AGT result, depending on the +/- AGT I adjust the source wire and segmentation (if need be), I then rerun the calculate to verify the new AGT result, if its OK, I go on to do the same steps for 698 mhz, and 473 mhz, then go back and check 585 mhz...

then I run the sweep 473 to 698, step=6...

is this correct?

thanks

when you say to run the AGT (for adjusting) on narrower bandwidth, what exactly do you mean by that,

OK, lets say you have a situation where AGT is close to 1 over say 470 - 600 mhz, but then goes unacceptably over/under 1 for higher frequencies. In that case, run a sweep for the 470 - 600 mhz range, then run another sweep after adjusting for AGT = 1 at the middle of the next range, the 600 - 698 mhz range of freqs, (if the AGT stays good within the acceptable range of course).

Ah, now I got you, this almost an art form!

is there no where we can see the AGT on a sweep one frequency after the other, or is it only after running each frequency far field separately , and look at it in the Main form?

Only on the Main window, AFAIK.

trying to get the hang of this here,


I tried optimizing a model using the UHF frequency sweep 473 to 698, after many attempts I could not get the gain above 11.5 dbi, (& SWR below 1.5), then I adjusted for AGT= 1, and lost a little gain, but then I redid the optimizing with the latest geometry, and a got a peak gain "jump" to 12.2 dbi this time, and drop in SWR to 1.3, so then I adjusted AGT=1 for this, and lost almost all of that gain jump, netting no more gain than in the previous optimization...

is that what the AGT was telling me, that the second run wasnt realistic (as compared to the first) and what could have caused this jump and AGT go out of whack so much, its not like the geometry changed all that much...


thanks

Yup....forcing AGT=1.0 keeps us all more or less HONEST wrt wild Gain claims......

do you know how or what the program is calculating to arrive at the AGT value

Most likely its an equation listed in the NEC manual part 1 (87 pages). The trouble is, I cant make it through just 2 pages of equations in that manual without falling asleep, heh.

yes, its all excellent bedtime reading for me anyway!

where is that part 1, I already read Part 3 (main parts), only 131 pages :eek: