This action might not be possible to undo. Are you sure you want to continue?
GUMMELPOON
MODEL
BIPOLAR
MODEL
DESCRIPTION EXTRACTION
PARAMETER
iC iB
1/RE
IKF 2,3*NF*vt 1 decade BF
CB'C'
ISE
1 decade 2,3*NE*vt
iB
RBB' B' iB'C'
IC
B
iB'E'
C C' RC vB(V)
CB'C'
iC'E'
IS
E' RE
E
F.Sischka Agilent Technologies GmbH, Munich
GummelPoon Toolkit B0_HEADR.WPS  06.07.01
Franz Sischka
GummelPoon Model
2
STRUCTURE OF THIS MANUAL
Introduction Operating Modes of the Bipolar Transistor The Equivalent Schematic and the Formulas of the SPICE GummelPoon Model A Listing of the GummelPoon Parameters A Quick Tutorial on the GummelPoon Parameter Extractions Proposed Extraction Strategy
CV Modeling Extraction of CJE, VJE, MJE, as well as CJC, VJC, MJC
Parasitic Resistor Mmodeling Extraction of RE Extraction of RC Extraction of RBM
Nonlinear DC Modeling Extraction of VAR and VAF Extraction of IS and NF Extraction of BF , ISE and NE Extraction of IKF Reverse Parameters NR, BR, ISC, NC and IKR
AC Small Signal Modeling, Parameter Extraction Extraction of RB, IRB and RBM Extraction of TF, ITF, and XTF Extraction of VTF Extraction of PTF Extraction of TR Modeling of XCJC Temperature Effects Model Limitations Appendices Linear Curve Fitting: Regression Analysis About the Modeling Dilemma Verifying the Quality of Extraction Routines Direct Visual Parameter Extraction of BF , ISE and NE Calculation of h21 of the GummelPoon Model Publications
GummelPoon Toolkit B0_HEADR.WPS  06.07.01
Franz Sischka
GummelPoon Model
3
This product has been developped to meet the local demands of European ICCAP users for more technical background information on extraction techniques and for the availability of extraction source code.
Published for the first time in 1990, it has been updated since then several times.
It is part of a series of supplementary modeling toolkits for the ICCAP users. These products feature source code and detailed technical description of the extraction routines. Please contact the author for further information.
The author would like to thank the many users for valuable inputs, and is hoping for fruitful discussions also in the future.
franz_sischka@agilent.com
GummelPoon Toolkit B0_HEADR.WPS  17.04.01
Franz Sischka
In this case.GummelPoon Model 4 ABOUT THIS MANUAL This manual is intended to explain the basics of modeling a bipolar transistor using the GummelPoon model as it is implemented in the simulation program SPICE of the University of California Berkeley (UCB) /see publication list/. independent of the measurement data! IMPORTANT NOTE: This manual and the underlying ICCAP model file GP_CLASSIC_NPN. as stated above.mdl.mdl the MASTER model file which is described in this manual and featuring the data management features of ICCAP 5. the data are exported into ICCAP mdm files (ASCII files) and imported into the master extraction file GP_EXTRACT_NPN. it covers the classical GummelPoon model without enhancements for also modeling the parasitic transistor.mdl for extraction.mdl.x.04. Therefore. all measurements are performed using the file NPN_MEAS_MASTER. i. Note: After you have become familiar with the modeling procedure itself. bipolar transistor modeling including the parasitic transistor. However.e. separating measurements from extractions: NPN_MEAS_MASTER.mdl a master file for measurement a master file for modeling as well as manyother ICCAP model files covering topics like: model parameter extraction using the tuner feature direct visual parameter extractions alternate modeling methods for DC.WPS  17. Then.and RFparameters. such model files are included in the file sets of this toolkit. You are also invited to get in contact with the author for assistance with such modeling problems. It is part of the GummelPoon Bipolar Model Parameter Extraction Toolkit.01 Franz Sischka .CV. you are encouraged to split the modeling into 2 parts: separate measurements and separate extraction strategy. i.mdl are intended to explain the basics of the GummelPoon modeling. GummelPoon Toolkit B0_HEADR. Please see the README macros in these ICCAP model files for more details.e. file GP_CLASSIC_NPN.mdl GP_EXTRACT_NPN. This toolkit includes the ICCAP model file GP_CLASSIC_NPN. This method allows to improve continuously the extraction strategy file.
Subcircuit model description. GummelPoon Toolkit B0_HEADR. open for user enhancements (HF modeling.). See also the file GP_MEAS_MASTER. and DC.01 Franz Sischka . All transistor pins are connected to SMUs for flexible measurements The transistor output characteristic and Sparameter measurements use a Base current stimulus rather than a BaseEmitter voltage in order to avoid 1st order thermal effects being visible. They can be easily modified to meet specific user needs. then parasitic ohmic resistors. However. selfheating might be present and affect the Gummel plots in the ohmic range. More chapters cover side aspects of bipolar modeling. which explain how to determine the model parameters from CV (capacitance versus voltage).GummelPoon Model 5 The ICCAP model file "GP_CLASSIC_NPN.WPS  17.04. The individual chapters follow always this scheme: explanation of the parameterdependent measurement setup explanation of the mathematical basics for the parameter extraction explanation of the parameter extraction explanation about how to use the ICCAP file. to finally high frequency measurements using network analyzers. parasitic pnp etc.mdl Organization of the chapters in this manual: There are 5 main chapters.mdl" features: The extractions are written using PEL (parameter extraction language) and are open to the user.
This is done by subtracting the Y matrix of the OPEN from the total measurement. packaging parasitics and other nonideal effects are not part of the model. It should be mentioned that in this case the SHORT itself has to be deembedded first from the OPEN parasitics! For packaged devices. However. The modeling procedures presented in this manual refer to already deembedded measurements. Please check the example files included in the file directory of this toolkit for examples. test probes that allow NWA calibrations down to the chip (like Cascade or Picoprobe probes) are commonly used. Therefore.GP: Introduction 1 INTRODUCTION : C O N T E N T S: Operating modes of the bipolar transistor The GummelPoon equivalent schematic The GummelPoon model equations List of the SPICE GummelPoon parameters A quick tutorial on the GummelPoon parameters Proposed global extraction and optimization strategy This manual describes the modeling of a bipolar transistor using the GummelPoon model as implemented in the simulator SPICE. Also. we need to use a test fixture. followed if required.by the subtraction of the Z matrix of the SHORT.01 Franz Sischka . It should be mentioned that the GummelPoon model itself covers only the internal part of a realtransistor. they can be added by using a subcircuit rather than just the standalone model. Deembedding means here to eliminate onwafer parasitics. Parasitic effects is specially important for network analyzer (NWA) measurements. the NWA has to be calibrated down to the ends of its cables using the calibration standards (SOLT) of the actual connector GummelPoon Toolkit B1_INTRO.WPS  17.04. In this case. For onwafer measurements. which are due to the test pads (OPEN dummy) and the lines from the test pads to the transistor itself (SHORT dummy). onwafer parasitics like a parasitic pnp transistor are not covered.
THRU). BE and BC layers are 'overcharged'. is described by the ohmic resistors. As a next step.GP: Introduction 2 type. Since the model is symmetrical. A file including such a procedure is included in the toolkit filesets. the test fixture has to be modeled (OPEN. See the example more_files/packaged_xtor_in_testfixture. The now known test fixture parasitics can be deembedded and the extraction techniques of this manual can be applied to the downstripped inner device. The saturation region. Operating Modes of the Bipolar Transistor There are four operating modes of a bipolar transistor as illustrated in figure 1. Fig. the DUT (device under test) is inserted into the test fixture and measured. but applied to the reverse measurements. for example.1: operating modes of a bipolar npn transistor GummelPoon Toolkit B1_INTRO.04. The DC and AC extraction procedures that are proposed in this manual cover mainly the forward region.mdl Please contact the author if you wish more info or help on deembedding. the region vCE<0. vBC REVERSE SATURATION (*) vBE FORWARD or NORMAL CUTOFF (*) NOTE: in the saturation range. Finally. SHORT.3V in the DC output characteristics.WPS  17.01 Franz Sischka . the reverse parameters can be extracted following the same ideas.
2b: GummelPoon large signal schematic of the bipolar transistor GummelPoon Toolkit B1_INTRO.04. neglecting the parasitic pnp transistor. B field oxide p+ p base E field oxide C poly n+ field oxide n+ CPI deep n+ CMU n well n+ buried layer substrate S Fig. It represents the physical transistor: a currentcontrolled output current sink.2b shows the large signal schematic of the GummelPoon model. see fig. This structure represents pretty much the physical situation of a bipolar transistor.WPS  17.2a.GP: Introduction 3 The GummelPoon equivalent schematic Fig. CB'C' iB RBB' B' iB'C' C' RC IC B C iB'E' CB'E' E' RE S iC'E' E Fig.01 Franz Sischka . and two diode structures including their capacitors.2a: physical situation for a bipolar transistor.
As said above. RBB' B' CB'C' =CMU IC i B B C' RC C CC'S' rB'E' CB'E' =CPI R0 gm*vB'E' S E' RE E Fig.04. This means. fig. S C B E Fig. However.2c).2b. the small signal schematic for high frequency simulations can be derived.01 Franz Sischka .2c is a pure linear model. the description of this manual does not cover this.WPS  17. It must be noted that the schematic after fig. ICCAP files for this type of modeling are included in the filesets of this toolkit.2c: AC small signal schematic of the bipolar transistor NOTE: XCJC effect neglected. See the macros in the model files instead. In order to do this.GP: Introduction 4 From fig. Such a schematic is used later for SPICE Sparameter simulations. It cannot be used to predict nonlinear highfrequency behavior of the transistor. GummelPoon Toolkit B1_INTRO. the DC currents are calculated and the model is linearized in this point (fig.2d depicts the subciruit used for modeling a npn transistor including the parasitic pnp.2d: subcircuit schematic when including the parasitic pnp. for a given operating point. RF simulators like HP_MDS or HP_ADS perform highfrequency simulations using the large signal model (harmonic balance simulations). In order to make the presentations of the schematics complete.
1 } BR NR vt + vBC ISC { exp[ . TEMPERATURE VOLTAGE: vt = k T q = 8.] .6171 E5 * (T /'C + 273.] .15) BASE CURRENT: iB = iBE + iBC (A) iB with = if/BF + iBErec + ir/BR + iBCrec rec: recombination effect (B) ideal forward diffusion current: if = vBE IS { exp[ .01 Franz Sischka .e. RE and RC.] .] .] .1 } BF NF vt vBE ISE { exp[ .04.] . schematic in fig_2 ) this gives: IS vBE .{ exp[ .{ exp[ .1 } NE vt vBC IS { exp[ .1 } NE vt iB = + IS vBC + .] . we assume no voltage drops at RB.GP: Introduction 5 The GummelPoon Model Equations For the reader's convenience all the GummelPoon equations are presented at a glance.1 } NC vt (G) GummelPoon Toolkit B1_INTRO.] .1 } NF vt (C) BE recombination effect: iBErec vBE = ISE { exp[ . i.WPS  17. vB'E'=vBE and vB'C'=vBC.1 } NC vt (F) ( see equiv.1 } NR vt (D) ideal reverse diffusion current: ir = (E) BC recombination effect: iBCrec vBC = ISC { exp[ . In order to make them better understandable.
.GP: Introduction 6 COLLECTOR CURRENT: iC = 1/NqB ( if .schematic in fig_2 ) or: ic = IS éæ vBE vBC ö æ ö ù * êç exp − 1÷ − ç exp − 1÷.01 Franz Sischka .ú Nqb ëè NF * vt ø è NR * vt ø û IS é vBC ù * êexp − 1ú Nqb ë NR * vt û − vBC é ù − ISC * êexp − 1ú NC * vt û ë with the base charge equation (I) NqB = q1S * 1 + 1 + 4q2 S 2 ( ) (J) for the modeling of nonidealities like the basewidth modulation: q1s = 1 vBE vBC 1− − VAR VAF (K) and the hilevel injection effect: q 2s = IS é æ vBC ö ù IS é æ vBE ö ù expç expç ÷ − 1ú + ÷ − 1ú ê IKF ë è NF * vt ø û IKR ê ë è NR * vt ø û (L) BASE RESISTOR: RBB = RBM + 3(RB − RBM ) with tan (z ) − z z * tan 2 (z ) (M) æ 12 ö i 1+ ç ÷ B −1 è PI ø I RB z= iB 24 2 PI I RB 2 PI = 3.14159 (N) GummelPoon Toolkit B1_INTRO..04.WPS  17...ir ) ir/BR iBCrec (H) ( see definition of iB above and equiv.
vBC / VJC ] MJC and CBE = CSBE + CDBE = TR + NR vt IS NqB vBC exp [ .01 Franz Sischka .44 VTF iB .GP: Introduction 7 SPACE CHARGE AND DIFFUSION CAPACITORS: CBC = CSBC + CDBC = (O) CJC = [1 .] NR vt (P) (Q) CJE = [1 . GummelPoon Toolkit B1_INTRO.] } 1.e. TFF + NF vt IS NqB vBE exp [ .vBE / VJE ] MJE with the transit time TFF if 2 = TF { 1 + XTF [ .] NF vt (R) (S) and the ideal forward base current if from the definition of equation (C).WPS  17.04.] if + ITF vBC exp [ . i.
75 MJC BC junction exponential factor .3 100 .base resistance at hi current RE emitter resistance RC collector resistance CBE: CJE BE zerobias deplet.for bias dependence of TF 0 VTF voltage describing VBC dependence of TF infinite ITF hicurrent parameter for effect on TF 0 PTF excess phase at frequency 1/(TF*2PI) 0 TR ideal reverse transit time 0 NOISE: KF flicker noise coeff.75 MJS substrate junction exponential factor 0 CAPACITOR FORWARD CHARACTERISTICS: FC forward bias depletion cap.33 CBC: CJC BC zerobias deplet.value 1.5 2 0 0 infinite infinite 0 infinite RB 0 0 Unit typ.E13 . extraction temp 0 1 27 27 F V 27 27 'C 'C GummelPoon Toolkit B1_INTRO. forward Early voltage reverse Early voltage forward current emission coeff.0001 25 5 10 1.3 .GP: Introduction 8 List of the SPICE GummelPoon Parameters Name DC: IS XTI EG BF BR XTB VAF VAR NF NR NE NC ISE ISC IKF IKR Parameter explanation transport saturation current temperature exponent for effect on IS energy gap for temperature effect on IS ideal forward maximum beta ideal reverse maximum beta forward & reverse beta temp.capacitance 0 VJC BC builtin potential .1E15 3 1.11 eV 150 .4 .7 1.0 1.E12 sec A A A A Ohm A Ohm Ohm Ohm F V F V V V OHMIC PARASITICS: RB zero bias base resistance IRB current at medium base resistance RBM min. reverse current emission coeff.coeff.5E12 .75 MJE BE junction exponential factor .6 .capacitance 0 VJE BE builtin potential .4 1 0 0 0 . BE leakage emission coeff.WPS  17.E3 A 0 deg 50.11 100 1 0 infinite infinite 1 1 1.5 TRANSIT TIME: TF ideal forward transit time 0 XTF coeff.5 2.33 XCJC fraction of BC capacitor connected to int.OPTIONS TNOM device meas.04.5 1.6 .05 . AF flicker noise exponent TEMPERATURE EFFECTS . BC leakage emission coeff.E15 A 3 1. and param.coeff.base 1 CCS: CJS zerobias collectorsubstrate capacacitance 0 VJS substrate junction builtin potential . .01 Franz Sischka .0 1.E12 sec 10 5 V 20.5 100 50 1. BE leakage saturation current BC leakage saturation current forward beta hi current rolloff reverse beta hi current rolloff SPICE default .1E12 1.TEMP device temperature for simulation /'C .E12 .
WPS  17.2 p slope: MJE 0. this chapter puts together the graphical equivalents of the parameter extraction techniques.8 p 3 1 CJE + Coffs 0 1 vBE (V) VJE FC*VJE Early voltage extraction iC(mA) 3 1 0 VAF 1 2 2 6 vC(V) GummelPoon Toolkit B1_INTRO.04. space charge capacitor modeling CBE (pF) 1.GP: Introduction 9 A quick Tutorial on the GummelPoon Parameters Although it is recommended to go through the individual extraction steps of the corresponding sections of this manual.6 p 1.01 Franz Sischka .
