Confused when using EZNEC...

[OZ6ABM]

I'm somewhat baffled by the calculations in EZNEC. I am using the demo version. When calculating a 2 or 3 element quad for 28 MHz it's telling me that the best SWR and dB gain is when the wire length of each loop is a half-wave.

Looking at other sources I see the loop is a suggested full wave (ARRL handbook for one), the problem is that when using a full wave in EZNEC the dB drops from 7 to about 3½ and the SWR goes from 1:1,2 to up to 1:8 ?

Another issue I'm trying to get to grips with is how the resonance of the antenna looks in the design frequency? Is there any easy way to see this before actually buidling and measuring ?

Thanks and hope someone can help

73
Robin
5P5R/OZ6ABM

[W8JI]

I'm somewhat baffled by the calculations in EZNEC. I am using the demo version. When calculating a 2 or 3 element quad for 28 MHz it's telling me that the best SWR and dB gain is when the wire length of each loop is a half-wave.

Looking at other sources I see the loop is a suggested full wave (ARRL handbook for one), the problem is that when using a full wave in EZNEC the dB drops from 7 to about 3½ and the SWR goes from 1:1,2 to up to 1:8 ?

Another issue I'm trying to get to grips with is how the resonance of the antenna looks in the design frequency? Is there any easy way to see this before actually buidling and measuring ?

Thanks and hope someone can help

73
Robin
5P5R/OZ6ABM

I have no such problems.

Can you post a link to the file you have, or did you try the built in Quad antenna that comes with EZnec? I'm not sure if the demo has the 15M quad or not. My EZnec and the demo version both handle all my quads perfectly.

[OZ6ABM]

I can email the file to you, but would need an email address to do that - your help would be appreciated. I actually created the quad from scratch.

73
Robin
5P5R/OZ6ABM

[AB8XE]

Are you using square corners or rounded? How many segments are you using?
There are a few modelling limitations with any nec simulation, especially with corners. The important thing is you spotted something unusual happening, instead of just accepting the results. There are a few traps and pitfalls to avoid , too many for a single web post. Google W4RNL's webpage and enjoy his informative writings on the art of antenna modelling.

[OZ6ABM]

OK I have managed to resolve the loop length issue ... I must have been sleeping because the negative coordinate values now resolve to give the correct length. I'm using square corners and only one segment per side (limit in the demo version of 20 segments). I'm still baffled on how to evaluate the resonance - and my complex mathematics is about 30 years old and shot.

[AB8XE]

It's a common technique to radically change the number of segments to verify your model is correct. If you change to 5 segments a side and see any major change, its a sign you didnt use enough segments.

[W8JI]

OK I have managed to resolve the loop length issue ... I must have been sleeping because the negative coordinate values now resolve to give the correct length. I'm using square corners and only one segment per side (limit in the demo version of 20 segments). I'm still baffled on how to evaluate the resonance - and my complex mathematics is about 30 years old and shot.

What do you want to do with resonance? You know EZnec plots the SWR in graph with frequency. That can be as many or few frequencies as you want. It can also tell you the source impedance at a single frequency. Resonance is where j op crosses zero. Like 35 j0

[OZ6ABM]

Well resonance is not necesarilly where lowest SWR is. So the fact that the graph dips at a specific frequency does not mean that the antenna is radiating best at that frequency. Put it another way, if I stuck a dummy load on a pole and saw SWR was low on the frequency of interest, that doesn't mean that it's a good antenna.

Ideally the frequency of resonance should be close to the lowest point of SWR (where the antenna is designed for). I have an AIM from Array Solutions and that plots both SWR and resonance and you can see the difference between the plots. However with the AIM you need the physical antenna.

[W8JI]

Well resonance is not necesarilly where lowest SWR is. So the fact that the graph dips at a specific frequency does not mean that the antenna is radiating best at that frequency. Put it another way, if I stuck a dummy load on a pole and saw SWR was low on the frequency of interest, that doesn't mean that it's a good antenna.

Resonance means nothing so far as antenna radiation anyway. Radiation efficiency of an antenna does not change much because reactance is zero. All that means is the feedline is matched, so resonance only really affects feedline losses. Pure reactance is lossless.

To find radiation efficiency of the antenna you must calculate the 3D pattern. Then, at the bottom of the main window, you will see an efficiency. If the efficiency is ever above 1.0 or 100 percent a guideline has been violated, so it is also a good indicator of a correct model (along with guideline check).

Another way to very roughly find peak efficiency is with gain. Gain is usually what we want to know anyway, but the only good way to know efficiency is to read the actual efficiency by doing a 3d plot.

Ideally the frequency of resonance should be close to the lowest point of SWR (where the antenna is designed for). I have an AIM from Array Solutions and that plots both SWR and resonance and you can see the difference between the plots. However with the AIM you need the physical antenna.

It is impossible for any instrument like that to measure efficiency. We can get an idea of efficiency only two or three ways:

1.) We know the absolute thermal rise of the system and applied power

2.) We know absolute radiated signal level in every direction and angle, and the applied power. The net radiated power and known applied power tells us the power wasted as heat in loss resistances.

3.) We enclose the antenna in a lossless or very low loss enclosure that is large enough to be outside the antenna Fresnel boundary. We then measure the resistance at the feedpoint when the box is in place, and the box removed. I knew this from theory at EM Radiation courses but it had no name at that time, but I think now it might be called a Wheeler Cap. Someone else more recently involved with this method might know what name this measurement method now has.

Note all of these require knowing either power or heat escaping the antenna. Reactance is never important except feedline or power factor correction losses in an external matching system.

EZnec tells you everything you need to know, but does not tell many useless things except perhaps "take off angle". Feedpoint or source impedance resonance is nothing but feedpoint resonance, and does not in any way in real life or the model indicate antenna efficiency.

73 Tom

[W5DXP]

I'm using square corners and only one segment per side (limit in the demo version of 20 segments).

Under "Outputs", click on "Segmentation Check". Does it say that the segments are too long? My EZNEC+4.0 says the segments are too long by about 125%. I used one segment per 1/4WL wire on a 10m full-wave loop.

With 1 segment per wire: Resonant Freq = 28.7 MHz, Feedpoint = 142 ohms, Gain = 7.99 dBi

With 9 segments per wire: Resonant Freq = 28.9 MHz, Feedpoint = 123 ohms, Gain = 8.7 dBi

Trying to design a functional beam using the demo version of EZNEC is somewhat problematic, IMO.