Wave Generation from Point Sources: Interference Patterns and Radio Aerials

[sophiecentaur]

How did you specify your model? What units were you using: wavelengths / m / inches?
It could be a GIGO thing. Happens to me all the time.
Seriously, what beamwidth were you expecting for a 2λ wide array?

 

[poor mystic]

"what beamwidth were you expecting for a 2λ wide array?"

Using a variant of the same routine with which I got a very surprising result, and simply looking at the graphed output, I get about 80 degrees bidirectional spread.
This is calculated at a distance of 250 wavelengths from the radiator axis.

 

[sophiecentaur]

It is normal to describe beamwidth in terms of angle between 3dB points and I imagine your 80^o figure is between first nulls.
For a 2λ wide parabolic antenna, the 3dB beamwidth is about 70λ/d - which would be 35^o, so what you got doesn't seem out of this world for that sort of aperture.
I suggest this http://www.kyes.com/antenna/navy/basics/antennas.htm" might give you some ideas about performance, based on conventional practice.
Of course, it is normal to calculate the far field values, rather than what you have done. It normally makes the sums easier, too.

 

[poor mystic]

No, I used the half power point just like everyone else would.
It probably seems arrogant of me but I'm not actually interested in conventional practice. I only like it when I can make people go "wow".

My next task then is to consider the feeder arrangement. I think that with a resonant feeder it might be possible to tap standing waves to power each notional dipole and provide a correctly-phased sum of signals to a receiver.
I expect I can think of a resonant coaxial feeder as a tank. Then the same current excites the entire circuit; the same potentials exist at every node. Each notional dipole receives equal power.
If I could clearly see how to calculate the impedance of each dipole I'd feel better. Naturally, I hope to find a simplifying trick.

I attach a plot showing a slice of the wave field for 21 dipoles.

 

[poor mystic]

I ought have some grace, and act embarrassed; sometimes it's good to look, then jump. I see that the "first null" is probably something you read on a meter, and means 3dB point. Is it a ham term?

I know it's usual to calculate far field but I couldn't locate my polar plotting routine, and had to grab a slice from a full-field plot. I didn't want to wait forever so I made the slice as close as I dared to hope might yet give a notion of far-field patterns.

Really astonishing wave fields are predicted to come from very many - say 500 - dipoles , or notional dipoles. I have generated graphs I fear may draw scorn and disbelief, you may generate them for yourself I guess.

I don't have much real experience of the sort I need in radio. I've gained an idea of how one matches a signal to this unfamiliar bearer (I understand information transmission practice very well everywhere but radio) and most of what is in the link is quite obvious to me, but knowledge of aerial impedances is something I do not presently have

I really think I should be up to the challenge of working things out on the hop if I need to, but some days it's like it has been all day for me today - I must have thunk too hard last night working out how to make the rectangular piece of copper into an array, and today thinking about a feeder.

Do you like the feeder?

 

[poor mystic]

please see attachment