Marconi's Penetrator 500 Eznec model

[Marconi]

SP 500 + Top Hat Modified

Well Bob, I missed a day due to another project, but here is the latest development in the evolution of my HyGain model.

I removed the wires from the bracket to the feed point and the pattern went to heck in a hand basket as I expected. I added just one wire back and the antenna responded nicely as I had seen earlier. However I noticed slightly more skewing. The plan was to split the source and force the feed point down to the physical feed point area on the antenna like you suggested.

This improved the Average Gain a bit and I think the results were just as I predicted in my post above, the actual gain went down...closer to what might be consider normal gain. It looks like, however, that the horizontal pattern increased a bit, but I'll have to look back at what was posted earlier to really see. The AG is still not good enough.

BTW, this split source feature I used also shows to have improved the raw match of the antenna, and that too might be expected. So, IMO the matching requirements look to be far simpler to achieve. The resistance ends up being very near 50 ohms, and that surprises me a little, but I think that is what Freecell use to tell us. I didn't believe him, and I didn't have any way of proving it, so I didn't make an argument, but it looks like maybe he was right.

I will try this on my I-10K and see what happens also.

Notice the split source indicator in the first image, and the improvement in the higher portion of the pattern, it is getting better:

View attachment SP 500 + Top Hat Modified.pdf

For you guys checking this out, my models does not necessarily represent what you will see in a real World installation, even though I modeled the Penetrator over what Eznec refers to as real Earth.

 

[Marconi]

SP 500 + Top Hat Modified #1

Bob, here is update #1 of the "SP 500 modified" model. The idea here was to fix the segments, as best I could to get them to equal lengths, and in this case it's about 3" each to accommodate the shortest wire in the model. I'm not sure if this is correct procedure, but I think this follows closely what I observed Cebik doing in his report on 5/8 vs. 1/4 waves, where I was able to duplicate his model results exactly.

What I did here changed the total segment count by over a hundred segments and that changed the results. It is also predictable, and it does change the gain as well. Among other things, the skewing was improved, the horizontal gain was reduced, and the vertical gain at maximum angle went up. So, to be fair when comparing this to the I-10K, I'll have to go back and do the same with that model too, and there I would expect those gain numbers to go up a bit as well.

So guys, just hang on and don't get all bent out of shape about these changing gain figures quite yet. Eznec doesn't have an auto optimize feature, so this may be the nature of tweaking such models, iteration by iteration.

View attachment SP 500 + Top Hat Modified #1.pdf

This is as far as I'm going to push tweaking this model unless I get some better dimensions for the overall length and radials. The Average Gain is just barley close enough, but I'll leave it at that.

 

[Marconi]

Fix to the I-10K won't be so easy

Heck Bob, I went to my model of the I-10K, that I posted the other day in comparison with the SP 500, with the idea to add segments as necessary and fix the feed point to a split source like I did above on the SP 500. Well, as usual, it ain't going to be so easy.

I noticed that I did not have the I-10K radiator raised up. I had never made the model to specs. The way my model is, with the radial attached to the base of the radiator, Eznec will not allow the split to occur at a joint between the radiator and the mast, because this is where there are multiple wires for the radials attached, and this is a "no-way-no."

I think I can fix it however, but it will require a whole new model to specs for the I-10K. I don't really look for much change, but if there is this could change everything that I've posted before regarding the I-10K.

It's OK for you to say, "...I knew the might be something wrong with Eddie's model of the I-10K."

 

[bob85]

from you work, have you formed any opinion on what could be causing the horizontal plot eddie?

 

[Marconi]

from you work, have you formed any opinion on what could be causing the horizontal plot eddie?

No Bob, it's hard to evaluate much while trying to tweak the model, there's too much going on trying to keep all the duck is a row, and just keep an eye out for the obvious.

The amount of horizontal RF noted in the pattern has always showed up around -20 db, and IMO that doesn't seem to be enough to make a response. Now, with modifications the pattern is more in the range of -30 db down, but it still shows up...and I've never seen it appear before.

There is a lot of current flow in this model, and in particular around this bracket, so maybe this current being under the radials has something to do with it. I just don't know, but I don't think this is like RFI on the mast. As a note the mast shows to have very little current flow at this height.

I also compared the current distribution in the tabular view to that I posted the other day. The antenna model the other day was not modified and the segments were considerably less than the latest model however. The distribution and magnitude on today's model looks much better. Not that I didn't expect change, but the big difference surprised me. If you would like a copy to compare then let me know.

I think I'll email Roy and ask him some questions.