HI,
I am working on designing a wide band Wilkinson power divider (2-6GHz), I designed three section power dividers with the same band but different shapes ( leaner , circular and rectangular structure please check attached file) using CST. I noticed that the insertion loss for all of them falls bellow -4.5dB in the upper band.
Could anyone suggest how to increase the insertion loss ?
Thanks
Hi Ammar, Hi all contribuors,
I can suggest another design (see the attached file). This design enhances the traditional Wilkinson design by adding two half-wave transmission lines to connect the integrated resistor. These additional lines of λg/2 electric lengths create the ideal framework for the isolation resistor, overcoming undesirable mutual coupling between the transformer arms and ireducing therfore losses.
I hope this is useful
Dear Ammar Hussein Ali,
You could increase the losses without creating nismatches to the ports. Therefore, you must act on the components with losses and, if it is necessary, you must act on the reactive components of the impedances as well.
You coul put an attenuator at each port; but I do not think that it is your case.
Hi
I think that the question is not correctly written
I you want to increase the insertion loss (meaning having more loss) you just have to add attenuation (attenuator) at each branch as Gifuni suggests, but I'm not quite sure that it is what you want.
I think you want to increase the magnitude of S21 (having mag(S21)>-4.5 dB). Increasing the magnitude of S21 is equivalent to decrease the insertion loss. Keeping in mind that the theoretical limit is -3dB in your case.
One way is to use an other substrate. A substrate with better performances, meaning less loss.
You can also use a substrate with a higher height, the lines will be wider and consequently the loss will be lowered. You can increase the thickness of the line as well.
Best Regards
Hi
Thanks a lot for your answers. You are correct Xavier. I ment decreasing S21.
I'll try to use other substrate with less losses as changing the thickness is not applicable as the line width will increase and it will not be possible to construct the last stage.
Regards. gards
Circulators do that, but they are not nearly as cheap, and they are cut for 1:2 frequencies.
Hi Ammar, Hi all contribuors,
I can suggest another design (see the attached file). This design enhances the traditional Wilkinson design by adding two half-wave transmission lines to connect the integrated resistor. These additional lines of λg/2 electric lengths create the ideal framework for the isolation resistor, overcoming undesirable mutual coupling between the transformer arms and ireducing therfore losses.
I hope this is useful
Hi Djilali,
Very interesting. I will try to redesign your power divider to make it work on my band and share the results here.
Thanks a lot.
Hi Ammar,
I am glad to hear that you're working on the new ,easy and compact design.
Share your results with us.
Thank you,.
Hellow Ammar Hussein Ali ,
1. I am a PhD student.
2. My research is about microstrip array antenna at frequency above 100GHz.
3. I designed at CST software Wilkinson 1:2 power divider for 110GHz that have insertion loss of (-5dB).
4. Can we cooperate academic?
5. Can you share your design at CST file?
You can sent it to my email:[email protected].
- Badges
- Science topic
- Similar topics
- Engineering
- Telecommunications Engineering