1 EE 435 Name Exam 1 Spring 2018 Instructions: The points allocated to each problem are as indicated. Note that the first and last problem are weighted more heavily than the rest of the problems. On those problems that need technology parameters that are not given in the problem, assume you are working in a 0.18µm CMOS process with process parameters μncox = 350μA/V 2, μpcox = 70μA/V 2, VTn =0.5V, VTp= -0.5V, COX= =8fF/µ 2 λ, = 0.01v -1,,0 = γ and AVT0= 20mV μm for both n-channel and p-channel devices. As a take-home exam, all work on this exam must be done individually. There should be no collaboration with anyone except for the course instructor, R. Geiger, should there be any questions. Since this is a take-home exam, please address any questions you may have by email to the course instructor. An immediate response can not be promised but I will check my email periodically throughout the weekend. You will be asked to make the following statement and provide your signature on the top of your solutions. All of the work on this exam is my own and I did not collaborate with anyone about this exam except possibly with the course instructor signature here Problem 1 (20 points) The operational amplifier has been designed with VEB = 150mV for all transistors with a total power dissipation of 20mW when biased with a single 5V supply (i.e. VDD=5V). The load capacitor is CL=20pF and the length of all transistors are 4µm. a) Determine the GB of the op amp b) What is the W/L ratio of M1? c) Determine an acceptable value for VB2 d) Express the dc gain in terms of the small-signal model parameters of the devices. e) Give a numerical value for the dc gain of this op amp. f) What is the 3dB bandwidth? g) What is the slew rate of the op amp? V DD M 5 M 6 V B2 V OUT M 3 M 4 C L V IN M 1 M 2 V IN I T V B3 M 11
2 Problem 2 (10 points) A large number of different op amp architectures were identified in class. A table summarizing some of the more popular possible structures is shown below. Common Source Current Mirror Differential Input Single-Ended Input Differential Output Single-Ended Output Stage 1 Tail Voltage Bias Tail Current Bias Common Source Current Mirror Differential Input Single-Ended Input Differential Output Single-Ended Output Stage 2 Tail Voltage Bias Tail Current Bias Internally Compensated p-channel Input Output Compensated n-channel Input a) Give the schematic of a two-stage op amp with a folded cascode first stage, a common source second stage, with differential inputs and single-ended output on the first stage. Use p-channel inputs on the first stage and n-channel inputs on the second stage. Use tail current bias on the first stage and tail voltage bias on the second stage. Use output compensation. Assume the total output capacitance (including any compensation capacitance) is CL. b) Give an expression for the dc gain of the amplifier in part a) in terms of the small signal model parameters and in terms of the practical design parameters. c) Give an expression for the GB of the amplifier in terms of the small signal model parameters and in terms of the practical design parameters.
3 Problem 3 (10 points) The magnitude and phase plots of a differential input, single-ended output all-pole operational amplifier are shown below. a) Determine the phase margin if this is used in a feedback amplifier with a feedback factor of β=0.05 b) Is the feedback amplifier stable? Why? c) What is the maximum value of β that can be used if the amplifier is to have a 75 o phase margin? d) If β = 0.05, what is the ideal dc closed loop gain if configured as a basic noninverting feedback amplifier and what is the percent closed-loop gain error due to the finite dc gain limitations of the op amp? e) How many poles does this amplifier have? 80 Magnitude Plot 60 Magnitude in db 40 20 0-20 10-1 1 10 10 2 10 3 10 4 ω -40 Phase in degrees -60 0-20 -40-60 -80-100 -120-140 -160-180 -200 Frequency in radians/sec Phase Plot Phase in radians/sec ω Problem 4 (10 points) β = 0.05 Generate the Nyquist plot for the amplifier of Problem 3 if
4 Problem 5 (10 points) A designer asked the layout technician to layout a differential pair using a common-centroid layout with two devices connected in parallel to form each transistor. The differential pair was used in the single-stage op amp shown below. The desired layout is shown on the left in the figure below with blue corresponding to the channel of transistor M1 and purple corresponding to the channel transistor M2. The drain and sources are on the right sides of the corresponding rectangles. But through a communication error, the actual layout that was obtained was that on the right. The dimensions on the axis are in µm. Assume µcox=100µav -2 and the threshold voltage at x=0 and y=0 was the nominal value of 0.5V. a) Determine the actual threshold voltage that would have been obtained for M1 and M2 if there is a positive gradient at +45 degrees relative to the x-axis of magnitude 4mV/µm. b) Determine the actual threshold voltage that will be achieved if the layout on the right were used (assume the same gradient as in part a) still exists). c) Determine the input offset voltage due to the gradient that would have been achieved if the common-centroid layout on the left were used and that which will be obtained if the undesired layout on the right were used. d) Repeat parts a)-c) if the gradient magnitude is the same but the direction is at 0 o relative to the x axis. y y 10 10 2 4 6 8 10 12 14 16 x 2 4 6 8 10 12 14 16 x 5V 20K 20K V OUT VIN M 1 M 2 V IN -5V I T 0.1mA
5 Problem 6 (10 points) An all-pole amplifier has two open-loop poles, one at 10Hz and the other at 1MHz and the dc gain of the amplifier is 80dB. If used in a feedback application in a closed-loop amplifier with β=0.5, determine the a) phase margin b) Q of the closed loop poles Assume the closed-loop gain satisfies the standard Black feedback expression As AFB where A(s) is the open-loop gain of the op amp. 1 As Problem 7 (10 points) Consider a feedback amplifier where the gain with feedback satisfies the standard Black feedback equation A FB As. Assume further 1 As A = s s 1 1 p kp 0 that the open-loop amplifier is a two-pole amplifier with gain A(s) 1 1 where k is the ratio of the open-loop poles. a) sketch a root locus plot of the closed-loop poles for 0<β<1. In this sketch, assume the pole ratio k satisfies the relationship k>>1. b) Determine the Q of the closed-loop poles if β=0.5, A0=10,000 and k=15,000. Problem 8 (20 points) A two-stage operational amplifier is shown. Assume VDD=5V. Assume that the tail current of the differential pair is 1mA. a) Size the devices M1 and M2 so that with a differential input voltage of 500mV, the deviation of the drain currents from linear is at most 2% and the random input offset voltage (1 σ value) is 15mV. b) Size M7 and determine V XX so that VEB7 is 150mV. c) Size M6 so that the power in the second stage is 10 times the power in the first stage. d) With the constraints given in the problem and in steps a)-c), identify the number of degrees of freedom remaining in the design of this circuit. With these constraints, complete the design leaving one degree of freedom to determine CC. e) Determine CC in your design so that the magnitude response is maximally flat. Assume CL=500fF. f) Determine the GB of your design
6 V DD M 3 M 4 M 5 M M V 1 2 1 V 2 C C V OUT C L V XX M 7 V XX M 6