Scribd
Upload
431 views

Cornell ECE 5790: RF Integrated Circuit Design Assignment 3

This document outlines 5 problems related to amplifier design. It provides instructions for calculating noise figure, gain, linearity measures, and designing low-noise amplifiers. Students are asked to simulate circuits, determine compression points and intermodulation distortion, compare differential and single-ended stages, and optimize an LNA design to meet noise, bandwidth, and power constraints. Designs should be submitted with plots showing the performance specifications have been achieved.

Uploaded by

Brimwoodboy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
431 views

Cornell ECE 5790: RF Integrated Circuit Design Assignment 3

This document outlines 5 problems related to amplifier design. It provides instructions for calculating noise figure, gain, linearity measures, and designing low-noise amplifiers. Students are asked to simulate circuits, determine compression points and intermodulation distortion, compare differential and single-ended stages, and optimize an LNA design to meet noise, bandwidth, and power constraints. Designs should be submitted with plots showing the performance specifications have been achieved.

Uploaded by

Brimwoodboy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

ECE 5790 Spring 2014 Assignment #3

Hamidreza Aghasi
Optional submission: Friday, March 22, 2014 Mandatory submission: Friday, March 29 , 2014

Problem 5 is the extra problem for those who target the mandatory submission due date. Good Luck!

Contents
Problem 1: Folded Cascode LNA (30 points) Problem 2: A1dB of highly nonlinear systems (20 points) Problem 3: Linearity of Dierential Pair(50 points) Problem 4: Low Noise Amplier Design (85 points) Problem 5: Noise Figure in multi-transistor circuits(20 points) 2 2 3 3 5

Problem 1: Folded Cascode LNA (30 points)


The LNA shown in Fig.1 is designed to operate with low supply voltages. Each inductor is chosen to resonate with the total capacitance at its corresponding node at the frequency of interest. Neglect channel-length modulation and body eect and the noise due to the loss in L2 (You need to model this loss components as a prallel resistance of RP 2 for calculating the gain and Noise Figure, but do not count the corresponding noise). a)Determine the noise gure of the LNA with respect to a source resistance RS assuming that L1 can be viewed as a resistance equal to Rp1 at the resonance frequency. Make sure the result reduces to a familiar form if Rp .(Hint:the equivalent transconductance of a degenerated common-source stage is given by gm/(1 + gmR1 ),where R1 denotes the degeneration resistance.)(10 points) b) Using BICMOS6hp process and your own values, simulate this circuit to achieve a noise gure smaller than 4dB at 2.4 GHz frequency. Make sure that the circuit achieves this noise gure for a BW of larger than 16 MHz. The total power consumption should be less than 10 mW and the inductors should have a quality factor of 20. This design is good enough for an IEEE 802.11.a standard requirement. For this design question please submit your clear plots/gures showing the performance for each specication.(20 points)

Figure 1: Folded Cascode LNA

Problem 2: A1dB of highly nonlinear systems (20 points)


Under certain conditions, we could model a nonlinear system by: y (t) = 1 x(t) + 2 x2 (t) + 3 x3 (t) + 4 x4 (t) + 5 x5 (t) As we saw in class, for a weakly nonlinear circuit, the 1-dB compression point can be calculated as A1dB = 0.145| 1 | 3 (2) (1)

This equation fails to explain the measured 1-dB compression point of highly nonlinear systems. In this case, we should take into account higher order nonlinear coecients. With similar technique that we used to obtain eqn.(2), nd the 1-dB compression point of a system with nonlinear coecients up to fth order. Page 2 of 5

Problem 2: A1dB of highly nonlinear systems (20 points)

Problem 3: Linearity of Dierential Pair(50 points)


In this problem, we are going to compare the linearity of bipolar ampliers in dierent structures. Lets rst assume we have a common emitter amplier as shown in gure 2. a)Calculate the 1-dB compression point of the amplier shown in gure 2.(3rd order nonlinearity)(5 points) b) How does the linearity change when we make a dierential pair as shown in gure 3?(Assume that only three nonlinear coecients are enough to characterize the behavior of the amplier. Also if you need any numerical value for the parameters in the problem, assume some typical values!)(10 points) c)(Bonus) What is the IIP3 of the circuit shown in gure 3? To do so, similar to our calculation in class, assume a two tone input and nd out the gain of fundamental and third-order intermodulation products and set them equal.(10 points) d)(Bonus) To further enhance the linearity, you design the structure shown in gure 4. For this circuit what should be the value of Ibias , in order that the amplier has the same transconductance as a conventional dierential pair with a tail current of IEE ? (10 points)

e) (Bonus) What is the 1-dB compression point of the circuit in gure 4?(10 points) f) (Bonus) What is the IIP3 of the circuit shown in gure 4?(5 points)

Figure 2: simple bipolar amplier

Problem 4: Low Noise Amplier Design (85 points)


In our class discussions, we saw that the Bipolar transistor exhibits a very nonlinear performance which makes it hard to design an LNA. In last question, we also compared the nonlinearity of a single-ended amplier and a dierential amplier. In this problem we are going to design a very good Low Noise Amplier using the BiCMOS6hp process. I should mention that if you had met these specications a few years ago, you had a very famous paper on LNA design. For this design we are going to use the structure shown in gure 5.

Before starting the design, lets rst review a few circuit problems.

Problem 4: Low Noise Amplier Design (85 points) continued on next page. . .

Page 3 of 5

Problem 4: Low Noise Amplier Design (85 points) (continued)

Figure 3: Bipolar Dierential pair

Figure 4: Modied Bipolar Dierential pair a) What is the small signal gain of the circuit shown in gure 5, neglecting the channel length modulation and body eect? For this part assume that each transistor is biased with a current of 5mA and all the transistors are biased 200 mV above the Threshold voltage.(In this part we assume zero coupling between inductors)(10 points) b) The combination of L1 and C1 is for the sake of segregation between the bias and RF signal. Assuming that for the matching, we have a direct connection from the input source to transistor, nd the values of Ls , Cf and Lg to have S11 < 20 dB at 3 GHz.(5 points) c) Compute the NF of each branch, assuming lossless passive components and neglecting the icker noise of transistors.(10 points) d)(Bonus) Assume that there exists a non-zero coupling between the drain inductors. Find the small signal gain in terms of the coupling coecient.(10 points)

Problem 4: Low Noise Amplier Design (85 points) continued on next page. . .

Page 4 of 5

Problem 4: Low Noise Amplier Design (85 points) (continued)

Figure 5: Proposed structure e)In this part, we are going to design an LNA(With new bias conditions of your choice), using this structure with the following specications: Gain 9dB

N Fmin < 4dB 3dB-Bandwidth from 2.5 4 GHz Power consumption < 20 mW S11,min < 13dB S12 < 5dB throughout the 3-dB BW IIP 3 > 2dBm Quality factor of inductors =20 (Use the library capacitor models) Your nal grade from part e is calculated from the following formula: Grade = 50 (9 Gain) (4 N F ) (P ower 20) (S11 + 13) (S12 + 5) (2 IIP 3) 2 1.5 (3)

For this design question please submit your clear plots/gures showing the performance for each specication.

Problem 5: Noise Figure in multi-transistor circuits(20 points)


Determine the Noise Figure of circuits in gure 2.79 (a and e) of the textbook with respect to a source impedance of Rs . Neglect the channel length modulation and body eect.

Page 5 of 5

You might also like