HO13_315aSP09_single_stage_OTA_1

# HO13_315aSP09_single_stage_OTA_1 - Single Stage OTAs Part 1...

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Single Stage OTAs Part 1 Boris Murmann Stanford University [email protected] Copyright © 2009 by Boris Murmann EE315A ― HO #13 B. Murmann 1 Basic Differential Pair OTA • Common mode feedback • Half circuit model • Return ratio analysis – Loop gain – Closed-loop gain • Noise N i analysis l i • Step response – Linear settling g – Slewing B. Murmann EE315A ― HO #13 2 Basic Differential Pair OTA VDD MP1a MP1b MPB Vom Vop - Vod + Vip Vim MN1a IT/2 MN1b IT • Suppose th t in the operating point Vip=Vim, i Vid=0 S that i th ti i t V i.e. 0 • What is the output common mode voltage Voc = (Vom+Vop)/2 ? EE315A ― HO #13 B. Murmann 3 Operating Point Sensitivity T D Operating Point 1 Operating Point 2 OC DD • The operating point is very sensitive to small changes in the device characteristics • Solution: Common mode feedback (CMFB) B. Murmann EE315A ― HO #13 4 CMFB • Common mode feedback loop adjusts ΔI such that VOC is very close to the desired voltage EE315A ― HO #13 B. Murmann 5 Idealized CMFB Implementation ΔI Voc = Voc,des + B. Murmann ΔI GCMFB = Voc,des EE315A ― HO #13 for GCMFB → ∞ 6 Ideal Balun xfmr xfmr • Useful for separating common mode and differential mode signal components p • Bi-directional, preserves port impedances • Uses ideal, inductorless transformers that work down to DC B. Murmann EE315A ― HO #13 7 CMFB Implementation • In practice, we won’t be able to let GCMFB p , ∞ for loop stability p y – Nonetheless, the loop will get us to within a few mV of where we need to be – And most importantly help absorb variations in the device characteristics • In the first few lectures on OTA design, we will use the idealized common mode f db k circuit ( shown previously) t avoid d feedback i it (as h i l ) to id distraction from the main design task • Practical CMFB implementation examples (using transistors) will follow later in this course B. Murmann EE315A ― HO #13 8 Differential Mode Small Signal Half Circuit • With the circuit at the proper operating point, we can analyze its small-signal behavior using a differential mode half circuit model • Note that (to first order) the CMFB loop does not influence the behavior of the differential mode signals B. Murmann EE315A ― HO #13 9 OTA with Capacitive Feedback • Let’s get started by placing our simple OTA into a capacitive feedback loop ( encountered e.g. i an SC circuit) f db k l (as t d in i it) • Questions – What is the phase margin? p g – What is the closed loop transfer function? – What is the total integrated noise? – How fast does this circuit settle (in response to a step)? B. Murmann EE315A ― HO #13 10 Half Circuit Model Gm = g mn Cftot Ro = rop rop Co = Cdbp + Cdbn Cx = Cgsn + Cgbn EE315A ― HO #13 B. Murmann 11 Return Ratio Analysis Cftot v x = β ⋅ vo CLtot ir β= Cftot Cftot + Cs + Cx vo vx Cs Cx Gmvx it Ro CL+Co C “Feedback factor” ⎛ 1 v 0 = −it ⋅ ⎜ R0 ⎜ sCLtot ⎝ T (s ) = − B. Murmann ⎞ ⎟ ⎟ ⎠ CLtot = CL + Co + (1 − β ) Cftot ⎛ ir 1 = β ⋅ Gm ⋅ ⎜ R0 ⎜ it sCLtot ⎝ ⎞ β ⋅ Gm R0 β ⋅ a0 = ⎟= ⎟ ⎠ 1 + sRoCLtot 1 + sRoCLtot EE315A ― HO #13 12 Frequency Response of T(s) T ( jω) T 0 = βGmR0 ωpo = Ro→∞ T0 1 R0CLtot ωpo T (s ) = βGmR0 βGm βGm = ≅ 1 1 + sRoCLtot + sCLtot sCLtot Ro βGm =1 j ωcCLtot ⇒ ωc ≅ β Gm CLtot for ωc Ro >> ω ω 1 ⇔ >> 1 sCLtot ωpo Ro is irrelevant for understanding high frequency behavior around ωc EE315A ― HO #13 B. Murmann 13 Phase Margin T ( jω) Ro→∞ T ( j ω) = T0 ωpo ωc ω ⎡T ( j ω) ⎤ ⎣ ⎦ 0° ωpo ω −45° −90° B. Murmann βGmR0 ω 1+ j ωpo T ( j ωc ) = βGm R0 ω 1+ j c