next up previous contents
Next: 6.4 Oscillator Up: 6. Examples Previous: 6.2 Simulation of a

6.3 Simulation of a 4H-SiC MESFET

An advanced RF SiC MESFET was investigated by means of numerical simulation with MINIMOS-NT. This device is fabricated by using epitaxial layers on semi-insulating substrates. The simulator has been extended by physically based models that permit the operation of the device to be examined and optimum device structures to be determined [12]. Good agreement between simulation results and measurements can be shown.

SiC-based devices have specific properties which allow them to be used in high temperature, high frequency, high power, and radiation hard applications. Due to the progress in SiC-related process steps the new devices have been developed in recent years [35,154]. Devices such as SiC MESFETs can be employed for microwave power amplifier and oscillator applications due to their excellent DC and RF performance [242]. In particular, the investigated 4H-SiC MESFETs structures are used as base station transmitters for cellular telephone systems and power modules for phased-array radars [166]. The devices are also attractive for higher operation temperatures.

In Figure 6.8, the cross-section of a SiC MESFET is depicted. The typical device parameters are the gate length $ L_\mathrm{g}$, the gate width $ W_\mathrm{g}$, and the thickness of the epitaxial layer $ a$. Due to the higher electron mobility, MESFETs in SiC are made of an n-type material. Furthermore, the mobility of 4H-SiC is twice than that of 6H-SiC [11]. In order to minimize parasitic capacitances, a MESFET is fabricated using epitaxial layers on semi-insulating substrates. The device has three metal-semiconductor contacts: two Ohmic contacts at the source and drain and a gate Schottky barrier.

Devices for microwave- or millimeter-wave applications typically have gate lengths in the range of $ {\ensuremath{L_\mathrm{g}}} = 0.1-1\,\mu$m. The channel thickness $ a$ is typically $ 1/3$ to $ 1/5$ of the gate length $ L_\mathrm{g}$. The spacing between the electrodes is up to four times $ L_\mathrm{g}$. $ W_\mathrm{g}$ and thus the cross-sectional area is directly related to the current handling capability.

For operation the drain contact is biased at a positive potential while the source contact is grounded. The current flow through the channel is controlled by negative DC and superimposed RF gate voltages. The RF signal modulates the channel current and provides an RF gain.

Figure 6.8: On the left, the cross section of a MESFET in SiC is shown and on the right a comparison of measured and simulated DC IV characteristics [12].
\includegraphics[width=0.49\linewidth ]{figures/mesfet.eps} \includegraphics[width=0.49\linewidth ]{figures/sic-iv.eps}

The 4H-SiC MESFET as shown in Figure 6.8 is analyzed by means of numerical device simulation for both DC and high frequency characteristics. For the calibration the specifications obtained from Cree's CRF-24010 4H-SiC MESFET [41] are used to define the simulated device. The charge carrier transport characteristics, the device dimensions, and the doping details are adjusted until good agreement between the simulated and measured IV characteristics is obtained.

The good agreement between simulated and measured steady-state IV characteristics is shown on the right of Figure 6.8. This MESFET produces a maximum channel current of about 450$ \,$mA/mm. The ability of the gate bias to turn the device off and on is good, as indicated by the channel current with zero and high reverse gate bias applied. Good turn-off characteristics are observed for reverse bias slightly greater than -10$ \,$V. The ability of the device to modulate current is given by the device transconductance which for this device is about $ \ensuremath{g_\mathrm{m}}= $160$ \,$mS/mm. The zero gate voltage drain current at $ \ensuremath{V_\mathrm{DS}}= $10$ \,$V is 0.42$ \,$A/mm. This device simultaneously shows a high breakdown voltage of 110$ \,$V and a low leakage current of only 100$ \,\mu$A/mm at 500$ \,$K.

After the successful calibration of the simulator, the small-signal simulation mode was used to obtain additional figures of merit. The frequency range under consideration was 100$ \,$MHz to 40$ \,$GHz. After adapting the input and output impedances, the agreement between simulated and measured data as shown in the left Figure 6.9 was obtained. It is important to note, that the RF results presented here were obtained at a high drain-to-source bias voltage of $ 40\,$V and at the gate quiescent voltage of $ \ensuremath{V_\mathrm{GS}}= $-9$ \,$V. In the right figure of Figure 6.9, the small-signal current and power gain are depicted, showing an $ \ensuremath{f_\textrm{T}}= $5.62$ \,$GHz and $ \ensuremath{f_\mathrm{max}}= $37.18$ \,$GHz at $ 0\,$dB.

Figure 6.9: Comparison of measured and simulated S-parameters in a combined Smith/polar chart with a radius of one (left). Small-signal current and power gain (right) [12].
\includegraphics[width=0.49\linewidth ]{figures/sic-s-parameter2.eps} \includegraphics[width=0.49\linewidth ]{figures/sic-cutoff-MAG-MSG250.eps}


next up previous contents
Next: 6.4 Oscillator Up: 6. Examples Previous: 6.2 Simulation of a

S. Wagner: Small-Signal Device and Circuit Simulation