Placing Ports for RF pad simulation in Sonnet
时间：03-25
整理：3721RD
点击：
Hi everyone,
I have issue with the type of ports to use in sonnet for my pad design. There are the 1) standard ports 2) co-calibrated ports and 3) Autognd ports
My pad is a one port circuit. so if I want to use the standard port, its either I put the ports (with port number 1) on the both sides of the pad as shown in the picture and then de-embed to the two sides of the pad.
The other option is to place a single port in the middle of the pad as shown below:
It is interesting to know that the simulation results for these two are actually different. What baffles me is that I was getting a negative capacitance value (i.e. it is inductive) in the case where the port is in the middle.
If I am to use co-calibrated port, I cannot use a single port because of interference of the port with a metal. So I used the first approach above where I place the ports by the sides of the pad with little or no de-embedding. I was getting positive capacitance value here but the resistance seems small (between 20-100ohms from 40 - 100GHz).
So, in short the port placements has significant effect on the simulation result. I think if the port is not properly placed, it could result in unreliable result like getting a negative capacitance value or a very low resistance value (e.g. 0.1 - 5ohms) or a very high Q like 100.
My questions are:
Can anybody tell me what I have been doing wrong.
Which approach is valid, a single port or two ports with the same port number?
What is the proper way of connecting the ports for an rf pad design in sonnet?
Thanks as I await your valuable response.
I have issue with the type of ports to use in sonnet for my pad design. There are the 1) standard ports 2) co-calibrated ports and 3) Autognd ports
My pad is a one port circuit. so if I want to use the standard port, its either I put the ports (with port number 1) on the both sides of the pad as shown in the picture and then de-embed to the two sides of the pad.
The other option is to place a single port in the middle of the pad as shown below:
It is interesting to know that the simulation results for these two are actually different. What baffles me is that I was getting a negative capacitance value (i.e. it is inductive) in the case where the port is in the middle.
If I am to use co-calibrated port, I cannot use a single port because of interference of the port with a metal. So I used the first approach above where I place the ports by the sides of the pad with little or no de-embedding. I was getting positive capacitance value here but the resistance seems small (between 20-100ohms from 40 - 100GHz).
So, in short the port placements has significant effect on the simulation result. I think if the port is not properly placed, it could result in unreliable result like getting a negative capacitance value or a very low resistance value (e.g. 0.1 - 5ohms) or a very high Q like 100.
My questions are:
Can anybody tell me what I have been doing wrong.
Which approach is valid, a single port or two ports with the same port number?
What is the proper way of connecting the ports for an rf pad design in sonnet?
Thanks as I await your valuable response.