Go to www.awrcorp.com
Back to search page Click to download printable version of this guide.

Second Generation Angelov (Chalmers) Model: ANGELOV2

Symbol

Summary

This is a further development of the well known Angelov/Zirath/Rorsman, or Chalmers, model. It is especially good for use with HEMT devices, but is also useful for MESFETs. Angelov2 includes improved capacitance and breakdown modeling, as well as modeling of thermal effects.

Although this model is similar in many ways to the original Angelov model, it is not identical, and, in general, parameters for the earlier model cannot be used in this one.

Equivalent Circuit

Parameters

Parameter Description Units Default Value
ID Device ID Text AF1
*IPK Current at peak Gm ma 50
*P1 I/V polynomial coefficient   1.0
*P2 I/V polynomial coefficient   0
*P3 I/V polynomial coefficient   0
*P4 I/V polynomial coefficient   0
*P5 I/V polynomial coefficient   0
*P6 I/V polynomial coefficient   0
*B1 P1 term   0
*B2 P1 term   3.0
*VPKS Gate voltage at peak Gm in sat Voltage -0.2 V
*VPK0 Gate voltage at peak Gm near 0 Vds   -0.4 V
*ALPHR Drain I/V knee parameter   2.0
*ALPHS Drain I/V knee parameter at saturation   0
*LAMBDA Drain-source resistance parameter   0
*LAMSB Surface breakdown parameter   0
*VTR Threshold voltage for breakdown Voltage 5.0
*VSB2 Surface breakdown parameter   0.0
*TAU Gate-drain time delay Time 0
*CGS0 Gate-source capacitance parameter Capacitance 0
*CGSP Linear part of Cgs Capacitance 0
*PC10 Gate-source capacitance polynomial coef.   0
*PC11 Gate-source capacitance polynomial coef.   1.0
*PC110 Polynomial coef. to model peaked Cgs   0
*PC111 Polynomial coef. to model peaked Cgs   0
*ADIV Term to model peaked Cgs (1= no peaking)   1
*PC20 Gate-source capacitance polynomial coef.   0
*PC21 Gate-source capacitance polynomial coef.   0.5
*CGD0 Gate-drain capacitance parameter Capacitance 0
*CGDP Linear part of Cgd Capacitance 0
*PC30 Gate-drain capacitance polynomial coef.   0
*PC31 Gate-drain capacitance polynomial coef.   0.5
*PC40 Gate-drain capacitance polynomial coef.   0
*PC41 Gate-drain capacitance polynomial coef.   1.0
*CDS0 Fixed drain-source capacitance (not scaled) Capacitance 0
*CDSW Scalable drain-source capacitance Capacitance 0
*CPG Gate-pad parasitic capacitance (not scaled) Capacitance 0
*CPD Drain-pad parasitic capacitance (not scaled) Capacitance 0
*ISG Gate diode current parameter Current 10-20 A
*NG Gate diode ideality factor   1.0
*RG Gate resistance Resistance 1.0 ω
*RS Source resistance Resistance 1.0 ω
*RI Intrinsic resistance Resistance 1.0 ω
*RD Drain resistance Resistance 1.0 ω
*RGD Gate-drain resistance Resistance 1.0 ω
*RCW RF drain-source resistance parameter Resistance 300 ω
*CRF Capacitance that determines Rds break frequency Capacitance 10-6 F
*LS Source inductance (not scaled) Inductance 0.0
*LG Gate inductance (not scaled) Inductance 0.0
*LD Drain inductance (not scaled) Inductance 0.0
*RTH Thermal resistance C/W   0.1
*TEX Temperature at which parameters were extracted Temperature 25 C
*TEMP Baseplate temperature Temperature 25 C
*TAU_TH Thermal time constant Time 1 mS
*TCIPK Thermal IDS IPK coefficient   0
*TCP1 Thermal IDS P1 coefficient   0
*TCCGS0 Thermal CGS0 coefficient   0
*TCCGD0 Thermal CGD0 coefficient   0
*TCRCW Thermal RCW coefficient   0
*TCCRF Thermal CRF coefficient    
*DTMAX Maximum temperature increase (C) in self-heating    
*TMIN Minimum device temperature    
AFAC Gate-width scale factor   1.0
NFING Number of fingers scale factor   1.0

* indicates a secondary parameter

Implementation Details

This is a further development of the well known Angelov/Zirath/Rorsman, or Chalmers, model. It is especially good for use with HEMT devices, but is also useful for MESFETs. Angelov2 includes improved capacitance and breakdown modeling, as well as modeling of thermal effects.

