Electronics notes / Transmission lines
To avoid some confusion: in power distribution there is a distinction between transmission line (everything between power plant and substation) and distribution line (substation to consumer). This distinction is not the sense we talk about below (though these are also transmission lines in the sense below).
Around electronics, transmission line takes on a more specific meaning of
dealing with frequencies high enough that
you have to start worrying about the wave nature of transmission
on the conductors you put it on.
You're probably already used to considering a conductor's basic imperfections,
such as "has a small resistance, so some voltage drop" and "thicker wires help".
At higher frequencies
- a pair of conductors also has capacitance between them that starts mattering
- a loop has inductance that starts mattering
...which are both more complex to deal with.
Transmission line introductions often point out that, while the effect is more continuous,
you can get a basic intuition by seeing a transmission line as a series of many LC circuits.
This illustrates that what's happening is much like a series of charge/discharge - a sluggish, impeded transmission.
It may also help intuit other side effects, like that
- how impedance mismatches is like a change in the L and C values somewhere in the middle, and how that means reflections, signal/power losses
- cabling also starts noticeably emitting the signal, causing loss (and interference issues),
For a sense of scale:
- A telegraph signal, or ~50Hz power, needs to start worrying when a single segment is is longer than a few miles.
- Audio and slowish digital communication dozen kHz runs into trouble at a few dozen meters.
- this is also why you can get away with ignoring transmission line details at slow speeds, roughly perhaps 30kbps.
- MHz-rate networking, and most PC video signals (dozens-of-MHz), above a few meters.
- which is why cabling is specific
- A ~1GHz cell phone signal on a generic cable is an issue over a centimeter or so
- which is why cabling is specific
- A ~5 GHz CPU needs to start worrying well below a millimeter
- (which is part of why smaller is better, why there is a speed above which things become too bothersome, and more)
It's also a reason oscilloscope probes are often short.
The footnotes to measuring high frequencies on oscilloscopes relate to the same theory.