MOSFET Small-Signal Model ■ Concept: Þnd an equivalent circuit which interrelates the incremental changes in iD, vGS, vDS, etc. Since the changes are small, the small-signal equivalent circuit has linear elements only (e.g., capacitors, resistors, controlled sources) ■ Derivation: consider for example the relationship of the increment in drain current due to an increment in gate-source voltage when the MOSFET is saturated-- with all other voltages held constant. vGS = VGS + vgs , iD = ID + id -- we want to find id = (?) vgs We have the functional dependence of the total drain current in saturation: iD = µn Cox (W/2L) (vGS - VTn )2 (1 + λnvDS) = iD(vGS, vDS, vBS) Do a Taylor expansion around the DC operating point (also called the quiescent point or Q point) defined by the DC voltages Q(VGS, VDS, VBS): 2 ∂i D 1∂ iD ( v gs ) + --iD = I D + ∂ v GS 2 ∂ v2 Q GS 2 ( v gs ) + … Q If the small-signal voltage is really Òsmall,Ó then we can neglect all everything past the linear term -∂i D iD = I D + ( v gs ) = I D + g m v gs ∂ v GS Q where the partial derivative is deÞned as the transconductance, gm. EE 105 Spring 1997 Lecture 12 Transconductance The small-signal drain current due to vgs is therefore given by id = gm vgs. iD = ID + id D G + + B vgs _ + _ VDS = 4 V S VGS = 3 V_ 600 500 iD id 400 vGS = VGS + vgs Q (µA) 300 vGS = VGS = 3 V gm = id / vgs 200 100 1 2 3 4 5 6 VDS (V) EE 105 Spring 1997 Lecture 12 Another View of gm * Plot the drain current as a function of the gate-source voltage, so that the slope can be identiÞed with the transconductance: D iD = ID + id G + + B vgs _ + iD(vGS, VDS = 4 V) 600 iD 400 VDS = 4 V S VGS = 3 V_ 500 _ id Q gm = id / vgs (µA) 300 200 100 3 2 1 vGS = VGS = 3 V 6 vGS (V) 4 5 vGS = VGS + vgs EE 105 Spring 1997 Lecture 12 Transconductance (cont.) ■ Evaluating the partial derivative: W g m = µ n C ox ----- ( V GS Ð V Tn ) ( 1 + λ n V DS ) L Note that the transconductance is a function of the operating point, through its dependence on VGS and VDS -- and also the dependence of the threshold voltage on the backgate bias VBS. ■ In order to Þnd a simple expression that highlights the dependence of gm on the DC drain current, we neglect the (usually) small error in writing: gm = 2I D W 2µ n C ox ----- I D = ------------------------- L V GS Ð V Tn For typical values (W/L) = 10, ID = 100 µA, and µnCox = 50 µAV-2 we Þnd that gm = 320 µAV-1 = 0.32 mS ■ How do we make a circuit which expresses id = gm vgs ? Since the current is not across the controlling voltage, we need a voltage-controlled current source: id gate + vgs drain gmvgs _ source _ EE 105 Spring 1997 Lecture 12 Output Conductance ■ We can also Þnd the change in drain current due to an increment in the drainsource voltage: ∂i D 2 W g o = ------------ = µ n C ox ------ ( V GS Ð V Tn ) λ n ≅ λ n I D 2L ∂v DS Q The output resistance is the inverse of the output conductance 1 1 r o = ----- = -----------go λn I D The small-signal circuit model with ro added looks like: id = gm vgs + (1/ro)vds gate drain + vgs _ source gmvgs id ro + vds _ EE 105 Spring 1997 Lecture 12 Backgate Transconductance ■ We can Þnd the small-signal drain current due to a change in the backgate bias by the same technique. The chain rule comes in handy to make use of our previous result for gm: ∂i D g mb = -----------∂v BS ∂V Tn g mb = ( Ð g m ) -----------∂v BS Q Q ∂i D = -----------∂V Tn Q ∂V Tn -----------∂v BS Q Ðγ n γ n gm = ( Ð g m ) ------------------------------------ = ------------------------------------ . 