This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: EEE 5320 BipolarvAnalog IC Design ' Lecthre # 2:3 5
“MMI/
. / Compensation Thought Questions 1. It is often stated that for O < [3 < l (where [i is the feedback factor B = T/ADM), B = l is the worst
case for stability. . Why?
 What situations can make stability worse than B = 1?
 Could there be situations with l [3 l < 1 where stability is worse than B: 1? 2. Show ways you could build an‘ op amp circuit where stability can be varied without affecting A“ ‘
and'without changing the compensation capacitor. 3. In most op amp ampliﬁers] there is a simple tradeoff between the closed—loop bandwidth and
the closed—loop gain. — What is that trade—off? _
 Why does that trade—off not apply to currentfeedback op amps? 4. The feedforward zero in a Millercompensated Op amp is in the right half of the s—plane, so it
hurts stability. (The zero keeps the gain from dropping as fast, while it is adding to the phase delay).  Show how adding a resistor R2 in series with the compensation capacitor can move the feedfor—
ward zero to improve the phase margin. ' — What value of R2 would move the zero to inﬁnite frequency?
~ What value would move the zero into the left half of the splane to cancel the non—dominant pole? — If increasing the value of R2 improves the phase margin, wouldn’t it be a good thing to increase
R2 to very high values? Why or why not? EEE 5320 Bipolar Analog IC Design Lecture '24
MN Compensation Thought Questions . P’/‘*" 1. It IS often stated that for 0 < B < 1 (where Bis the feedback factor B: T/ADM), B: 1 IS the worst
case for stability. — Why?
 What situations can make stability worse than B = 1?  Could there be situations with l B l < ‘1 where stability is worse than B :1?
0< 18K 1 $7 I916 is Féﬂ/{A(K06 emf/ex), 7%“ {34 7: wont—W,
[92W If 341/94 % étjé/eﬂi‘ T, 447?, $460. . PM” ﬂw (”W Mai/6“”) 7%” sag/,6, wax/Le
Lomsa 7%“ Kn 1/0/14? J(°/(9~J.e/V 64504 Mg“; WMZ fO/C’J 7409/ 0(2414017“ f/M avak EEE 5320 Bipolar Analog IC Design ; Lecture 24 2. Show ways you could build an op amp circuit where stability can be varied without affecting Am ‘ '
and without changing the compensation capacitor.  ('31. ﬁx Smquhmjug RA— w {35 W (447»; Van/«a4 ﬁi {EB Look/44?
W T, wit—L ﬂatpédgl x439. ﬁght WMa'rg, T Maia—@768“), %% rmPNt/‘e 5&6.ij w rt tie/[4:44, c/oxeo/F/aﬁf Iowan/[é ‘W/ s/m HAM WW . 3. In most op amp ampliﬁers I there is a simple tradeoff between the closed—loop bandwidth and
the closedloop gain.  What is that tradeoff?
 Why does that trade—off not apply to current—feedback op amps? 2(f‘zL
It 272‘ ’ a
k'%—“ 6:: / 50 ,McW—ng Ayn 5: 400 rs WWW
wanna/7 pafwﬁMﬂ
+9 wt Q’CBCJ. k lsle‘édrrw/Q (\D 4 I99), EEE 5320 Bipolar AnalogIC Design ‘ . .1 A ' '_ , Lecture :4
P 3/3 4. The feedforward zero in a Millercompensated op amp is in the right half of the splane, so it
hurts stability. (The zero keeps the gain from dropping as fast, while it is adding to the phase delay.)  Show how adding a resistor R2 1n series with the compensation capacitor can move the feedfor
ward zero to improve the'phase margin. ~ What value Of Rz would move the zero to inﬁnite frequency?
 What value would move the zero into the left half of the s plane tocancel the non dominant pole?  If 1ncreasing the value of R21mproves the phase margin, wouldn’t 1t be a good thing to increase
Rz to very high values? Why or why not? %' a sled/7 WW Mm 52 +91%, egg 1% f/W
Zf7fSCLQWH/L1/‘lﬁ {WMMJ’MMWJC 9%; ﬁlm (444/ A”) 4M 7% if: Mu—gémawqe loo/e. I; 1%in W0 “WM 1am A1494 ATM/K7 2752 PZWW
w “ff/47 , bfﬁéﬂm i. we 79, «mm 1;“ WW
w /( MA? 112 WW . 19% mm fame/7
ﬁwwm/J / rel/m1 (we/ca Add at Manda”! WW :4) (( abrﬂ
“Tm/J 71/47 s/me 776: Lﬂefmvaﬂ ,: «Lb/7‘?“ Page 143/, ﬂag/of. ...
View
Full Document
 Fall '10
 Dr.RobertFox

Click to edit the document details