Hi. Great tutorial. I've a question from the 4:50 onwards. When we have drawn a bounding layer, substrate is cut accordingly but air is still in a rectangular shape. If I want to put air only where substrate is, so that I can see reflections caused by reduced air, how I'll do that? Secondly, what if I have two boards of different substrates, placed back to back in a single metal cavity? How can I assign two substrates to a single layer board within a single stack up definition. Sorry for long question, and thanks in advance.
1st part, the Bounding area layer can be of polygonal shape as I showed in the video so you will have air modeled accordingly near to your substrate and if you have conductors very close to the substrate edge they will see the abrupt truncation of the field due to dielectric and air interface very close to them so you will have the desired effect modeled there. I think you got confused looking at the overall Airbox which will always be a rectangular kind of shape no matter which tool you use as per my knowledge. 2nd part, You can do so using Smart Mount to bring over 2 PCB technology and assemble them in a master design and here your master design can be a simple Air dielectric based substrate...Watch the Smart Mount video that I posted with one of the Top10 features of ADS2021 video playlist.
Thank you. This is the first time I've heard about the bounding layer option. I've always been puzzled as to why the substrate gets extended to the edge of the radiation box. I have a general question though about the size of the radiation box for simulating microstrip structures. I've seen that the box should be at least lambda/4 away but I see dramatic changes in the results of the microstrip filter I am simulating by making just small changes in the box size > lambda/4. Also, in this video I saw that you had the substrate wall boundary set to "perfect conductor" The default is "open". Is perfect conductor more relevant to a microstrip filter design? When should I choose open?
You should use "open" which is the default option to have absorbing boundary condition. You can use perfect conductor is you want to model a metallic enclosure like a box around your structure. For detailed help, I would recommend you to contact local Keysight support team for assistance.
Hi! thanks for the tutorial! how can make a non-metelized rectangular hole at center of a substrate. I have a 2 layer PCB. pcb traces are on sub1 and at the center of top substrate (sub2) there is a rectangular cute.
Hello, thanks for an informative tutorial. Could you please give some suggestions to proceed with the case as described- I have designed some components with the help of magnetic wall boundary condition within certain substate lateral extension and using direct port setup configuration. All these components are designed with 50ohm characteristic impedance and gives the best results. Now if i want to combine these designed components in the layout, and want these components to behave as they behaved independently, is there any unique technique of doing that? I'll need magnetic walls around every components in the circuit. But in setup I can select only one common magnetic wall boundary around the entire circuit and that would change the behavior of individual components. Have I messed up with designing components in the way as described?
The kind of behavior you are describing would need some sort of full 3D drawing & simulation tool to model the kind of boundary condition you are describing. You could do something like this in ADS using the bounding area layer around your structure but having a proper full 3D tool like EMPro, HFSS, CST etc might make job little simpler for you however I have no experience in dealing with circuits like these so can't help much.
@@BhargavaAnurag Thanks for the response. I tried creating the boundaries around the components in the similar way as you showed in the video. Is there any way we could define this bound layer as PMC? I don't see any option to select material for this bound layer in the substrate window. I think it just cuts some amount of substrate as per bound shape.
For radiating/guiding structure design, what would be the rule for (a) & (b) (a) Mesh -> Initial Mesh -> Global -> Target Mesh (b) Port Surface Scaling & Port Surface Boundary (under Physical Model -> Advanced) (c) Bounding area layer making the PCB size finite, is it anyway different than finite ground (ground plane as strip in 2 layer substrate definition) simulation? It looks like the simulation box/radiation boundary is touching ground plane, so no back radiation is considered here unlike 2 layer structure. (d) For finite PCB size, will substrate lateral and vertical extension be followed similar to infinite ground plane case (as 1/2 wavelength and 10 times substrate thickness respectively) as well?
