next up previous contents
Next: 5.4 Model Parameters used for Simulation Up: 5.3.2 Fitting Procedure Previous: 5.3.2.1 Simulation of the Transfer Characteristics


5.3.2.2 Simulation of the Output Characteristics
 
In  Figure 5.24 the measured output characteristics of HEMTref is shown. The left dark shaded area indicates the linear region of the device were the electrons do not reach their saturation velocity yet. The light shaded area indicates the active region in which the device is usually operated. It is characterized by the output conductance g0. and ID for a certain DVGS, i. e. gm. At about VDS = 2.0 V a slight kink can be observed. According to [63] it is caused by impact ionization.
 
 

 
Figure 5.24 Measured output characteristics of HEMTref. The characteristics with the highest current is obtained for VGS = 1.0 V. The remaining curves are separated by DVGS = 0.2 V.
 
 

A strong increase in g0 in the right dark shaded area indicates breakdown due to impact ionization. Inclusion of impact ionization in the simulation leads to very unstable convergence. Therefore it is not considered in the simulations of this thesis. For comparison of measurements and simulation only data with negligible impact ionization can be used. Figure 5.25 shows the simulated output characteristics indicated by circles along with measured data. As depicted the characteristics agree very well for VDS < 2.0 V and VGS < 0.6 V. The discrepancy between simulation and measurements for VDS < 1.0 V and VGS > 0.2 V is related to the interface model and the transport model in the channel. With the applied models and the corresponding fitting parameters it was not possible to reproduce this part of the output characteristics very well.
 

 
Figure 5.25 Measured (lines without symbols) and simulated (lines with circles) output characteristics of HEMTref. The characteristics with the highest currents are obtained for VGS = 1.0 V. The remaining curves are separated by DVGS = 0.2 V.
 
 

It can be observed that for low electron concentrations in the channel, i. e. small VGS the output characteristics is modeled quite well. For higher carrier concentration, i. e. large VGS, the current increases only slowly with VDS. This indicates that the interface model does not describe high currents properly for small voltage drops over heterojunction barriers and/or the hydrodynamic transport model in the channel underestimates the velocity for high carrier concentration in a certain range of electric field.

For larger VDS g0 in  Figure 5.25 is underestimated because impact ionization is not included in the simulation. For high VGS g0 is overestimated which is most likely due to temperature effects as the device heats up and the transport properties are deteriorated. This in return reduces the current.

Thermal effects can be reduced in the measurements if VGS is only applied during short duty cycles. In  Figure 5.26 DC and pulsed measurements of an output characteristics according to [64] are compared. It appears that temperature effects can account for more than 10 % change in ID at VDS = 5.0 V.
 

 
Figure 5.26 DC and pulsed measurement of output characteristics of a power HEMT taken from [64].
 



next up previous contents
Next: 5.4 Model Parameters used for Simulation Up: 5.3.2 Fitting Procedure Previous: 5.3.2.1 Simulation of the Transfer Characteristics

Helmut Brech
1998-03-11