When differential signaling was invented, and why the fuck computers

>>62197552
>3 wires (tx rx gnd)
>4 wires (tx+ tx- rx+ rx-)

>>62197580
Technically 4 wires if you're running power too and 6 wires for differential.

>>62197580
If serial vs serial yes, but if you compare to IDE for example, things get quite advantageous for a fast serial connection, unless it was too expensive to do on a decent clock rate in the 80's.

USB uses it

>>62197634
Yep.
Yet things like serial ports, parallel ports, PS/2 ports etc.. existed before it, none taking advantage of the differential signaling.

>>62197603
2 wires are enough for power + full duplex data

>>62197552
>When differential signaling was invented,
Ok, sounds like the beginning of an interesting story, or maybe the setup to a good question, let's see where this goes.
>and why
...

>>62197655
That stuff was invented when speeds were low and every extra transistor mattered.

>>62197552
because it's not necessary at normal serial data rates and it requires a differential amplifier, increasing the cost for no gain
learn to cost/benefit analysis anon.

>>62200089
>increasing the cost for no gain
We use them in industrial control automation because the differential signals handle electrically noisy environments better than the alternative.