6.2.1 DC Characteristics

The investigations are based on two power pseudomorphic HEMTs manufactured on the same wafer which differ only in their gate lengths of L_G = 220 nm and 500 nm. The simulations are performed with the simulation set up obtained by the fitting procedure on HEMT_ref. The model parameters are given in Table 5.2 . First the simulation of the device with L_G = 220 nm was fitted to the measurements only by adjusting the location and density of the active doping in the supply layer, the interface charge density, and the gate­to­channel separation. All adjustments are made well within realistic ranges. Then the device with L_G = 500 nm was simulated with the same set of simulation parameters but the corresponding lateral spacings.
 

The measurements and simulations of the transfer characteristics and the transconductance g_m of the two HEMTs are compared in  Figure 6.22 and  Figure 6.23. Although only the simulation of the device with L_G = 220 nm was fitted to the measurements also the results for the device with 500 nm agree very well.
 

Again the current near V_T is overestimated. Despite this small discrepancy it is shown in both the simulation and the measurements that V_T differs by about 110 mV. The lower average electron velocity due to the longer gate manifests itself also in a reduction of g_m as shown in  Figure 6.23. A more positive V_T and a lower g_m results in a significantly lower I_{D max} for the device with L_G = 500 nm. However, there is also a small but important detail in which simulation and measurement do not agree completely. Although the simulated I_{D max} for the device with L_G = 220 nm is slightly underestimated the simulated I_{D max} of the HEMT with L_G = 500 nm is somewhat overestimated. A possible reason for this behavior can be explained as follows.

The current saturation for large V_GS is mainly governed by real space transfer, the velocity in the barrier layers, and the interface charge density between the semiconductor and the passivation. The real space transfer and the velocity changes with the gate length but the interface charge density is kept constant. Thus, the relative impact of the interface charge density and the transport properties on the total current is changed with the gate length. The C_G of HEMT_ref in Figure 5.22 revealed an underestimated carrier concentration and an overestimated carrier velocity for V_GS > 0.6 V. This is obviously getting significant due to the change of the relative impact of the parameters. This is another indication that the discrepancy between measurement and simulation of HEMT_ref for V_GS > 0.6 V is not solely due to thermal effects but also calls for improvements of the HD model for AlGaAs.

On the other hand the two devices already cover a very wide range of technically relevant gate lengths, thus, based on the verification of the simulation results the capability of predictive simulations can considered to be very substantial.
 

• 6.2.2 Dependence of Recess Geometry on the E-Field Distribution
• 6.2.3 RF Characteristics
• 6.2.3.1 Reduction of the Gate Length
• 6.2.3.2 Contributions to the Gate Capacitance
• 6.2.3.3 Dependence of C_G on Passivation Thickness
• 6.2.3.4 Dependence of C_G on T-gate Cross Section
• 6.3 Millimeter Wave HEMTs
• 6.3.1 DC Characteristics
• 6.3.1.1 Dependence on the Gate-to-Channel Separation
• 6.3.1.2 Dependence on the Gate Length
• 6.3.2 RF Performance
• 6.3.2.1 Dependence on the Gate Length
• 6.3.2.2 Contributions to the Gate Capacitance
• 6.3.2.3 Dependence of f_T on d_GC
• 6.3.2.4 Dependence of f_T and f_max on L_R
• 6.3.2.5 Dependence on the Passivation Thickness
• 6.3.3 Optimization of Millimeter Wave HEMTs
• 6.4 Comparison between Low Noise, Power, and Millimeter Wave HEMTs
• 6.5 Influence of Backside Doping on the C_G(V_GS) Characteristics