This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: ECE 327: Electronic Devices and Circuits Laboratory I Notes for Lab 6 (Digital-to-Analog Converter (DAC) Lab) 1. Strictly speaking, this laboratory is not about digital-to-analog conversion . A DAC converts an abstract numerical quantity to a physical quantity (e.g., a voltage) Current-summing DAC uses binary-weighted currents switches select which currents to sum In practice, single-pole double-throw (SPDT) switches are used (i.e., no floating) Can generate currents with binary-weighted resistors * Turning on and off currents quickly causes switching transients (i.e., introduces noise) * Instead, switch constant current flow toward and away from sum junction * Binary-weighted resistors are difficult to implement (especially for high number of bits) R 2 R ladder used to make a more clever current-summing DAC ( Figure L6-1 ) * Need R , 2 R , and switch per bit Much easier to match R and 2 R (helps with linearity refer to INL and DNL ) * DAC input impedance is 2 R regardless of number of bits and switch states Each new bit just splits off more current Constant current means no switching transients Multiplier DAC (mDAC) like a digital potentiometer * Ratiometric : Output is a ratio of input (usually power supply) Principle used in relaxation oscillators to make them insensitive to power supply changes Frequently used with sensors for same purpose Use zero-order hold (ZOH) to construct analog signal from digital time series * Result is pulse-amplitude modulated (PAM) version of digital signal * Low-pass filter to smooth edges if necessary (sometimes system is enough of a filter) Oversampling can increase filter performance and decrease its complexity and cost Oversampling: reconstructing a time series at much higher frequency by interpolation Alternative: Oversample and use 1-bit DAC (looks sexy on shiny metallic label) 1-bit DAC: Each point is represented as one of two analog states ( cheap to implement) Filtering the quickly switching output produces the desired analog signal Marketers make empty/nonsensical claims of quantization noise reduction (real benefit: cost ) Pulse-density modulation (PDM, - , - , SDM) is frequently used * Simple low-pass filter (LPF) restores analog signal * LPF can be omitted if load has appropriate frequency response Pulses are often used in digital communication pulse-code modulation (PCM) represents binary c odes as serial high and low p ulses * e.g., compact disc digital audio (CDDA) pit-land transitions encode 1 * a sequence of pulses (e.g., 7 subsequent pulses) represents one codea sequence of pulses (e....
View Full Document
- Fall '08