2 Ð 2 φ p Ð V BS 2 Ð 2 φ p Ð V BS The ratio of the Òfront-gateÓ transconductance gm to the backgate transconductance gmb is: 2qε s N qε s N a C b(y=0) g mb 1 a --------- = ---------------------------------------------- = --------- ------------------------------------- = -------------------C ox 2 ( Ð 2 φ p Ð V BS ) gm C ox 2C ox Ð 2 φ p Ð V BS where Cb(y=0) is the depletion capacitance at the source end of the channel -gate source channel Cb(0) depletion region bulk EE 105 Spring 1997 Lecture 12 ,, MOSFET Capacitances in Saturation fringe electric field lines gate drain ,,, , ,, ,,, source n+ n+ Csb qN (vGS) overlap LD overlap LD Cdb depletion region In saturation, the gate-source capacitance contains two terms, one due to the channel chargeÕs dependence on vGS [(2/3)WLCox] and one due to the overlap of gate and source (WCov, where Cov is the overlap capacitance in fF per µm of gate width) 2 C gs = --- WLC ox + W C ov 3 In addition, there are depletion capacitances between the drain and bulk (Cdb) and between source and bulk (Csb). Finally, the extension of the gate over the Þeld oxide leads to a small gate-bulk capacitance Cgb. EE 105 Spring 1997 Lecture 12 Complete Small-Signal Model ■ The capacitances are ÒpatchedÓ onto the small-signal circuit schematic containing gm, gmb, and ro Cgd gate id + drain vgs Cgs gmvgs Cgb gmbvbs ro _ source _ Csb vbs Cdb + bulk ■ p-channel MOSFET small-signal model the source is the highest potential and is located at the top of the schematic + source gmbvsb Cgd _ gate vsb gmvsg Cgs vsg ro −id drain Cgb _ bulk Csb Cdb EE 105 Spring 1997 Lecture 12 Circuit Simulation ■ Objectives: ¥ fabricating an IC costs $1000 ... $100,000 per run ---> nice to get it ÒrightÓ the Þrst time ¥ check results from hand-analysis (e.g. validity of assumptions) ¥ evaluate functionality, speed, accuracy, ... of large circuit blocks or entire chips ■ Simulators: ¥ SPICE: invented at UC Berkeley circa 1970-1975 commercial versions: HSPICE, PSPICE, I-SPICE, ... (same core as Berkeley SPICE, but add functionality, improved user interface, ...) EE 105: student version of PSPICE on PC, limited to 10 transistors ¥ other simulators for higher speed, special needs (e.g. SPLICE, ■ RSIM) Limitations: ¥ simulation results provide no insight (e.g. how to increase speed of circuit) ¥ results sometimes wrong (errors in input, effect not modeled in SPICE) ===> always do hand-analysis Þrst and COMPARE RESULTS EE 105 Spring 1997 Lecture 12 MOSFET Geometry in SPICE ■ Statement for MOSFET ... D,G,S,B are node numbers for drain, gate, source, and bulk terminals ,, , Mname D G S B MODname L= _ W=_ AD= _ AS=_ PD=_ PS=_ MODname speciÞes the model name for the MOSFET NRS = N (source) PS = 2 × Ldiff (source) = W NRD = N (drain) PD = 2 × Ldiff (drain) = W L Ldiff (source) Ldiff (drain) W AS = W × Ldiff (source) AD = W × Ldiff (drain) EE 105 Spring 1997 Lecture 12 MOSFET Model Statement .MODEL MODname NMOS/PMOS VTO=_ KP=_ GAMMA=_ PHI=_ LAMBDA=_ RD=_ RS=_ RSH=_ CBD=_ CBS=_CJ=_ MJ=_ CJSW=_ MJSW=_ PB=_ IS= _ CGDO=_ CGSO=_ CGBO=_ TOX=_ LD=_ Parameter name (SPICE / this text) SPICE symbol Eqs. (4.93), (4.94) Analytical symbol Eqs. (4.59), (4.60) Units channel length Leff L m polysilicon gate length L Lgate m lateral diffusion/ gate-source overlap LD LD m transconductance parameter KP µnCox A/V2 threshold voltage / zero-bias threshold VTO VTnO V channel-length modulation parameter LAMBDA λn V-1 bulk threshold / backgate effect parameter GAMMA γn V1/2 surface potential / depletion drop in inversion PHI - φp V DC Drain Current Equations: ( V GS ≤ Ð V TH ) I DS = 0 KP I DS = -------- ( W ⁄ L eff )V DS [ 2 ( V GS Ð V TH ) Ð V DS ] ( 1 + LAMBDA ⋅ V DS ) 2 2 KP I DS = -------- ( W ⁄ L eff ) ( V GS Ð V TH ) ( 1 + LAMBDA ⋅ V DS ) 2 ( 0 ≤ V DS ≤ V GS Ð V TH ) ( 0 ≤ V GS Ð V TH ≤ V DS ) V TH = V TO + GAMMA ( 2 ⋅ PHI Ð V BS Ð 2 ⋅ PHI ) EE 105 Spring 1997 Lecture 12 Capacitances SPICE includes the ÒsidewallÓ capacitance due to the perimeter of the source and drain junctions -n+ drain (area) (perimeter) CJ ⋅ AD CJSW ⋅ PD C BD(V BD) = -------------------------------------------- + ---------------------------------------------------MJ MJSW ( 1 Ð V BD ⁄ PB ) ( 1 Ð V BD ⁄ PB ) Gate-source and gate-bulk overlap capacitance are speciÞed by CGDO and CGSO (units: F/m). Level 1 MOSFET model: .MODEL MODN NMOS LEVEL=1 VTO=1 KP=50U LAMBDA=.033 GAMMA=.6 + PHI=0.8 TOX=1.5E-10 CGDO=5E-10 CGSO= 5e-10 CJ=1E-4 CJSW=5E-10 + MJ=0.5 PB=0.95 The Level 1 model is adequate for channel lengths longer than about 1.5 µm For sub-µm MOSFETs, BSIM = ÒBerkeley Short-Channel IGFET ModelÓ is the industry-standard SPICE model. EE 105 Spring 1997 Lecture 12