Why digital?
The world is analog, but the machine is digital. This topic is why almost every modern engineering system keeps its "brain" in 1s and 0s.
Built from first principles. Opening topic of the course.
Before you start
What you need first
- Nothing — this is the first topic. Just curiosity about how the electronics around you actually work.
What you'll be able to do
- Say what a signal is, and tell analog from digital.
- Explain in your own words why digital won.
- Describe quantization and the bits-vs-resolution trade-off.
What do we mean by "signal"?
A signal is any physical quantity that carries information over time — a voltage on a wire, air pressure on a microphone, light on a camera sensor, the angle of a steering wheel.
Every signal falls into one of two families — analog or digital. This whole course is about the digital kind, so let's see the difference first.
Analog: smooth and continuous
An analog signal varies continuously: between any two values there are infinitely many in-between values. A thermocouple voltage, a microphone's output, a volume knob, a brake-line pressure sensor — all analog.
Digital: a small set of fixed levels
A digital signal takes only a small, fixed set of allowed values — in modern electronics almost always two:
0— a low voltage (e.g. 0 V).1— a high voltage (e.g. 3.3 V or 5 V).
Why two? A switch is on or off; a transistor conducts or it doesn't. Two states are the easiest thing to build reliably out of physics.
Why digital won
- Noise immunity. A "1" is anything above a threshold, a "0" anything below. As long as noise doesn't cross the threshold, the original value is recovered exactly.
- Perfect copying. A digital file copies bit-for-bit, a million times, with no loss. (Try copying an audio cassette ten times.)
- Cheap storage & transmission. SSDs, hard drives, Ethernet — all natively digital.
- You can compute on it. Logic gates and software process digital signals far more flexibly than analog circuits can.
- Programmability. One general-purpose digital chip can be reprogrammed to do almost anything; an analog circuit is wired for one job.
The price we pay: quantization
To store an analog quantity digitally we must quantize it — round it to the nearest allowed step. With \(n\) bits you get \(2^n\) levels:
- 8 bits → \(2^8 = 256\) levels
- 12 bits → \(2^{12} = 4096\) levels
- 16 bits → \(2^{16} = 65{,}536\) levels
✏️ Try it yourself
(a) Classify each as analog or digital: a microphone's voltage; a doorbell button; a
mercury thermometer; the data on a USB cable.
(b) How many levels does a 10-bit ADC have?
Recap — the whole topic on one screen
| Idea | What you own now |
|---|---|
| Signal | A physical quantity carrying information over time |
| Analog | Continuous — infinitely many values; vulnerable to noise |
| Digital | A few fixed levels (usually two: 0 and 1); noise-immune |
| Why digital won | Noise immunity, perfect copying, cheap storage, computable, programmable |
| Quantization | Rounding to \(2^n\) levels; more bits = finer but costlier |