Hi Riju, Too many questions in a single message but I will try my best to respond to these: For a). Automatic usually should work for you. If it doesn't and you need a faster convergence doing custom setting with lower mesh size will result in faster mesh convergence as you would change the starting point of the mesher. I would say lamda/10 or lamda/20 could be good starting point but be aware that if you end up creating very small mesh size to begin with then memory required will be much higher. For b). Again 3 & 3 is a good starting point but if you want larger port area for TML/Waveguide Port then 5-5 is another option to consider but mostly 3 should be fine in most cases unless you want to excite higher order modes etc For c). Bounding area is different than finite ground per say, you can have large substrate but limit the ground conductor size which you referred as finite ground. In my case I have used Cover at the bottom hence it is PEC boundary condition applied on Lower Z side hence you don't see backlobe but if you make it like a proper 2 layer stackup definition and limit the ground and use bounding area layer to limit the substrate etc you will see backlobe while performing FEM. We will cover this when we talk about Coax Fed Patch etc in future tutorials. For d). You need to decide lateral extension based on the size of the real world PCB fabrication you will eventually use to make sure there is no distortion to circuit performance incase you cut the PCB very close to the transmission line as it can happen in some cases etc. For vertical extension it depends on what is the box height in which you will place your circuit and if you want to model the box then vertical extension should be defined as per box height and the boundary type should be used as PEC instead of Open. Hope this clarify your queries....!!
@@BhargavaAnurag Thanks for explaining in great detail. Will wait for the coax fed patch. For (d), for general case (not modeling the box), what value of vertical extension is taken when the antenna has finite ground (proper 2 layer stackup) & bounding area are used?
Hi! Thanks for the wonderful effort. Please do a tutorial about how to test and simulate active components (Transistors to be more specific :) )!!!!!!!!!!!!!!!!!!!!!!!!!!1
awesome,thanks
Great, keep it up!
Hello thank you for your video can you please explain to us how to simulate GCPW with FEM simulation.
Process is same as any other FEM simulation. You just need to ensure proper G-S-G pin pairing for CPW Port creation...😊
Hi. Great tutorial.
I've a question from the 4:50 onwards. When we have drawn a bounding layer, substrate is cut accordingly but air is still in a rectangular shape. If I want to put air only where substrate is, so that I can see reflections caused by reduced air, how I'll do that?
Secondly, what if I have two boards of different substrates, placed back to back in a single metal cavity? How can I assign two substrates to a single layer board within a single stack up definition.
Sorry for long question, and thanks in advance.
1st part, the Bounding area layer can be of polygonal shape as I showed in the video so you will have air modeled accordingly near to your substrate and if you have conductors very close to the substrate edge they will see the abrupt truncation of the field due to dielectric and air interface very close to them so you will have the desired effect modeled there. I think you got confused looking at the overall Airbox which will always be a rectangular kind of shape no matter which tool you use as per my knowledge.
2nd part, You can do so using Smart Mount to bring over 2 PCB technology and assemble them in a master design and here your master design can be a simple Air dielectric based substrate...Watch the Smart Mount video that I posted with one of the Top10 features of ADS2021 video playlist.
@@BhargavaAnurag thank you, I think it's time to get the newer version of ADS. I'll try it for sure.
Thank you. This is the first time I've heard about the bounding layer option. I've always been puzzled as to why the substrate gets extended to the edge of the radiation box. I have a general question though about the size of the radiation box for simulating microstrip structures. I've seen that the box should be at least lambda/4 away but I see dramatic changes in the results of the microstrip filter I am simulating by making just small changes in the box size > lambda/4. Also, in this video I saw that you had the substrate wall boundary set to "perfect conductor" The default is "open". Is perfect conductor more relevant to a microstrip filter design? When should I choose open?
You should use "open" which is the default option to have absorbing boundary condition. You can use perfect conductor is you want to model a metallic enclosure like a box around your structure. For detailed help, I would recommend you to contact local Keysight support team for assistance.
Hi!
thanks for the tutorial!
how can make a non-metelized rectangular hole at center of a substrate.