04.01 Franz Sischka .3*NE*vt vB(V) IS beta BF IKF.GP: Introduction 10 forward beta parameter extraction iC iB IKF 2.WPS  17.3*NF*vt 1decade BF 1/RE ISE 1decade 2. RE NE ISE vBE GummelPoon Toolkit B1_INTRO.
04.GP: Introduction 11 Base resistor parameter extraction extrapolated for infinite frequency j*IMAG rBE rBB'+RE rBB'+1/gB'E'+RE(1+ß) REAL iB frequency this gives: RBB(Ohm) RB IRB RBM RBM IRB iB(A) GummelPoon Toolkit B1_INTRO.01 Franz Sischka .WPS  17.
GP: Introduction 12 Transit time parameter determination First model the TFF trace without VCE effect log (fT) 1 fT = 2*PI*TFF log (iC) TFF(psec) XTF TF(1+XTF) Effect of the space charge capacitors ITF TF theoretical curve isothermically measurable range iC(mA) then model the dependence on VCE: log (fT) 1 fT = 2*PI*TFF vCE log (iC) GummelPoon Toolkit B1_INTRO.WPS  17.04.01 Franz Sischka .
current one. cv: Extract the CV parameters CJx. Extract IS. NF. optional also for CCS. XTF and ITF. RBM and RC from flybackmeasurements. well below ohmic effects show up. Extract IKF from the ßcurve. Finetune all DC parameters in all DC setups. BR. ITF and VTF parameters. For example. optimize the TF. XTF. optimize the CV parameters. During the extraction process. Sparameters: Deembed the measurement data. all resistors Rx=1m means no ohmic effects and so on. ISE and NE. This will firstly get rid of old parameter values which belong to the last modeling and not to our actual. the default parameters are those which reduce the complexity of a model completely.WPS  17. GummelPoon Toolkit B1_INTRO. VAF=1000 means: no Early effect. XTF. Go back to the rBB' setup and optimize the Sparameter fitting of the RB. ISE. and thus. perform optimizer finetuning. NE and BF from the forward GummelPoon plots. dc: Extract VAR and VAF from the output characteristics. IRB and RBM. and secondly. again in the rBB' setup. ITF (lowest vCE). Optimize the Sparameter fitting of VTF (all vCE). Optimize Sparameter fitting of TF. Optimize RE in the upper region of the GummelPoon plots (iB and iC). IKF=1000 no knee current.and BCcapacitance. Measure again Sparameters. Fineoptimize VAR.to Hparameters and get a frequency f20dB from the 20dB/decade of h21. frequency swept. ohmic parasitics: Extract the parasitic resistors RE. half the current out of the Collector. but now with the constant frequency f20dB and swept iB and swept vCE and extract TF. we get more and more parameter values.01 Franz Sischka . as well as VTF. the model becomes more and more complex and accurate. Optimize the GummelPoon plot for IS. Emitter grounded. or: Extract them from overdriven Sparameter measurements: high current at Base. Then.GP: Introduction 13 Proposed Global Extraction And Optimization Strategy First reset the model parameters to default (Window Model Parameters). IRB and RBM from S11 measurements with swept frequency and base current as a secondary sweep. Optimize BF and IKF in the ßcurve at high bias.04. MJx and VJx for the BE. VAF and BF in the output characteristics setup. NF. Extract the base resistor parameters RB. If required. Transform S. Finally: Resimulate all setups and check the fitting quality in the verify setups.
This strategy may vary a bit depending on the actual data. make sure the parameter values in the ICCAP parameter list are all set to typical values that you will expect later for your parameter extraction. modeling should be done in those regions where the transistor will be operated later. you might consider that your physical device under test may not be so well represented by the GummelPoon model! To 'synthesize' such pseudomeasured data in ICCAP. change it back to 'B' again and hit again <RETURN>. perform a simulation for every setup in your model file. change the setup output data type to 'S'.GP: Introduction 14 Macro 'extract_n_opt_ALL' in ICCAP file gp_classic_npn.mdl contains an example for such a modeling strategy based on the measurement data included in the file. contact tour design engineer colleagues and ask them about the specific operating range. In this way you are sure that your extraction strategy works well for a perfect GummelPoon transistor. simply modify the macro to meet your local requirements. and to check the extraction routines on these data. before defining the measurement ranges. See also the appendix. to transform these simulated data into measured ones and to try to get the (known) parameters back again. NOTE: A smart way of defining or verifying the most appropriate extraction strategy is to synthesize quasimeasured data from simulation results. Then reset the parameter values to default and try your extraction strategy. hit <RETURN>. This means to simulate all setups using a given parameter set.01 Franz Sischka . If you have afterwards problems during the curve fitting. In this case.04. As a general rule. GummelPoon Toolkit B1_INTRO.e.WPS  17. LAST NOT LEAST: Before performing your measurements. Now you have identical data in both measured and simulated arrays. i.
VJS. the Sparameter measurements are modeled. as well as CJE. MJC. followed then by the parasitic resistor modeling and the nonlinear DC curves. Finally. VJC. We follow this idea and begin with the CV modeling. GummelPoon Toolkit B2_CV.GP: Space Charge 1 C V M O D E L I N G.independent of the other model parameters.04. P A R A M E T E R E X T R A C T I O N C O N T E N T S: The GummelPoon CV equations Extraction of CJC. MJE and CJS.01 Franz Sischka . MJS Some comments on CVmodeling Since the CV parameters are like for most bipolar models. they are usually extracted first.WPS  17. VJE.
.) NR VT NqB NR VT (CV1) with equation (H) from the introduction chapter.. vBE and vBC are the stimulating voltages. GummelPoon Toolkit B2_CV.GP: Space Charge 2 The GummelPoon Capacitor Equations Provided that: vBE = vB'E' and vBC = vB'C' .WPS  17..44 VTF (CV3) and the ideal Collector current if from (C) CSBi models the space charge and CDBi the diffusion capacitance between Base and Emitter or base and Collector respectively.vBE / VJE ] + CDBE diC dvBE = + TFF (H) CjE = MJE [1 .04.exp ( .vBC / VJC ] + CDBC diC TR dvBC = + (H) CjC = MJC [1 . (S) are: CBC = CSBC CjC MJC [1 . and CBE = CSBE CjE MJE [1 .vBC / VJC ] TR IS vBC + .exp ( .vBE / VJE ] TFF + NF VT IS vBE .01 Franz Sischka .] } 1.) NqB NF VT (CV2) again with equation (H1) from the introduction chapter and additionally with TFF if 2 = TF { 1 + XTF [ . the capacitors in the GummelPoon model given in the introduction chapter with equations (O) .] if + ITF vBC exp [ .
capacitance BE builtin potential BE junction exponential factor These parameters model the BaseEmitter and the BaseCollector space charge capacitance. MJC and CJS. Measurement setup: Measurement result and extraction techniques (pF) C BE 1. FC*VJE VJE Mesuring and modeling the BaseEmitter capacitance GummelPoon Toolkit B2_CV.2p slope: MJE C(v) 0.e.WPS  17. (CJC. i. Therefore this chapter covers only the modeling of the BaseEmitter capacitance. The second terms with the TFF and TR parameters will be modeled later by Sparameter mesurements. MJS is the same) CJE VJE MJE BE zerobias deplet.6p . the Collector is left open while the Emitter is open during the measurement of the BaseCollector capacitance. VJC. VJE. the modeling formula is the same. MJE.01 Franz Sischka .04.8p 3 1 CJE 0 1 vBE (V) CV meter . VJS. the first term in (CV1) and (CV2). open 1. In both cases.GP: Space Charge 3 MODELING THE SPACE CHARGE CAPACITORS: Extraction of CJE. For the measurement of the BaseEmitter capacitance.
WPS  17. that eliminates by its measurement principle parasitic capacitances to ground.5 GummelPoon Toolkit B2_CV. the small parasitic package capacitors CBC . The equation: The behavior of the space charge capacitor is given by equations (CV4a) and (CV4b): for vBE < FC * VJE : C SBE = C JE æ v BE ö ç ç1 − V ÷ ÷ JE ø è M JE (CV4a) and else: C SBE = with CJE : VJE : MJE (1 − FC ) C JE (1+ M JE ) é v ù * ê1 − FC * (1 + M JE ) + M JE * BE ú VJE û ë (CV4b) space charge capacitance at vBE = 0V builtin potential or pole voltage (typ.7V) : junction exponential factor. this effect can be avoided by applying an AC short to the open transistor pin versus ground (big capacitor). CBE. the measurements of CSBC and CSBE are always an overlay of the other capacitances CSBi ( i = E. CV meter . default 0.5um): (linear pn junction (> 5um): MJE = 1/2) MJE = 1/3) FC : forward capacitance switching coefficient. C SBC package capacitors CBC CCE C SBE C(v) CBE .04.and CCE .C ) and. when measuring packaged devices.GP: Space Charge 4 NOTE on the influence of the remaining capacitances of the open pin: as one of the transistor pins is left open. When using a capacitance meter like the Agilent4284.01 Franz Sischka . determines the slope of the cv plot (abrupt pn junction (<0. 0. The total measured capacitance is therefore CSBi in parallel with the parasitic ones. This means that the measurement results are always too big.
Therfore.vBE / VJE ] (CV5) This equation can be interpreted as a linear function according to the ideas of linear regression analysis: y = b + m x (CV6a) b = ln(CJE) (CV6b) m = MJE x = ln[1 . depending on the actual value of VJE. we only use the measurement data from the negative bias..01 Franz Sischka . Following (CV6d).m. this procedure is repeated with an incremented VJE. Once the best regression coefficient is found. A linear curve is fitted to this transformed 'cloud' of stimulating and measured data. This is done using a starting value for the unknown parameter VJE (e. In the next step. the three main equations of a linear regression are b=f(xi.b).9999.9. Thus we get from (CV6c): MJE = . together with a fitting quality factor r²=f(xi.yi.m(VJE_opt) and from (CV6b): CJE = exp [ b(VJE_opt) ] Validity of this extraction: The parameter extraction for the space charge capacitor is valid only for stimulus voltages vBE below FC * VJE . the stimuli data of the forcing voltage vBE are nonlinearily converted too. and we get another pair of m(VJE) and b(VJEC). FC_default = 0.e. the regression line fits better or worse the transformed data 'cloud'. The logarithmic conversion of (CV4a) yields: ln(CSBE) = ln(CJE) MJE ln[1 .. Thus we get the yintersect b(VJE) and the slope m(VJE) for the actual value of VJE.WPS  17.2V).0.5. I. See also the appendix.04. How to proceed: the measured values of CSBC are logarithmically converted according to (CV6a). For a good fit.GP: Space Charge 5 Determination of the CV parameters: For simplicity. But now the regression coefficient r2 will be different from the earlier one.yi) and m=f(xi. r²~0. GummelPoon Toolkit B2_CV. 0. These two arrays are now introduced into the regression equations (see appendix) as corresponding yiresp. the iteration loop is exited and we finally get VJE_opt as well as the corresponding b(VJE_opt) and m(VJE_opt). xivalues.vBE / VJE ] (CV6c) (CV6d) with y = ln(CSBE) and Linear regression means to fit a line to given measurement points.g.yi).
The measured Sparameters are deembedded. X_HIGH. Then simulate with the extracted parameter values. a network analyzer can be used instead of the CV meter as well. package or pad capacitances.mdl for more details. Click 'Copy to Variables' under 'Options' in that plot. This transform applies a data transformation and regression analysis to the data inside the box.g. using simulation or the substitute transform calc_cv. If they are not known and therefore cannot be deembedded (calculated out of the measured data). we might end up with a mix of parameter values obtained during our last transistor modeling and today! open setup "/gp_classic_npn/cv/cbe_bhi" ( means modeling of CBE.01 Franz Sischka . This may happen especially to VJC and MJC. NOTE: try also macro 'extract_n_opt_CV' Some comments on CVmodeling In practice there is always an overlay of this capacitance with some parasitic ones. Y_HIGH perform transform "br_CJE_VJE_MJE" (box regression). See ICCAP file: 1_gummel_poon/more_files/s_to_cv. Y_LOW.07. converted to Y parameters and the CV traces can be calculated out of their imaginary parts. the extracted CV parameter values may have no physical meaning. In this case.2 GP: Space Charge 6 WHAT TO DO IN ICCAP: Since this is our first parameter extraction step. perform a measurement. see ICCAP Window: 'Model Parameters' Otherwise. click a box into plot "cvsv" (capacitance vs. do the same for the capacitor CBC in setup 'cbc_bhi'. the Base is biased and Emitter and Collector are grounded.WPS  06. voltage) to select the measurement data used later for extractions. If there are resolution problems with fFcapacitances and CV meters. This will cause ICCAP to save the box corners in the 'cbe_bhi' Setup Variables X_LOW. we first reset all parameter values to default. e. GummelPoon Toolkit B2_CV. with Base contact at high voltage pin).
Also. RB 1 MODELING THE RESISTORS C O N T E N T S: Extraction of RE Extraction of RC Extraction of RBM from DC measurements An alternate method to calculate the ohmic parasitic resistors from sparameter measurements ________________________________________________________ The methods given below are considered as standard extractions.01 Franz Sischka . the other model parameters are still not yet knwon. no simulation or optimization is performed in the setups of DUT prdc in file gp_classic_npn.WPS  19.3 GP: RE. Therefore. they are tuned in order to fit the ohmic regions: RE in the forward Gummel plot (iC and iB vs vBE) and RC in the reverse plot (iE and iB vs vBC). In the Gummel plots. these parasitic resistor parameters are finetuned in the setups dc/fgummel and dc/rgummel. But the parameter values are pretty often merely a 'first guess'. RC.mdl Instead. GummelPoon Toolkit B3_OHMIC.12.
20 iB (m A ) RE = ∂v CE ∂i B transformed measured data: RE . When we apply a Base current and have the Emitter pin grounded.01 Franz Sischka . If we derivate vCE with respect to iB. Measurement of the open Collector voltage ('flyback method') and the transformed measurement data in the RE domain (delta(vCE) / delta(iB)) Extracting the parameters: The ohmic emitter resistor is physically located between the internal Emitter E' and the external Emitter pin E. RC. visu_RE GummelPoon Toolkit B3_OHMIC.WPS  19. we get the equivalent RE for each operating point. we get a voltage at the open Collector that is proportional to the Base current through this Emitter resistor.12. RB 2 MODELING THE EMITTER RESISTOR Extraction of RE Measurement setup: v C E (m V ) 50 Measurement result: iC=0 30 123456 RE 1 vCE iB 10 0 10 . The value of RE is then the mean value of the flat range in this plot.3 GP: RE.
RB 3 WHAT TO DO IN ICCAP: .3 GP: RE. since . The values of the ohmic parasitics will be finetuned later in the GummelPoon plots GummelPoon Toolkit B3_OHMIC. Do not simulate or optimize this setup.run transform visu_RE and enter '1' (data transform) this will derivate the measured data and display the calculated effective RE against the stimulus iB.12.01 Franz Sischka .click a box around the most constant range of measured data and click ‘Copy to Variables’ .reexecute transform visu_RE to extract the RE value (enter '1' for this operation mode).WPS  19.measure the setup rb_re . .the other DC model parameters are not known yet . RC.the GummelPoon model cannot represent ‘unconventional’ measurement conditions like the actual flyback method.
