Simple microstrip simulation and meshing

How to use openEMS. Discussion on examples, tutorials etc

Moderator: thorsten

Post Reply
jockeosth
Posts: 13
Joined: Wed 11 Jan 2017, 20:24

Simple microstrip simulation and meshing

Post by jockeosth » Fri 05 May 2017, 21:59

Hi!

I have a simple microstrip with lumped ports. For some values on the resolution of the mesh and/or widths of the strip the simulator converge and for others it get stuck.
Probably I miss something about the mesh? For example changing the strip_width from 240 to 185 makes the simulator stuck. Any advice?
I try to use the -1/3 2/3 rule in the meshing.
Best regards
Joakim
Attachments
Microstrip1.m
(3.66 KiB) Downloaded 193 times

jockeosth
Posts: 13
Joined: Wed 11 Jan 2017, 20:24

Re: Simple microstrip simulation and meshing

Post by jockeosth » Sat 13 May 2017, 20:12

Hi!

Late replay to my own post. Problem was I forget to add a grid point in the middle of the line where the lumped ports connect.
Moreover, I added an airbox with lambda/4 length at f_max. It seems to improve the results as suggested in the http://openems.de/index.php/FDTD_Mesh page.
If I remember correctly I read somewhere in the forum that the PML ideally should be lambda/4 in length too. Is this correct? In that case if I use a PML_8 the optimal mesh size should be lambda/32?

Best regards
Joakim
Attachments
Microstrip1.m
(3.81 KiB) Downloaded 210 times

thorsten
Posts: 1388
Joined: Mon 27 Jun 2011, 12:26

Re: Simple microstrip simulation and meshing

Post by thorsten » Sat 13 May 2017, 22:30

Hi Joakim,

I'm happy you found a solution, for some reason I think I missed your question. As for the PML, it does not need to have a specific width, just the 8 cells with a reasonable resolution.

About your file. You should make your lumped port cover the entire width of the MSL, no need to have it only as a line in the middle.
Furthermore, why not use the MSL port? Have a look at the MSL notch filter tutorial as a reference. This port is a bit more difficult to use, but it is much more accurate and has more features....
If you configure it to have a lumped R you can still place it at the beginning of a substrate and do not need to terminate it in the PML as in the tutorial...

regards

jockeosth
Posts: 13
Joined: Wed 11 Jan 2017, 20:24

Re: Simple microstrip simulation and meshing

Post by jockeosth » Sat 13 May 2017, 23:39

Thanks for your reply Thorsten.

I will change the lumped port to cover the width of the MSL. I used the lumped port because it is easy to use and later on I will expand the design to cover non-symmetrical striplines too. But I will consider to add the MSL port for better performance in MSL cases :)

Another question pop up in my mind, when using the 1/3 2/3 rule, will the "effective" width of the structure (MSL in this case) be smaller due to this?

I have some interest in making simple SI simulations with crosstalk between transmission lines and from S-parameters calculate the time-domain response. I will try to put some example code in the forum when finish :)

Best regards
Joakim

thorsten
Posts: 1388
Joined: Mon 27 Jun 2011, 12:26

Re: Simple microstrip simulation and meshing

Post by thorsten » Sun 14 May 2017, 13:58

Another question pop up in my mind, when using the 1/3 2/3 rule, will the "effective" width of the structure (MSL in this case) be smaller due to this?
Yes, because the FDTD method otherwise is not as accurate modeling the flat metal lines and the fringe fields on the edges.
But the important part is that it is cut short a third of a mesh cell at that edge. Thus you always have it defined with the proper width, but the mesh cells offset 1/3 inside the flat metal line, such that during the discretization it will end up shorter...

br
Thorsten

Post Reply