Open pad wafer level deebedding process [help]

[covcst]

Hi everybody,
I am new to RF measurement at wafer level and a little bit confused regarding the de-embedding procedure using open pad deembedding for a 1 port network using a network analyzer.
What I have learnt is as follows:-

1. Get the S11 parameter of the device with the GSG pads.
2. Get the S11 parameter of the open pad GSG structure.
3. Convert both the S11 parameters to their respective Y11 parameters.
4. Y11(device only) = Y11(device with GSG pad) - Y11(only GSG pads).
5. Convert the Y11(device only) to its S11(device only).
6. Get the smith chart of the device characteristics itself....

Someone plz tell me whether I have understood it correctly, if I'm wrong please rectify me......

 

[covcst]

Anyone help plz......

 

[streamlet]

Yes, you are right.

The most simple de-embedding method includes two measurements: a measurement in which only the interconnect pattern surrounding the DUT (so called "open") is measured, and a DUT measurement which of course includes the interconnect parasitics. This correction procedure only corrects for the parallel parasitics, in other words, it is implicitly assumed that series parasitics in the interconnect lines are negligible.

Hi everybody,
I am new to RF measurement at wafer level and a little bit confused regarding the de-embedding procedure using open pad deembedding for a 1 port network using a network analyzer.
What I have learnt is as follows:-

1. Get the S11 parameter of the device with the GSG pads.
2. Get the S11 parameter of the open pad GSG structure.
3. Convert both the S11 parameters to their respective Y11 parameters.
4. Y11(device only) = Y11(device with GSG pad) - Y11(only GSG pads).
5. Convert the Y11(device only) to its S11(device only).
6. Get the smith chart of the device characteristics itself....

Someone plz tell me whether I have understood it correctly, if I'm wrong please rectify me......

 

[covcst]

thanks streamlet...I just wanted to confirm it.....

 

[covcst]

Hi everybody,
I have just simulated a mimcap in HFSS and modeled it as 1 port network. I have also extracted the S11 and Y11 parameters which are attached here. But, when I am going for de-embedding procedure, I am not getting the desired result, as after de-embedding Q factor is decreasing...But de-embedding means getting rid of pad parasitics which is supposed to increase Q factor of the cap.....
Also, resistance is seen to be increasing....
Somebody plz help to explain this apparent anomaly..........