1.6 KiB
A-D Converters
==Not on the upcoming exam==
==HW7 due in 3 weeks==
==HW8 is individual. Fill out one for each of your group members==
Analog signals are inherently not digital. In electrical systems, analog signals are represented by a voltage that is proportional to its intensity.
Digital signals are discrete. A digital representation is the result of a number of limited precision to a continuously variable quantity.
Digitizing
periodic sampling to approximate the value
More samples is more better-er
Sampling rate must be at least twice as high as the highest frequency you want to represent.
Bare minimum spec, and you can go higher.
Hardware
hardware exists to convert analog to digital
it can be a standalone component, and is built into many microcontrollers
There’s different ways that it can be done, but we will focus on successive Approximation ADC
ex.
Voltage Range = 0-10
I have a voltage in that range, we successfully divide by two until we find essentially the same number.
We’re inputting a 7.65V source.
We’re gonna cut it in half, then we compare to the source. If its higher, write a 1, take the divided number, and set as the lower bound. If its lower, write a 0, take the divided number, set as the upper bound.
{(10+0)\over2}=5<7.65;\ 1\
{(10+5)\over2}=7.5<7.65;\ 1\
{(10+7.5)\over2}=8.75>7.65;\ 0\
{(8.75+7.5)\over2}=8.125>7.65;\ 0\
{(8.125+7.5)\over2}=7.8125>7.65;\ 0\
{(7.8125+7.5)\over2}=7.65625>7.65;\ 0\
{(7.65625+7.5)\over2}=7.578125<7.65;\ 1\
{(7.65625+7.578125)\over2}=7.578125<7.65;\ 1\
Final result is 11000011
Bit-depth of the microcontroller creates the resolution.