Input impedance in HFFS

Hi,

Anyone know why the input impedance of an antenna changes when in case of a lumped port the plane used for the lumped port is changed in dimensions?

I am simulating a planar antenna with a slot that is fed with alumped port of 1x0.5 mm (plane). If I change the dimensions to 1x0.1 or other than the input impedance of the antenna is changed considerably...

Does anyone an explanation...


tony

Hi
lumped gap source is an approximated way to define a port
for a structure
You will assign the impedance rather have HFSS calculating it

If you lokk at the field excitation of a lumped port
you will find a Efield constant on the surface of the port
and it could be that if you do not use a reasonable surface for the lumped port
your results can change

Rugbyfun

Hi Rugbyfun,

I understand but sometimes the port of the antenna is inside the model and I think it is only possible to use a lumped port in that case...
Is it then most accurate to use a port as small as possible?


tony

And how to know the input impedance, especially using microstrip feed? whether should I take accout of the microstrip or not? Thanks!

in this case I have no microstrip connection to the antenna but the connection is made directly to the antenna structure...so I have to use a lumped port here

For lumped ports there are two main sources of error.

1.) The port forces a constant E field. There is no phase change so at high frequency this is obviously wrong. Keeping the port smaller than 0.1 lambda will minimize this error.

2.) Port discontinuity error. For example on a microstrip line, a wave port solves for the mode that propagates along the microstrip and then excites that mode. There is very little difference in the excitation and the propagating mode. However placing a lumped port between the microstrip and ground will excite an infinite number of modes. All of these modes but one decays exponentially. If any power couples to the structure through any of these decaying modes, you will get wierd results. This can be prevented by keeping the port smaller than 0.1 lambda.

To sum up, keep the lumped ports smaller than 0.1 lambda. The smaller the better really.

dear Willey,

I am keeping the lumped port as small as I can, I use much smaller dimensions then 0.1 lambda. For example going from 0.1 mm to 0.01 mm for a interested frequency of 3 GHz still gives cinsidderable difference in Z values.


Thanks for your reply,

tony

Put the width of the port same as the dimension of ur dipole...the length can be less than 0.1 lambda...