4 x 17 LFA yagi assistance please

[VK4OX]

Hi again Leigh,
Your plot of Sun noise vs degrees for the last 3.8 degrees of Sun time is interesting and, perhaps, confusing. It would appear that there is a "hole" in the main lobe suggesting a phasing problem with one or more individual yagis in the array.. Then again it could be just the vagaries of ground gain coming into play... Too many variables clouding the issue.

It looks to me like you will have to take the array down. If that is so, then my previous comments apply... I'd like to see a plot of sun noise with the array elevated at least 60 degrees above the horizon.... Looking at ZL1RS's computer simulations, the second sidelobe is at about 55 degrees relative to the main lobe. Elevating the array to 60 degrees would get the second sidelobe out of the ground noise/ground gain zone and would minimise errors.

I'm a bit worried about the IC910's S meter .. 5 S units (15dB or 30dB??) of Sun noise seems like a lot to me.. I think An RMS voltmeter across the speaker audio is a better indicator. Analogue voltmeters should have the familiar -20dB...0dB...+5dB scale giving 25dB range without switching ranges.. The 0..+6 dB section of the scale is spread over half the meter's scale giving good resolution to about +/- 0.5dB in this part of the scale. If you do not have access to this type of meter, then a 0-10dB step attenuator in 1dB steps in series with a 0-50dB step attenuator in 10 dB steps could be inserted between the array and the IC 910 (no external preamp?), giving you 0-60dB in 1 dB steps.

Take a baseline measurement at night... This establishes your "cold" sky measurement, "N". As the sun rises and traverses the sky, the second sidelobe, first sidelobe and boresight, you will get various values of "S" (Signal) + "N" (Noise) i.e., S+N.

What you require is the ratio S/N. S/N =(S+N)/N -1. So you have to preform that mathematical operation on your data.

For example, suppose you get the following raw data:

Second sidelobe 0.1dB above N

First sidelobe 1.0dB above N

Main Lobe 10dB above N

It would appear that the first sidelobe is only 9dB down relative to the main lobe and the second sidelobe only down 9.9dB, but the following "corrections" must be applied:
(S+N)/N = 10dB. Converting this to a numerical ratio of noise powers, P1/P2 = antilog(10/10)=10. (S+N)/N - 1 = 10-1 =9.
Converting this back to decibels 10log(9)= S/N = 9.54dB


(S+N)/N = 1dB.. Converting this to a numerical ratio of noise powers, P1/P2 = antilog(1/10) = 1.2589. (S+N)/N - 1 = 0.2589.
Converting this back to decibels: 10log(0.2589) = S/N = -5.868 decibels


(S=N)/N = 0.1dB. Converting this to a numerical ratio of noise powers P1/P2 = antilog(0.1/10) = 1.02329 (S+N)/N - 1 = 0.02329.
Converting this back to decibels: 10log(0.02329) = S/N = -16.327dB

So the first side lobe is down by (9.54 +5.86) = 15.4dB relative to the main lobe and the second sidelobe is down (9.54 +16.33) = 25.87dB

All this reminds me just how easy it is to do the sums and the computer simulations. The hard work is always the hardware!

Good luck.

73, Adrian. VK4OX.

[VK4APN]

Leigh VK2KRR

Would be interested to know what the issue was with the array and are you happy with it now?

Also, could you detail the feed arrangement on the loops? I.e. a pawsey stub, pseudo pawsey stub or something else.

Tnx Paul vk4apn

[VK2KRR]

Tnx Adrian and Bob for the extra comments.

Paul, I am just going to live with it a bit and see how it goes, I've already spent way to much time sorting things out with it. It certainly doesn't have the f/b ratio claimed. I am probably just lucky my location is already fairly quiet, it does pick up bigger signal from the rear than what the single 17b2 Cushcraft yagi did, and also off the side. But having said that, the main lobe is quite pointy to where even 10 deg can make a decent difference in signal and does appear to have a little more gain than the single 17b2, hard to get a good gauge on this with Tropo variability, need a bit more time.

There is no pawsey stub, I believe this has been replaced with the forward most side of the loop having an aluminium (conductive) mounting clamp. There is also a 4 loop choke wound into the coax near the feed.

[VK2OMD]

... There is also a 4 loop choke wound into the coax near the feed.

This is commonly known as an Ugly balun.

The graph above shows the common mode impedance of an Ugly balun using dimensions somewhat similar to those described.

It can be seen that it has stunning common mode impedance over a narrow frequency range, and if its self resonance is engineered to coincide with the operating frequency, performance is excellent.

The real question about their effectiveness for a narrow band application then comes down to where their self resonance is located.

Like almost all balun sellers, InnovAntennas does not publish measurements of common mode impedance (not that I have been able to find).

Owen

[VK2KRR]

Owen, very little information came with the antennas, so you know as much info as me about any measurements they may have conducted.
I was simply told to wind 4 turns or coax around a household aerosol can to make the choke balun, no specific dimension were given.

[VK2OMD]

Owen, very little information came with the antennas, so you know as much info as me about any measurements they may have conducted.
I was simply told to wind 4 turns or coax around a household aerosol can to make the choke balun, no specific dimension were given.

Ah yes, Justin's 'standard aerosol' can trick!

Ok. I know he sells these things (Ugly baluns that is, not standard aersol cans), and I suspected from some of your facebook posts that yours may have been wound by yourself as part of making a one piece tail for the antenna to power divider.

Well, I have illustrated the issue with Ugly baluns, they can be superb... if tuned.

I see signs that Justing doesn't consider baluns a very important element of an antenna system. Though he makes a lot of his designs having low sidelobes (low noise), he doesn't seem to acknowledge that the model patterns are very dependent on symmetric current drive of the antenna. BTW, he is not the only antenna designer / manufacturer / seller that does that.

The graph I produced above is for a 4 turn coil on a 50mm diameter former with similar dimensions to RG8 cable... and my suspicion is that it is resonant way below the band. Of course, I would defer to actual measurment, and if I get some time, I might wind one (though not on a 'standard aerosol can') and measure it. (I can't duplicate Justin's or yours because neither have specific dimensions and it matters.)

In the text under that pic, he does say "Now roll the can until you have 4 to 5 turns (50/70Mhz 2 is OK for 144MHz and 1 tight turn on 430Mhz." Perhaps my estimate is not too far off!

None of this discussion about Ugly baluns is to offen the large (huge) number of hams who use Ugly baluns (http://www.hamuniverse.com/balun.html) and are convinced that they are the best thing since sliced bread, so save your breath, I know they can be very good over a narrow frequency range. Indeed I have used carefully tuned Ugly baluns as a single band balun on HF.

Back to your antenna, my guess is that the baluns do not account for your reports, the operative word is guess... and problems aren't solved quickly and directly by guessing your way to the answer. I would not be getting the crane out to pull them down though!

Owen

[VK2KRR]

Owen, thanks for your comments, all noted. I wont pull them down at this stage. Like to test them through some good tropo openings and see how they perform at that stage.

[VK2KRR]

One thing I may have to do regarding the H frame assembly is to cut maybe 300mm from the bottom of the main mast that sits into the rotator, there is just to much stick out above the top of the tower before the horizontal H frame tubes begin. Its making the whole H frame to unstable when the winds pick up and may help make the yagis more level, as is currently an issue of the front of the array pointing down slightly.
The other thing with this is where the H frame is currently sitting, the lower two antennas are about level with the top of the tower, so this would drop them down 300mm below the top of the tower level (tower top is 700mm square). Could have a bad effect on performance, only one way to find out I guess.