Thanks Rob, this is very nice work! Due to ground proximity (using 6m poles on SOTA), we sometimes need to shorten the wire lengths for resonance - if we could also have the ability to control the wire length parameter it would be useful.
Thanks again and 73
Geoff ZL3GA
1 Like
Re sometimes: This effect can be explained by the fact that when calculating the antenna length, it is not the mechanical height of an antenna, but rather the electrical height (relative to the wavelength) that is decisive.
This is automatically taken into account in the NEC engines and the tools based on them such as EZNEC, but not in various online calculators for linked dipole antennas …
In practice, this effect is noticeable, for example, in multi-band antennas, which are also used for harmonic frequency bands. For example, a 40/20/10m EFHW whose length is adjusted on the 40m band for a perfect VSWR will therefore be too short at the same mechanical antenna height on the harmonic bands (< on 20m and << on 10m).
1 Like
Hi Geoff,
Thanks for your kind comments! As far as needing to adjust the length of the antenna is concerned, the designer pages allow the user to alter the base, or main, frequency of the antenna in question, and decent results can be had on the multiband-type antennas (EFHW or half-square) that way by varying the frequency. It has to be understood that achieving good VSWR on several bands with these types of antenna is always going to be a compromise, no matter how you arrange the antenna.
Some other ways in which the performance of an antenna can, and will, be affected are:
- changing the height at which the antenna is erected - generally, the higher, the better;
- changing the angles of the main antenna sections - for example, for a simple dipole, the optimum included angle is often quoted as 90°, which will result in a feedpoint impedance of 50 ohms;
- choosing an optimal ground-type for the antenna - for example, if you will regularly set the antenna over damp rich soil (perhaps for POTA/WFF activations), then set that kind of ground as your default when designing the antenna;
- making sure the feedline length is optimized for the antenna - i.e. presenting no high impedances at the design frequencies;
- keeping the antenna as far away as possible from buildings or other structures which may affect (absorb, reflect) the RF radiation.
The only real test of a new antenna is when it’s set up in the field and tested with a VSWR meter or, say, a nanoVNA. The best tip here is: cut the antenna long when building it, set it up and then carefully snip it to a length which provides an acceptable performance - this has always been the case since antennas were first used over a hundred years ago. And use an antenna tuner if necessary.
Finally, the sota-antennas.com site is NOT a site to provide off-the-shelf, exact and perfect solutions to those seeking to build an antenna - the user should always understand that the various designers in the site are there to help the user to
- experiment with the settings and options to discover how a virtual antenna will behave under those constraints;
- choose an optimum set of constraints for their choice of real antenna … and then to test and adjust that antenna out in the field.
The user, with his or her own skillsets, is the final arbiter in the process.
Cheers, Rob
2 Likes
I’d just like to point out to anybody not familiar with this new site, that the calculations for the propagations patterns, VSWR curves, antenna currents and Smith charts in the site designers are all performed by a NEC engine - specifically NEC2 - running on the server, and hence the results should compare favourably with those returned by other NEC-based utilities such as EZNEC.
Cheers, Rob
2 Likes
First of all: thank you Rob for this tool!
I have a question (probably already answered somewhere):
Is it assumed that uninsulated wire (bare copper) is used for the antennas?
73 Fric YU1WC
1 Like
Basically yes, but with reservations. Chief among these are the fact that I am no expert in this field!
From what can be gathered online, opinion seems to be divided between insulation having NO effect on antenna electrical length, and it having a small lengthening effect of ~1.5% (see e.g. this copy of a document originally from L. B. Cebik, W4RNL: https://rudys.typepad.com/files/insulated-wire-and-antennas.pdf, page 6). FWIW, the ARRL Antenna Book makes no mention of wire insulation in its’ treatments on wire antennas.