01 Franz Sischka .12. with additionally measuring vBE /T. (vBEvCE) iB 100 iC=0 60 123456 vCE 123456 RBM=27 Ohm vBE 40 80 120 1/iB iB .3 GP: RE.IRB. or the nonlinear RBB' = f(RB. This method is applied now. The theoretical values of the measured voltages are: vCE = VT * ln(1/AI) + iB * RE AI: reverse current amplification in common Base and vBE = iB * RE + iB * RBM + vB'E' Subtracting these equations and dividing by iB yields: GummelPoon Toolkit B3_OHMIC. Applying these three methods to the same transistor 'will generate typically three different values' for the Base resistor (!). An interesting method to determine RBM is to use the REflyback method.Base resistance at high current There are several methods to determine the Base resistor: either the constant part of it (RBM) from pure DC measurements.RBM) from a s11 plot or from noise figure measurements.WPS  19. RC.Zimmer/. visu_RBM iB Measurement setup and determination of RBM out of transformed measured data. RB 4 Extraction of RBM from DC measurements RBM min.
cannot be determined by this method. NOTE: when RB becomes measurable DCwise (the 'ohmic' range in the GummelPoon plot). its value is typically already lowered to the value of RBM. .click a box around the most constant range of measured data and click ‘Copy to Variables’ . In a final step.3 GP: RE.vCE iB = const iB + RBM i.01 Franz Sischka .reexecute transform visu_RBM to extract the RBM value Again. we then apply a loop to these data. in which a line is fitted to two adjacent points. RB 5 vBE . parameter RB (the higher Base resistor value for lower Base bias). The incremental yintersect is then displayed against the stimulus iB. this is usually not the case.run transform visu_RBM and enter '1' (data transform) this will calculate the local Base resistor for each bias point. since the other DC model parameters are not known yet NOTE: If a sensitivity analysis for a GummelPlot shows a reasonable impact of the Base resistor to the forward and reverse Base current. as described above. However. NOTE: See also the appendix chapter 'direct visual parameter extraction' WHAT TO DO IN ICCAP: .12. a regression analysis applied to these transformed measured data will give the yintersect RBM. GummelPoon Toolkit B3_OHMIC.the measurement of setup rb_re is reused .WPS  19. an optimizer run on these two curves simultaneously might make sense to obtain a guess on the actual value of RBM. This means. Therefore. and display the RBM value against the stimulus iB. and the yintersect is calculated. we simply set RB=RBM. RC.e. do not simulate or optimize this setup.
the same flyback method like for RE is applied. If there is a parasitic pnp transistor present.measure the setup rc .3 GP: RE. The only difference is that the Collector pin is grounded.12. RB 6 MODELING THE COLLECTOR RESISTOR For the extraction of RC.run transform visu_RC and enter '1' (data transform) this will derivate the measured data and display the calculated effective RC against the stimulus iB.01 Franz Sischka . The values will be finetuned later in the setups fgummel and rgummel. do not simulate or optimize this setup Note: Try also and study macro 'extract_resistors" NOTES: as mentioned above. For details on alternate DC modeling methods of the parasitic resistors. Again.reexecute transform visu_RC to extract the RC value. GummelPoon Toolkit B3_OHMIC. these 'classical' extractions of the ohmic model parameters are used to get a good estimation about the parameter values. . see also the publications of /Berkner/ and /MacSweeny/.WPS  19. WHAT TO DO IN ICCAP: . RC. and the Emitter pin is left open and its voltage is measured. See the corresponding model file of this toolkit. this method will not give accurate RC values.click a box around the most constant range of measured data and click ‘Copy to Variables’ .
RB 7 An alternate method to calculate the ohmic parasitic resistors from sparameter measurements Since the fitting of the Sparameters is the goal of a good transistor modeling.01 Franz Sischka . it makes sense to think about extracting the ohmic parameters from Sparameter measurements also.WPS  19. NETWORK ANALYZER PORT 1 PORT 2 GROUND Ib Ib/2 GROUND Ib/2 GROUND Bias for Parasitic Extraction Lc Rc Ib 2 GROUND rd Lb Rb Rce Vc rd Vt rd = Ib/2 Ib Ve Re Le GROUND GROUND GummelPoon Toolkit B3_OHMIC. With the known value of the Base current. The following figures sketch a reliable way to do that. The basic idea is to overdrive the transistor and to reduce its effect to simple diode characteristics ('hot' measurement). the remaining resistor values can be calculated easily.12.3 GP: RE. RC.
RC.rd / 2 with rd= VT / Ib Lc= Imag(Z22)/w . take the mean value from the lowest frequency.0 Resistance (ohms) Rc 0.0 5 G H z FRE Q 6 .Le Le= Imag(Z12)/w Lb= Imag(Z11 )/w . RB 8 Bias for Parasitic Ext.0 5 G H z GummelPoon Toolkit B3_OHMIC. The plot below gives an example: 2.0 0 .Re Re= Real(Z 12) Rb= Real(Z11) .WPS  19.Re . continued rd << RCE Lc Ib 2 GROUND Rc Lb Rb rd / 2 Vc=Ve Re Ib GROUND Vt rd / 2 = Ib Le GROUND Lb Port 1 Rb rd / 2 Rc Re Lc Port 2 Le GROUND GROUND Z11 Z12 Z21 Z22 = rd/2+Rb+Re+jw(Le+Lb) Re+jwLe Re+jwLe Rc+Re+jw(Le+Lc) Rc= Real(Z 22) .Le The resistor values are finally displayed versus frequency and their values are obtained as a simple mean value.12.3 GP: RE. If there is a frequency drift.01 Franz Sischka .
GummelPoon Toolkit B4_DC.WPS  17. this approach can easily lead to the 'infinite loop' mentioned above! ICCAP does not need the simplified equations. These three plots have a certain context between each other. ISE and NE Extraction of IKF Extraction of the remaining reverse parameters NR. it can easily happen that if the Gummel plot itself is nicely fitted. ISC. Because. Then. but between 2V and half the value of the maximum vCE of the output plot. the applied CollectorBase voltage is set to vBC = 0V. if the Gummel plot fits. Most often. we apply a vCE that is not zero. 0. The optimizer in ICCAP always uses a true simulator like SPICE in the background that includes the complete GummelPoon equations. if we use vBC = 0 for the Gummel plot.01 Franz Sischka . We can take a smarter approach.2 . The reason for this is that it simplifies the modeling equations (H). Otherwise. We do this for the following reason. Neglecting this context can easily lead to one of the famous. one for forward and another for reverse mode.. It is based on a measurement of iB and iC simultaneously.04.. Once the Gummel plot is fitted for this special voltage. This can be explained as follows: Let's consider the forward Gummel plot. Differently from the commonly used method mentioned above. NC and IKR Three measurements are required in order to extract the DC parameters: > an output plot including both.. there is no reason for having vBC = 0 for the Gummel plot.5 to vCEmax). we leave this setup for the moment. forward and reverse operation. We first measure the forward output characteristic and extract VAR and VAF. in this case. > and two socalled Gummel plots.(L) drastically. while extracting the DC parameters or other parameters. the output characteristics doesn't match and so on. A final finetuning is then easily achieved by adjusting VAF and BF. the following output plot simulation already hits the measured curves exactly in the middle of the output characteristic. this means that the output characteristic fits in the saturation range (vCE ~ 0. and measure the forward Gummel plot.GP: DC 1 NONLINEAR C O N T E N T S: DC MODELING Extraction of VAR and VAF Extraction of IS and NF Extraction of BF . versus vBE and is typically plotted halflogarithmically. socalled 'infinite modeling loops'. BR. However. Therefore.9V) and not int the disired linear range (vCE ~ 0..
The relationship between iC and iB leads to the beta plot. considering a cut through this plot for a fixed vCE (4V in the example). if we extract the DC forward parameters from a GummelPoon measurement that is biased like this.01 Franz Sischka . all measurements fit together.DC1. so does the output characteristic.WPS  17. log(iB. and using this value of vCE when measuring the GummelPoon plot. the output characteristic is measured and VAR and VAF are extracted. GummelPoon Toolkit B4_DC. Therefore.iC) iC iB beta vCE=4V vBE vBE iC iB vCE vCE=4V Fig. First. if beta fits. Therefore.DC1: Proposed context of the DC forward measurement setups.GP: DC 2 An illustration of this idea is presented below in fig. Then.04. also plotted against linear vBE instead of the usual logarithmic iC (which is the same for iC below the ohmic effects in the GummelPoon plot) and again highlighting the corresponding output data points by buttons. which we were starting from. we have data points that refer directly to our previous output characteristics measurement with the corresponding vBE .
mdl in the more_files directory. by forcing iB instead of vBE . GummelPoon Toolkit B4_DC. which are not included in the standard GummelPoon model. However.04.. It is recommended in this case to apply a vCE as low as possible for the Gummel plot (below the thermal runaway). if the output characteristic drifts off when forcing vBE .GP: DC 3 For the output characteristic. The ohmic effects are in this case overlaid by the thermal selfheating. because it could be affected by selfheating as well. or no good fitting at all.WPS  17.DC2. Note: iC is scaled identically for both plots! Because.DC2: Forcing iB rather than vBE for the output characteristics prevents from obtaining measurement curves including the selfheating effect.01 Franz Sischka . and we will either get wrong model parameters for RE and IKF. we should be careful when measuring the Gummel plot. Note:Such a check is implemented in file data_mgmt/BIP_MEAS_MASTER.mdl iC iB iC vBE vCE vCE Fig. but well above the saturation region of the output plot. we also prevent from measuring thermal selfheating effects. This is sketched below in fig. we should also measure the same Collector currents values with a corresponding vBE as well. Note: See also ICCAP file bip_output_char_i_or_v.
we need to go ahead and extract the remaining DC forward model parameters from the Gummel plot. Only then. 1 2 2 6 vC(V) Fig. we come back to this setup and finetune the VAR/VAF values by optimization later.WPS  17. we consider VAF. with the correct BF etc. this may affect the other model parameters. Now. followed by VAF.. VAR/VAF can be determined with only little overlay of the other (actually still unknown) parameters.01 Franz Sischka . and before optimizing the fit of this setup. This typically happens for VAR<5. 3 iB 123456 iC 1 0 vCE iB . let's discuss the theoretical background of the Early voltage extraction.DC3: Measurement of the output characteristic GummelPoon Toolkit B4_DC.04. Therefore.GP: DC 4 MODELING THE OUTPUT CHARACTERISTIC EXTRACTION OF VAR VAR VAF AND VAF reverse Early voltage forward Early voltage Modern fast bipolar transistors exhibit small values of VAR. Therefore. measurement setup: iC(mA) measurement result: . As will be shown in this chapter. For an easier understanding. the simulation of the output characteristic can fit the measured data (!). Due to the simplifications of the GP space charge model implementation in SPICE. After the Early voltages are extracted. the extraction of the nonlinear DC parameters is best started with the extraction of VAR.
Thus (DC4) becomes: iC = IS NqB IS IKF vBE exp[ . (L) of the introduction chapter with the Base charge equation q NqB = 1S * 1 + 1 + 4q 2S 2 for the modeling of nonidealities like the Basewidth modulation: ( ) (DC5) q1S = 1 vBE vBC 1 ..01 Franz Sischka .04.] NF VT (DC5) and (DC4) : q2S = (DC6) GummelPoon Toolkit B4_DC. therefore the terms vBC { exp[ . we have to start with some simplifications: We consider only the forward active region. the Base Collector voltage is vBC < 0V.1} Ni VT for Ni = NR resp. (H) ..GP: DC 5 The equation: Provided that: vB'E'=vBE and vB'C'=vBC .1) NF vt IS BR vBC .] NF VT vBE exp[ .] NR vt vBC { exp[ .( exp[ .] . the GummelPoon model describes iC by iC = IS NqB vBE {( exp[ .] .(DC4) and (DC4) may be neglected.1 } (DC4) In order to handle this complex formula.1)} NR vt 1 } vBC { exp[ . Here.1 } + IS IKR vBC { exp[ .VAR VAF (DC3) and the hilevel injection effect: q2S = IS IKF vBE { exp[ .WPS  17. Ni = NC in equ.] NF vt .] NR vt .] .] NC vt  ISC  1 } (DC4) see equ.
DC3 showed iC versus vCE.GP: Equ. with vBE = f(iB).iB) as shown in fig. GummelPoon Toolkit B4_DC.04. Therefore.(DC6) in (DC5) yields: DC 6 NqB = 1 æ v 2ç 1 − BE VAR è æ æ v öö I * ç 1 + 1 + 4 S expç BE ÷ ÷ I KF è n F VT ø ÷ V ö ç ø − BC ÷ è VAF ø (DC7) (DC7) in (DC5) gives: iC = æ v v ö * ç 1 − BE − BC ÷ VAR VAF ø æ v ö è I 1 + 1 + 4 S expç BE ÷ I KF è n F VT ø æ v ö 2 * I S expç BE ÷ è n F VT ø (DC8) Fig.WPS  17.DC3.01 Franz Sischka .vCE (DC9) (DC9) in (DC8): iC = æ æ 1 1 ö v CE ö * ç 1 − v BE ç + ÷ ÷+ è VAR VAF ø VAF ø æ v BE ö è IS 1+ 1+ 4 expç ÷ I KF è n F VT ø æ v ö 2 * I S expç BE ÷ è n F VT ø (DC10) with typically vBE << VAR and vBE << VAF we get: æ v ö 2 * I S expç BE ÷ è n F VT ø æ v ö * ç 1 + CE ÷ iC = VAF ø æ v ö è I 1 + 1 + 4 S expç BE ÷ I KF è n F VT ø or æ v ö 2 * I S expç BE ÷ è n F VT ø iC = 1+ 1+ 4 æ v ö IS expç BE ÷ I KF è n F VT ø * 1 ( VAF + v CE ) VAF (DC11) Thus we got iC = f(vCE. (DC8) has to be rearranged using vBC = vBE .
Have also a look into "/gp_classic_npn/dc/routput/READ_ME". and that the curves do not match. NOTE: As you will find out with your own measurements. Then.(DC11) are pretty straight forward. This means that we should be sure that the output characteristics measurement has been taken in the linear range of the GummelPoon plot. the Collector current of the output characteristics measurement (equation DC11). See equation (DC18) of the next chapter. open setup "/gp_classic_npn/dc/foutput". WHAT TO DO IN ICCAP: open setup "/gp_classic_npn/dc/routput". becomes zero for vCE = VAF. VAF is rather a function of the bias current than a constant. i. An alternate method could also be to determine VAF out of the delta of two Gummel plot curves iC(vBE) for two different CollectorEmitter bias voltages.. GummelPoon Toolkit B4_DC. Depending on the type of transistor.04.WPS  17. perform a measurement. sigma(VAF) can range up to VAF/2 ! The reason is that the assumptions in equations (DC5). Then. with a maximum iC well below the ohmic effects. perform a measurement.e. How to proceed: VAF is the xaxis intersect of the tangent fitted to the linear region of the output characteristics. Therefore an estimation of VAF by using only 1 tangent may be sufficient.GP: DC 7 Extracting the parameter: We consider all assumptions from above valid(!!). perform transform "br_VAR" (extract VAR) simulate with the extracted value of VAR. Do not be confused about the simulation result. perform transform "be_VAF" (extract VAF) simulate with the extracted value of VAF. The standard deviation of the values of VAF found by applying tangents to all slopes in the output plot is most often very big.01 Franz Sischka . it is only important that the slopes of simulated and measured curves match! We will have a much better fitting after the extraction of the other DC forward parameters. when an optimizer run is performed later (after the extraction of the remaining DC forward parameters). Please note again that the IC_CAP optimizer calls the simulator which includes the full set of model equations and therefore finds the correct final value of VAF. Because all other DC parameters are still set to default.
01 Franz Sischka .WPS  17.DC4: Measurement of the Collector current vs. measurement setup: 123456 123456 iC iB vBE .04. are the only ones that are dominant in the measurement setup given in the figure DC4 below.3*NF*vt vBE(V) æ v BE v CB IS ç 1 − + ç VAR VAF è ö ÷ ÷ ø Fig. NF determines the slope and IS the yintersect of the halflogarithmically plotted iC(vBE).GP: DC 8 MODELING THE COLLECTOR CURRENT: EXTRACTION OF IS AND NF IS NF transport saturation current forward current emission coefficient These 2 parameters. BE voltage GummelPoon Toolkit B4_DC. measurement and extraction principle: . iC IKF 1/RE 1decade 1decade 2*(2. together with the already known Early voltages. vCE=const.3*NF*vt) 2.
( exp[ .01 Franz Sischka ..VAR VAF (DC14) and the hilevel injection effect (half the slope of log(iC) vs. no ohmic and no IKF effects. i.1 } (DC15a) IS vBC + . vBE for high iC) q2S = if IKF + ir IKR . (L) from the introduction chapter: iC = 1/NqB ( if .] . especially the Basewidth modulation factor q1S = 1 vBE vBC 1 .] IKR NR VT or: q2S = IS vBE .1) NF VT vBC { exp[ . we get for forward biasing vBE vBC vBC ~ < << 0.] IKF NF VT .{ exp[ ..7V << VAR 0 V VAF.] NR VT 1 } vBC .e. Firstly.GP: DC 9 The equation: Provided that vB'E'=vBE and vB'C'=vBC .] NC VT 1 } (DC12) with NqB = q 1S * 1 + 1 + 4q 2 S 2 ( ) (DC13) for the modeling of the charge dependencies.04.WPS  17. we start again with the iC formula (H) .] . (DC13). the formula reminds to apply the following series approach for small values of x : GummelPoon Toolkit B4_DC.ir ) IS NqB IS BR ir/BR iBCrec or: ic = vBE {( exp[ . and Therefore (DC12) simplifies to: æ v ö I i C = S expç BE ÷ N qB è N FvT ø (DC16) Let's have a closer look to equ.1 } (DC15b) In order to determine IS from (DC12) for small vBE.1)} NR VT ISC vBC { exp[ .{ exp[ .
