Frostman
Newbie
- Joined
- Dec 12, 2024
- Messages
- 1
- Helped
- 0
- Reputation
- 0
- Reaction score
- 0
- Trophy points
- 1
- Activity points
- 22
Hello everyone,
I started using and learning Ansys EDT with simple designs. In the attached pictures I have my LC planar component. The geometry is pretty small, to give an idea of the sizes, the circle's radius is 5.31 um and the gap with the ground is 20 um.
The idea is to design a lumped resonator with a frequency of 6.8 GHz assuming a kinetic inductance of 200 nH (negliging the geometrical inductance). Since the target wavelength is higher than the geometry, I used Q3D to get the capacitance of the pad to the ground and it is -2.73002 fF. Assuming L = 200 nH and C = 2.73002 fF, I get a frequency of 6.811 GHz which is enough for me.
In this case, I removed the strip line as shown in the Q3D simulation.
Q3D simulation design
My idea was also to simulate the same design using HFSS Eigenmode and see if I get a similar frequency. In this case, I kept the strip line and I assigned a Lumped RLC Boundary with L = 200 nH, while ground and island have a Perfect E Boundary. In this case, I get 5.82 GHz which is 1 GHz different from the Q3D simulation.
HFSS Eigenmode simulation design
My question is: Is it normal to get this big difference between these two approaches? If I had to choose probably I trust Q3D since we are in a quasi-static condition, but I would like to read a more expert answer from you.
Why do I do that? I would like to have an estimation of the geometrical inductance and see if the condition of neglecting it holds for large kinetic inductance or if I should consider it in estimating the frequency from the Q3D simulation. My approach would be the following one: if I get similar frequencies from Q3D (assuming only the kinetic inductance) and HFSS Eigenmode (assigning an inductance of 200 nH for the Lumped RLC Boundary) I can, in principle, set the inductance for the Lumped RLC Boundary to zero, perform an Eigenmode simulation and from the frequency of the mode that I get I can retrieve the geometrical inductance knowing C from the first Q3D simulation.
I hope I well explained my problem, in particular my approach to retrieving the geometrical inductance. I do not if it is the right approach that I discussed here, probably it is possible to do it most easily, but these are my first approaches with Ansys EDT and I would like to learn how to use it in the right way.
Thank you for your time!
I started using and learning Ansys EDT with simple designs. In the attached pictures I have my LC planar component. The geometry is pretty small, to give an idea of the sizes, the circle's radius is 5.31 um and the gap with the ground is 20 um.
The idea is to design a lumped resonator with a frequency of 6.8 GHz assuming a kinetic inductance of 200 nH (negliging the geometrical inductance). Since the target wavelength is higher than the geometry, I used Q3D to get the capacitance of the pad to the ground and it is -2.73002 fF. Assuming L = 200 nH and C = 2.73002 fF, I get a frequency of 6.811 GHz which is enough for me.
In this case, I removed the strip line as shown in the Q3D simulation.
Q3D simulation design
My idea was also to simulate the same design using HFSS Eigenmode and see if I get a similar frequency. In this case, I kept the strip line and I assigned a Lumped RLC Boundary with L = 200 nH, while ground and island have a Perfect E Boundary. In this case, I get 5.82 GHz which is 1 GHz different from the Q3D simulation.
HFSS Eigenmode simulation design
My question is: Is it normal to get this big difference between these two approaches? If I had to choose probably I trust Q3D since we are in a quasi-static condition, but I would like to read a more expert answer from you.
Why do I do that? I would like to have an estimation of the geometrical inductance and see if the condition of neglecting it holds for large kinetic inductance or if I should consider it in estimating the frequency from the Q3D simulation. My approach would be the following one: if I get similar frequencies from Q3D (assuming only the kinetic inductance) and HFSS Eigenmode (assigning an inductance of 200 nH for the Lumped RLC Boundary) I can, in principle, set the inductance for the Lumped RLC Boundary to zero, perform an Eigenmode simulation and from the frequency of the mode that I get I can retrieve the geometrical inductance knowing C from the first Q3D simulation.
I hope I well explained my problem, in particular my approach to retrieving the geometrical inductance. I do not if it is the right approach that I discussed here, probably it is possible to do it most easily, but these are my first approaches with Ansys EDT and I would like to learn how to use it in the right way.
Thank you for your time!
