What’s so complex about USB-C that we couldn’t have had this technology 20 years ago?
In: 1727
There is nothing complex. This was a question of money and patents and jostling by tech companies. There are roughly 2 (maybe 3) groups of tech companies that have banded together and combined their IPs to make standards like SATA and USB and even within those groups there was active debate because of financial implications for how these things play out. We’re at where we are because governments have gotten increasingly frustrated at this useless jostling and begun to get involved and impose standardization.
The only thing keeping us from a USB-C type interface 20 years ago was the capitalist principle of profit first.
Put simply:
1. Faster clocks in transceivers.
2. Better line encoding/scrambling.
3. Reduced crosstalk
Each pulse of the clock allows for a detection of signals, 1s and 0s. (It may be more precise to say detection of signals or signal-change)
In slower olden times, the clocks were slow and could only catch so many. Additionally, the ‘on-the-wire’ signalling was prone to cross-talk. This is when the electromagnetic force of one carrier line affected the signaling of other lines in the same cable.
As clocks grew faster, parallel signaling was also just increasingly a pain to work on. Once an inbound signal was ‘past’ the transceiver, PCB designers were already trying to minimize the amount of traces they had to run to move that data around.
This resulted in mainboard signaling methods that derive from things like the ‘media independent interface’ (MII and its various flavors).
At the same time, hardware people were realizing the advantage of twisted pair cabling as it reduced crosstalk. Back in the days of 10Base2 and 10BaseT (old networking standards), they used a primitive line encoding called Manchester that simply described how the computer’s 1s and 0s went on the line. I can’t recall if they already had a scrambler by then, but essentially the second problem was that treating a block of 1s and 0s the exact same on the wire is bad.
Think about an email that has a bunch of blank spaces. Lets pretend they’re all 00s.
If the wire actually sent that as all 000000000000000000000000s, then the clock on the other side might actually get screwed up. You see the clocks on both ends of a line (whether USB or ethernet, etc) have to regularly time their clocks to make sure they don’t miss those signals or transitions.
Scrambling is the act of looking at the data to send and actually altering the code so that there are more 0s and 1s. 8b/10b 64b/66b are examples of these scrambling functions that preserve clock.
But to put it all together, they couldn’t keep reducing the wires and devices until all of this codependent technologies had come around, all interlocked and evolving at the same time.
With regards to lightning, USB-A/C whatever, that broadens into topics of which vendors have influence at a given time and is more market and politics driven.
Latest Answers