IKR in reverse (DC15a).04. we consider the lower and medium current ranges well below the Effect of IKF or the influence of the ohmic Resistor RE . a transformation can be applied to the measured data in order to obtain a linear context between the measured values of iC and the stimulating values of vBE in (DC19): A log10 conversion of (DC19) gives: log (ic) = log (IS) + vBE NF VT log(e) or: log (ic) = log (IS) + 1 2.3026 NF VT vBE (DC20a) This can be considered as a linear form: when setting: = b +   y = log (ic) b = log (IS) y m      GummelPoon Toolkit B4_DC. the terms vBx < VAx can be neglected and we obtain: æ v BE ö i C = I S expç çN v ÷ ÷ è F Tø (DC19) NOTE: compare (DC19) with the measurement result given in the figure DC4 above. On the other hand. the other coefficient q2S begins to contribute for high Collector currents above IKF in forward operation resp. Therefore. For the modeling of IS and NF. Extracting the parameters: Following the curve fitting techniques given in the chapter on regression analysis in the appendix. (DC16) simplifies to: æ æ v ö v v ö i C = I S ç 1 − BE − BC ÷ expç BE ÷ è VAR VAF ø è N F vT ø or. because vBC=vCB æ v BE v CB ö æ v BE ö i C = IS ç ç1 − V + V ÷ ÷ expç çN v ÷ ÷ è è F Tø AR AF ø (DC18) For bigger values of the Early voltages.WPS  17.01 * x  (DC20b) Franz Sischka .GP: DC 10 1+ x ≈ 1+ x / 2 what means for our case: NqB = q1S * (1 + q 2S ) (DC17) NqB from (DC17) is split into two parts: q1S represents a lowering of the Collector current for increasing Early voltages (DC14). and reduces the Collector current as well. This can be seen in the iC Gummel plot as a curve shift to lower Collector currents.
04.3026 m VT) (DC21b) Validity of the extraction: vBE between 0. click a box into plot "ibic_vbe" around a linear range for the IS/NF extraction click 'Copy to Variables' (check how the box bounds are exported into the setup variables X_LOW. where the halflogarithmicly plotted data represent a straight line. optimize with transform "bo_IS_NF" Have also a look into "/gp_classic_npn/dc/fgummel/READ_ME". From comparing (DC20a) with (DC20b) we know how to resubstitute the parameters out of 'b' and 'm': Is = 10 b (DC21a) and NF = 1 / (2.GP: DC 11m = 1 / (2. Y_LOW. NF). we obtain yintersect 'b' and the slope 'm' of the straight fitted line.3026 NF VT)  x = vBE and How to proceed: We select a subrange of the measured data. GummelPoon Toolkit B4_DC. simulate with the extracted parameter values. which refers to X_LOW etc. Applying these formulas. Then the logarithmically converted iCi of this subrange are interpreted as y. Y_HIGH) perform transform "br_IS_NF" (box regression IS.2V [no noise] and 0.7V [no high current effects] WHAT TO DO IN ICCAP: open setup "/gp_classic_npn/dc/fgummel".WPS  17.and the linear vBEi values as xdata for the regression formula. X_HIGH. perform a measurement.01 Franz Sischka .
01 Franz Sischka . In the case of NF this would mean to start with æ v ö i C = I S expç BE ÷ è N FvT ø to convert it logarithmically in order to obtain 1 ln(i C ) = ln(I S ) + * v BE N F vT This is the mathematical representation of the halflogarithmic Gummel plot for iC. See also appendix A for more infos. The parameter NF is proportional to the slope and we have therefore to differentiate ln(iC) with respect to vBE and obtain: ∂ ln(i C ) 1 = ∂v BE N FvT Solved for NF gives NF = 1 ∂(ln(i C )) VT * ∂( v BE ) Therefore. we get NF build the mean value from this range vBE(V) Fig. In directory "visu_n_extr" of this toolkit you will find more ICCAP model files that follow the idea of direct visual extraction. if we display the calculated NF (what is the 'effective NF' for every measured data point) versus vBE.04.GP: HINT: DC 12 Transforming the measured data such that the model parameter can be displayed directly against the stimulating voltage or current is another smart way to determine model parameters. GummelPoon Toolkit B4_DC. if the model is able to fit the measured data at all (if there is a constantly flat range) and then to easily extract the parameter as the mean value of that flat range.WPS  17.DC5: Direct visual extraction for parameter NF This allows us to check.
GP: DC 13 Some comments on ic_measured above IKF: ________________________________________________ Referring to many graphical parameter extraction methods and also to fig. From (DC17) we get neglecting the Earlyeffect: NqB = or solved for NqB: NqB = 1 + iC I KF i C * N qB 1 * 1+ 4 2 I KF Let us consider the two cases: ic KF _______________________________ < I                 ic > IKF ________________________________ here is NqB ~ 1 and here is NqB ~ ic IKF vBE IS exp[ .01 Franz Sischka .DC4 above. the iC and the iB curve in the same way. In practice. and therefore (DC22b) is not so well suitable for extracting IKF.04. This means that the effect of RE cancels out for the beta curve ß=iC/iB. the overlaying parameter RE is affecting basically both.] NF VT ic = ______ √IKF IS (DC22a) (DC22b) Interpreting the result: From (DC22b) we learn that the ic curve has half the slope for currents above IKF (see fig. However.WPS  17. Therefore.] 2 NF VT and introducing this into (DC16) gives finally: ic = or: IKF ic ic = vBE IS exp[ .DC4).] NF VT vBE exp[ . On the other hand. parameter IKF affects only iC. there is always an overlay with the ohmic resistor RE. IKF is commonly extracted from the beta curve of the transistor ! GummelPoon Toolkit B4_DC. some more background information is given for the model curve of ic above IKF. however.
same measurement setup as in fig.DC4 extraction principle: iB 1 RE 1decade BF knee! 2. Therefore we will not follow the graphical extraction method. the three parameters of this chapter are most often introduced with their corresponding influence on the different ranges of the iB curve in DC6.WPS  17. but develop another method instead. ISE AND NE BF ISE NE ideal forward maximum beta BE leakage saturation current BE leakage emission coefficient In the literature.GP: DC 14 MODELING THE BASE CURRENT: EXTRACTION OF BF . Also. This is especially true for BF in the beta plot (overlaid from IKF and NE). BE voltage GummelPoon Toolkit B4_DC.04.3*NE*vt 1decade vB(V) Fig. In practice.DC6) is not visible.01 Franz Sischka . there is most sometimes an overlay of the influences.DC6: measurement of the Base current vs.3*NF*vt ISE 2. We will derive a formula for the 3 parameters directly from measured data that has been taken from the range around the 'knee'. modern transistors have pretty low recombination effects for the BE diode: the 'famous knee' (see finger pointer in fig.
e.GP: DC 15 The equation: Provided that vB'E'=vBE and vB'C'=vBC .1 } NE VT vBC + ISC { exp[ .] NF VT + vBE ISE exp[ . Extracting the parameters: This equation (DC25) is one of the few cases during the bipolar modeling.DC4. GummelPoon Toolkit B4_DC.{ exp[ .01 Franz Sischka . (A) .1 } BR NR VT vBE + ISE { exp[ .] . ISE and NE have to be calculated and then set to zero.1 } BF NF VT IS vBC + . then iB = iBE + iBC see equ..04.(DC23) to: iB IS = BF vBE exp[ .] .1 } NC VT (DC23) We assume once again that: and vBE vBC ~ < 0. the Collector current iC with neglected high current effectsinto (DC24) yields the pretty simple form: æ v BE i i B = C + ISE * expç ç N *V BF T è E ö ÷ ÷ ø (DC25) We will use both iC = f(vBE) and iB = f(vBE) from the simultaneously measured currents of the GummelPoon measurement of fig.] NE VT (DC24) Introducing (DC19) i.] . (G) of the introduction chapter = IS vBE . The solution of this system of equations finally gives these 3 parameters. (At least.. We now have iB as a function of vBE as desired.WPS  17.7V 0 V << VAR This simplifies equ.{ exp[ . Therefore the partial derivations of the curve fitting error in (DC25) versus BF. the author had not sufficient intuition!).] . where a nonlinear transform applied to the measured data doesn't give a straight line.
) .] NE VT .Σ .exp( .04. Step by step: The partial derivation of (DC29) versus BF is: 1 N iCi2 .GP: DC 16 As iB ranges from pico.Σ .exp() BF i=1 iBi2 NE VT N 1 2vBEi + ISE Σ . This is unfortunately not possible for NE .01 Franz Sischka .exp() i=1 iBi2 NE VT N iCi Σ i=1 iBi = 0 (DC30) and versus ISE: 1 N iCi vBEi .exp() = 0 i=1 iBi NE VT (DC31) GummelPoon Toolkit B4_DC.] NE VT 1 + Erel = i + (DC27) or: Erel = i iCi iBi BF + vBEi exp[ .similar to VJ of the space charge capacitor . iCi and vBEi. Therefore it is expanded by the individual error Ereli for every data point of index i: iCi iBi BF ISE iBi ISE iBi vBEi exp[ .Σ BF i=1 iBi2 + ISE N iCi vBEi Σ .WPS  17.until the sum of individual errors according to (DC29) is minimized. we will have to minimize the relative error between measured and fitted curve.to milliAmpère.] NE VT (DC26) Equation (DC26) is only approximately true for the real measured data iBi. Thus we get from (DC25) after dividing by iB: 1 = iC iB BF + ISE iB vBE exp[ .1 ] iBi NE VT ! = Minimum E tot = (DC29) It can be shown that the parameters BF and ISE can be separated out of the partial derivations with respect to BF and ISE with a reasonable effort.1 (DC28) Using least means square techniques we now have: N 2 N iCi Σ Erel = Σ [ i=1 i i=1 iBi BF ISE vBEi 2 + .exp() i=1 iBi2 NE VT N 1 vBEi . This parameter has to be iterated .
the range around the 'KNEE' (see fig. BF and ISE of this minimized error is the final parameter extraction result.01 Franz Sischka . GummelPoon Toolkit B4_DC.exp() i=1 iBi i=1 iBi2 NE VT How to proceed: A subset of the measured data iBi and iCi .04. The triplet of NE. Next a suitable starting value for NE is selected ( e. NOTE: the complexity of (DC32) and (DC33) illustrates that transforming measured data to a linear context and applying linear regression techniques is often a much smarter approach for parameter extraction. NE = 1) and the error according to (DC29) is calculated. NE is then incremented until this error becomes a minimum.exp() i=1 iBi2 NE VT and (DC31) by N iCi2 Σ i=1 iBi2 These two new equations are added and their sum is solved for ISE: N 1 vBEi N iCi2 N iCi N iCi vBEi Σ .e.WPS  17.exp() Σ [ Σ .exp() Σ Σ Σ .exp() ] i=1 iBi2 NE VT i=1 iBi2 i=1 iBi2 NE VT (DC32) Now we can also separate BF from (DC30): BF = (DC33) Validity of the extraction: vBE above measurement resolution and below high current effects.GP: (DC30) is expanded by  DC 17 N iCi vBEi Σ . i.g.DC6) are selected and introduced into equations (DC32) and (DC33).exp() i=1 iBi NE VT i=1 iBi2 i=1 iBi i=1 iBi2 NE VT ISE = N 1 2vBEi N iCi2 N iCi vBEi 2 Σ . N iCi2 Σ i=1 iBi2 N iCi N iCi vBEi Σ ISE Σ .
due to the shift of the iC Gummel plot. In this case. NE). perform transform "bo_ISE_BF_NE" (box optimization ISE. GummelPoon Toolkit B4_DC. Therefore. the Base current recombination effect does not occur.01 Franz Sischka . BF. click "Copy to Variables". without the assumption of big Early voltages. Equation (DC18). perform transform "br_ISE_BF_NE" (box regression ISE.04. is better in this case. Have also a look into "/gp_classic_npn/dc/fgummel/READ_ME".GP: DC 18 WHAT TO DO IN ICCAP: in setup "/gp_classic_npn/dc/fgummel". correct the measured Collector current values to i* C _ measured = i C _ measured v 1 − BE VAR Note: VAR<<VAF. you may also try the tuning function in "tune_ISE_BF_NE" If there is no 'knee' with your measured transistor. It would lead to a too low extracted value of BF. the Collector current formula of (DC19) inserted into (DC25) is not quite correct. so this correction is sufficient before inserting them into equations (DC32) and (DC33). NE). click a box into plot "ibic_vbe" around the 'knee' at low vb.WPS  17. simulate with the extracted parameter values. switch off the Base current recombination effect in the GP model. BF. This can be done by setting ISE to a very small value (ISE=1E30) and the slope parameter NE to a flat slope (NE=2). Note: For low values of VAR.
and of IKF for the decrease. IKF reduces already beta. This helps a lot in better understanding the influence of the parameters ISE and NE on the increase of beta.beta) iC iB beta Fig.WPS  17. GummelPoon Toolkit B4_DC. This effect describes a decrease of the proportionality of log(iC) versus vBE. we display it always versus the same stimulus vBE. log(iC. This is shown in fig. DC4. From a modeling standpoint. it can be best extracted out of ß=iC/iB.04. the iB and iC curve.DC7: Plotting beta together with iC and iB for better understanding of the context of the model parameters It becomes also clear why for some transistors. Therefore.GP: DC 19 MODELING THE CURRENT AMPLIFICATION AT HIGH CURRENT EXTRACTION OF IKF IKF forward beta high current rolloff Referring to fig. Unfortunately. as already mentioned. the beta plot should not be considered isolated from its origin. iB and iC. IKF models the Webster pushout effect. together with iB and iC. this effect is also overlaid by RE. or referring to the beta plot. Therefore. DC8. However. IKF only affects iC. while RE affects mainly both.01 Franz Sischka . before it can reach the value of the BF for increasing bias. BF seems not to affect the maximum of the beta trace at all: there is no parallel region between log(iC) and log(iB) in the Gummel plot.iB.
thermal selfheating has to be taken into account.01 Franz Sischka . estimate or extract RE first! Hint: avoid thermal selfheating effects!.DC3.. NOTE: for best results.DC8: Beta from the measurements of fig.50mA. because it displays a measurement result versus a stimulus and not another measurement result.WPS  17.04.. provided we neglect highcurrent effects. DC pulsed measurements with pulse widths about 1us should be used in this case (the HP4142 offers only pulse widths ≥100us!).GP: DC 20 same measurement setup as for fig. But this way of plotting log(ß) is more straightforward. To avoid this.DC4 beta BF NE ISE mainly IKF but also RE vBE Fig. Note: log(ß) is plotted versus vBE .DC4 and fig. This becomes visible if the betaplot for a forward and reverse vBE sweep looks different at high vBE. For Collector currents >10. The conventional plot using log(ß) versus log(iC) is identical to it. GummelPoon Toolkit B4_DC.
04.WPS  17.beta) iC iB beta vBE GummelPoon Toolkit B4_DC.GP: DC 21 Before we start with the extraction of IKF. vBE slope VAF vBE vBE log(iB) iB iC iB nonideal BE diode recombination effect Mainly RE affects the slope of both.iB. we are now ready to understand the following schematized Gummel plots. vBE down for higher VAF Webster effect reduced log(iC) vs.01 Franz Sischka . They characterize at a glance the different effects for the Base and Collector current in the GummelPoon modellog(iC) log(iC) iC iC Early effect shifts log(iC) vs. iC and iB at higher vBE bias vBE vBE log(iC.
