[SOLVED] hybrid quadrature ring coupler for 30GHz implementation

[yefj]

Hello , in regular quadrature coupler we have on the output two ports which equal in power and 90 phase difference.
In high frequencies its problematic because the wavelength is very small.
so i thought to use a ring coupler shown below .but he problem is the output is 180 degrees where as i need 90 degrees.
Is there a way to convert the ring configuration bellow so i will have exactly 90 phase outputs and isolation property like in the branch line coupler?

 

[PlanarMetamaterials]

There are only a few differences between the couplers; the phase difference is one of them. They are both on the same order of electrical size, and in general, the ring hybrid is just slightly larger.

If you are just looking for a bigger quadrature hybrid, I suggest investigating replacing the 1/4 wavelength sections with 3/4 wavelength sections. I would think that this should function similarly.

 

[yefj]

Hello planar meta material,you think i should do the rat race as shown below?
if my input is at port one why in this configuration ports 3 and port 2 will be at 90 degrees phase difference and 4 will be as isolated?
Thanks.

 

[PlanarMetamaterials]

No, I was suggesting you do that with the quadrature hybrid - since in my interpretation of your problem, you are worried that the quarter-wave device is too small. Is that an accurate assessment of your issue?

 

[yefj]

Hello PlanarMetamaterials, i am thinking that production tolerances will ruin the regular quadrature coupler.
Thats why i trying rat race ,because its less vurneble to the production tolerance because of its shape
If i am correct could you please help me understand if its possible to convert the rat race of 180 into 90 degree? keeping the isolator property?
Thanks.

 

[PlanarMetamaterials]

I'm not sure why its shape is relevant to you. The line widths have the same values in both cases... do you mean in the images you showed one is rectangular while the other is circular?

The quadrature hybrid arms can be whatever shape you want, so long as they are the correct lengths/widths. There will be various tradeoffs to each shape. I've designed circular ones; they're generally considered to possess higher bandwidth than the regular rectangular ones.

 

[yefj]

Hello Planarmetamaterials,I want to design a ring coupler as shouwn below.
How whould you reccomend to change the length of the valuues shown in the red arrows so i will have 90 degrees between 3 and 2 and pot 4 will be isolation( like in quardarure coupler)?
Thanks.

 

[PlanarMetamaterials]

Hi yefj,

I'm still not sure what your issue with the quadrature hybrid is, but you can list the differences between the rat-race and the quadrature hybrid by inspection:

1. The λ/2 section between ports 3 and 4 should be reduced to λ/4.
2. This section between ports 3 and 4 should have its impedance as Zo, an increase from Zo√2.
3. The section between ports 1 and 2 should have its impedance as Zo, an increase from Zo√2.

 

[yefj]

Hello planarmetamaterials ,so every section is lambda/4 and every width is Zo as shown bellow.
then it will act as quadrature branch line coupler?
Thanks.

 

[PlanarMetamaterials]

Hello planarmetamaterials ,so every section is lambda/4 and every width is Zo as shown bellow.
then it will act as quadrature branch line coupler?

No, two of the branches in the quadrature hybrid coupler have impedance Zo√2: in your diagram, the section between ports 1 and 3, and between ports 2 and 4.

 

[yefj]

Hello planarmetamaterials, so if i understand correctly you suggest:
between 1 and 3 sqrt(2)*z0
between 3 and 4 z0
between 4 and 2 sqrt(2)*z0
between 1 and 2 z0
as shown in the diagram below.Is that your meaning?
Thanks.

 

[yefj]

UPDATE:
Thank you very much it worked

 

[PlanarMetamaterials]

Great, I'm glad you got what you wanted.