Hi all,

I have a very simple question about a simple dipole (not a folded one, just simple). I have used several simulation tools (FEM, FDTD...) to simulate just a dipole but never was never successful to get correct results. I just designed my dipole in computer as two dipoles each 20 cm which are connected through a 2 mm gap using an 1 A AC source and simulated up to 3 GHz. In all my simulations the first resonance occur in 750 MHz (which is theoretically correct). The problem is that the impedance between two points (gap) is extremely low, while it should be around 73 ohm.

I just wanted to ask you experts, if I have simulated my dipole correctly. I used two copper parts (very very thin) and 20 cm long. Then put a 2 mm gap between and simply connected two gaps using an AC source. Then, connected two points with a resistor (75 ohm) parallel to the source. The resistance in each dipole is also neglected since it is very thin. I calculated impedance between two points of the gap by probing two points and then abs(V1-V2) in MATLAB.

Have I modeled my dipole correctly? Can anybody let me know whay I get incorrect (very low) input impedance?

Thank you very much,

Sherry

 

Why the 75Ω terminating resistor?

Wouldn't a 0.4m total length dipole resonate at about 375MHz?

I use MMANA which does just fine at antenna modelling.

 

I am very sorry I wrote wrong. It is actually a 0.2 m dipole where each dipole is 0.1 m. Well, in all dipole configurations a coaxial feedline with 75 ohm input impedance has been used. I just tried to make model as real as possible.
How you define source in your simulations? Did you use a AC source which simply connects two dipoles with no resistors?

Thanks,

Sherry

 

Understood.

The 75Ω source model would be in series with the source - not in parallel.

I normally model the dipole as a single continuous element with the source embedded at the center [feed] point.

If you want to try the MMANA program you can download it from the web - via this link for instance.

http://www.smeter.net/antennas/mmana.php

I'm pretty sure it has a simple dipole model in the various example models included in the setup.

 

Hi Sherry:

I do all my antenna modeling with Lawrence Livermore NEC-2 Numeric Electric Code. It's a very mature and extremely reliable algorithm. (and, best of all, it's FREE!)

Have you modeled your antenna as a thin wire? If it is a "fat" antenna, the feedpoint impedance will drop precipitously....you will have a huge capacitance across the gap.

Model it with at least a 1000:1 length to diameter ratio and see if that fixes it.


eric

 

I don't know how it would work in simulation - I suppose it depends on the accuracy of the simulator - but in 'real world' effects a dipole which is directly centre fed from coaxial cable will not work well.

A centre fed dipole is balanced, coax is unbalanced. Somewhere you are effectively shorting a 'live' feed point to ground, which means the coax will have standing waves on the screen and other nasties; the coax becomes part of the antenna rather than a none-radiating feeder.

You should either use balanced feeder, or a balun transformer at the dipole, or simply make the dipole solid with the coax screen connected to the centre and the feed point tapped slightly out along one leg (say starting at 10% and move to get the best match).

 

I don't know how it would work in simulation - I suppose it depends on the accuracy of the simulator - but in 'real world' effects a dipole which is directly centre fed from coaxial cable will not work well.

A centre fed dipole is balanced, coax is unbalanced. Somewhere you are effectively shorting a 'live' feed point to ground, which means the coax will have standing waves on the screen and other nasties; the coax becomes part of the antenna rather than a none-radiating feeder.

You should either use balanced feeder, or a balun transformer at the dipole, or simply make the dipole solid with the coax screen connected to the centre and the feed point tapped slightly out along one leg (say starting at 10% and move to get the best match).

What you describe is a gamma match, which also needs some series capacitance, unless you shorten the dipole to compensate. This gets complicated very quickly, especially at UHF!

Programs like Spice are notoriously bad at modeling antennas, and I'm guessing the OP's originaly simulation was something Spice based.

NEC has excitation sources that are specifically suited to antennas. Even Mininec would handlle this elegantly.

Eric