] + ISE exp[ .exp[ .GP: DC 22 The equation: The current amplification is defined as: ß = iC iB (DC34) Provided that: vB'E'=vBE and vB'C'=vBC and further vBE << VAR < 0 V and vBC We introduce (DC24) for iB and(DC16) for iC into (DC34): iC iB IS vBE .] NF VT we get: iC iB iC vBC ( 1 .exp{ ..] NqB NF VT IS vBE vBE .04.01 Franz Sischka . This gives: iC iB iC vBC vBE ( 1 .[ .WPS  17.] BF NF VT NE VT = (DC36) Divided by vBE IS exp[ .exp[ ...+ . see equ.DC18.] BF NF VT NE VT = (DC35) We further introduce an approximation for NqB.) IS exp[ ..exp[ .] + ISE exp[ .] IKF VAF NF VT IS vBE vBE ..]} BF IS VT NE NF = (DC37) GummelPoon Toolkit B4_DC....) IKF VAF 1 ISE vBE 1 1 .
* iC This formula is again more ore less true for the measured data iCi and iBi with the stimulating voltage vBE.. we proceed as in the chapter of the determination of the iB parameters: iC vBC ( 1 ..]} BF IS VT NE NF (DC39) E = N N iBi Σ Ereli2 = Σ { i=1 i=1 iCi How to proceed: In order to keep things simple. 10A...WPS  17.. e. Thus we have to introduce again an individual error Ereli for each measured data point of index i: iBi 1 + Ereli = ..[ .+ .[ ...) IKF VAF 1 ISE vBE 1 1 .01 Franz Sischka . and then divided by 2 in every iteration.exp{ .IKF VAF 2 * . Have also a look into "/gp_classic_npn/dc/fgummel/READ_ME"...]} BF IS VT NE NF (DC38) Finally the total error for all measured data ( 1 . until the total error given in (DC39) is minimized.IKF VAF 1 ISE vBEi 1 1 .+ .[ .g.]} BF IS VT NE NF 1 iB = .. and also in some alternate methods on the IKFextraction in the 'direct visual parameter extraction' model file. N ) is (least means square): iCi vBCi 1 .* iCi iCi vBCi 1 . Thus IKF is set to a starting value.1} 1 ISE vBEi 1 1 .04.exp{ . (DC39) is solved for a best IKF by iteration. Finetuning is then done by the optimizer. GummelPoon Toolkit B4_DC.GP: DC 23 Extracting the parameters: As we want to consider again relative errors. then perform transform "e_IKF" (extract IKF) and check the simulation result...+ .. WHAT TO DO IN ICCAP: open setup "/gp_classic_npn/dc/fgummel" and plot "ibic_vbe" (beta is the rightaxis data). run transform 'o_BF_IKF_RE' for finetuning the parameters of this setup..exp{ .
WHAT TO DO IN ICCAP: open transform README in setup "/gp_classic_npn/dc/rgummel" and follow the modeling sequence given there Last not least. macro 'extract_n_opt_DC' includes a suitable automated modeling strategy for both DC forward and reverse. ISC. The strategy used in this macro covers that effect by looping a bit between forward and reverse extraction and optimization. BR. that the reverse Early voltage is affecting the forward modeling. IKR The reverse modeling can be performed like the forward modeling. due to its low value. Just correct the macro if required. GummelPoon Toolkit B4_DC.04.WPS  17. This sequence may be different for your actual transistor.GP: DC 24 EXTRACTION OF THE REMAINING REVERSE PARAMETERS EXTRACTION OF NR. Included in this example is also the interesting and pretty often recognizable effect.01 Franz Sischka . NC. Simply exchange Emitter and Collector.
See also the chapter on the limitations of the GummelPoon model at the end of this manual.01 Franz Sischka .WPS  17.mdl under directory 'more_files' in this toolkit file collection. For more details. rbeta iB iE a flat range! vBC vBC vx (a steeper slope of log(iB) versus vBC for vBC < vx. N and RS. you might consider replacing the GummelPoon recombination modeling (parameters ISC and NC) by a external diode with its parameters IS. GummelPoon Toolkit B4_DC. something that is not included in the GummelPoon model).GP: DC 25 Note on reverse GummelPoon modeling If your reverse beta curve and reverse Gummel plot look like below. refer to file rgummel_special.04.
If you need additional support for deembedding. i B RBB' B' CB'C' C' IC B RC C CC'S' rB'E' R0 CB'E' gm*vB'E' S E' RE E Fig. It can be derived from fig.2b of the introduction chapter as a linearization at each bias point of the transistor. GummelPoon Toolkit B5_AC. Please contact the author. It is assumed that the measurements have been made on the pure semiconductor device without being affected by packaging parasitics. Once again the key to meaningful AC measurements and thus modeling is a good network analyzer calibration with excellent standards and a correctly defined calibration kit data in the network analyzer.GP: AC mdlg overview 1 A C S M A L L S I G N A L M O D E L I N G OVERVIEW We are now ready to consider the basics of modeling for frequencies higher than 100MHz.04.WPS  17. It should also be mentioned that the probe pins have to have an excellent frequency performance within the transistor measurement frequency range. This means to find the proper semiconductor behavior out of the distorted measurement by deembedding.AC1: AC small signal schematic of the bipolar transistor NOTE: XCJC effect neglected. If this is not possible. bond pads or other parasitics. deembedding techniques have to be applied. please refer to the additional toolkits for ICCAP. calibration and for better understanding Sparameters.01 Franz Sischka . Let us go first for the AC small signal equivalent schematic.
01 Franz Sischka .GP: AC mdlg overview 2 The following equations give the values of the internal elements in fig.vB'C' / VJC ] (AC4) ________________________________________________________ gm The transconductance gm finally is using equ.WPS  17. (H) d iC d vB'E' + d iC . d vB'C' (AC5) GummelPoon Toolkit B5_AC. ________________________________________________________ g0 g0 = The output conductance is: diC / dvB'C' diB / dvB'C' (AC2) ________________________________________________________ CB'E' or CPI including the delay time effect modeled by TFF is given in equ. CB'E' simplifies to CB'E' = TFF d iC d vB'E' ¸ TFF gm (AC3) ________________________________________________________ while CB'C' or CMU (MU or µ stands for 'mutual' simplifies because of vBC < 0 at forward operation to: CB'C' = CjC mJC [1 . They represent the linearized DC.AC1. As a first order estimation.and CVequations at the DC operating point.04. gB'E' vB'E': gB'E' = From equ.(B) in the introductory chapter we get from the derivative of iB versus IS BF NF VT vB'E' exp[ .] NE VT + (AC1) where the second term can most often be neglected for operating points of iC above 1 uA.(P) and (R3) of the introductory chapter for the particular operating point voltages.] NF VT ISE NE VT vB'E' exp[ .
It is organized like this: .RBB2.derivation of the small signal schematic for the parameter extraction .RBB1. The measurement setup for the rBB' characterization is given below in fig. we further assume: vBE = vB'E' and vBC = vB'C'.Base resistance at hi current It is assumed that XCJC = 1.short introduction into the basics of the Smith chart . we refer to fig.discussing the expected frequency dependence of rBE.GP: Base Resistor 1 MODELING THE BASE RESISTOR rBB' Extraction of RB. Network Analyzer port1 port2 BIAS TEE BIAS TEE iB_DC VCE_DC Fig. considering rBB' constant . at medium Base resistance min. We simplify it to cover mainly the imput impedance.01 Franz Sischka .04. and that ß is the DC current amplification. This chapter explains how to model the Base resistor from S11 data. This figure explains the two cases: frequency > 0 Hz and frequency > ∞ Hz GummelPoon Toolkit B5_RBB.RBB1: Measurement setup for the rBB measurement To begin with. Simulations and optimizer runs after the parameter estimation will eliminate this simplification.AC1 from the previous chapter.WPS  17. This leads to the schematic of fig. IRB and RBM RB IRB RBM zero bias Base resistance curr.enhancing the schematic for the biasdependent omic resistor rBB' As an approximation to keep the equations simpler.
S11 = 0 for R = 50 Ohm .RBB2. we consider first the hypothetical case that rBB' is no function of bias.GP: Base Resistor frequency > 0 Hz CB'C' and RC omitted iB 2 frequency > infinite.50 S11 = R + 50 R . R j50Ohm 1 50Ohm 1 j Fig.50 S11 = R + 50 j with the NWA measurement impedance of 50 Ohm.RBB2: Simplified AC BaseEmitter input impedance scheme for low and high frequencies In order to evaluate the schematic and the device parameters of fig.RBB3: Note on the Smith chart transformation (S11 = 1 for R = infinite .01 Franz Sischka . we have to consider the measured S11 data.04. Note: In order to get familiar with the problem. As a reminder. This is best done by displying them in a Smith chart. GummelPoon Toolkit B5_RBB. we can use S11 instead of H11 for the RIN modeling as well and our measurement result should look like figure RBB4 . a Smith chart transforms the right side of the complex resistor plane R into the area of a circle of radius '1' using the transform R .WPS  17. S11 = 1 for R = 0 Ohm) Therefore. the Base resistor is considered as a constant. ohmic resistor RBB'. CB'C' and RC omitted iB rBB' B' rBB' B' B gB'E' C B gB'E' C i=0 CB'E' ß*iB E' RE E' RE E E Fig. In other words.
biasindependent RBB' Ideally. For infinite frequencies. so that the influence of a uncertain value of RE is minimized using this method. CB'E' will act more and more like a short and eliminate the influence of resistor 1/gB'E'. GummelPoon Toolkit B5_RBB. RIN should look like a circle.GP: Base Resistor Complex resistor plane for the interpretation of RIN j*IMAG RBB'+RE 3 Rin RBB'+1/gB'E'+RE(1+ß) REAL frequency and the same measurement result displayed in a Smith chart: j 1 1 j Fig. assuming an ohmic. Now the BaseEmitter capacitance has completely shorted 1/gB'E' and thus the transconductance gm became 0 as well. For higher frequencies. This means that the transistor has no beta any more. RIN should hit the xaxis at RBB' + RE (effects of CB'C' and RC omitted!). the value of RBB' can be estimated quite accurately.WPS  17. This method is advantageous because the estimation of the Base resistor is affected only by the parameter RE.RBB4: idealized input resistor curve. S11 / Smith chart = RBB' + RE (frequency > infinite) (RBB1) As RE is known from DC measurements. Moreover. there is mostly RBB' >> RE .04.01 Franz Sischka . The starting point at DC is RBB'+1/gB'E'+RE(1+ß).
Fig. bias dependent one. overlying this DC bias dependency with the frequency dependance from above.04. low iB E B C high iB E B C RBint RBM the voltage drop on RBint decreases the internal voltage B'E'. the channel moves to the base contact and the overall RBB' is reduced to RBM. we end up with S11 curves like sketched in fig. In order to separate rBB' with the proposed method. GummelPoon Toolkit B5_RBB. I. The base majority carriers will tend to cross the BE junction closer to the external base contact. Realworld measurement curves will look like these curves at low frequencies only. we expect a lower Base resistor for higher bias.01 Franz Sischka . where the potential is higher. which is then assumed to be equal to rBB'+RE. we must fit circles to the S11 curves at low frequencies and then calculate the xintersect from an extrapolation of the circle for infinite frequency. the more iC is extending its flow area closer to the internal Base contact due to current crowding. biasdependent Base resistor rBB'.WPS  17.RBB7.RBB6: idealized input resistor curve with a real.e. In reality. effects of CB'C' and RC omitted.GP: Base Resistor 4 So far we considered rBB' to be simply ohmic.e. This means. i. This is shown in fig. rBB' is modeled more complexly.constant. This is due to the overlay of more second order effects. The GummelPoon model combines both resistors into a single. The sketch below depicts that. One separate resistor from the outer Base contact to the inner Base contact (ohmic RBM) and a biasdependent part from the inner Base contact to inside the inner Base. the higher iB.22a: current crowding leads to a bias dependent Base resistor Now. extrapolated for infinite frequency j*IMAG r in rBB'+RE rBB'+1/gB'E'+RE(1+ß) REAL iB frequency Fig.RBB6. This means.
They are centered to the xaxis. Finally. and therefore ß.x02 This again can be considered as a linear form (!) with ylin = b + m xlin where x2 + y2 = ylin r2 .WPS  17. This contradicts general rule to always concentrate on meaningful measurements close to the operating range for good parameter extractions. iB should also be as high as possible. an optimization (of the S11 parameters) of this setup should only be applied after the fitting of these transit time parameters.RBB7: Idealized ohmic Base. to keep the rBB' influence dominant over RE (1+ß).GP: Base Resistor rBB' + RE (Ohm) 5 RE+RB frequency > infinite RE+RBM iB(A) Fig. Unfortunately.x02 (RBB3a) = b 2 x0 = x = m xlin and (RBB3b) (RBB3c) (RBB3d) GummelPoon Toolkit B5_RBB. is as small as possible. The suitable circle formula is:: (x . which will be extracted next. Moreover. Also.x0)2 + y2 = r2 (RBB2) + 2 x0 x or x2 + y2 = r2 . Therefore. so that iC > IKF.04.01 Franz Sischka .RBB7 is often overlaid by the parameters TF.plus Emitterresistor versus iB Notes on some limitations of this extraction strategy: Because of the influence of 1/gB'E' on the range where the circle must be fitted to. the trace of Fig.effect by 1/gB'E' . The extraction strategy: Circles must be fitted to the lowfrequency sections of interest. iB should be as high as possible to not dominate the r BB' . ITF. this range of iB typically is also not the operating one. the rBB' measurement could now be dominated by thermal effects. XTF and VTF.
GP: Base Resistor 6 This means: The measured data xi and yi are introduced into equ.WPS  17. Next the ylin(i) are plotted versus the xlin(i) and a straight line regression is applied. using (RBB3c).01 Franz Sischka .RBM] z tan^2(z) (RBB4) with z= æ 12 ö iB 1+ ç ÷ −1 è PI ø IRB æ 24 ö iB ç 2÷ è PI ø IRB 2 see model equations (M) and (N) of the introducion chapter. This means we are now ready to consider the formula for RBB in the GummelPoon model: The equation: The nonlinear Base resistor is described in the GummelPoon model as: rBB' = RBM tan(z) . Fig.r After all these preconsiderations. we get: x0 = m / 2 and from the yintersect b using (RBB3b): __________ r = √ (b + x02) Finally the left circle intersection with the xaxis (for the frequency > infinite) for our rBB'extraction is: rBB' + RE = x0 .z + 3 [RB . From the slope m. we are now able to generate the trace of RBB out of the measured Sparameters.04.(RBB3a).RBB8 shows the plot of this equation: GummelPoon Toolkit B5_RBB.
GP: Base Resistor rBB' (Ohm) 7 RB RBM IRB iB(A) Fig. we now have to fit the model curve from fig.RBM) / 2 GummelPoon Toolkit B5_RBB. zero bias Base resistance When iB > infinite then z > PI/2 and 3 [tan(z) .value RBM + (RB . min. How to proceed: When iB > 0 then z > 0 and therefore [tan(z) .z] z tan2(z) > 1/3 We get from (53) solved for RB: RB = rBB' [ iB > 0 ] .RBB8 to the measured data of fig.01 Franz Sischka .z] z tan2(z) > 0 what gives from equation (RBB4) solved for RBM: RBM = rBB' [ iB > infinite ] .21.RBM] 0.Base resistance at hi current Finally when iB = IRB then z = 1.51 ¸ (RB + RBM) / 2 That's why IRB is the current where the Base resistor is half its max.RBB7 (after subtraction of RE).04. Thus rBB' = RBM + [RB .RBB8: Base resistor curve as implemented in the model This means: after we got rBB' from the measurement.WPS  17.
