EE 472 ndash; Senior Project.ppt

“Noise Characterization of A/D Converter on PIC Microcontroller” PIC Microcontroller PIC: Programmable Interface Controller Modifiable by end user: - Software (C code) Hardware (modification to circuitry) Common use in robotics and control applications Potential use in data acquisition or measurement Resolution What is maximum resolution? For 5V, maximum resolution ~ 5 mV 1 LSB represents 5mV increment A/D Converter Noise Quantization noise (error due to discretizing) Thermal noise Reference voltage variations Etc. What is the sum of all these contributions? Oversampling and Averaging Artificially enhance resolution Cost efficient alternative to paying for higher resolution BUT …. Tradeoff throughput (High resolution means slower conversion speed) More CPU intensive Requirements Must be able to approximate sum of all noise contributions as white noise White noise has uniform power spectral density Two ways to check for this: (1) Histogram analysis (2) Take FT of autocorrelation Data Measurement Built-in potentiometer / DC power supply Labview collects digital output values Histogram analysis Serial Interface Serial interface PIC chip with PC Receive streaming data from output pin Approaches taken: (1) Visual Basic – failed effort (2) Labview – better Visual Basic Approach Problems with Visual Basic Buffer check requires too much time Causes lag in data acquisition Too slow for collecting large number of data points Labview Waveform Waveform Close-up Another Waveform Histogram – 88 Histogram - 11 Histogram – 255 DC Histogram – 18 * EE 472 – Senior Project Hang B. Lee, Applied Physics EE Advisor: Prof. Robert Grober EE 472 – Senior Project Project Goal Understand noise characteristics of A/D Converter Investigate methods to minimize or eliminate noise Artificially enhance resolution – software methods EE 472 – Senior Project 10-bit resolution 8 channels for analog input