Data Transmission

Hello Everyone!
I'm not sure if this is the right place to pose this question, but I'm thinking that someone might have some useful information. I've been trying to research how data is transmitted over a network; i.e. ethernet. Basically, I want to find out 'low-level' information, below software and protocols. I guess that this crosses into the field of electronics, but I think it is an interesting topic that is often overlooked.

Just to illustrate, what I'm researching is akin to the way that electricity travels through circuits in order to light up your room; except that instead of providing light, the end result would be a sum of data.

Any resources/links/info would be greatly appreciated.

Thanks,
piirx

 

Log in or Sign up to hide all adverts.

As far as I know data is transmitted digitally across wires instead of analog. Basically the data is transferred using current, or the absence of current. The absence of current corresponds with 0 and the presence of data corresponds with 1 (like binary). The NICs (as well as any routers or switchs) on either end of the transmission will then recieve the signals until the NIC finds that the packet of data has been filled by the sender. The NIC then translates and checks the information before it sends it on to the software that manages the information.

So I guess after all of that rambling, I have said one thin, computers send information digitally.

Please Register or Log in to view the hidden image!

 

Log in or Sign up to hide all adverts.

Log in or Sign up to hide all adverts.

Hello,
Thanks for the info, that helps to clarify a bit. It would make more sense for them to vary voltage rather than current, but as you guys mentioned, the principle is the same.

There's a bit more that I'd like to find out though. For example, how is the start and end of a transmission signified? The same goes for 'pulses' between 1's and 0's (I.E. how do you distinguish getting two 0's, versus just one?)

Also, any links to more information and/or technical schematics would be greatly appreciated. I heard that IEEE posts some nice docs on 802.x, so I'll check those out too.

Thanks,
piirx

 

I don't know if this was what you're looking for or if it'll help ya, but, here it comes:

As someone said, zeros and ones are represented by voltage levels(ttl standarts use 0-0.8 V to represent a digital zero, and 1.8 to 5V do represent digital ones, the range between 0.8 and 1.8 is considered some kind of error range of the components used, and should not be used), for long lines of communications, it's used lines with a greater voltage, line 12 V (because it's more difficult to have interference that changes from 0.8 to 12V

Please Register or Log in to view the hidden image!

)

Let me give a try at the transmission part, you have the line at low level, nobody is transmitting anything... so one computer should make the "start bit", it could be just one bit, or a sequence of bits... So there is a change, something like 10 (two bits), so what you have in time will be:
0000000000000000000000010...

for the next 8 bits, your protocol could be treating like if it was data, so you fill an entire byte, you can know the duration of the bit by the time, lets say that each bit should be at the line for 3 ms, so, you can know now extacly what is the bit at the line in a given time, does'nt matter if was two zeros, two ones, etc once you have the byte filled, the line goes to low again, until a new start bit is found... and you know that to transmit a byte, you'll need at least 30ms, cause you have 8 bits plus two start bit (you could have stop bits too, and crc checks)... so the velocity of the line will be something close to 30 Bytes/sec.

An other aproach, it's to use the change in the line, and not a level in a given moment. So you Establish taht for your protocol, a change from the level zero, to one, means one, and a change from the level one to zero, means zero... So you can always be sure of the value, but again, you must have the time that the line stays at each level, so no mistakes are commited...

 

Hello,
Thanks AgamemnoN, that helps a lot. That's actually somewhat similar to what I was thinking originally, but it's nice to have it confirmed.

I was planning on doing an experiment to see how this works in action, so I'll keep you guys informed.

Thanks again,
piirx