Lect. 10: Small signal model (4.

6)
Decompose all signals into Large Signals and Small Signals vGS= VGS + vgs iD = ID + id vD = VD + vd

Large signals provide biases. Small signals are related to amplification ( vd = Av vgs). What is Av=vd/vgs?

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

What is Av (in saturation)?
1 W vD = VDD iD RD = VDD n Cox (vGS VT ) 2 RD 2 L With vGS = VGS + vgs , what is vD = VD + vd ?
2 1 W vD = VDD nCox (VGS + vgs ) VT RD 2 L 1 W dv VDD nCox (VGS VT ) 2 RD + D |VGS vgs 2 L dvGS

(VD or vD with vgs = 0)

(vd or vD due to vGS ) f (X ) + df |x = X a) dx

(Taylor expansion to the first order: f ( x + a ) |x = X vd = Av = dvD W |VGS vgs = nCox (VGS VT ) RD vgs dvGS L

vD = VD+vd

vd W = nCox (VGS VT ) RD v gs L

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Circuit model for small signals only (Linear circuit model).

id vd

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Small signal model for NMOS vGS=VGS + vgs, vDS=VDS + vds iG = IG + ig, iD=ID + id ig, id as functions of vgs, dvs
ig = 0 From iD = 1 W nCox (vGS VT ) 2 with vGS = VGS + vgs 2 L
2

di W 1 nCox (VGS VT ) + D |VGS vgs iD 2 L dvGS W id = nCox (VGS VT ) vgs = g m vgs L (g m : transconductance)

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

id vd gmvgs

Small signal circuit: Linear!

vd = - g m vgs RD = n Cox W (VGS VT )vgs RD L

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Different expressions for gm

From iD = gm = diD dvGS

1 W nCox (vGS VT ) 2 2 L W |VGS = nCox (VGS VT ) L

W = 2 nCox I D ; L 2I D = VGS VT

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

How accurate is the small signal model?
iD = 1 W nCox (vGS VT ) 2 2 L With vGS = VGS + vgs di W 1 nCox (VGS VT ) + D |VGS vgs 2 L dvGS
2 2

iD

W W 1 = nCox (VGS VT ) + nCox (VGS VT ) vgs L L 2 But iD =

2 2 W W 1 1 nCox (vGS + vgs VT ) = nCox (vGS VT ) + vgs L L 2 2 2

W W W 1 1 = nCox (VGS VT ) + nCox (VGS VT ) vgs + nCox vgs 2 (Nonlinear Effect) L L L 2 2 Small signal analysis is valid only if W W 1 nCox vgs 2 << nCox (VGS VT ) vgs or vgs << 2(VGS VT ) L L 2

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Small signal model for NMOS with channel-length modulation NMOS I-V with =0 to 0.1 by 0.01 increment. vGS=5.0V

iD =

1 W k ' (1 + vDS )(vGS Vt ) 2 2 L

vGS=2.5V

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Small signal model for NMOS with channel-length modulation

iD =

1 W k ' (1 + vDS )(vGS Vt ) 2 2 L i i iD = D vGS + D vDS vGS vDS 2iD iD W = k ' (1 + vDS )(vGS Vt ) = = gm L vGS Vt vGS iD 1 W = k ' (vGS Vt ) 2 vDS 2 L

iD , =0 =

1 r0

(r0 : output resistance, VA = r0 I D ) id = g m vgs + vds r0

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Ex. 4.10 Vt=1.5V, k (W/L)=0.25mA/V2 1. Determine bias conditions. 2. Derive small signal circuit model. 3. Voltage gain? 4. Input resistance? 5. Max. vi?

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

1. Determine bias conditions. Vt=1.5V, k (W/L)=0.25mA/V2 ID=0.5X0.25 (VD-1.5)2 VD=15-RDID ID=1.06mA, VD=4.4V gm= k (W/L) (VGS-Vt) = 0.725mA/V

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Voltage Gain Small-Signal model

vo Av =

g m vi ( RD RL ) vo = g m ( RD RL ) = 3.625 vi

What if VA is finite? (VA = 50V ) r0 =

RD=10K, RL=10K, RG=10M

VA = 47 k ID

Av = g m ( RD RL r0 ) = 3.3

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

RD=10K, RL=10K, RG=10M Input Resistance

Rin = ii =

vi ii

vi vo v v = i (1 o ) RG RG vi vi R = G = 2.33 M ii 1 Av

Rin =

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

For Max. vi, vDS>vGS-Vt (saturation)

VDS AV vi ,max = VGS + vi ,max Vt vi ,max = 0.34V

How good is the small signal approximation for above condition?
From iD =
2 1 W nCox (vGS + vgs VT ) 2 L 2

W W 1 W 1 = nCox (VGS VT ) + nCox (VGS VT ) vgs + n Cox vgs 2 L L 2 L 2 1 W 1 1 vgs x0.34 nCox vgs 2 2nd-order term 2 2 2 L = 0.059 = = = first-order term C W (V V ) v (VGS VT ) 4.4-1.5 n ox GS T gs L

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory

Lect. 10: Small signal model

Various small signal models

( model)

(T model)

Electronic Circuits 1

High-Speed Circuits and Systems Laboratory