About on-wafer inductor testing and deembedding

hi I use a 2-port vna to do on-wafer inductor testing(off-wafer iss calibration & open-short deembedding). When the frequency goes beyond 10GHz, the Q-factor seems to be unbelievably high and disordered. I have no idea whether it comes from inaccurate deembedding model or testing error.

And I find out that for the open structure, the (-Y12) which denotes the 2 port coupling and substrate leakage is a negative value rather than a positive one, which is a main reason for the Q factor error. but how it comes?

can anyone help me? thanks!

It seems parasitic oscillation.

then what's the condition of parasitic oscillation? I have several "open" layout for different devices, but it all happens.

Here are a couple of interesting articles that relate to this subject:

1. Rob Groves, Jing Wang, Lawrence Wagner and Ava Wan, "Quantitative Analysis of Errors in On-Wafer S-Parameter De-embedding Techniques for High Frequency Device Modeling," IEEE Bipolar/BiCMOS Circuits and Technology 2006, October 2006.

2. J. C. Rautio, and R. Groves, "A potentially significant on-wafer high-frequency measurement calibration error," IEEE Microwave Magazine, December 2005, pp. 94 - 100.


Good luck,

--Max

Thank you guys. I'vd figured it out. The negative Y12 is a real phenomena, not testing error.