Column Control DTX

Using IC-CAP to Extract a Novel Model for Bulk FinFETs

Case Studies

 In recent years, Fin Field-Effect Transistors (FinFETs) have generated much interest, not only for digital applications, but also for RF solutions. Hangzhou Dianzi University created a novel compact model for bulk FinFETs that accommodates self-heating behaviors. Keysight Technologies’ IC-CAP software was used to accurately extract the necessary model parameters. In Figure 1, the simulation results of the model with self-heating effect (SHE) show excellent agreement with measurements up to 50.2 GHz.

 “In our study, we needed to capture the network parameters of FinFET devices from kHz to several GHz. IC-CAP controls the entire measurement system and intuitively displays the measurement data on a graphical interface. The parameter extraction procedure was realized through PEL/Python programming. The calling ADS simulator function, as well as the built-in optimizers in IC-CAP, were effective in ensuring accurate model simulation results and highly efficient model generation.”

Challenge

Due to the miniature dimensions and limited thermal conductivity of silicon material, FinFET performance is affected by the self-heating effect. Self-heating causes an increase in the device temperature and manifests itself as a reduction in on-current and long signal delays. While such manifestations can adversely impact the FinFET, these self-heating related FinFET device characteristics are rarely considered or physically modeled in conventional small-signal FET device models.

Solution

In answer to this challenge, researchers at Hangzhou Dianzi University proposed a small-signal model for bulk FinFETs. The model, and general thermal equivalent circuit, is shown in Figure 2.

To create the model, a parallel combination of resistance-inductance networks originating from the self-heating was first obtained and a new transfer gain model developed. An analytical method for accurately extracting the model parameters was then proposed and implemented. The Keysight IC-CAP device modeling software was used to execute the model extraction procedure.

Results

To verify the small-signal model, Hangzhou Dianzi University researchers fabricated a 4-Fin N-MOSFET with a total number of fingers, NF = 32, and gate length L = 16 nm, using SMICs 14-nm bulk FinFET technology. During the model parameter extraction, small-signal S-parameters of the device, from 100 kHz to 3 GHz and 200 MHz to 50.2 GHz, were measured and de-embedded (open + short). The Keysight E5072 network analyzer was used for the measurement, while the E8363B network analyzer was used for de-embedding. The initially determined Csub and Rsub from zero-bias measurements are given in Figure 3.

In the final step, the random optimization algorithm implemented in Keysight IC-CAP was used to make minor adjustments to the obtained model parameter values. Excellent agreement was achieved between the measured and simulated admittance parameters (Y-parameters) in the 100-kHz to 50.2-GHz range.

×

Please have a salesperson contact me.

*Indicates required field

Preferred method of communication? *Required Field
Preferred method of communication? Change email?
Preferred method of communication?

By clicking the button, you are providing Keysight with your personal data. See the Keysight Privacy Statement for information on how we use this data.

Thank you.

A sales representative will contact you soon.

Column Control DTX