Hawaslsh
Full Member level 3
- Joined
- Mar 13, 2015
- Messages
- 175
- Helped
- 5
- Reputation
- 10
- Reaction score
- 7
- Trophy points
- 1,298
- Location
- Washington DC, USA
- Activity points
- 3,623
Hello,
I've been playing around with the 2D extractor within Q3D in ansys to try and gather some information about capacitance in some stripline geometries. In one case I have a single signal conductor in an asymmetric stripline geometry and I can correctly assign the proper material properties, signal lines, ground references.
In this case I can assign the center square as a signal line, and the upper and lower rectangles as reference ground. Solving only for the capacitance in the geometry I get a single value in the solution matrix which is the capacitance between the signal line and the two ground references. The number I get matches quite well with the equations from various text books and I tend to believe those results.
However, my real aim is to model the capacitance between two conductors in a stripline geometry and simulate the capacitance between the two square traces, as well as the overall capacitance seen by a single trace. I naively though I could assign the two traces in the center as both signal lines, and the upper and lower conductors as ground references. When only solving for capacitance the results matrix yielded a 2x2 matrix. I thought one diagonal of the matrix would represent the capacitance one trace sees to the entire structure, with the other would be the capacitance seen from one signal trace to another. However, I always get a negative number when looking at the values for the capacitance between traces.
So despite the capacitance between plates being negative (if that number even represents that), it does get very close what I would expect for capacitance between the signal traces. To try and prove that to myself I adjusted the parameters of the same simulation to be more representative of a parallel plate capacitor. In this case two 4mm plates, separated by 0.25mm in a silicon dielectric. I made the traces pretty thin to minimize fringing capacitance. The simple formula for a parallel plate neglecting fringing would give1685pF/m, which is fairly close to the -1816pF/m.
Does anyone know if the method I am using to assign both center conductors as signal traces is correct? Does anyone have experience with the 2D extractor in Q3D and understand why I am seeing a negative number for the capacitance between objects? I've poked around in the help PDF but haven't found anything that would immediately explain what is happening. In reality these lines would be lines of a coupled resonator (think coupled line filter), so not truly in parallel, and not differential either.
Happy to provide more context, Thanks in advance of any help or advice.
I've been playing around with the 2D extractor within Q3D in ansys to try and gather some information about capacitance in some stripline geometries. In one case I have a single signal conductor in an asymmetric stripline geometry and I can correctly assign the proper material properties, signal lines, ground references.
In this case I can assign the center square as a signal line, and the upper and lower rectangles as reference ground. Solving only for the capacitance in the geometry I get a single value in the solution matrix which is the capacitance between the signal line and the two ground references. The number I get matches quite well with the equations from various text books and I tend to believe those results.
However, my real aim is to model the capacitance between two conductors in a stripline geometry and simulate the capacitance between the two square traces, as well as the overall capacitance seen by a single trace. I naively though I could assign the two traces in the center as both signal lines, and the upper and lower conductors as ground references. When only solving for capacitance the results matrix yielded a 2x2 matrix. I thought one diagonal of the matrix would represent the capacitance one trace sees to the entire structure, with the other would be the capacitance seen from one signal trace to another. However, I always get a negative number when looking at the values for the capacitance between traces.
So despite the capacitance between plates being negative (if that number even represents that), it does get very close what I would expect for capacitance between the signal traces. To try and prove that to myself I adjusted the parameters of the same simulation to be more representative of a parallel plate capacitor. In this case two 4mm plates, separated by 0.25mm in a silicon dielectric. I made the traces pretty thin to minimize fringing capacitance. The simple formula for a parallel plate neglecting fringing would give1685pF/m, which is fairly close to the -1816pF/m.
Does anyone know if the method I am using to assign both center conductors as signal traces is correct? Does anyone have experience with the 2D extractor in Q3D and understand why I am seeing a negative number for the capacitance between objects? I've poked around in the help PDF but haven't found anything that would immediately explain what is happening. In reality these lines would be lines of a coupled resonator (think coupled line filter), so not truly in parallel, and not differential either.
Happy to provide more context, Thanks in advance of any help or advice.