ωpo ⎛ ω ⎡T ( j ω) ⎦ ⎤ = − tan −1 ⎜ c ⎣ ⎜ ωpo ω=ωc ⎝ ⎞ ⎟ ≅ −90° ⎟ ⎠ PM ≅ 180° − 90° ≅ 90° EE315A ― HO #13 14 Closed Loop Transfer Function A (s ) = • vo T(s ) d = A∞ + vi 1+ T ( s ) 1+ T ( s ) Need to find A∞ and d – Let’s start with A∞ Let s Gm → ∞ Cftot Cs vi vx vo ix Cx Ro CL+Co Gmvx ⇒ v xm → 0 ⇒ i x → 0 0 = v i sCs + v o sCftot A∞ = vo vi =− Gm →∞ EE315A ― HO #13 B. Murmann Cs Cftot 15 Finding d at Low Frequencies • Capacitors are open circuits d0 = B. Murmann vo vi =0 Gm = 0 EE315A ― HO #13 16 Low-Frequency Closed-Loop Gain A0 = A∞ T0 d + 0 1 + T0 1 + T0 A∞ = − ⇒ A0 = − • Cs Cftot T0 = βGmRo d0 = 0 Cs 1 1 Cftot 1 + βGm Ro Error in low-frequency closed-loop g q y p gain ε0 = A0 − A∞ A∞ ε0 = ⎛ A0 T 1 1⎞ 1 −1= 0 −1= − 1 ≅ ⎜1− ⎟ − 1 = − 1 A∞ 1 + T0 T0 ⎝ T0 ⎠ 1+ T0 ε0 ≅ 1 T0 EE315A ― HO #13 B. Murmann 17 d at High Frequencies ieq = v i Cs ⋅ sCftot = v i β ⋅ sCs Cs + Cx + Cftot d= B. Murmann vo vi = Gm = 0 Ceq = (1 − β ) Cftot ieq C 1 =β s v i s Ceq + CL + Co CLtot ( EE315A ― HO #13 ) 18 High-Frequency Closed-Loop Gain (1) A( s ) ≅ A∞ ( ≅− • T(s ) d + 1+ T ( s ) 1+ T ( s ) Cs Cftot βCs βGm C s 1 − s ftot sCLtot CLtot Cs Cs 1 − z Gm + =− =− βGm βGm Cftot 1 + s CLtot Cftot 1 − s 1+ 1+ p sCLtot sCLtot βGm Pole frequency: ωp ≅ βGmRo βGm ≅ ωc ≅ ≅ T0 ⋅ ωpo CLtot RoCLtot As expected. EE315A ― HO #13 B. Murmann 19 High-Frequency Closed-Loop Gain (2) Cftot s Cs Cs 1 − z Gm =− A( s ) = − Cf 1 + s CLtot Cf 1 − s p βGm 1− s • Zero frequency: ωz = Gm Cftot ωz CLtot = ωp βCftot usually >> 1 • Therefore, the closed-loop -3dB frequency is approximately ω−3dB ≅ ωp ≅ B. Murmann βGm CLtot EE315A ― HO #13 20 Putting it All Together s Cs 1 1 − z A( s ) ≅ − Cf 1 + 1 1 − s T0 p T0 = βGmR0 p≅− βGm CLtot z≅+ Gm Cftot β= Cftot EE315A ― HO #13 B. Murmann Cftot + Cs + Cx 21 Noise Analysis (1) R≅ 1 βg mn Neglecting finite output resistance of the MOSFETs 2 vo 1 = 4kT γ n g mn + γ p g mp ⋅ R Δf j ωCLtot ( ) γ p g mp ⎛ = 4kT γ n g mn ⎜ 1 + γ n g mn ⎝ B. Murmann EE315A ― HO #13 2 ⎞ R ⎟⋅ ⎠ 1 + j ωRCLtot 2 22 Noise Analysis (2) 2 vo ∞ γ p g mp ⎛ p = ∫ 4kT γ n g mn ⎜ 1 + γ n g mn ⎝ 0 γ p g mp ⎛ = 4kT γ n g mn ⎜ 1 + γ n g mn ⎝ ⎞ R ⎟⋅ ⎠ 1 + j ωRCLtot 2 df ⎞ 2 1 ⎟⋅R ⋅ 4RCLtot ⎠ γ p g mp ⎞ ⎛ 1 1 = 4kT γ n g mn ⎜ 1 + ⋅ ⎟⋅ γ n g mn ⎠ βg mn 4CLtot ⎝ ⎛ ⎜ γ p g mp 1 kT = γn ⎜1+ β CLtot γ n g mn ⎜ ⎜ ⎝ B. Murmann ⎞ ⎟ ⎟ ⎟ ⎟ ⎠ EE315A ― HO #13 23 Noise Analysis (3) • For low noise – Make gm2 as small as possible, i.e. use small gm/ID for current source device • Issue: Smaller gm/ID means larger “Vdsat” i.e. less available voltage swing – Maximize feedback factor β β= Cftot Cftot 1 1 1 = ≅ = Cggn g 1 + Cs + Cx 1 + Cs + Cx 1 + A∞ + mn 1 + A∞ + Cftot ωT Cftot Cftot Cftot Want ωt ∞ (short channel) B. Murmann EE315A ― HO #13 24 Noise in Differential Circuits • In differential circuits, the noise power is doubled (because there are two half circuits contributing to the noise) • But, the signal power increases by 4x – Looks like a 3dB win? DRsingle ˆ V2 ∝ o kT C DRdiff ( 2Vˆ ) ∝ o 2 kT C 2 =2 ˆ2 Vo kT C • Yes, there’s a 3dB win in DR, but it comes at twice the power dissipation (due to two half circuits) • Can get the same DR/power in a single ended circuit by doubling all cap sizes and gm EE315A ― HO #13 B. Murmann 25 OTA1 Behavioral Model Parameters: Gm, a0, fT, γ OTA1 in + vx - Gm 2π fT Rnoise = Cin Rnoise 1 in Gmvx Ro noiseless 1 γ ⋅ Gm 2 i n = 4kT 1 Rnoise Δf = 4kT γ ⋅ Gm Δf CCCS B. Murmann a0 Gm Cin = - Ro = EE315A ― HO #13 26 ...
View Full Document

## This note was uploaded on 08/13/2009 for the course EE 315 taught by Professor Borismurmann during the Spring '09 term at Stanford.

Ask a homework question - tutors are online