Johnny Novice to the rescue!

[Steve - WB3HUZ]

Looks like you have good coverage into my part of VA. Must be the pink part of the 3D pattern or is that only for coverage to W1IA. I can never remember. 

Yep, almost identical to mine, except mine's multi-color! 

If you want to see something really neat, check the 7000 kc pattern vs. the 7300 kc pattern. Midway up the band, it crosses the line from broadside to endfire.

The issue with that pattern in the daytime is where the nulls materialize. I've included a VOACAP shot of typical performance at 9 AM on a July day with a 0 sunspot index. Not exactly the best bearing for a null from my location!

It does great overseas, though, so I don't want to mess with it. I'll take another approach for daytime operation.

In Derb's case, though, secondary 40 meter radiators might be just the ticket for 40.

[Detroit47]

What software do you folks use for antenna modeling? I always read what you gentlemen have to say about antennas.
 Thanks Johnathan N8QPC

[KA1ZGC]

Johnathan,

We're all using packages built around the NEC engine. There are numerous versions out there, but they're all wrappers around the same module. EzNEC is popular. The package I'm using is called 4nec2x.

I like that package because it will feed NEC's output pattern directly to VOACAP, which is a propagation modelling package used by VOA. I find that helpful to view what areas you stand to gain and lose by changing various properties. It helped convince me to leave my 200' flat-top alone.

The interface between NEC and VOACAP needs a lot more work, though, and you need to be able to convert cartesian to polar in your head to make it work. That's why I don't recommend it to everyone. The NEC part of it works just fine.

There are some guys who just type up files in the NEC input syntax and hand-feed the standalone engine. The engine itself was originally run on a mainframe and fed by a card reader, so the syntax is very simple. It's all documented online.

There are tons of NEC packages out there, but the NEC engine itself is the only important part.

That's my opinion, anyway.

[KA1ZGC]

Looks like you have good coverage into my part of VA. Must be the pink part of the 3D pattern or is that only for coverage to W1IA. I can never remember. 

Yep, looks like you're just off my southwestern null. The 3D shot of the pattern is looking down on the ant from the norteast. The lobe facing you is just to the left of the Z-axis label.

So it looks like I've got better than 6dBi @ -6dBk comin' atchya.

Speaking of W1IA, it wuz good tryin' to hear ya on 75 the other night. Too bad Tina kept steppin' on ya in dem pink pumps, caw mawn.

[KA1ZGC]

One last post on this tangent.

Just for comparison, I ran some long-haul models under conditions like we have now. This shows how my northeastern null is mostly in the waters of the Mediterranian.

The nulls are well-placed overseas, but not here at home.

This also shows that it pays to be a good scientist and run the same models numerous times at different levels of detail. The 10000 km great-circle map looks a whole lot more optimistic than the 5000 km map looks, using the exact same numbers.

Of course, these are mere predictions of monthly averages at a given time of day, so they're no substitute for s-meter readings, but they're handy in showing you likely effects of changes to your system.

At any rate, these maps show why I'm not going to do anything to this antenna to correct the issues on the previous map. All I'm going to do to this antenna is raise it to lower the takeoff angle of those European lobes and start banging out some serious Old World AM on 40.

Derb, if I know the height and length of your ant, I can chart out your feedpoint resistance, reactance, and impedance across huge swaths of HF. I already know how much feedline you've got. Sounds like you just need to append some feedline. I can find out how much you need to get a reasonable match on whatever bands you want. You can always attach it temporarily to see if it's going to help or harm your performance before you commit to it.

[Steve - WB3HUZ]

This also shows that it pays to be a good scientist and run the same models numerous times at different levels of detail. The 10000 km great-circle map looks a whole lot more optimistic than the 5000 km map looks, using the exact same numbers.

Indeed. With the modeling programs now available, many available free of charge, one can really check things out before cutting wire or tubing.

Check out MNANA and 4Nec2 (like Thom has). Both are free.

TL Details or W9CF''s transmission line applet are good for checking impedances and losses on transmission lines.

Just  watch out for those Pink Pumps!

[KA1ZGC]

While we're at it, I also want to get a plug in for one last package.

The NEC packages are great for generating antenna patterns. The ITS suite (REC533, VOACAP, ICEPAC) are great for taking that pattern and casting it into an ionospheric propagation model*.

If you're looking to model tropospheric propagation between 20 MHz and 20 GHz, and don't mind caressing some of the data by hand, then SPLAT! comes in handy:

http://www.qsl.net/kd2bd/splat.html

I've used it for everything from coverage analysis of my Skowhegan repeater, to modelling performance of a homemade 920 MHz paraflector, to coverage analysis of customers' proposed facility upgrades, to determining post-transition DTV antenna requirements. I even used to test the viability of a 23 MHz link between Skowhegan and Monticello.

You need to understand your true receive criteria before you'll get useful numbers, but once you know what you're doing, it's a powerful tool. A small crash-course in Longley-Rice path loss analysis goes a long way.

