Lab 02 – EE 421L

Authored by Chandon Esplin,

EsplinC2@UNLV.Nevada.edu

09/09/2016 

  

Design of a 10-bit digital-to-analog converter (DAC) 

 

Return to:    C. Esplin EE 421 Labs

 

Pre-Lab Objectives

Gain and demonstrate intimate knowledge of both Analog-to-Digital & Digital-to-Analog Converters [ADC & DAC], in preparation of, the design of a 10-bit DAC. Reinforce the habit of backing up and storing previous versions of both completed work and the technical libraries used. Further develop and expand upon foundational skills with Cadence software.

 

Pre-Lab Tasks

1.     Provide narrative of Pre-Lab

2.     Provide differing simulations (from example) and discuss results to demonstrate knowledge of ADC & DAC

3.     Explain method of determining the Least Significant Bit (LSB, the minimum voltage change on the ADC’s input to see a change in the digital code B[9:0] of the converter. Provide simulations in support of explanation.

 

Pre-Lab Data

           (1) Narrative

 

Creating back-ups of previously completed work prior to reading through laboratory

 

Downloading the entire user created CMOSedu directory, zipping, and dating.

*Note: This process proved to be time consuming and excessive for back-up purposes.  Next back-up will consist only of user altered files and directories.

 

Downloading, uploading, and unzipping lab2.zip to cds.lib in design directory

The previously created directory CMOSedu is the target destination to upload lab2.zip. Once unzipped, the cds.lib file will need the statement DEFINE lab2 $HOME/CMOSedu/lab2 added before Cadence will recognize the newly added converters.

 

Load schematic of sim_Ideal_ADC_DAC & run simulation

     

The above simulation shows the input voltage Vin, an analog signal, deconstructed into digital components and reconstructed back into an analog output voltage Vout.

 

(2) Provide & Discuss Simulations

 

Altering the amplitude of Vin to a higher voltage than Vref

Vref limits the bounds of Vout, as seen above when the amplitude of 3 volts exceeds the 0-5 limit, clipping occurs.

 

Increasing Vref to 10 volts with a relative increase to Vin

There is a limit to where Vref can be increased without diminished returns.

 

Vref set equal to Vin offset (2V)

Further demonstration of the constraint Vref has over the output voltage.

 

(3) Explain: Determination of Least Significant Bit

As found in CMOS Circuit Design, Layout, and Simulation 3rd edition

The least significant bit is given by:                            Using the initial example values of:

                                                      n=10 & Vref =5 V

 

                               The Least Significant Bit:

 

                      4.883mV

 

Simulation of Vin set just above LSB viewed below

As seen in this simulation, the resolution of the reconstructed signal is down to one bit.

 

 

Lab: Design of a 10-bit digital-to-analog converter (DAC)

 

Experiment(s):

·      Design - 10-Bit DAC with 10kΩ n-well Resistors

·      Determine Output Resistance

·      Time Delay - Driving a Load

·      Symbol Creation

·      Simulations of Various Load Types

        

Results:

·          Design - 10-Bit DAC with 10kΩ n-well Resistors

10-Bit DAC Schematic

 

 

Output Resistance Determination

Parallel to Series

Combining 2R in parallel with 2R produces R in series with R. This parallel/series reduction and recombination continues, resulting in a total resistance of R.

 

Delay, Driving a Load

Time Delay Circuit

Symbol Creation

Created DAC Symbol

 

Simulations of Various Load Types

Resistive

  

 

Capacitive

   

 

Resistive-Capacitive

   

 

Discussion:

         In reality, the switches used within the converters are implemented with transistors. Discuss the outcome if the resistance of the switched is not small relative to R.

 

If left unaccounted for, the addition of a new series resistance will alter the parallel/series decomposition and reconstruction. This chain reaction will cause the previously even distribution of bits to be altered. The end result will be larger sources of errors and failure to obtain the number (1024) bits as designed.

 

 

 

 Return to:   C. Esplin EE 421 Labs