integration line + cpw
Seems to be a basic question. But I donno the answer. So please help me in defining ports for CPW. How do I do this?. By drawing rectangles as usual ....fine. BUT how to draw TERMINAL LINE or INTEGRATION line as there are two GNDs on either side of signal line
thanx
i've done a finite ground CPW and i drew the integration line on one port from the ground to the center line on one side, and one the second port i drew the integration line from the other ground to the center line. I verified the results i got with an ADS result and it matched pretty good. you can always try veryfying with ADS if you have access becasue it is easy to simulate simple structures in ADS
You can use the next method from Ansoft( look at the picture):
define it as a waveport not a lumpport
hiall,
Thankyou verymuch for the help. Its clear now. BUT (sorry if its irritating!..forgive me.im very new to HFSS) how to define TERMINAL line or INTEGRATION line. Terminal line---from center strip to side strips? --this may require 2 such lines(I understand that terminal line defines the + and - of the ports)
INTEGRATION line---- (which is the DIRECTION of field pattern) again two lines from center to side strips?
please clarify
thanx
cv
hi,
the problem of terminal/integration line applies to strip line also.
hope some1 replies
Hi,
Please go through the previous posts,,,,,posted by me....I hope you will find the answer....
Define integtration line from ground conductor to the strips.....direction shouldnt matter, as it will give you the same magnitude of the field.....you can also change teh excitation of the sources bu going to escitaion---->edit sources.....
In my case, to feed CPW using wave port you can draw the port like the following pic.
W_port > 3*(2*space + trip_w)
H_port > 4*substrate_h
Have a nice day
There is a waveport tutorial in Ansoft OTS explaining this.
A few advanced remarks on CPW ports:
1) By default the waveport outerboundaries are perfect E. Do not make your port too high or too wide, otherwise you may excite different modes (waveguide modes). Try limiting yourself to lambda/2.
2) In very high frequencies where you also simulate the actual 3D via configuration that shorts the grounds, and have very strong radiation above the trace - do the following: Open the Setup1 options, click the advanced tab and check the use Absorbing Boundary Conditions on ports (ABC). Make sure that the port will cut two vias in the middle (that short the left ground to the bottom groundplane and the right ground to the bottom ground plane). Draw the port such that it is as wide as the vias below the trace and as about half lambda wide above the trace.
Explanation: When selecting the ABC checkbox the port boundaries absorb the radiation (default is a Perfect-E box). ABC is more accurate, but without shorting the 3 grounds together you will end up with 4 nets in the port (left ground,trace,right ground,bottom ground). The waveport first mode may not be the CPW mode. Therefore you need the vias crossing the port and shorting the grounds (now you're back with two nets in the port and a CPW excitation).
3) In both cases use the ZPI impedance calculation in which the arrow defines just the +/- sign of the field (and not the Z0 calculation, so it does not really matter to which ground it is pointing). Be consistent with all ports in the model. For example the line should always start with the trace and end with the ground. If you fail to be consistent then ang_deg(S12) will be wrong by 180degrees (+/- sign).
Itai