The package is very much in its infancy: there's no graphic user interface, even on the Windows version. There is a lot of data you need to acquire on your own (the website has the links to all that), and some you need to generate by hand (all explained in the manual). It's not for the faint-hearted, but the numbers have been pretty close to real-world measurements as long as I've used it.

Armed with a NEC engine, the ITS suite, and SPLAT!, you can cover all your HF/VHF/UHF/SHF needs* and take much of the guesswork (and "so-and-so said such-and-such") out of antenna design.

Okay, I'm done now. Really.

--Thom
Killer Agony One Zipper Got Caught
*Note that the ITS suite is increasingly less reliable the further you go below 4 MHz. The engineer in charge of VOACAP's development explains in an online writeup somewhere why the VOA model doesn't work for medium-wave propagation. A reliable medium-wave model has been published in the last few years by an FCC engineer and has received rave reviews, but not yet coded by anyone to my knowledge.

[Steve - WB3HUZ]

*Note that the ITS suite is increasingly less reliable the further you go below 4 MHz. The engineer in charge of VOACAP's development explains in an online writeup somewhere why the VOA model doesn't work for medium-wave propagation. A reliable medium-wave model has been published in the last few years by an FCC engineer and has received rave reviews, but not yet coded by anyone to my knowledge.

That's true. VOACap and it's deritatives are considered pretty much NG for 160 meters.

For MW stuff, most BC ops are mostly concerned with groundwave propagation, which has been pretty well predicted for 80-90 years.

[KA1ZGC]

That's true, broadcast MW stations don't care about skywave (apart from how much their coverage area is eaten by QRM at night, but there's nothing they can do about that). Probably that's why nobody really delved into MW skywave models.

However, MW skywave models have undergone a lot of refinement in recent years. I read up on one a few months back. Seems the Earth's magnetic field plays an important role in skywave propagation at those wavelengths. That's not a (discernable) factor at HF and above, so those models don't address it.

The model receiving the most attention was written by a Dr. Wang (or is it Wong?) from the FCC. I tried to relocate the writeup I found before but am coming up empty. It's out there somewhere, anyway.

I might find the hardcopy I was going to use to try coding the model myself, before I decided I'd really rather be out chasing women. 

[Steve - WB3HUZ]

Yea, Dr. Long Wang, Phd.

The electron gyro frequency has a big effect on absorbtion of skywave signals at MF, most especially the upper MF range. The frequency varies across the globe but is in the 1400 kHz range, most often. Horizontally polarized waves suffer far more absorbtion than vertically polarized waves. So, if you want to work DX on 160, ya better use a vertical (usually).

[K1JJ]

The electron gyro frequency has a big effect on absorbtion of skywave signals at MF, most especially the upper MF range. The frequency varies across the globe but is in the 1400 kHz range, most often. Horizontally polarized waves suffer far more absorbtion than vertically polarized waves. So, if you want to work DX on 160, ya better use a vertical (usually).

Yep, it appears to be true in the real whirl.

W8JI talks about his experience putting a dipole at 300' for 160M.   That's over 1/2 wavelength high and would normally be very competitive with any vertical system.  He found it to be way down from his 160M vertical array for DX.

I had up a NE 3el 160M parasitic vertical array with 90 radials per vertical. I compared it against a 160M dipole at 190'. Into Europe, the verticals were always louder than the dipole by at least 10-15db.   This is surprising cuz on any other band, verticals work poorly here and I've given up on them.

So if a vertical at Tom Vu's QTH (poor ground conductivity) works better than a horiz dipole for 160M, you KNOW something else is going on.

T

[KA1ZGC]

Found it!

Recommendation ITU-R P.1147-4:

http://webs.uvigo.es/servicios/biblioteca/uit/rec/P/R-REC-P.1147-4-200708-I!!PDF-E.pdf

In the very first sentence of the introduction, they discuss the model being specific to vertical antennae. I'd forgotten all about that.

They state the model applies at 150-1700kHz, but the margin of error on 160 is probably still within reasonable amateur tolerances.

[Steve - WB3HUZ]

This doesn't get very specific, but the report states:

Measurements made in the United States of America suggest that Fig. 3 (hourly loss factor) is likely
to be accurate for frequencies near 1 000 kHz in a year of low solar activity. As the frequency
deviates in either direction from about 1 000 kHz, particularly during transition hours, appreciable
errors may result.

[KA1ZGC]

All that's really saying is that the curves for absorbtion-loss changes at sunset and sunrise provided in Fig. 3 are shaped differently on frequencies removed from 1 MHz. The endpoints are still going to be 0 and 30, and the shift duration will probably be the same (all other things being equal), just a different shape to the curve.

For a narrow swath like 160M, a reasonable shape-fudge factor could probably be determined, either theoretically or empirically; but the error itself may be benign for our purposes. I personally wouldn't expect tremendous accuracy out of any model in gray-line conditions, anway. That's the only time the shape of those curves really matters.