On Chip Transformer Simulation

[soroosh1234]

Hi,
I'm trying to create a transformer on SiGe .13um. When I add 4 ports that are referenced to the ideal ground in momentum everything is fine and I'm gettign my desired performance.
But when I add the ground plane which the grounds of the G-S-G pads are connected, I lose my performance and it seems the return current path to the ground has some inductance. The distance of the ground plane with my outer turn is is around the radius of my transformer.
I wanted to know how should I define the ports to correctly account for the ground ports of GSG probe.
And also what are the design guideline of a transformer.
One terminal of the primary is connected to VCC and one terminal of secondary is connected to ground.

Your help is really appreciated.

 

[BigBoss]

If you supply some sketches/schematics/simulation set-up, we'd better understand your case.

 

[soroosh1234]

It is the same as transformer feedback LNA.

 

[BigBoss]

It is the same as transformer feedback LNA.

View attachment 139728

Post your layout and simulation set-up.Show us what you did exactly..

 

[volker@muehlhaus]

But when I add the ground plane which the grounds of the G-S-G pads are connected, I lose my performance and it seems the return current path to the ground has some inductance. The distance of the ground plane with my outer turn is is around the radius of my transformer.

That diameter/distance sounds ok, as far as coupling ground <> coil is concerned, but of course the physical return path size with its length adds inductance. This is real.

If you use "ideal" grounds for return, that has no physical size and adds no inductance, but this ideal ground doesn't exist in reality.

Screenshots of your layout/model would be useful to understand what you did.