<<

Analysis and Design of Analog Integrated Circuits Lecture 8

Cascode Techniques

Michael H. Perrott February 15, 2012

Copyright © 2012 by Michael H. Perrott All rights reserved.

M.H. Perrott Review of Large Signal Analysis of Current Mirrors

Vdd Δ V2 I I 1 2 1 μ W2 2 λ nCox (VGS2-VTH) (1+ 2Vds2) I2 2 L = 2 I 1 W 2 1 μ C 1(V -V ) (1+λ V ) 2 n ox GS1 TH 1 ds1 M L1 2 ΔV M1 Vds2 > Vdsat2 1 V +ΔV V +ΔV Δ Δ Δ Δ Vss=0 TH 1 TH 2 But, VTH+ V1=VTH+ V2 V1 = V2 λ I2 W2 L1 (1+ 2Vds2) I2 = λ I1 W1 L2 (1+ 1Vds1) M2 in Saturation Current Mismatch setting due to Vds based on difference M2 in geometry

Note: for accurate ratio, set L1 = L2

Vds2 Vdsat2

M.H. Perrott 2 The Issue of Vds Mismatch in Current Mirrors

V λ dd I2 W2 (1+ 2Vds2) = λ I1 W1 (1+ 1Vds1) I1 I2 Current Mismatch setting due to Vds based on difference geometry V V ds1 ds2 Note: we are assuming L = L M1 M2 1 2

. Issue: Current I2 can vary significantly as a function of the drain voltage of M2

- We often want a tightly controlled current set only by I1 and sizes . How do we improve the matching performance?

M.H. Perrott 3 Cascoded

I Rth R ref d3 thd3

Ibias Vbias V M bias 3 M3

M ro1 2 M1

. Offers increased output resistance - Reduces small signal dependence of output current on the output voltage of the current source - From Lecture 6, we derived:

. Output resistance boosted by intrinsic gain of M3, gm3ro3

. But how do we reduce the influence of large signal Vds mismatch between M1 and M2?

M.H. Perrott 4 Match Vds of Current Mirror Devices With Proper Bias

Vdd

I1

I2 λ I2 W1 L4 (1+ 1Vds1) Recall: = V I1 W L (1+λ V ) M3 M2 o 4 1 4 ds4 V > ΔV W/L W/L ds2 Current Mismatch Δ Δ VTH+ V VTH+ V setting due to Vds based on difference

M4 M1 geometry Δ W/L W/L Vds1 = VTH+ V V +ΔV V +ΔV Vss=0 TH TH

. Key transistor for determining I2 is M1

- Why is M2 less important? . Above biasing approach provides a much closer match between Vds1 and Vds4 W1 1+λVds1 W1 I2 = I1 I1 W4 1+λVds4 ≈ W4 M.H. Perrott 5 The Drawback of Basic Cascode Bias Approach

Vdd

I2 I 1 M1 and M2 M1 in saturation I2 in saturation M2 in triode V M3 M2 o Δ W/L W/L Vds2 > V M1 and M2 Δ Δ in triode VTH+ V VTH+ V

M M V 4 1 V V +2ΔV o Δ 1 TH W/L W/L Vds1 = VTH+ V V +ΔV Δ calculation of V1 is nontrivial Vss=0 TH VTH+ V

. Output voltage range is reduced

- Now Vo must be > VTH + 2V - What will happen to the of the current source if the output voltage is too low? - Can we improve the voltage range? M.H. Perrott 6 Improved Swing Cascode

I2 M1 and M2 in saturation

Vdd

M1 and M2 in triode no wasted voltage region I1

Δ 2VTH+3 V Vo I2 Vdsat1+Vdsat2 M3 M5 αW/L W/L Vo V +2ΔV M2 V +2ΔV V +ΔV TH Δ TH TH W/L Vds2 > V Δ VTH+ V

M4 M6 M1 W/L W/L V = ΔV W/L ds1 Δ Δ Δ Vss=0 VTH+ V VTH+ V VTH+ V

. Key idea: set size of M3 such that Vds1 = V

- Assuming strong inversion for M1 and M3:

M.H. Perrott 7 Alternative Implementation of Improved Swing Cascode

I2 M1 and M2 in saturation

Vdd M5 M6 M7 M1 and M2 no wasted voltage region Wp/Lp in triode Wp/Lp Wp/Lp I1 I1 I2 V 2ΔV o Vo M3 M2 Δ αW/L W/L Vds2 > V I1 Δ Δ VTH+2 V VTH+ V

M4 M1 Δ W/L W/L Vds1 = V V +ΔV Δ Vss=0 TH VTH+ V

. Set  as on previous slide . Note: both implementations share a common problem

M.H. Perrott 8 The Issue of Current Mismatch

I1 I2

V +2ΔV M2 λ TH I2 W2 (1+ 2Vds2) Recall: = I W (1+λ V ) Δ 1 1 1 ds1 VTH+ V Mismatch M4 M1 due to Vds Δ Δ Vds4 = VTH+ V Vds1 = V difference W/L W/L

. The improved swing approach causes a systematic mismatch between I2 and I1

- Key issue: Vds1  Vds4

. Can we fix this problem?

