The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Semiconductor Devices >> nonlinearity of poly-poly or MIM caps
https://designers-guide.org/forum/YaBB.pl?num=1223964187

Message started by vivkr on Oct 13^th, 2008, 11:03pm

Title: nonlinearity of poly-poly or MIM caps
Post by vivkr on Oct 13^th, 2008, 11:03pm

Hi,

There is a small nonlinear component even in linear caps such as poly-poly or metal-metal, which is important when designing very precise analog circuitry. Typically, the nonlinearity is much less than 100 ppm/V.

Intuitively, I would expect the nonlinearity to be quadratic, i.e. C(V) = C0 + C2(V^2), as I would imagine that the polarity of the applied voltage has no (or very little) influence on the nonlinearity, and only the magnitude matters. However, I often see a linear term here, and

C(V) = C0 + C1(V) + C2(V^2)

The linear term appears to be not insignificant. Why would there be a linear term at all?

Regards,
Vivek

Title: Re: nonlinearity of poly-poly or MIM caps
Post by rf-design on Oct 14^th, 2008, 1:32am

I assume that the linear term is steeming from a space charge effect similar to a MOS cap. Typical linear caps in submicron are obtained by canceling two opposite nonlinear terms. It could also be a extraction artifact where the linear term was not set to zero in a more general extraction routine.

http://www.ece.nus.edu.sg/stfpage/elelimf/pub(updated_20Nov2004)/Kim%20VLSI%2004.pdf

Title: Re: nonlinearity of poly-poly or MIM caps
Post by thechopper on Oct 14^th, 2008, 7:14pm

Hi Vivek,

I'm just guessing, but since usually the parasitics on the cap bottom plate is different from the one on the upper plate I would tend to think the sign of the applied voltage might be important. If that was the case then it makes some sense to have a non-insignificant linear term in the capacitor model as a function of the applied voltage...

Regards
Tosei

Title: Re: nonlinearity of poly-poly or MIM caps
Post by vivkr on Oct 21^st, 2008, 10:57pm

Thanks to everyone for the very useful answers and references. Now things seem a bit clearer.

Regards,
Vivek