Project 3: Stability, Programmability, and further Amplifier Design

The goal of this project is to understand stability in amplifier design, as well as its implications for realistic analog design (e.g. parastic capacitors). Further, this project explicitly allows using Floating-Gate (FG) circuit techniques in the designs. You will have two designs to consider for this project. Both designs require analysis, simulation, and IC layout.

You will use your previous IC process as well as simulation parameters for this design.

In both cases, you must demonstrate stability. You must show the step response for your amplifier connected as a unity gain amplifier. Stability is defined as Q <= 1, which is equivalent to at least 45 degrees of phase margin. For both designs, Area and Power must be calculated and minimized.

Transimpedance Amplifier Specific Items

Folded-Cascode Specific Items

the design is a differential input, single-ended output amplifier. Some specifications are

In addition each group will be assigned for additional specifications; these specifications will be sent by email to the class. These specifications include

Part of the grade will be on the creativity of the design, as well as how well the report is written up.

The noise measurement is input referred, that is, if one has an ideal noise voltage source, it would appear in series with either the input to the positive or negative terminal. Typically from SPICE for a noise measurement, you will get noise / rt(Hz) at the output (we are only assuming thermal noise for this project), so an approximate measurement is to divide by the lowest in-band gain (you have a given -3dB frequency, so take the lowest gain in that region) to get your input referred noise / rt(Hz).

Remember that you will need to provide all biases on chip. Address these issues in terms of area, power, and temperature effect issues. You will need to bias the components in your design, and you are allowed to choose one resistor (as desired for your biasing) for the design. One strong suggestion would be to use a bootstrap current source generator, and use this circuit to bias any cascode elements, if desired.

Your presentation will include the table of all specifications and what specifications you achieved with the resulting data. You need to insert your target numbers for your unique specifications (4 of them; I will check that they are correct). Of course, you need to have supporting evidence for your specifications.

Additional Materials for Project 3

Reading Materials that might be useful

Lecture Boards