Continue to Site

# Coaxial Coupler: What is the best way to design?

Status
Not open for further replies.

#### sinatra

##### Full Member level 4
Hello all.
I would like to design a coaxial coupler. I will use it as a reflectometer.
Is there any paper where it is shown how to calculate the coupling coeficient between the center conductor of a coaxial line (air dielectric) and a second conductor beside it (far from the center)?

Can anyone help?

Thank you
S.

Hi,

did you consider coupling two equal coaxial cables trough a longitudinal slot in the outer conductor? You can vary the slot width to adjust the coupling and length to tune the central frequency. You can even use two semi-rigid cable and solder them once you achieved the desired performance as a first experiment. I think this will work better than two inner conductors, one central, the other offseted having the same cylindrical outer conductor, if I understood you.

flyhigh

You can make one emperically. Since the coupling factor is probably small, size the center conductor and rectangular box for the impedance you want.

Then the coupled line is small diameter and close to the outer conductor. You can adjust the spacing to the wall and the wire diameter to get the coupling you need.

Scientifically the coupling is a function of the common mode impedance, differential mode impedance, and length. From the coupling you want you can get the ratio of these two impedances. Then from equations for coaxial and for parallel wires in a shield you can back out the wire diameters and spacings.

There is the question of do you really need a directional coupler. It has the advantage of making the driving pulse lower amplitude on your receiving circuit but the disadvantage is the rising frequency response which will distort the pulse shape. You could easily put your receiving circuit directly across the line and have some form of protection circuit, like a diode shunt, to keep it from being damaged or saturated by the driving pulse.

Thanks Flatulent and Flyhigh.
In fact I have already made some couplers empiricaly as you described.
It works alright just making a standard coaxial transmission line and then adding another wire offset from the center conductor. This is the standard way to make directional couplers that are used in reflectometers.

But I was in fact looking for a paper that gives aproximate or closed formulas relating the coupling and the geometrical dimensions of the transmission line. Something similar as the paper from Kinshining and Jansen for the case of microstrip directional couplers.

Unfortunatly the coaxial couplers are much older. The papers relating coaxial transmission lines are also old and sometimes difficult to find......

Do you have some paper about coaxial couplers?????

Greetings

S.

Ps: Flatulent, I'm working in the frequency domain, so pulse distortions would be no problem... directivity isolation and match are the main parameters

I think that what you are looking for is given in many textbooks. I saw the original paper on the subject. You assume TEM mode and then the common mode and differential mode capacitances between the conductors per unit length can be calculated from the geometry. From these the coupling can be calculated.

You could also make a hybrid with transformers much like that used in the old telephone days. You can probably get a decade of frequency range and have a set to cover different decades.

Status
Not open for further replies.