Problem: Varactor Based Phase Shifter Design

[waqasbee4]

I want to design a Phase Shifter at 10 GHz using FR4.
I am using branchline coupler followed by two varactors at port 2 and 3 (port 1 is input and 4 is output).

Although FR4 is not suitable for high frequencies but I am getting S11 and S22 below -14db at 10 GHz

The problem is in Phase of S21,
I am getting no phase change at 10 Ghz in whole biasing range

Instead I am getting phase change at 2 GHz (around 40 degrees phase shift)

varactor diode supports upto 12Ghz frequency.

I cannot figure out any mistake

Need help urgently!

 

[vfone]

The shape in the picture is not a branch line coupler.
To bias the varactors, instead using an inductor connected to the middle of the coupler, try using two resistors (~100k)

 

[biff44]

that might be a branchline coupler if all the sections of line were around quarterwavelength long at 10 ghz.

But squinting my eyes and trying to envision a branchline hybrid out of what you drew, it looks like you have one of the varactors on the wrong port.

 

[waqasbee4]

Its is a 180 degree coupler/ rat race coupler (branch line not circular!)
I am using microstrip inductor for blocking ac and otherwise it will travel to my power supply

 

[vfone]

Its not a rat race coupler either.

Looks like it is a version of a Schiffman phase shifter.
**broken link removed**

The 100k resistor (or two) would block the AC also, and more than this, do not have cut-off frequencies or parasitic resonant frequencies as an inductor.

 

[waqasbee4]

I am going to use 100k Resistors and will inform about results.

Fingers Crossed!

 

[waqasbee4]

I checked the design using 100k resistor

Now although S-Parameters are Still fine

S11:-22 dB
S22:-18 dB
S21:-6.4 dB

But now phase shift is restricted to 1 GHz only (Previously 2.2 GHz)

Please Advise!

Need urgent help

 

[vfone]

I don't have an answer, having only the information you give.

The quadrature phase shifter can provide almost any desired phase shift, but the loads should present purely reactive impedances.

I am not sure you are using the right design for the coupler. Try to make a standard branch line coupler (as in the picture)

 

[waqasbee4]

Initially I tried Standard Branch line coupler but i was unable to achieve desired S11 at 10 Ghz
Standard Branch Line Coupler has less bandwidth, my design is a modification of standard branch line coupler.
By adding one segment to it, its bandwidth increases

So my modified design performs same (with some compromise on coupling)