If despite all of these efforts the transformed measured data do not match the curve of fig. If you need these high bias values.RBB8 can be derived. the same 'freq'. after the deembedding. This is both done by ICCAP DUT 'nwa_meas'. it can be quite complex to obtain a reasonable S11 plot from which a rBB' curve like that one in fig. i.optimize after you are finished with the TFF parameter extractions. Then. PLease note that the stimuli must be identical in all these setups. an OPEN structure on the wafer.open setup "/gp_classic_npn/nwa_extr/rbb". set RB=RBM and model the Base resistor bias independent.01 Franz Sischka . the deembedded data are exported from this setup and reimported partially (depeding on the required subdata for the individual modeling steps) into the setups of DUT 'nwa_extr'.check the plot "rbbvsib" (rbb versus ib) .GP: Base Resistor 8 WHAT TO DO IN ICCAP: first of all. (see proposed optimization sequence given there). the modeling steps are: . Here.when measuring packaged devices). If this occurs.perform transform "calc_RBB" to convert the measured Sparameters to rBB' (set model variable DEMO=0 in order to obtain the check of the upper frequency limit for the rBB extraction) . see also transform READ_ME NOTE: As you might experience. reduce the bias for both the rBB and forward Gummel setup.mdm file. the network analyzer has to be calibrated.simulate the setup with the so far determined DC and CV parameters . You should then remeasure your SHORTOPENLOADTHRU calkit standards using the ICCAP DUTs 'CAL_xxx' for documentation.simulate with the extracted parameters. 'ib' and 'vc' Inputs must be used. . while setup 'freq_n_bias' is used to measure all data for the HF modeling. The deembedding is then performed in setup 'de_embed' and transform 'S_deemb'. Note: avoid thermal selfheating (esp. This will show up if the fitting of the forward Gummel plot of iB for high vBE becomes worse when the fitting of the S11 plot is improving during finetuning of RE and RB.perform transform "e_RB_IRB_RBM" . GummelPoon Toolkit B5_RBB. consider using pulsed measurements (DC bias pulse width around 1us). . resp. Finally. the setup 'dummy_open' is used to measure the OPEN structure.WPS  17. an empty package. .RBB8.04.import the data from the .e. has to be measured for the deembedding of the inner transistor from the outer parasitics.
01 Franz Sischka . (P).. and TFF is modeled with the formula: TFF if 2 = TF { 1 + XTF [ .. ITF. GummelPoon Toolkit B6_TFF.(R). and XTF Extraction of VTF Modeling the Diffusion Capacitor CDBC Section 1: TF XTF ITF Extraction of TF. This chapter covers the modeling of CDBE. the already CVmodeled CSBC dominates over CDBC in equ. and XTF ideal forward transit time coefficient for bias dependence of TF highcurrent parameter for effect on TF For forward active operation of the transistor. while in equ. TF is the ideal forward transit time modeling the 'excess charge'.) } 1.44 VTF see (S) with the ideal forward Base current if from equation (C). S in the introduction chapter).] if + ITF vBC exp ( .11.(R). The parameters XTF.6 GP: Transit Time 1 TRANSIT TIME MODELING C O N T E N T S: Modeling the diffusion capacitance CDBC Section 1: Section 2: Extraction of PTF Extraction of TR Extraction of TF. ITF. In this operating mode. CDBE is described by the biasdependent transit time TFF in equ. and ITF cover the operating point dependence from the DC bias iC ~ if . the AC behavior is modeled by CBC and CBE (see equations O . the more important term is CDBE.WPS  30. while VTF describes the dependence from vCB ~ vCE.
this formula is extended to TFF (iC . however. it is mentioned that the transit time after equ.6 GP: Transit Time 2 Preconsiderations concerning the measurement: Like in the previous chapter.p / p01(iC. /Sinnesbichler p. Therefore we can neglect the zero p01 against the pole pp1. GummelPoon Toolkit B6_TFF. In some other publications. v CE ) = 1 − (RC + RE + RB / ß ) * CBC 2 * PI * f T 1− pole (iC . like e. v CE ) = 2 * PI * f T 1− pole (iC . it can be shown that the transistor's h21(f)parameter behaves frequency wise like a lowpass filter with the transfer function h21(f) = ß 1 .g.WPS  30.106/. p. the parameter estimation is again performed using a simplified model. the TFF used for modeling is RC*CBE smaller than the value converted from fT. and the transit frequency for h21(f) = 1 is simply fT 1− pole i C . whereas the parameter finetuning is finally done during an optimizer run using the full set of SPICE model equations.vCE) 1 + p / pp1(iC. Note: In many publications. v CE ) = 2 * PI * f T 1− pole (iC .(TFF1) is TFF (iC . there is p01 > pp1. v CE ( )= 2 * PI * TFF (i C . v CE ) = 2 * PI * f T 1−pole (i C . with the goal of a direct extraction of the TFF parameters followed by a postoptimization. v CE ) In practice.01 Franz Sischka . Referring to appendix B.198/ TFF (iC . v CE ) 1 − RC * CBC In this case.11. v CE ) 1 (see (11) of appendix B) or inverted: TFF (i C . v CE ) 1 (TFF1) where fT1pole is a function of the bias current iC and the bias voltage vCE . the additional terms can be neglected and the simple equation (TFF1) is sufficiently correct. v CE ) 1 − (CBE + CBC) ⋅ vT − (RCX + RE) ⋅ CBC iC or after /B.Ardouin.vCE) with p = j * 2PI * f Typically.
This frequency is now used as a fixed frequency f20dB for the Sparameter measurements of this setup. It shows the transit frequency with and without simplification (Appendix B). iC. This measurement frequency can be found when transforming the measured rBB' Sparameters to Hparameters (using the TwoPort function).6 GP: Transit Time 3 Now back to the parameter extraction: We will first consider the extraction of the parameters TF. v CE ) = h 21 (i C . this fixed frequency should be from a 20dB/decade range of MAG[h21(freq.WPS  30. v CE ) * f −20dB or f T 1−pole (i C . From the dBplot of ABS(h21(f)) versus log(frequency) we determine a frequency where the slope fits a 20dB/decade rolloff. for the whole biasdependent array of h21. a special measurement for fT(iC) and later for fT(iC.01 Franz Sischka .e. GummelPoon Toolkit B6_TFF. but it can show up like a second 20dB/decade slope shifted in frequency.11. The underlying DC bias values are a swept iC and a constant and small value of vCE (to neglect the VTF effects). vCE) at a fixed frequency higher than the the 3dB frequency. This means more parasitics are subtracted than present in reality. As we assume a 1pole lowpass for h21. what gives the biasdependent array of TFF to be fitted.TFF1 shows logh21 as a twodimensional function of the Collector current iC and the frequency freq. Then. The dependence of vCE is neglected for simplification. NOTE: if your MAG[h21] does not follow the 20dB/decade law. these Sparameters are converted into Hparameters and we get for the constant gainbandwidth product of this assumed onepole lowpass filter: 1 * f T 1−pole (i C . there is probably a socalled overdeembedding. v CE ) = h 21 (i C . vCE)] . VTF will be covered later. vCE) is needed. ITF and XTF. In other words.TFF2 is valid for all DC bias conditions. This means. does not affect the slope. underdeembedding. The opposite deembedding problem. i. Therefore it is sufficient to measure a h21(iC. Fig. v CE ) * f −20dB (TFF2) fT1pole after equ. the gainbandwidth product is a constant. This new array fT1pole is then introduced into (TFF1).
Usually.TFF2 shows such a curve. available from the author.TFF1: log h21 as a function of log (iC) and log(f) for vCE=const.TFF2: fT as a function of iC and vCE Note: for a correct modeling. For more detailed examples about how the RF power level might affect the fT curve. Under this condition and considering the nonlinear diode characteristic at the Base of the transistor. chapters 3 and 4. and the ICCAP examples on nonlinear RF modeling. log (fT) 1 fT = 2*PI*TFF vCE log (iC) Fig. iC)  4 this is the DC beta curve! log (freq) fT log (iC) real (iC) fT 1pole (iC) Fig. also including the dependence of fT from vCE . This is the typical diagram published in many data sheets.Wijnen. see literature P. fT is plotted against iC .WPS  30.v. the rectified AC signal will contribute to the DC bias! A flat trace of the fT curve at low Collector current is an indicator for that effect. GummelPoon Toolkit B6_TFF. Fig.11.6 GP: Transit Time log  h21(freq.01 Franz Sischka . check the fT curve at low iC for socalled selfbiasing! This effect occurs if the RF signal power at the Base is in the range of the DC bias power.
] if + ITF vBC exp [ . we can further simplify and get finally: iC 2 TFF = TF { 1 + XTF [ . which will eliminate this small error. we start with: if 2 + XTF [ .WPS  30.vCB TFF = TF { 1 + XTF [ .] exp [ .44 VTF TFF = TF { 1 (TFF3) see (S) with the ideal Collector current if from equ. we will use a final optimization on the Sparameter curves. if is the ideal Collector current. After the extraction of the parameters of this section. Performing the measurement: GummelPoon Toolkit B6_TFF.] } 1.11.we can set if = ic .(C).6 GP: Transit Time 5 Preconsiderations concerning the model equation: As cited at the beginning of this section. vBC as well as vBE in (C) are the DC bias voltages at the operating point.44 VTF If we choose vCB ~ 0. If we consider currents below IKF .] } iC + ITF 1. Therefore the curvetobefitted is: iC 2 .01 Franz Sischka . In this equation.] iC + ITF (TFF4) Validity of (TFF4): iC<IKF and linear forward operating point with vCB~0 or as small as possible.
vCE) = h21(iC.TFF3: measurement setup for the TFF parameter measurement First the network analyzer is set to a constant frequency on the 20dB/decade rolloff of h21(f). More details were given above.6 GP: Transit Time Measurement setup: 6 Network Analyzer set to a constant frequency at the 20dB/decade rolloff of h21 port1 port2 BIAS TEE BIAS TEE iB_DC VCE_DC constant and small (~1V) Fig. GummelPoon Toolkit B6_TFF.vCE) * f20dB (see TFF2) Then we calculate TFF = 1/(2 PI fT1pole) as the biasdependend total transit time.01 Franz Sischka . Next the transit frequency for a 1pole lowpass model is calculated as given in (TFF2): fT1pole(iC.WPS  30.11. The used test frequency had been estimated from the S_to_H parameters of the RBB measurement.
TFF4: The theoretical transit time TFF (bold) as a function of iC for vCB = 0 compared to the theoretical trace.6 GP: Transit Time 7 Extracting the parameters: Figure TFF4 highlights the theoretical trace of TFF of equation (TFF4). Due to these overlay and measurement problems.(TFF4) can be obtained from measurements without being overlied and distorted by thermal selfheating effects.WPS  30. As sketched in fig. This method is explained below.01 Franz Sischka . GummelPoon Toolkit B6_TFF. The real measured curve is overlaid by the space charge capacitor effects for low collector currents. For a detailed description of this effect. This can be seen also from equation (CV1) in the chapter on CV modeling.TFF4. this is mainly because it is not possible to force such a high Collector current that the trace of equ. but the extraction results may be not much better. see /Berkner 1993/ TFF(psec) TF(1+XTF) measured curve (effect of the space charge capacitors) ITF TF theoretical curve (from the TFF formula) iC isothermically measurable range Fig. This figure shows the theoretical curve in addition to the measured one to make things more clear. There exist some more complex strategies.11. it had been found that a pretty simple and straightforward extraction technique can be applied that gives nevertheless quite reasonable results.
correct this estimation by ITF=5*ITF(equ.TFF6). ITF and XTF is performed simultaneously with parameter VTF in a single extraction routine. and no Collector voltage in quasisaturation! XTF The behavior of TFF was given in fig. NOTE: this ITF extraction method follows the idea of the Base resistor IRB parameter extraction! WHAT TO DO IN ICCAP: the extraction of the parameter TF. since the end of the TFF trace is often not measurable.11. select a vCE ~ 1V for the extraction. So XTF is estimated from the trace of TF at max. It is difficult to measure for a higher Collector current due to thermal limitations. Due to the Collector current limitations. i. see the optimizer strategy at the end of this chapter.TFF5) can improve the starting conditions for the optimizer.01 Franz Sischka .1 (TFF5) This usually gives a pretty good firstorder estimation. See next chapter GummelPoon Toolkit B6_TFF.6 GP: Transit Time 8 How to proceed: TF is extracted as the minimum value of TFF. ITF Referring to the same measurement restrictions as above. Note that a prerequisite is vBC=0.TFF4.e.WPS  30. a good firstorder estimation of ITF is related to the max. Collector current measured: (TFF6) ITF = MAX(iC_meas) / 2 Again. an estimation correction like XTF=10*XTF(equ. For more details. applicable Collector bias current under the assumption that it would be TFF at infinite current: MAX(TFF) = TF (1 + XTF) or XTF = MAX(TFF) TF .
WPS  30.Extraction of VTF VTF voltage describing the vBC dependence of TF Finally. we consider also the vCE sweep. Measurement setup: Network Analyzer set to a constant frequency at the 20dB/decade rolloff of h21 port1 port2 BIAS TEE BIAS TEE iB_DC VCE_DC swept TFF (psec) vCE VTF iC(mA) Fig.01 Franz Sischka .6 GP: Transit Time 9 Part 2 .11.TFF5 The transit time TFF as a function of iC and vCE GummelPoon Toolkit B6_TFF.
] = TFF2 and finally: vCB2 .WPS  30. Notes: From the preconsiderations given above.11.vCB1 exp[ . So. that we do not forget to also optimize the Sparameter fittings for all bias conditions after that (setup rbb). iC represent curves which had been extrapolated from the Sparameters.] 1.44 VTF VTF = v CB2 − v CB1 æ TFF1 ö 1.44 * lnç ç TFF ÷ ÷ è 2ø (TFF7) After the extraction of these four parameters for CDBC . the plots RBB vs.vCB1 1. all the AC extraction methods need absolute clean measurements and elimination of parasitics by deembedding techniques.44 VTF] exp [ vCB2 / 1. we will next run an optimization to improve the fitting of the fT plot. the Sparameter measurement is "the real world" and the fitting should be optimized in this domain ! Again. iB and TFF vs. However. it is very important. Otherwise no curve fitting might be possible or the parameters obtained might make no physical sense.01 Franz Sischka .44⋅ VTF or TFF1 TFF2 This gives: TFF1 ln[ .6 GP: Transit Time 10 VTF can be obtained from (TFF3) for a fixed value of iC: − v CB TFF = ⋅const ⋅ e 1.44 VTF = exp [ vCB1 / 1. GummelPoon Toolkit B6_TFF.44 VTF] = vCB2 .
"calc_TFF" and "ic" check the plots "ftvsic" and "TFF" (TFF versus calculated Collector current iC) perform transform "e_TF_ITF_XTF_VTF" and simulate with the extracted parameters for optimization. execute "o_VTF" to opotimize the fitting for the biggest vCB bias finally.6 GP: Transit Time 11 WHAT TO DO IN ICCAP: open setup "/gp_classic_npn/nwa_extr/tf_ib_vcb". call "o_TF_ITF_XTF_VTF to optimize all bias points" 2.mdm file of setup 'nwa_meas/de_embed' perform transforms "S_to_H".11. IRB and RBM setup by running "o_RB_IRB_RBM" 3.01 Franz Sischka . see below. HOW TO OPTIMIZE THE SPARAMETER SETUPS IN ICCAP: for the data supplied with the file 'gp_classic_npn. GummelPoon Toolkit B6_TFF. execute finally "o_TFF". then.) in setup "/gp_classic_npn/nwa/tf_ib_vcb". "calc_ft". it has been observed that the fTfitting affects also the rBBfitting. see also macro "extract_n_opt_NWA" NOTE: Depending on the device type. run the optimizer transforms "o_TF_ITF_XTF" to optimize for the smallest vCB bias..) go back to setup "rbb" and optimize the Sparameter fitting of the RB. what optimizes all TFF parameters at all bias conditions of setup "rbb".WPS  30. select a 20dB/decade frequency from the plot 'mag_h21vsf' of setup 'rbb' (middle mouse click) and enter it to model variable F20dB import the deembedded data of this 20dB/decade frequency into setup 'tf_ib_vcb' from the exported .mdl'.) in the same setup. this Sparameter sequence is best: 1.
