DOC PREVIEW
MSU ECE 410 - lab5

This preview shows page 1 out of 4 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 4 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 4 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Lab 5: Introduction to Parametric Analysis Due: Thursday March 05, 2009 Summary: Lab 5 introduces the use of simulation tools to perform parametric analysis, wherein designers can observe the impact of varying specific circuit/device parameters. Learning Objectives: 1. Learn how to use and set design variables and how to create cells that use passed parameters. 2. Learn how to perform parametric analysis and use it to solve a simple optimization problem. 3. Observe the effect of varying transistor widths on the delays of an inverter and the effect of varying input voltage slope on inverter delays. Resources: Tutorial D Preliminary Notes: To ensure consistency of results, always use a 3V square wave with a rise and fall time of 0.05ns as the input to your inverter unless otherwise specified by the instructions. A width of 5ns and a period of 10ns are recommended, but may be any values that allow the output to fully rise and fall. Procedure: 1. Complete Tutorial D. Before continuing, be sure you have a good understanding of the following concepts: specifying and setting design variables and passed parameters, and parametric analysis. 2. Construct and analyze an inverter with equal high-to-low and low-to-high propagation delays. a. Using parametric analysis on an inverter schematic, determine the approximate Wp/Wn ratio that equalizes the propagation low-to-high (LH) and propagation high-to-low (HL) delays. You must choose a value that that will result in a PMOS width that is multiple of λ/2 (0.15μm). Record the value of Wp/Wn, tHL, tLH, and calculate the percentage difference in the two times (|difference|/average). Save the plot showing the inverter propagation delays for the β that equalizes tHL and tLH. Refer to this Wp/Wn value as the “measured β value” for the rest of the assignment. b. Create a new inverter layout with the PMOS width changed to match the measured β value. Extract the layout (including parasitic capacitances) and run transient simulation using the extracted view. Use the same input stimulus file that you used in Step 2. Record the propagation HL, propagation LH delays, and calculate the percentage difference in the two times. Save the plot showing the inverter propagation delays from the extracted view. c. Create a new inverter cell that uses the following two passed parameters: “Beta” and “f” (inverter size). “Beta” should be used as a multiplier for the PMOS transistor width, and “f” should be used as a multiplier for both the PMOS and the NMOS widths. In other words, the width of the PMOS will be the product of “Beta”, “f”, and the specified minimum width 1.5μm, and the width of the NMOS will be the product of “f” and the specified minimum width 1.5μm. d. Create a new symbol for the new inverter. If you use a copied symbol, you will not be able to pass parameters to it. The symbol shape may be left as the default box.3. Determine how to set transistor sizes to minimize output transition delays. a. Create a new schematic. Place two of these new inverters in series, followed by a 20f F capacitor. For both inverters, pass the “measured β” value that you determined in step 2. For the first inverter, pass an “f” value of 1. Parameterize the passed value for “f” of the second inverter. Your schematic should look similar to the following figure: Note: Beta value was picked at random in this example. The red text above the symbols does not appear by default. The shape of the inverter symbol does not matter here. b. Using parametric analysis, determine the value of “f” for the second inverter that minimizes the end-to-end propagation delay (from the input of the first inverter to the output of the second inverter) in the circuit. Limit the range of “f” to between 1 and 3. The measurement should be to the nearest tenth (0.1). Do not worry if this “f” value results in transistor widths that are not multiples of λ/2 (0.15μm). Record this value of “f” and the corresponding propagation LH and HL delays. Save the plot showing the inverter propagation HL and LH delays for the optimal value of “f”. 4. Analyze the effect of input rise and fall times on output transition delays. a. Create a new schematic. Place a pulse voltage source followed by the new inverter using your measured β value and an “f” value of 1. Specify a 3V voltage pulse with a 5ns width, 10ns period, and parameterized rise and fall times. The rise and fall times should be set using the same design variable name, so that they are identical. Your schematic should look similar to the following figure:0b. Using parametric analysis, sweep the voltage pulse rise/fall time from 0.05n to 0.50n in 11 linear steps. Record the average inverter propagation delay (the average of propagation HL and LH delays) for each of the 11 different input transition times. Note: Do not include an input voltage source in your stimulus file here. It is only necessary to set vdd to 3V. c. Plot inverter propagation delay vs. inverter input rise/fall time using a program like Excel, Mathematica or MATLAB. Make sure the data is connected by lines – either point-to-point, or using a fitted line. This step can be performed after completing the lab assignment. d. Find the line of best fit (i.e. slope and y-intercept) for the inverter propagation delay vs. input rise/fall time using a simple linear regression. In Excel, this can be accomplished by using the SLOPE and INTERCEPT functions on the data. 5. Build a schematic for a two-input AND gate by instantiating your previously-designed NAND gate followed by the new parameterized inverter cell. Pass the measured β value and an “f” value of 1 to the inverter. It is not necessary to make a symbol or layout. Simulate the AND gate for a 11 >>10 falling transition and measure the propagation delay. a. Using the results from your linear regression in step 4d and the timing analysis for the NAND done previously in Lab 5, calculate the predicted 11 10 falling propagation delay for the AND gate. Compare this to the measured value by calculating the % error. It is not unreasonable to have high accuracy here, assuming you have carefully measured all necessary delays and have done the calculations correctly. 16. Print the Lab 5 Grading Sheet and check off your lab with a TA by 4pm on Friday. 27. Construct a report of this assignment using the Guide to Writing Lab Reports. Include appropriate tables or lists of


View Full Document

MSU ECE 410 - lab5

Documents in this Course
hw4-s08

hw4-s08

2 pages

Exam 1

Exam 1

5 pages

lab3

lab3

2 pages

lab4

lab4

4 pages

Lab 5

Lab 5

4 pages

Ch12

Ch12

34 pages

lab2

lab2

2 pages

lab7

lab7

3 pages

Ch11

Ch11

43 pages

Lab 6

Lab 6

4 pages

lab2

lab2

3 pages

Ch7

Ch7

32 pages

lab7

lab7

2 pages

Lab 4

Lab 4

3 pages

lab1

lab1

3 pages

Ch3-5

Ch3-5

69 pages

Load more
Download lab5
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view lab5 and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view lab5 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?