Combiner Phase Performance

Sem z

Hello

 

I have designed 4:1 RF combiner PCB in FEKO and for some reason the phase results (I look on S51, S52, S53 and S54 phases values) look very strange.

They also not symmetric, although the design is symmetric.

Attached the design.

What can be a reason for these strange results.

Thanks.

Jaehoon

Please also attach your cfx file, if possible.

Sem z

Hello

 

Attached the CADFEKO design file.

Jaehoon

After reviewing your .cfx file, I found that the inner and outer conductors of the Waveguide port 5 are connected to the same transmission line as below.

I don't know why you changed the direction of the coaxial feeder, but please make sure the outer conductor should be connected to the ground plane.

Sem z

Hello

 

I have fixed the connector.

I still see quite strange results where:

1. S11, S22, S33 and S44 values are relatively high, although S51, S52, S53 and S54 values are ok

2. There is a significant phase difference between the channels, although the combination network is symmetric

 

What can be a reason for these issues?

 

Best Regards

 

Semion

Jaehoon

Sem z said:

Hello

 

I have fixed the connector.

I still see quite strange results where:

1. S11, S22, S33 and S44 values are relatively high, although S51, S52, S53 and S54 values are ok

2. There is a significant phase difference between the channels, although the combination network is symmetric

 

What can be a reason for these issues?

 

Best Regards

 

Semion

I simplified your simulation model with infinite planar layer and microstrip ports as attached.

The simulation results show not-good matching characteristics and  some phase difference among the ports. (The combiner geometry is not perfectly symmetry. Please redraw to modify these.)

Please make sure the transmission line's impedance are well designed to be matched with 50 Ohm and please locate the center of the geometry to the origin in order to use "Geometrical symmetry".

If you apply the symmetry for your geometry, then you'll have more balanced results, because of the balanced mesh.

After that, please add the coaxial cable as a feeder. Then you can concentrate on the feeders.

Thank you,

Jaehoon

Sem z

Jaehoon Kim_22316 said:

I simplified your simulation model with infinite planar layer and microstrip ports as attached.

The simulation results show not-good matching characteristics and  some phase difference among the ports. (The combiner geometry is not perfectly symmetry. Please redraw to modify these.)

Please make sure the transmission line's impedance are well designed to be matched with 50 Ohm and please locate the center of the geometry to the origin in order to use "Geometrical symmetry".

If you apply the symmetry for your geometry, then you'll have more balanced results, because of the balanced mesh.

After that, please add the coaxial cable as a feeder. Then you can concentrate on the feeders.

Thank you,

Jaehoon

Hello

 

Thanks for feedback.

How is it possible to see the port input impedance in FEKO, to check its 50 ohm matching?

Jaehoon

Sem z said:

Hello

 

Thanks for feedback.

How is it possible to see the port input impedance in FEKO, to check its 50 ohm matching?

You can add "Voltage source" to the microstrip/edge/wire port in order to check the input impedance at the port in POSTFEKO.

However, the waveguide port only provides the reflection coefficient due to its characteristic.

Thank you.

Sem z

Hello

 

I have redesigned the combiner to be like Wilkinson bridge.

Attached its design.

I would have expected that the phases between the 4 input ports and the output port will look the same but they look very strange, like in the attached picture.

Also the Sxx of the ports is quite big. 

What can be the reason for both issues?

Thanks.