GummelPoon Toolkit B6_TFF. Therefore it is advisable to use an optimizer run in order to get the value of PTF.TFF5)versus vBE /T.Zimmer/. It can be obtained when plotting the phase of h21 of the TFF measurement from above (fig. As a limitation of the method.01 Franz Sischka .6 GP: Transit Time 12 Extraction of PTF PTF excess phase at frequency 1/(2PI * TF) Implemented into the model as a 2nd order allpass Besselfunction. PHASE(h21) simulated 140 measured 150 160 700m 800m vBE Fig.11. this parameter can be used to add some extra phase to the RF simulation curves.WPS  30. the measurable range is again only covering a small part of the desired curve.TFF6: Measured and simulated phase(h21) before optimization.
for the modeling of transistors in digital applications. An optimizer run on the Sparameters of this setup is used to obtain the parameter value from a typical starting value..g. 1V NOTE: TR can also be obtained from pulse measurements using an oscilloscope. TR. TR is the dominant parameter in such a setup.6 GP: Transit Time 13 Extraction of TR TR ideal reverse transit time The reverse transit time of the GummelPoon model is modeled by one parameter only.. NOTE: After /Sinnesbichler/. NOTE: the reverse biased Sparameter can be measured for these example DC bias levels: 1st order DC sweep: vCE = 1 . TR can be optimized in the S22 and S12 plot of the reverse biased Sparameters.WPS  30. GummelPoon Toolkit B6_TFF. TR can be obtained from an h21 measurement versus frequency at a typical reverse operating point.7 .TFF7: The reverse transit time TR.. MAG(h21) 4 2 0 log(freq) Fig. often more important than TFF and the junction capacitances.. e.01 Franz Sischka . Therefore. TF * 100. 5V 2nd order DC sweep: vBE = 0. the TRmodeling is a must. In many cases.11.
it is either Re or XCJC! GummelPoon Toolkit B6_XCJC. if the geometry of the device is known. if S12 becomes 'big' for high frequency. including the Emitter area. However. This parameter is difficult to determine from CV measurements. where AE is the Emitter area. Shift of Sxx vs.02 Franz Sischka . meaning the CCB capacitance is tied completely to the inner Base. it can be calculated pretty easily after XCJC = 1 .5) S21: XCJC S12: XCJC Generally speaking. the capacitance is between the outer Base and the Collector. increasing XCJC (XCJC=0 > XCJC=0. this parameter is finetuned after the other HF parameters have been determined.7 GP: XCJC Modeling 1 MODELING OF THE PARAMETER XCJC This parameter distributes the CCB junction capacitance between the inner and the outer Base contact.WPS  12. increasing XCJC (XCJC=0 > XCJC=0.AE/AB. Its default value is XCJC=1.04.5) S11: XCJC S22: XCJC Shift of Sxy vs. Usually. For XCJC=0. and AB the total area of the Base. The following Sparameter figures show the effect of XCJC for device modeling.
7 GP: XCJC Modeling 2 Note: fmax may also be used to model the effect of XCJC. the Sparameters and the transformed rBB' and TFF curves. Sparameters rather than the rBB' or the ft and TF plots. as depicted below: Although there is no efect of XCJC on ft.02 Franz Sischka .04. i. fmax is heavaily affected. This is the real world and the fitting there might be more important than the fitting of the rBB' plot with all its limitations (extrapolated S11 at infinite (!) frequency) or the fT plot (again extrapolated from H21 from StoH parameter transformation).e. you should try to optimize what you have measured. Of course. the best modeling result is a good fit in all domains. XCJC=0 XCJC=1 A final remark on the SParameter Extraction and Optimization Strategy When you run into problems when fitting both. GummelPoon Toolkit B6_XCJC.WPS  12. the transformed rBB' and TF curves.
(Temperatures in 'K).02 E4 TEMP2 TEMP + 1108 Ni = TEMP 1.) [ IS(TEMP) / IS(TNOM) ] TNOM TEMP ( .( VT TNOM .04.) [ IS(TEMP) / IS(TNOM) ] TNOM XTB TEMP 1 / NC ( .01 Franz Sischka .1151 exp[ .1) ] BF(TEMP) = BF(TNOM) BR(TEMP) = BR(TNOM) ISE(TEMP) = ISE(TNOM) ISC(TEMP) = ISC(TNOM) VJE(TEMP) = VJE(TNOM) VJC(TEMP) = VJC(TNOM) GummelPoon Toolkit B7_TEMP.WPS  17..) TNOM XTB TEMP ( .) TNOM XTB TEMP 1 / NE ( .45 E10 ln [] Ni 1.5 1.) TNOM XTB TEMP ( .( .) + 2 VT TNOM 1.16  7.) TNOM q EG 1.45 E10 ln [] Ni EG TEMP exp[ .45 E10 ( . used auxiliary variables: VT = k TEMP q EG = 1.+ ) ] 2k TEMP TNOM temperature dependant modeling parameters: IS(TEMP) = IS(TNOM) XTI TEMP ( .) + 2 VT TNOM TEMP ( .GP: Temperature Modeling 1 MODELING OF THE TEMPERATURE EFFECTS The parameters given below are modified when the selected simulation temperature TEMP is different from the extraction temperature TNOM.
Like in the forward region. the slope of ß above the knee current IKR has a fixed value of "1" and also the output characteristics saturation range is modeled inflexible. especially versus vCE. It is still very useful especially when enhancing it with external parasitics like inductors. VJC and VJS must be greater than 0.01 Franz Sischka . this is most often overlaid by RE.5V) lacks of specific parameters. The modeling of the saturated region in the output characteristics (VCE < 0. FORWARD DC MODELING The parameter IKF models the begin of the decrease in beta. They have no temperature coefficients. Currently. REVERSE DC MODELING: The reverse DC modeling suffers from a separate parameter IS.4V to insure convergence for temperature analysis up to 200'C. However. new models like the VBIC or the HiCum come more and more into play. the GummelPoon model is a good compromise between accurate modeling and a limited amount of parameters. GummelPoon Toolkit B8_LIMIT. Therefore the model cannot cover modern transistors in this range (quasisaturation). TEMPERATURE MODELING The TNOM value of VJE. APPLICATIONS IN INTEGRATED CIRCUITS There is no parasitic transistor inncluded in the model Conclusion: disregarding these limitations. parasitic diodes or lateral pnp transistors.04. AC MODELING: The TF modeling. No reverse breakdown effects are included in both BaseCollector and BaseEmitter diode.WPS  17. Ask the author for the corresponding toolkits. is not physical and often not accurate The TR parameter is not a function of current or voltage like TFF.GP: Model Limitations 1 Limitations of the GummelPoon Model OHMIC EFFECTS: The Collector and Emitter resistance parameters are constant and not functions of current or voltage. As a limitation of the model the slope of ß above the knee current IKF has a fixed value of "1" on a loglog scale. Thus NR sometimes has to be misused for better fitting the reverse iE versus vBC plot.
01 Franz Sischka . ISE and NE Calculation of h21 of the GummelPoon Model GummelPoon Toolkit B9_APPEN.WPS  17.04.GP: Appendices 1 A P P E N D I C E S C O N T E N T S: Linear Curve Fitting: Regression Analysis About the Modeling Dilemma Verifying the Quality of Extraction Routines Direct Visual Parameter Extraction of BF .
these measured values were plotted. the fitting error is: E i = Y( x i ) − y i (1) and for all data points: E = i=1 å E 2i = N i =1 å [Y( x ) − y ] i i N 2 (2) This error shall be minimized. We obtain a system of equations.01 Franz Sischka . But in general. this is a pretty smart way to get the curve fitting parameters easily without much calculations.e. NOTE: This regression analysis is simple for a straight line fit. This is known as regression analysis. Unfortunately. (3) GummelPoon Toolkit B9_APPEN.WPS  17. As will be seen in the diode example later. a nonlinear regression analysis can be quite complicated. The fitting will be done by varying the coefficients of the fitting curve of equation (2). we obtained the array {xi . Referring to an individual measurement point.GP: Appendices 2 LINEAR CURVE FITTING: REGRESSION ANALYSIS ICCAP File: $ICCAP_ROOT/examples/demo_features/4extraction/basic_PEL_routines/1fit_lin. This means.mdl Let's assume we made 'N' measurements yi at the stimulating points xi. A curve Y(x) shall be fitted to this array of measured data points using least square curvefitting technique. I.and the xivalues. This problem can be solved if we use a suitable transformation on the measured data. we have to differentiate E partially versus the curve coefficients and to set the results to zero. A linear curve with the equation y(x) = m x + b shall be fitted to these points.04.yi}. measured data is nonlinear. This situation is depicted below. The minimum of the total error E depends on the values of these coefficients. yi}. Provided we have got an array of N measured data points of the form {xi. solve it. and get the values of the coefficients for a best curve fit. Subsequently. This means that the measured data is transformed to a linear context between the yi.
GP: Appendices
3
y * *
yi
*
m
* * * *
*
1
fitte d line y = mx + b
b
*
xi
data points
x
Fig_1: linear regression applied to measurement points The error of the ith measurement is:
Ei = [ m xi + b ] yi (4a)
Using the least means square method following equ.(2) yields:
N E = N = ! + b yi ] 2 = Minimum (4b)
Σ Ei2
i=1
Σ [ m xi
i=1
Partial differentiation versus slope 'm' gives:
N 2
Σ
i=1
[ m xi
+
b

yi ] xi
=
0
(5)
and versus yintersect 'b':
N 2
Σ
i=1
[ m xi
+
b

yi ]
=
0
(6)
We obtain from (5) after a rearrangement:
N m N + b N =
Σ xi2
i=1
Σ xi
i=1
Σ yi xi
i=1 N
(7)
and from (6):
N m Σ xi i=1 + N b =
Σ
i=1
yi
(8)
Multiplying (7) by N and (8) by Σ xi and adding these two equations allows the elimination of the coefficient 'b', and we can separate the slope 'm':
GummelPoon Toolkit B9_APPEN.WPS  17.04.01
Franz Sischka
GP: Appendices
N m [ ( N N N = i=1 N N
4N N
Σ xi )2
i=1
Σ xi2 ]
i=1
Σ xi Σ yi
i=1
Σ xi yi
i=1
(9)
or:
N N
m
=
Σ xi Σ yi  N Σ xi yi i=1 i=1 i=1 N N 2 ( Σ xi )  N Σ xi2 i=1 i=1
(10)
and from (8) for the yintersect 'b':
N b = [ N m
Σ yi
i=1
Σ xi ] / N
i=1
(11)
with 'm' according to (10). With equations (10) and (11), we determined the values of the two coefficients of the linear curve which fits best into the 'cloud' of measured data. Finally, a curve fitting quality factor r2 is defined. Its value ranges from {0 < r2 < 1}. The closer it is to 1, the better is the fit of the linear curve.
N 1 x 2 å i − N i = 1 r 2 = m 2 N 1 y 2 å i − N i = 1
( (
N x i å i = 1 N y i å i = 1
) )
2
2
(12)
with 'm' from (10)
GummelPoon Toolkit B9_APPEN.WPS  17.04.01
Franz Sischka
GP: Appendices
5
About The Modeling Dilemma
Using ICCAP for the extraction of model parameters offers a lot of flexibility in terms of creating userdefined models and implementing the corresponding extraction routines. But when developping a new extraction strategy, we may run into two major problems: do the routines extract the parameters correctly? and is the model able to fit the measured device at all? This appendix proposes a method that allows us to  verify the quality of the extraction routines  check the fitting of our model to the measured data and to perform the parameter extraction simultaneously. This modelfitcheck method is also called 'Direct Visual Parameter Extraction' Both issues are pretty important in order to obtain reliable parameters and thus satisfying simulation results of the complete circuit.
GummelPoon Toolkit B9_APPEN.WPS  17.04.01
Franz Sischka
Simulate this setups using these parameter values. 4. we should obtain the right parameters. 7. The trick is to 'synthesize' quasimeasured data out of a set of parameters and to apply then the extraction routines to these data. Apply the extraction routineundertest and check the quality of the extracted parameters. i.WPS  17.GP: Appendices 6 Verifying the quality of extraction routines Let's start with the first problem: assumed. we can be sure that the extraction works correctly. This can be done as follows: 1. Define a measurement setup in ICCAP. we would know the parameters tobeextracted in advance. but only simulation data. but instead real realistic values!) Change the 'Output' data type to 'S' (simulated only). This is true if the measured data have the same shape like the model equations! If not. The array behind that output is no longer onedimensional. If we now apply the extraction to realworld measured data. Provided we get the parameter values back within a good tolerance. Example: an output characteristic for an Earlyvoltage extraction. And this leads us to the second part of this paper: GummelPoon Toolkit B9_APPEN. we might have to choose another model or go for subcircuit modeling. Thus the simulated data of step 4. 6. 2. is now converted to measured data! Reset the model parameters by clicking 'Reset to Defaults' and simulate the setup using the default parameters. ICCAP doubles now the data field to measurement and simulation data.04.01 Franz Sischka . 5. Change the 'Output' data type back to type 'B'. Select a 'typical' set of parameters (no default values like 'zero' or 'infinite'.no measurement data any more.e. it is simple to perform such a check. for which the extraction routines shall be tested. it would be easy to check the validity of the extraction routines! Using ICCAP. 3.
01 Franz Sischka . If it is there. we know that the xintersect of lines fitted to an output characteristic of a bipolar transistor should hit always the same point.04. in which range the parameter is dominant and can therefore be used for finetuning with the optimizer. The advantage of using this method is that we can see clearly. we start with some basic equations that refer to figure 1: GummelPoon Toolkit B9_APPEN. and if we display the result of this operation versus the collector voltage (first order sweep). If there is no flat range. We only have to check if there is a flat region in the transformed data domain or not. ISE and NE or: Checking if the model can fit the measured data at all This task can be solved by transforming the measured data to a domain where the parameter itself can be displayed against the measurement stimulus. if the model is able to fit the measured data at all.GP: Appendices 7 Direct visual parameter extraction of BF . we will obtain a plot of the 'equivalent' Earlyvoltages of adjacent measurement points. the model cannot fit the measured data. we can extract the parameter very simply by calculyting the mean value of the flat region. the Early voltage.WPS  17. If we apply an ICCAP PEL (parameter extraction language) program to calculcate the xintersect of a line that is fitted to two adjacent measured points. As an example. And we know at the same time. We could vary the parameter as much as we like and would not achieve a fit of the simulated to the measured data! For the application of this method.
y1 y0 = x2y1 .WPS  17.x2y1 y2 .x1 Some more usefull formulas to calculate xintersect x0 starting with y2 x0+x2 y1 x0+x1 yintersect y0: y2y0 x2 y1y0 x1 = = we get: x0 = x1y2 .x1y2 x2 .x1 These equations are implemented to the model files of directory "visu_n_extr". The following plots give some examples on how to apply this idea to the parameter extraction of a bipolar transistor using the GummelPoon model.04. BSIM etc.01 Franz Sischka .GP: Appendices 8 y y2 y1 m y0 x0 x1 x2 x Fig_1: on the definition of some basic equations for the direct visual parameter extraction Assumed we have y = m*x + y0 where m: slope y0: yintersect Then it is: x0 = b/m and: y2 . as well as to other models like Statz. GummelPoon Toolkit B9_APPEN. Curtice. It should be mentioned that this method can be applied to all the parameters of this model.y1 m = x2 .
RC get RC from the flat range at high Ic iB=2mA Ic Ic Fig_3: RC can be calculated from measured vC at stimulated iB and iC the "visualized" RC is the yintersect of two adjacent data points Franz Sischka GummelPoon Toolkit B9_APPEN.04.GP: Appendices 9. RE Fig_2: RE is the derivative of the flyback measurement the "visualized" RE is the yintersect of two adjacent data points from that flyback plot.WPS  17. . iC=0 RE measurement setup: 123456 vCE iB d ( vC) RE = d(iB) REFlyback Method . RC measurement setup: 123456 vCE iC ∂v RC = CE − R E ∂i C RCFlyback Method iB=const.01 .
24 in the rBB chapter in order to estimate not only RBM but also IRB and RB (see plot on the right). we might expect a flat range. In this case.WPS  17. see fig. Next. the yintersects of two adjacent data points in this transformed data plot are drawn versus 1/iB again. 1/iB visu_RBM iB Fig_4: visual extraction of RBM and to estimate IRB and RB possibly.GP: Appendices 10 RBM (after the method of /Zimmer/. From the curve above. but in reality we get a curve like the left one below: In our example. GummelPoon Toolkit B9_APPEN.04. we might display the visualized data versus iB. and interprete the result simular to fig. RBM ranges from 5 to 12 Ohm.01 Franz Sischka .10 of the chapter on the DC resistor modeling): measurement setup transformed measurement result: (vBvC) iB 100 iC=0 60 123456 vCE RBM=27 Ohm 40 80 120 1/iB 123456 iB vBE .
! Determine VAF out of the xintersect of a line through two adjacent measurement points: X=vc Y=ic.04.M i=1 WHILE i < SIZE(Y)1 VAF[i]=ABS(X[i+1]*Y[i1]X[i1]*Y[i+1])//(Y[i+1]Y[i1]) i = i + 1 END WHILE Output Characteristics VAF Fig_5: the Early voltage is calculated from the xintersect of a line through two adjacent data points GummelPoon Toolkit B9_APPEN. iC vCE iB .WPS  17.01 Franz Sischka .GP: Appendices 11 VAF measurement setup: 123456 .