The designer pages in this site use a NEC2 engine, which has no native support for wire insulation, unlike the IS card option offered by NEC4. Nonetheless, NEC2 can be “persuaded” to include a factor for wire insulation by using the LD card to model the insulation (which, by its’ very nature, is distributed along the wire’s length) as a lumped inductance (see e.g. Insulated Wires: The NEC-2 Way - Section “Insulation through LD2”).
I have not yet considered including such an option since the effects are evidently rather small, and probably negligible - especially considering that the wire for a new antenna should in any case be cut longer (a few percent) than calculated to allow for tuning and adjusting in the field.
Personally, I use an insulated wire (DX-WIRE FL, hargezogene Kupferlitze für Drahtantennenmit Polyethylen PE Isolation) from DX-Wire for my own home-brew antennas and have had no problems cutting them to size while using a VSWR meter.
HTH, cheers, Rob
2 Likes
Fantastic site, thank you for your hard work!
A few ideas:
- Allow for sharing of a specific design. I.e. a link that could be posted that would allow a design to be reproduced
- A gallery of favorite configurations with comments?
For example different vertical EFHW legs optimized for minimizing or maximizing lobes etc. - With a more open version (e.g. closer to the underlying engine) it could be fun to have weekly “pattern challenges” trying to design an antenna to meet some metric as an ongoing competition.
And also a usability (and possible server load) report:
The VSWR plot can take a long time to come up. I just counted and it took 8 seconds after the ALT-AZ plots populated for the VSWR window to populate. Before realizeing there was a delay, I fooled with parameters/recalculated many times to try to get it to work. I only figured out that this was the problem trying to take a screen shot for you and seeing that the VSWR plot populated while I was getting my screenshot tool ready. Maybe add an “VSWR loading…” indication or something similar? I suspect I’m not the only one who has had you re-run the simulation looking for a plot like that.
Again, love it, and thanks again for sharing this with the community.
73
Tony KM6AM
2 Likes
You’re welcome!
So, how much are you prepared to pay me for implementing these very-nice-to-have options, while the site is still in the process of being developed? Of course, I’m joking… But this site is still very much in development, let’s not try to run before we can walk.
Anyways, to your first point, there are options already in place in a couple of the designer pages (the linked-dipole and delta-loop designers thus far, if my memory serves me right) for antenna dimensions to be exported in CSV format - just click on the “Details…” button, and then “View CSV…” button to view the details, which can be copied in whole. These dimensions can be used in e.g. EZNEC to model the same antenna there.
To your second point, there are options in place in each of the designers to save the antenna one is working on to a “Favourites” window (browser local storage holds them for you), and one can build up a set of favourites in this way. In addition, it’s possible (click the “Examples…” button in the “Favourites” window) to load a few “default” antenna designs from the database and to import one or more of these into the favourites window. Double-click on a favourite, or click the “Load >>” button to load it into the main design area.
Indeed, and this is covered in a few places in the help or the about page. In addition, there is a little
image beside the “Bands range” controls to be found in a couple of the designers - it’s not there for decoration! Hovering the mouse over the little image will (should!) cause a popup displaying this message, which should be self-explanatory:
The reason theses plots take so long is that, basically, we’re not dealing with a dedicated optimized super-fast executable program running on the user’s machine, but are instead dealing with a distributed set of tools in the internet, some of which actually run as command-line utilities on the server. For VSWR plots over several bands, a separate set of results (in essence a complete, but pared down to basics, antenna) for each frequency from plot start to plot end has to be calculated, the end results built into a NEC text file which then has to be opened, analyzed, data-reduced and passed back to the front-end running in the browser. It’s the part which calculates the data which takes so long - the rest is very fast.
FYI, here’s an example of a NEC output text file for a 40-meter EFHW calculated over the bands 40m through 10m: https://www.sota-antennas.com/sample_nec_output.txt … each block starting with “--------- FREQUENCY --------” is a full, but pared down, antenna calculated at that frequency.
It is what it is - I’m not particularly pleased with the long run-times myself, and have done my best to minimize them as far as is possible, having reduced typical run-times from over a minute to just a few seconds: that’s it.