Although this model is similar in many ways to the original Angelov model, it is not identical, and, in general, parameters for the earlier model cannot be used in this one.

Equations:

Below are the equations for this model.

Drain current

The drain current is given by

where

and

Many of the above parameters are related directly to the peak current and transconductance of the device. See the references for the original Angelov model for further information.

Soft Breakdown

When breakdown is included, the above expressions are modified as

Capacitances

The model uses a capacitance formulation to determine the reactive gate-to-source and gate-to-drain currents. This approach allows the model to be consistent with time-domain (SPICE) formulations. In this implementation, the gate charge is formulated as a single function of gate and drain voltages, so the gate current is

The first term in the above equation represents gate-to-source current, and the second, gate-to-drain current. This approach simplifies parameter extraction, because no transcapacitances are needed and the charge derivatives are identical to the small-signal capacitance.

The resulting capacitance functions are given by the following expressions:

and

The parameter ADIV is used to account for a peak and decrease in capacitance, with increasing gate bias voltage, that sometimes is observed in pHEMTs. In most cases, ADIV = 1 and the expressions are symmetrical. It is important to recognize that this capacitance formulation is entirely different from the original Angelov model, and the parameters of the original model are not transferable to this one.

Drain Dispersion

The model uses a simple approach to account for nonlinear, frequency-sensitive drain-to-source resistance, often called drain dispersion. The drain-to-source resistance is described by a current source having the following I/V characteristic:

where

is given by (2). The parameter RCW, which has units of resistance, is only approximately the drain-to-source resistance in current saturation; it is best viewed as a model parameter used to fit the measured small-signal drain-source resistance.

This element has a large inductor in parallel to provide a return for dc currents generated in the element and to force its dc voltage to be zero under all conditions.

Thermal Model

The thermal model modifies the parameters IPK, P1, CGS0, CGD0, RCW, and CRF as follows:

where

and T is the instantaneous temperature, determined from an electrothermal equivalent circuit. The electrothermal circuit consists of a thermal resistance and capacitance. The total power dissipation in the device, not just the drain dissipation, is used to determine temperature. The temperature coefficients are chosen so that, in most cases, they will be positive quantities; however, occasionally one or more may be negative.

RCW and CRF. The temperature dependence of RCW and CRF requires some explanation. In general, RCW (the RF drain-to-source resistance) decreases with temperature so, for accurate power-amplifier analysis, its temperature dependence should be included. Although CRF is a nonphysical component, it models the transition between the DC and RF drain-to-source resistance regions. This transition increases in frequency as temperature increases, so CRF should be made temperature dependent when necessary to model this phenomenon. In circuits where there are no frequency components near the transition frequency (which is usually on the order of 1 KHz to 1 MHz), this phenomenon need not be modeled, so TCCRF can be set to zero.

The temperature increase, ΔT, is calculated from an electrothermal equivalent circuit consisting of a thermal resistance, RTH, and a capacitance, C. C is calculated from the thermal time constant, TAU_TH. = RTH AzA C. ΔT is limited to 300 C in a numerically acceptable manner. The power used to calculate ΔT includes all RF and DC power dissipated in the device, not just the drain dissipation.

AWR Extensions

In the original model, ΔT was temperature in Celsius degrees. In the AWR® implementation, it is as shown above. Setting MDLFLAG=0 changes this to the original formulation, in effect setting TEX = 0 C.

Layout

This element does not have an assigned layout cell. You can assign artwork cells to any element. See “Assigning Artwork Cells to Layout of Schematic Elements” for details.

References

There are no references for this model. This model was implemented before publishing. The above information comes from personal communication from Prof. Angelov. The breakdown modeling is a joint effort of I. Angelov and S. Maas.

For further information about the model, contact Cadence AWR Technical Support.

Legal and Trademark Notice