3*NF *vt 1decade BF 1/RE ISE 1decade 2. GummelPoon Toolkit B9_APPEN. iC 123456 iB 123456 iC iB vBE . interpretation of the GummelPoon plot for parameter extraction: iC iB IKF 2. vCE=const.GP: Appendices 12 Let's study also the application of the "visual" method for the GummelPoon plot: measurement setup: measurement result: GummelPoon Plot .WPS  17.04.3*NE *vt vB(V) IS Fig_6: a typical GummelPoon measurement and its interpretation.01 Franz Sischka .
WPS  17.M) i=1 WHILE i < SIZE(Y)1 ISE[i]=(X[i+1]*Y[i1]X[i1]*Y[i+1])//(X[i+1]X[i1]) i = i + 1 END WHILE IS Fig_8: IS from the yintersect of a line through two adjacent measurement points GummelPoon Toolkit B9_APPEN.GP: Appendices 13 NF NF NF = 1 ∂ (ln(i C )) VT * ∂ ( v BE ) Fig_7: NF can be obtained from the inverse derivative of the iCGummelplot IS !Determine IS out of the yintersect of a line through two adjacent measurement points: X=vb Y=log10(ic.01 Franz Sischka .04.
This is exactly the meaning of the transformation results in fig_9 and fig_10! Franz Sischka GummelPoon Toolkit B9_APPEN.WPS  17.GP: Appendices 14 NE NE NE = 1 ∂ (ln(i B )) VT * ∂ ( v BE ) Fig_9: NE as obtained from the inverse derivative of the iBGummelplot ISE Determine ISE out of the yintersect of a line through two adjacent measurement points: X=vb Y=log10(ib. This means that the parameters ISE and NE will not contribute a lot to the curve fitting and may be difficult to extract. the measured data does not show a big recombination effect on the ib(vbe) curve.01 .M) i=1 WHILE i < SIZE(Y)1 ISE[i]=(X[i+1]*Y[i1]X[i1]*Y[i+1])//(X[i+1]X[i1]) i = i + 1 END WHILE ISE Fig_10: ISE from the yintersect of a line through two adjacent measurement points As can be expected from an inspection of fig_6.04.
such that IKF can be extracted from the 'knee' of the 'strippedoff' icGummelplot! !stripoff the effect of RE on the Gummelplot !iCmeas = IS*exp(vBEint/(vt*NF) = IS*exp((vBEext(iC+iB)*REiB*RB)/(vt*NF)) ! i. we have to 'stripoff' the effect of RE on the GummelPoon iC curve: ___________________________________________________________________________ ! this extraction assumes that RE is already extracted properly. It eliminates the effect of RE on the GummelPlot.M) Yib=ABS(ib.01 Franz Sischka .WPS  17.GP: Appendices 15 IKF In order to be able to visualize the effect of IKF (see fig.IS*(exp(X//(MAIN. calculated out of IS and NF iC meas.6 in the introduction chapter).curve Fig_11a: offstripped effect of RE on the GummelPoon iC curve GummelPoon Toolkit B9_APPEN.NF*VT))) ! calculate the ideal ic Gummel curve ! as a reference tmp1= MAIN.RBM*Yib)//(MAIN.e.RE*(Yic+Yib)+MAIN.04.NF*VT))) ___________________________________________________________________________ RETURN tmp+j*tmp1 GummelPoon Plot of iC for high vB ideal iC curve.curve strippedoff from RE original iC meas.M) ! calculate the strippedoff Gummel Plot tmp = Yic*(exp((MAIN. multiplying iCmeas by exp((vBEext(iC+iB)*REiB*RB)/(vt*NF)) ! will give iC without the influence of RE and RB !! X =ABS(vb) Yic=ABS(ic.
IKF = Y1[index] vB(IKF): when deviation (ideal / off_stripped) = SQRT(2) then: goto iC(vBE) and get IKF=iC(vB(IKF)) ideal_iC_curve offstripped_iC SQRT(2) Fig_11b: calculating IKF from the ratio ideal_iC_curve / strippedoff iC GummelPoon Toolkit B9_APPEN.WPS  17.GP: Appendices NOW CALCULATING IKF 16 X = ABS(vb) Y = IMAG(strip_off_RE)*(REAL(strip_off_RE))^1 Y1 = REAL(strip_off_RE) i=SIZE(Y)1 index = 0 WHILE i > 0 IF Y[i] < SQRT(2) THEN index = i i = 0 END IF i = i .1 END WHILE PRINT index ! calculuate IKF out of iC(vBE) at that index MAIN.01 Franz Sischka .04.
IS//MAIN. strippedoff from NqB 200 beta 0 vBE Fig.12: beta plot and beta offstripped from base charge effects NqB for BF extraction GummelPoon Toolkit B9_APPEN.IKF*exp(vbe//(MAIN. beta 600 400 beta.01 Franz Sischka .04. beta" extraction for BF: !Calculation of ß = iC/iB vbe = vbve vbc = vbvc beta = ABS(ic) / ABS(ib) ! with some simplification is: ! beta ~ BF/nqb ! nqb = q1//2*(1+SQRT(1+4*q2)) ! with q1 ~ (1vbe//VARvbc/VAF))^1 ! q2 ~ IS//IKF*exp(vbe//(NF*VT)) ! VT : temp.VAF)^1 q2 = MAIN.WPS  17.voltage (a model variable) q1 = (1vbe//MAIN.VARvbc//MAIN.NF*VT)) nqb=q1//2*(1+SQRT(1+4*q2)) BF=ic//ib*nqb RETURN BF get BF from the offstripped curve from a range where beta is max.GP: Appendices 17 BF the idea is to stripoff the base charge effect (NqB) from the beta plot and then apply the often cited "max.
dummy SWITCH = dummy PRINT "calculating the x_intersect .M[i] N = N+1 END IF i = i+1 END WHILE MAIN. Let us conclude with an example of a PEL program used for this method.04. It is the VAF transform of fig_5.GP: Appendices 18 This method of direct visual parameter extraction can be applied to all GummelPoon parameters.M Y = SMOOTH3(Y) ! link to stimulus data ! link to measured data ! smooth measured data SWITCH = SWITCH ! a model variable LINPUT "data transform (1) or extraction(1) ?"."Display Plot") LINPUT "click a box and rerun this transform to extract VAF". Literature: HPEESOF. For more examples see the ICCAP files in directory "visu_n_extr".WPS  17.SWITCH.VAF = result//N ! export parameter value PRINT "MAIN. The program performs either the data transformation (SWITCH=1) or performs the parameter extraction (SWITCH=1): X=vc Y=ic." tmp = Y ! a lousy array declaration i=1 WHILE i < SIZE(Y)1 tmp[i]=ABS(X[i+1]*Y[i1]X[i1]*Y[i+1])//(Y[i+1]Y[i1]) i = i + 1 END WHILE tmp[0] = tmp[1] tmp[SIZE(tmp)1] = tmp[SIZE(tmp)2] ! watchout for the array bounds IF SWITCH == 1 THEN ! sumup all parameters within the box i = 0 N = 0 result = 0 WHILE i < SIZE(X) IF X[i] > X_LOW AND X[i] < X_HIGH THEN result = result + tmp.VAF ELSE MENU_FUNC("Visu_va". GummelPoon Toolkit B9_APPEN. Characterization Solutions Journal.VAF = ". Spring 1996.MAIN.dummy RETURN tmp END IF NOTE: Included in your "bipolar toolkit" is a directory called "visu_n_extract" that contains ICCAP model files with suggestions on direct visual extraction for most of the GummelPoon parameters..01 Franz Sischka ..
9.vB' / rBB' (1) Output: 0 = (iC .WPS  17.ß / pCB'C 1  0  rB'E 0 ß rB'E 1  iB rB'E  0 0 1 1 / rB'E  0 (1): (2): (3): (4): GummelPoon Toolkit B9_APPEN. using some small simplifications:   CB'C  iC B ____ B'  <.1 / rBB'  iB  1 / pCB'C 0 . since h21 = iC / iB for RL = 0 The equations: Input: vB = iB rBB' + vB' iB = vB / rBB' .ß iB' + iB) rB'E using rB'E Internal 2: iB' = vB' / rB'E This gives the matrix iC vB iB' vB'   0 1 / rBB' 0 .C o____oiB' oo rBB'   V   __   __  / \ ^ / \ ¦ \ /  CB'E  rB'E \ / ¦ / \   / \ ¦ iC'=ß iB'  CAC \___/ iB   \___/ V             Note: Collector is ACwise shorted.schematic of fig.04.GP: Appendices 19 Calculation of h21 of the GummelPoon Model We start from the small signal equival.ß iB') / pCB'C + vB' (2) (3) = rB'E // CB'E (4) Internal 1: vB' = (iC .01 Franz Sischka .
ß / pCB'C (1)   + 1 / rB'E    rB'E 1  rB'E ß rB'E  .ß / pCB'C 1    .rB'E { .ß / pCB'C 1  .ß / pCB'C 1     1 1 / rB'E  1 / pCB'C 1 1 / pCB'C .ß / pCB'C 1    iB rB'E 0 ß rB'E 1     0 0 1 1 / rB'E   0 1 / rBB' 0 .iB rB'E iC = gives for h21: iC h21 = iB = ß .GP: Appendices Solved for iC: 20 iC =  iB 1 / rBB' 0 .ß / pCB'C 1     rB'E 0 ß rB'E 1     0 0 1 1 / rB'E  Solving for the 2nd column of the nominator and the 2nd column of the denominator:  0 .04.(1)} pCB'C rB'E 1 1 ß rB'E ß rB'E ....+ rB'E + .1 / rBB'  rB'E ß rB'E 1   0 1 1 / rB'E iC =      Now solving for the 1st column of the nominator and the 3rd row of the denominator:  .1 / rBB'   1 / pCB'C 0 .1 / rBB'    0 0 .WPS  17.1 / rBB' iB rB'E ß rB'E 1     0 1 1 / rB'E  1 / pCB'C .} pCB'C rB'E pCB'C pCB'C ß / rB'E .01 Franz Sischka .{ .pCB'C 1 / rB'E + pCB'C = GummelPoon Toolkit B9_APPEN..
01 Franz Sischka .WPS  17.GP: Appendices Finally resubstituting 1 / rB'E = 1 / rB'E + pCB'E 21 yields h21 = ß /rB'E pCB'C 1 / rB'E + p(CB'C + CB'E) or h21 = ß pCB'C rB'E 1 + p(CB'C + CB'E) rB'E (5) what is depicted below: log  h21  20dB/decade 40dB/decade log (2PI*freq) 1 (CB'C'+CB'E')*rB'E' ß CB'C' * rB'E' Now we are ready to calculate the transit frequency fT: from (5) follows for h21 = 1 : = ß2 4 PI2 fT2 C2B'C r2B'E 1 + 4 PI2 fT2 (CB'C + CB'E)2 r2B'E GummelPoon Toolkit B9_APPEN.04.
GP: Appendices or: f T = 22 gB'E 2PI 2 (CB' C + CB' E )2 − CB 'C ß2 − 1 (6) (7) (8) with: gB'E' CB'C and: CB'E ~ TFF NF VT iC (9) ~ = diB' dvB'E CSBC(vBC) ~ iB' NF VT ___________________________________________________________ SIMPLIFICATION: In order to keep things a little simpler for parameter extraction.(5) is modified a bit.iC NF VT iB' GummelPoon Toolkit B9_APPEN. neglecting the zero (at high frequencies) against the pole(low frequencies): ß 1 + p(CB'C + CB'E) rB'E h21 ~ (10) Calculating again the transit frequency for this simplified h21 yields: ß2 = 1 + 4 PI2 f2T1pole (CB'C + CB'E)2 r2B'E 1 2PI ß 2 −1 f T1 − pole = (CB' C + CB' E )2 2 r B' E CB'E > CB'C ~ ß > 1 (9) ~ (7) ß 2 PI CB'E rB'E ß TFF NF VT 2 PI .04.01 Franz Sischka . equ.WPS  17.
04. we get the pretty simple form fT1pole ~ 1 2 PI TFF 23 (11) Or solved for TFF: 1 2 PI fT1pole TFF = (12) what is the wellknown formula for the TFF parameter extraction.01 Franz Sischka .GP: Appendices since ß = iC / iB'.WPS  17. GummelPoon Toolkit B9_APPEN.
March 1994.827 Simplification of DC characterization and Analysis of Semiconductor Devices.7. ISBN 0471874930 I. Tektronix Publication No. no. Bell Syst.Zimmer: Contribution à la modélisation des transistors haute fréquence. p. 1977 D. 1976 H.Techn. New York. Vector Measurements of High Frequency Networks.Jespers: Process and Device Modeling for Integrated Circuit Design.Getreu: Modeling the Bipolar Transistor. H. 17. Cazenave.01 Franz Sischka .10. 062284100. HewlettPackard. McGrawHill.G. Meresse. Xtract Supplement Manual. NATO Advanced Study Institute. P. Thèse à l'Université de Bordeaux I. vol.11.Letters.Jackson: Analysis and Design of Digital Integrated Circuits. Dom..27.J.21. G. Application Note AN 12024: Advanced Bipolar Transistor Modeling Techniques.Hodges. Zimmer.Engl.Massobrio: Semiconductor Device Modeling with SPICE.R. IC_CAP marketing group T.C. John Wiley&Sons. 10. HP application note 3831 EEsof. 1988.A. Publication HP59580387. 49 [1970].G. Belgium.91. McGrawHill. LouvainlaNeuve.1895 Paul Schmitz. Electron.1992.Meyer: Analysis and Design of Analog Integrated Circuits. ISBN 0070021074 F.Poon: An Integral Charge Control Model of Bipolar Transistors.van der Wiele. New York. 1984. p.WPS  30. April 1989 GummelPoon Toolkit B9_PUBL. 'Simple Determination of BJT Extrinsic Base Resistance'. R. H.K. W.Antognetti.L. 1983. HPEEsof.G.Gray. ISBN 0070291535 P.Gummel.11 GP: Publications 1 P U B L I C A T I O N S P.
Germany TF modeling: Kendall et al.Ning. Großsignalmodellierung von SiGeHeterobipolartransistoren für den Entwurf von Millimeterwellenschaltungen. Düsseldorf. PhD Thesis University of Delft. Oregon F. October 1011. December 2001 Paras.11. pp.. vol. Halbleiter Elektronik Frankfurt/Oder GmbH. GummelPoon Toolkit B9_PUBL. published in the proceedings of the European ICCAP User's Group Meeting 1994.Berkner: A Survey of CDMethods for Determining the Series Resistances of Bipolar Transistors Including the new ∆ISubMethod.ICCAP User Meeting in The Haague. publshed 1995 by Cascade Microtech. Inc.MacSweeny et.WPS  30. Proceedings of the 1996 Europ. Colmar.: Modeling and Parameter Extraction for Lateral Bipolar Devices Using ICCAP SubCircuits. D. France DC and paras.resistor modeling: D.resistor modeling: J.al. Dissertation (PhD Thesis) Universität München. Netherlands T. Contribution à la modélisation et à la caractérisation en hautes fréquences des transistors bipolaires a heterojunction Si/SiGe. published in the proceedings of the European ICCAP User's Group Meeting 1993. Frankfurt/Oder. Esslingen. SMI GmbH. IEEE Transactions on Electron Devices.409412. ISBN 3183327090 B.Tang: Method for Determining the Emitter and Base Series Resistance of Bipolar Transistors. VDIVerlag. April 1984 CV and TF modeling: J.: Direct Extraction of Bipolar SPICE Transit Time Parameters Without Optimization.11 GP: Publications 2P. PhD Thesis University Bordeaux. June 2223.X.01 Franz Sischka . École doctorale de sciences physiques et de l'ingenieur. published in the proceedings of the US ICCAP User's Group Meeting 1993.ED31.Ardouin. van Wijnen.Sinnesbichler.Berkner: Parasitäre Effekte bei der SPICEModellparameterbestimmung für integrierte Bipolartransistoren. Beaverton. 1992. On the Characterization and Optimization of HighSpeed Silicon Bipolar Transistors.
Germany Mr. Ireland.WPS  30.Dermot MacSweeny from the NMRC in Cork. the author would like to especially acknowledge Mr. Infineon. GummelPoon Toolkit B9_PUBL. Munich.01 Franz Sischka .11 GP: Publications 3 For providing valuable feedback on the extraction methods presented in this document.11. Jörg Berkner.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.