SvenB
Junior Member level 1
- Joined
- Nov 28, 2014
- Messages
- 15
- Helped
- 4
- Reputation
- 8
- Reaction score
- 3
- Trophy points
- 3
- Location
- Delft, The Netherlands
- Activity points
- 137
A compiled Matlab application can be in fact distributed and used without a Matlab license.
I'm looking forward to the addition of surface wave modes.
I see you have the excitation set up as common or differential, are you planning on generalizing this for a multi-modal environment?
I compiled a standalone version of the program. It is attached to this message. It seems that it works properly. Check it out and let me know what you think!
- Your code seems to simulate with zero thickness. This is fine for wide flat metal (typical PCB case) but for on-chip cases the metal height becomes relevant. It changes loss for microstrip and it changes capacitance between metals for coupled lines and coplanar.
Yes, that's true. The labels are there in order to load quickly some pre-programmed settings.- It seems that we can edit parameters, but keep the old material text label that no longer applies?
I will add it!- You could add SiO2 as lower slap material, which is very common as inter-metal dielectric in RFIC stackups. Permittivity is usually specified between 3.9 and 4.2, with small or zero loss tangent.
This can be implemented easily. That may be nice to add in the future.- You have included silicon as a dielectric with loss tangent. The bulk silicon that we have as bottom medium isn't accurately described by tand, because it has relevant (DC) conductivity. Typical values are in the range 1 .... 10S/m but it depends on the actual technology.
Indeed, the metals are simulated with zero-thickness with a surface-impedance assuming that the metal is electrically thick (t >> skindepth).
A change in the surface impedance can be accounted for, in the case of a finite thickness, which I will implement in a future version. A change in field distribution however, and thus capacitance, cannot be accounted for unfortunately.
Yes, that's true. The labels are there in order to load quickly some pre-programmed settings.
This is actually discussed in the Help document provided with the program and good described by the paper of Rautio mentioned in the Bibliography of the Help document (This paper described the model in Sonnet). Indeed at this point the program does not account for this. I was thinking by implementing this by looking at the ratio of the dielectric constants of the slab to estimate the ratio between the top and bottom current and with that alter the effective surface impedance. What do you think?If you simulate the conductor as a thin sheet and calculate the equivalent surface impedance, you need to make an assumption if there is skin current only on one side (e.g. bottom) or both sides (top and bottom).
Sure I agree with you, I will take care of it.To avoid confusion with text label and value no longer matching: How about an explicit "user defined" setting, or alternatively take out the predefined material name when the user changes the values?
I am not sure if I understand your question correct, but the spectrum, of a CPW for example, will show me two poles; one pole for the common excitation and one pole for the differential excitation. These are the two modes the user can choose to track. For other structures, only the main propagating mode will be tracked.
This is actually discussed in the Help document provided with the program and good described by the paper of Rautio mentioned in the Bibliography of the Help document (This paper described the model in Sonnet). Indeed at this point the program does not account for this. I was thinking by implementing this by looking at the ratio of the dielectric constants of the slab to estimate the ratio between the top and bottom current and with that alter the effective surface impedance. What do you think?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?