"T" divider in stripline
I'm designing a T divider (S12=S13= -3 dB and S11< -25 dB).
Measuring the isolation between the output ports (2 and 3) I see it's -6 dB and s22=s33=-6 dB.
Is it possible? I have been reading that on a 3 ports component ("T" divider) it's not possible to have the s11 = s22 = s33 = 0.
To have output isolation I have to use a Wilkinson Divider or a Rat race or a branch line.
Please can you confirm this?
Regards
Lupin
thats cool. lol its better than the wanted results.. cheer up :P
You do not describe your circuit, but a basic T junction, to be matched at the input, has to do an impedance transform from a load of 50 ohms to an impedance at the junction center of 100 ohms. The two 50 ohm output loads, transformed to the junction center, look like two 100 ohm resistances in parallel--hence S11=0.
So power coming into the output port gets to the junction center, it sees three paths to exit the device:
1) reflect back to the output connector you are feeding
2) go to the 50 ohm load at the "input" connector at port 1
3) go into the 100 ohm load that the other output connector appears like at the junciton center.
So the power that arrives at the other output connector is 6 dB down.
You would probably also notice that the S22 and S33 are poor.
In order to get a better isolation from one output connector to the other output connector, the signal needs to take two paths and you need to configure the two paths so the two signals arrive at the other output connector out of phase by180 degrees, and of equal magnitude.
biff44 do you have the basic theory of that or can you give me the picture or diagram....
thanks
David Pozar discusses this in his book "Microwave Engineering" It might be helpful to look over his comments.