But the tangential E field is only zero for the dielectric - conductor interface right...so that is only happening for the ground plane...in all other directions how can we set Et = 0 boundary condition?
Nice work. I'm curious as to why such a large area used for the analysis/simulation for each of the ports. Shouldn't they only be the size of the trace to ground on each port? Also, the return loss seemed poor for the centre frequency of 10Ghz. Is the ring radius/width a bit off? It would be great if you could show/share the calculations behind W,Wr,Rout,Rin etc, as the ring dimensions is the most critical for uptimum performance. Thanks again.
Thanks for your comment. To determine the extension of ports for microstrip structure, you should use the designed "Macro" for this purpose in CST to configure your simulation setup in the best way. The return loss has a good value (-25 dB) at 10 GHz. However to have the resonance at 10 GHz, you should perform an optimization on your structure dimensions. Often, the theoretical design do not match the simulation results because there are some simplifying assumptions in the theoretical formulation. You can use this textbook to design such a coupler: Microwave Engineering [David Pozar] [4th Edition]
@@EMSpectrumLab Sir, there is a question. Like considering your example and notations port 4 is 180 degree out of phase wrt port 1 right? So in s4,1 at operating the phase in CST should close to 180 degree right?
@@t012_amoghjoshi7 No, this coupler has different scenarios. If the port 1 is input, port 2 & port 3 are in-phase and the power splits equally through these ports. In this case the phase of S21 & S31 should be equal in the design frequency. In my design, this concept is verified in 10.26 GHz. For more information you can read the textbook: Microwave Engineering - David Pozar - 4th Edition - Section 7.8.
@@EMSpectrumLab 0:08 at the start of the video you state that your line impedance is 50 Ω, seems you have calculated microstrip line length to accommodate 50 Ω (parameter L) but without checking the tick on S parameter settings->Normalise to fixed impedance ->50 there might be some discrepancies. Have you tried your thingy (coupler) in real life?
This point is not necessary when your ports are set to 50 Ohm impedance (as I considered in this design). If the value of port impedance is 50 Ohm exactly, with or without normalization to 50 Ohm in the solver setup you should have the same results.
Why did you give all boundaries as Et = 0 12:27
For this structure, you should set "PEC" boundary conditions which means "Et=0" on all boundaries.
But the tangential E field is only zero for the dielectric - conductor interface right...so that is only happening for the ground plane...in all other directions how can we set Et = 0 boundary condition?
@@athiramanoharan2904 It is assumed that the coupler structure has been placed in the metallic box, so all wall boundaries should be "PEC".
Nice work. I'm curious as to why such a large area used for the analysis/simulation for each of the ports. Shouldn't they only be the size of the trace to ground on each port?
Also, the return loss seemed poor for the centre frequency of 10Ghz. Is the ring radius/width a bit off?
It would be great if you could show/share the calculations behind W,Wr,Rout,Rin etc, as the ring dimensions is the most critical for uptimum performance. Thanks again.
Thanks for your comment. To determine the extension of ports for microstrip structure, you should use the designed "Macro" for this purpose in CST to configure your simulation setup in the best way.
The return loss has a good value (-25 dB) at 10 GHz. However to have the resonance at 10 GHz, you should perform an optimization on your structure dimensions. Often, the theoretical design do not match the simulation results because there are some simplifying assumptions in the theoretical formulation. You can use this textbook to design such a coupler: Microwave Engineering [David Pozar] [4th Edition]
Good chanel! Thanks!
I would be grateful if you subscribe to my channel and stay connected with me😉
@@EMSpectrumLab actually! I have done. Keep doing. I susport your work!
How you calculate rin rout length and width
You can use an online calculator: wcalc.sourceforge.net/cgi-bin/ic_microstrip.cgi
Great Work. May i know the format of loading parameters file in the very start of video Please?
The file format is .txt. Before recording the video, I entered the parameters in CST and then exported them as a text file.
How to measure whether it is 180 degree coupler or not?
You can plot the phase of S-parameters versus frequency to investigate it.
@@EMSpectrumLabYes I tried but not getting the accurate value. So I thought might be I am doing something wrong.
@@EMSpectrumLab Sir, there is a question. Like considering your example and notations port 4 is 180 degree out of phase wrt port 1 right? So in s4,1 at operating the phase in CST should close to 180 degree right?
@@t012_amoghjoshi7 No, this coupler has different scenarios. If the port 1 is input, port 2 & port 3 are in-phase and the power splits equally through these ports. In this case the phase of S21 & S31 should be equal in the design frequency. In my design, this concept is verified in 10.26 GHz. For more information you can read the textbook: Microwave Engineering - David Pozar - 4th Edition - Section 7.8.
@@EMSpectrumLab Ok, Thank You Sir. One more doubt, what should be the phase of s21 and s31, theoretically?
12:33 seems you forgot normalize impedance to 50 Ω
What do you mean "normalize impedance to 50 Ohm"?
@@EMSpectrumLab
0:08 at the start of the video you state that your line impedance is 50 Ω,
seems you have calculated microstrip line length to accommodate 50 Ω (parameter L)
but without checking the tick on
S parameter settings->Normalise to fixed impedance ->50 there might be some discrepancies.
Have you tried your thingy (coupler) in real life?
This point is not necessary when your ports are set to 50 Ohm impedance (as I considered in this design). If the value of port impedance is 50 Ohm exactly, with or without normalization to 50 Ohm in the solver setup you should have the same results.