There actually are commands in the Javascript which attempt to replace the normal mouse cursor with a “wait” cursor while these processes run - I’ve used these for years successfully, but it would appear that the latest browser versions seem to ignore the command completely 
. I’ll try to come up with a better “loading…” indication, but it will only be showing an estimate.
Cheers, Rob
Thanks Rob. I will say the “suggestions” from my perspective are 100% nice to have imagining cool ways your work could be built upon down the road. Like I can totally see someone here using your tool as a great way to share an antenna design. I made a post a few months back with eznec results where I did some optimization on the vertical/horizontal sections of an efhw trying to smooth out lobes. The suggestion came about because your tool would have been a better way to share that finding with the community. If it were someday possible to link into a specific model then that would just be cool:-). But that’s probably scope creep anyway relative to what you’ve set out to accomplish (and have done so exceedingly well!)
Re load times, the processing time is what it is, honestly doesn’t bother me in the least. And I admit I started playing without reading directions. I think for me the main issue was that the window for vswr was totally blank with just a border and looked explicitly like it was a broken link. If you just put in a place holder text in the plot frame saying “VSWR calculating, be patient” I would have been free from confusion. I think I won’t be the only one concluding something is broken as is now, the pop up goes away, two of the plots show up, but the last one is just… empty…
Points taken, and - of course - they were already in the works, I’m neither blind nor lazy, and I’ve been doing this for 25 years. I only have so much time in the day for anything at all, I’ve already spent 12 full hours today on dev in the site and writing replies here - I’m bushed.
Hope the site can be of some use to you, if only in a limited way.
“Have a great day”, Rob
That is beyond the remit of my site. In other words, that’s something you’re going to have to realize for yourself and your “community”. I see you’re already familiar with EZNEC, so that should be an easy task for you.
1 Like
If I didn’t think the site was awesome I wouldn’t have taken time to provide feedback! No criticism implied or intended. I have also been stuck doing some UI design (as a non expert) and found user feedback beyond invaluable, so many things I thought were obvious turned out not to be, one of which was a similar lack about process feedback during a step that could take up to 30 seconds to process a large block of data. I haven’t made a site like yours myself but appreciate the huge amount of work you must have taken!
1 Like
There’s a note at the top of this thread which needs to be fully appreciated. It runs thus:
This was intended to convey the message that the site is not yet finished, and will likely have a few, shall we say, unpolished edges. Like all helpful texts, it can only actually convey that message when it is read and understood.
I’m grateful that so many actually did take the time to read it, and did understand it.
I only made any suggestions because you had posted that 
I read it as an invitation to constructive criticism/suggestions! My apologies if I was incorrect in that assumption! Thanks again for this, it will be a value to the community even if you leave it exactly as is right now.
You are most welcome, my friend.
1 Like
This has bugged me also for a while now, but I left it to one side while concentrating on more pressing issues. Your remarks have given me impetus to come up with a workable solution to this, and I think I’ve managed to come up with something which should prove acceptable. At the moment, it’s only been implemented in the EFHW page, but the other pages will follow shortly… And I also fixed a bug in the EFHW Smith chart.
Cheers, Rob
2 Likes
Rob,
Thank you very much for this great tool, very useful and inspiring!
73 Stephan
2 Likes
As promised, I’ve managed to get the OCFD (off-center fed dipole) antenna page running:
New page OCFD Antennas now included in the sota-antennas website
There are still one or two dialogs (Help, Export, etc.) not yet available in this new page, but the “important” stuff is running well enough. Not covered/included are Carolina Windom antennas - which are a type of OCFD antennas - but they may make an appearance in the page at some time…
Enjoy!
Cheers, Rob
1 Like
Another addition to the sota-antennas.com site - a Moxon antenna designer:
for those who like their antennas beam-shaped. The page is brand new, and at the moment a couple of popup dialogs need attention, but the designer itself works as it should.
Cheers, Rob
2 Likes