M.H. Perrott 9 Techniques to Reduce Current Mismatch

I1 I2

M3 V +2ΔV M2 W/L TH W/L Δ Δ VTH+ V VTH+ V

M4 M1 Δ Δ Vds4 = V Vds1 = V W/L W/L

. Systematic mismatch between I1 and I2 is greatly reduced by using the above circuit (now Vds1 ≈ Vds4)

- Note that gate bias on M2 and M3 may be provided by previously discussed circuits

. Additional techniques for accurately matching I1 and I2

- Set L1 = L4 >> Lmin

. Note: set L2 = L3 ≈ Lmin for lower area and capacitance

- Set W2/W3 = I2/I1 so that V2 = V3 M.H. Perrott 10 Another Common Cascode Bias Topology

Vdd M5 M6 M7 Wp/Lp Wp/Lp Wp/Lp I1 I1

M8 W/L

M9 Δ VTH+2 V W/L

M10 I2 W/L M 11 M3 M2 W/L W/L W/L I1 M Δ Δ 12 VTH+ V VTH+ V W/L

M13 M4 M1 Δ Δ W/L Vds4 = V Vds1 = V W/L W/L . Key issue: needs two bias current branches M.H. Perrott 11 Utilizing a Simple Resistor to Achieve One Bias Branch

M5 M7 Wp/Lp Wp/Lp I1

Δ VTH+2 V

Δ V RB I2

M3 M2 W/L W/L I1 Δ Δ VTH+ V VTH+ V

M4 M1 Δ Δ Vds4 = V Vds1 = V W/L W/L . Issue: poly resistor is large and won’t track NMOS devices across temperature and process variations M.H. Perrott 12 Better Approach: Use PMOS Device In Triode Region

M5 M7 Wp/Lp Wp/Lp I1

Δ VTH+2 V

M6 Δ Wp/Lp V I2

M3 M2 W/L W/L I1 Δ Δ VTH+ V VTH+ V

M4 M1 Δ Δ Vds4 = V Vds1 = V W/L W/L . Much smaller, better tracking with NMOS devices than resistor M.H. Perrott 13 Wilson Current Mirror

I2 R I1 thd2

M2

M3 M1

. Relies on feedback in its operation . Using Hybrid- analysis

- Output resistance comparable to cascode current source . This circuit is rarely used these days

M.H. Perrott 14 Enhanced Cascode Current Source

I I bias bias2 Iref

M4

M3

M 2 M1

. Offers output resistance comparable to double cascode current source . As with Wilson mirror, analysis is tricky due to source/gate coupling - Using results shown in the following slide:

M.H. Perrott 15 Thevenin Resistances for CMOS Transistor Feedback Pair

R R R RC A C A R M D thd D Rthd 4

M4

M3 vgs4 gm4vgs4 -gmb4vs4 ro4 S M3 Rths RB S ro3 -gmb3vs3 gm3vgs3 vgs3 Rths

vs4 RB

vs3=0

M.H. Perrott 16 Basic Cascode

RD V out M2 M2 R ths2 is2 d2 RG M1 α i v Vin 2 s2 Rth RD out RS d2

s2 M1 R RG g1 ths1 is1 d1

vin Rth v Av1vg1 α i R g1 g1 1 s1 thd1

s1

General Model RS

. Allows improved frequency response (discussed later) . Reduction to two-port will be done in several steps M.H. Perrott 17 Eliminate Middle Sections

RD V out M2 M2 R ths2 is2 d2 RG M1 α i v Vin 2 s2 Rth RD out RS d2

s2 M1

RG g1 d1

vin Rth v G v R g1 g1 m1 g1 thd1

. Calculation of Gm1 same as for amp . To reduce further, note that

M.H. Perrott 18 Resulting Two-Port Similar to Common Source Amp

RD V out M2 M2 d2 RG M1 G v v Vin m1 g1 Rth RD out RS d2

M1

RG g1

v R in thg1 vg1

. Key difference: drain impedance much larger

M.H. Perrott 19 Slight Twist to Cascode Amplifier

Vdd RL Vout RL

Vout

Ibias 1 is1 ro1 gm1+gmb1 Vbias Iin is1 M1

i Vin in ro4 ro2 M4 M2 M3 Vss=0

. What is the difference between this amplifier and basic cascode amplifier?

. What are the constraints in setting Vbias? . What is the maximum output voltage swing?

M.H. Perrott 20 Constraints on Vbias and Output Range

Vdd

RL

Vout

Ibias

M1 Δ > V1 Vbias Iin Δ VTH+ V1

Δ Δ Vin > V4 > V2 M4 M2 M3 Vss=0

. To keep M2 and M4 in saturation

. To keep M1 in saturation

M.H. Perrott 21 Calculation of Maximum Output Range

Vdd

RL

Vout

Ibias

M1 Δ > V1 Vbias Iin Δ VTH+ V1

Δ Δ Vin > V4 > V2 M4 M2 M3 Vss=0

. Minimum Vbias allows the maximum output range

. Resulting output range

M.H. Perrott 22 Variation on a Theme: Enhanced Cascode

Ibias1 Ibias2 R1

Vout

M4 Input Source M3

I M in Rs 2 M1

. We can turn the enhanced cascode current source into an amplifier

- Inject a current input at the source of M4 . Key aspects of small signal analysis can be done using Thevenin method - Simply leverage Thevenin resistance formulas shown on Slide 16

M.H. Perrott 23 Small-Signal Analysis of Enhanced Cascode Amp

Rout Ibias1 Ibias2 R1 R1

Vout Vout

M4 M4 Input Source Input Source M3 M3

1 Iin Rs Rth Rin Iin Rs M2 M d1 1 gm2

. From Thevenin resistance calculations on Slide 16: - is quite low

- Output impedance is probably determined by R1

. This amplifier is useful for extracting a current signal while keeping the source voltage nearly constant M.H. Perrott 24