help: Sonnet sim found inductance for a 50Ohm CPW

I simulated a 500um long, supposedly 50 Ohm coplanar waveguide (signal 40um, ground 150um, gap 15um, substrate 300um epsi=9.0). Then in "view response" window, I used "inductance 2" equation to measure its series inductance. For a matched transmission line, the inductance should be close to zero, but the simulation gives me 0.19nH, which is significant.

Anybody knows why?

Hello alexneverhurts,

It may be due to modeling errors...
There are different types of CPW structures, see the below link...

The following is a list of some of the more popular configurations:

1. Ungrounded Coplanar: No metal on the bottom of the substrate (discussed in the section above).

2. Grounded Coplanar: The bottom of the substrate has a metal ground plane which is electrically connected to the two outer conductors.

3. Floating Coplanar: Metal exists on the bottom of the substrate, but it is not connected to the two outer conductors. Thus the bottom metal is electrically "floating" at a different potential from the two outer conductors.

4. Infinite Coplanar: The two outer conductors are considered infinitely wide. Usually, there is no metal on the bottom of the substrate (case #1), but any of the above three configurations can have finite or infinite outer conductors.

5. Shielded Coplanar: Outer conductors are shorted to the side-walls of a rectangular waveguide. Thus, all four sides of the waveguide are at the same ground potential as the coplanar outer conductors. Waveguide dimensions are part of the design

Therefore, the method you use to simulate your coplanar line depends on how you are exciting it. Each of the above cases should be simulated differently:

1. Ungrounded Coplanar: Add an extra air layer under the dielectric to move the bottom of the Sonnet box far enough away so it has no effect. The sidewalls of the box should also be moved far away (at least several line widths). If the sidewalls and box bottom are far away, using push-pull ports should give the same answer as shorting the outer conductors to the box wall.

2. Grounded Coplanar: Do not use push-pull ports. Just connect the two outer conductors to the sidewall of the box (the same sidewall to which the center conductor is connected). This electrically connects them to the metal ground plane on the bottom of the substrate.

3. Floating Coplanar: Use push-pull ports because the metal on the bottom of the substrate is not connected to the two outer conductors.

4. Infinite Coplanar: Since the two outer conductors are considered infinitely wide, this cannot be simulated with Sonnet. However, you can approximate this case by truncating the two outer conductors to a finite width. Do not use push-pull ports unless you have a floating coplanar line as described in #3. Instead, connect the ground conductors to the box sidewall without ports. This is because the error introduced by truncating the outer conductors can be partially compensated for by allowing the current that should have traveled on the outside edge of the outer conductor to travel through the box sidewalls.

5. Shielded Coplanar waveguide: In this case, create the Sonnet box to be the same dimensions of your coplanar waveguide box. Do not use push-pull ports. Instead, connect the ground conductors to the box sidewall without ports.

Have you followed above steps, before measuring the inductance...

http://www.sonnetsoftware.com/suppor...r_waveguid.htm


---manju---