I have a 2 layer PCB. pcb traces are on sub1 and at the center of top substrate (sub2) there is a rectangular cute.
You can add a dielectric via and define material as Air. This kind of geometry can be easily created and simulated by FEM in ADS.
@@BhargavaAnurag
Thank you sir!
Hello, thanks for an informative tutorial. Could you please give some suggestions to proceed with the case as described-
I have designed some components with the help of magnetic wall boundary condition within certain substate lateral extension and using direct port setup configuration. All these components are designed with 50ohm characteristic impedance and gives the best results. Now if i want to combine these designed components in the layout, and want these components to behave as they behaved independently, is there any unique technique of doing that? I'll need magnetic walls around every components in the circuit. But in setup I can select only one common magnetic wall boundary around the entire circuit and that would change the behavior of individual components. Have I messed up with designing components in the way as described?
The kind of behavior you are describing would need some sort of full 3D drawing & simulation tool to model the kind of boundary condition you are describing. You could do something like this in ADS using the bounding area layer around your structure but having a proper full 3D tool like EMPro, HFSS, CST etc might make job little simpler for you however I have no experience in dealing with circuits like these so can't help much.
@@BhargavaAnurag Thanks for the response. I tried creating the boundaries around the components in the similar way as you showed in the video. Is there any way we could define this bound layer as PMC? I don't see any option to select material for this bound layer in the substrate window. I think it just cuts some amount of substrate as per bound shape.
For radiating/guiding structure design, what would be the rule for (a) & (b)
(a) Mesh -> Initial Mesh -> Global -> Target Mesh
(b) Port Surface Scaling & Port Surface Boundary (under Physical Model -> Advanced)
(c) Bounding area layer making the PCB size finite, is it anyway different than finite ground (ground plane as strip in 2 layer substrate definition) simulation? It looks like the simulation box/radiation boundary is touching ground plane, so no back radiation is considered here unlike 2 layer structure.
(d) For finite PCB size, will substrate lateral and vertical extension be followed similar to infinite ground plane case (as 1/2 wavelength and 10 times substrate thickness respectively) as well?
Hi Riju,
Too many questions in a single message but I will try my best to respond to these:
For a). Automatic usually should work for you. If it doesn't and you need a faster convergence doing custom setting with lower mesh size will result in faster mesh convergence as you would change the starting point of the mesher. I would say lamda/10 or lamda/20 could be good starting point but be aware that if you end up creating very small mesh size to begin with then memory required will be much higher.
For b). Again 3 & 3 is a good starting point but if you want larger port area for TML/Waveguide Port then 5-5 is another option to consider but mostly 3 should be fine in most cases unless you want to excite higher order modes etc
For c). Bounding area is different than finite ground per say, you can have large substrate but limit the ground conductor size which you referred as finite ground. In my case I have used Cover at the bottom hence it is PEC boundary condition applied on Lower Z side hence you don't see backlobe but if you make it like a proper 2 layer stackup definition and limit the ground and use bounding area layer to limit the substrate etc you will see backlobe while performing FEM. We will cover this when we talk about Coax Fed Patch etc in future tutorials.
For d). You need to decide lateral extension based on the size of the real world PCB fabrication you will eventually use to make sure there is no distortion to circuit performance incase you cut the PCB very close to the transmission line as it can happen in some cases etc. For vertical extension it depends on what is the box height in which you will place your circuit and if you want to model the box then vertical extension should be defined as per box height and the boundary type should be used as PEC instead of Open.
Hope this clarify your queries....!!
@@BhargavaAnurag Thanks for explaining in great detail. Will wait for the coax fed patch.
For (d), for general case (not modeling the box), what value of vertical extension is taken when the antenna has finite ground (proper 2 layer stackup) & bounding area are used?
Hi!
Thanks for the wonderful effort. Please do a tutorial about how to test and simulate active components (Transistors to be more specific :) )!!!!!!!!!!!!!!!!!!!!!!!!!!1
Hi Ali, That's upcoming shortly...stay tuned