antennas

I have to connect together two air dielectric coaxial cables with same external diameter, but different diameter of the inner conductors.
My question is: how can I simulate this kind of connection?
I have a microwave CAD but unfortunatly the program doesn't work on this kind of component (I think is possible simulate it with a capacity, but I don't know how to relate the capacity to the geometrical dimension of the lines).
Could someone help me?
Thank you.

The coaxial cable impedance is dependent on the outer/inner diameter relationship, see

http://www.madisoncable.com/RFCoaxialCables.htm

for formulas.

Hi max,

You have a junction between two lines of different characteristic impedance. If the size of the junction is much less tahn the wavelengths in the lines (this is surely your case), tou can predict theoretically the behaviour of the union (reflections, input impedances, etc) considering ideal lines. You can use formulas or Smith Chart.
Regards

Z

Zorro is absolutely correct if you don't need to be concerned with higher order modes, that is, when this junctions is far away from other discontinuities. Otherwise, those 3-D field solvers like MWStudio, HFSS, etc. can simulate your problem accurately.

Since the impedances of the two cables are differenent (OD/ID ratios different with same dielectric constant) , you will get more reflections from that source than from the small shunt capacitance from the diameter differences.

you may like to try xfdtd or empire...They are good...

if you just care about the junction, you can check some basic formula. I think this is accurate enough for the case with same meterial in two Tl.

I have looked this up in an old 1955 standard text and they say that the effective shunt capacitance will be under 1 pF in even extreme diameter differences and usually much smaller. This will cause a shunt reactance of several thousand Ohms in the VHF region. This will produce less transmission imperfections than the manufacturing tolerances of the cables.

I think you can also adopt the same techniques as Koul's books on finlines to solve your problem --- mainly mode-matching method....Very easy to write....

Mode matching also has its "relative convergence" problem. If there is no other discontinuities, lumped model is the way to go.

I'm looking for design equation about RF losses in squared or rectangular coaxial lines..

Can anyone help me?

Thanks in advance
Lupin

check this out