Electrical specs for communication technologies

Hi there,
I'm interested in learning electronics coupled with computer science.
I'm mostly interested to know the voltage sent on different types of
wires, the speed or frequency at which information is sent through such
wire and so on. Basically, I would be interested to build an
electrical oscilloscope for an ethernet connection. Or anything as
funky as this.

But my thirst doesn't stop just here, I'm also interested about the
other basic wires that exists on the back of my computer, such as the
printer port, serial port, PS/2 port and what other port is there?...
Well, I do have USB but this may be much more complicated to deal
with... might still be very much interesting though!

I've tried google, but I either find stuff about electronics or about
computer science, but nothing about the electronics of computer
science, or if I did, it wasn't talking about this topic...

Thanks,
Simon

On 16 Apr 2006, in the Usenet newsgroup comp.os.linux.networking, in article
<(E-Mail Removed) .com>, (E-Mail Removed)
wrote:

> I'm mostly interested to know the voltage sent on different types of
>wires,

The specification for the particular interface should give a clue. For
example, the printer port is TTL - meaning the transmit specs are a low
being 0.4 Volts max at up to 16 milliamp sink, a high being 2.4 Volts min
at up to a source of 800 ua, and the receive specs being a low of 0.8 Volts
max at 1.6 milliamp, and a high being 2.0 Volts min at a maximum of 40 ua.
Serial ports are defined by EIA-232D, PS/2 is TTL, and so on.

>the speed or frequency at which information is sent through such
>wire and so on.

Depends on the port and application

>Basically, I would be interested to build an electrical oscilloscope for an
>ethernet connection. Or anything as funky as this.

Why build? Scopes are widely available, and far better performing than
anything a homebuilder could make. As for scoping an Ethernet connection,
you'll also have to understand Miller Code (NRZ) which is the modulation
scheme used. By the way, a scope really doesn't show that much of interest
on Ethernet - a packet sniffer is FAR more useful.

>But my thirst doesn't stop just here, I'm also interested about the
>other basic wires that exists on the back of my computer, such as the
>printer port, serial port, PS/2 port and what other port is there?...
>Well, I do have USB but this may be much more complicated to deal
>with... might still be very much interesting though!

Your best bet might be to get the datasheets for a computer chipset, as
most manufacturers provide "example" schematics of how to use their product
and that example is often virtually identical to what the motherboard
manufacturer actually does. In the early 1980s, you could buy the IBM
databooks for the IBM PC and PC-XT, and these included schematics. As this
was still the age of 74xx series TTL (there was some 74LS and 8000 series
used, but) and you could also get the data books from the chip manufacturers
(such as the National TTL data book I grabbed those number above from), and
these plus several buckets of coffee allowed you to understand what the
signals looked like, AND how the whole thing worked.

> I've tried google, but I either find stuff about electronics or about
>computer science, but nothing about the electronics of computer
>science, or if I did, it wasn't talking about this topic...

Well, yeah. Most generalist discussions don't speak about signal levels
or timing - you're told it's RS-232 or IEEE-488 or some such spec, and
that's the end of it. The reason is simple. You don't care what the specs
are any more. You aren't building the hardware that generates or receives
the signals (or if you are, you already know which spec sheet to read, and
have a copy) - you buy a part that does it for you. Look at the modern
motherboard - there might be a 3 to 10 large chips that do everything
except milk the cows. Then look at an IBM PC System Board (what IBM called
the motherboard originally) with... <counts chips> up to 97 small chips
and two dip switches. The PC-XT motherboard was even more complex, while
the original PC-AT board was 128 chips, and the Type 2 had only 103.

Old guy

Thanks a lot for your reply.

The reason for my "Ethernet oscilloscope" is not to gather specific
information about some network (or at least not yet).

I'm more interested in learning electronics related to networking.
That's my main goal.

I may also be interested in building applications with electronic
adapters to have some internet connection work from a printer port or
anything as funky as this (or as useless as it may appear to you).

Thanks a lot for the info, I'll try to get all I need that way.
Thanks,
Simon

(E-Mail Removed) wrote:
> Thanks a lot for your reply.
>
> The reason for my "Ethernet oscilloscope" is not to gather specific
> information about some network (or at least not yet).
>
> I'm more interested in learning electronics related to networking.
> That's my main goal.
>
> I may also be interested in building applications with electronic
> adapters to have some internet connection work from a printer port or
> anything as funky as this (or as useless as it may appear to you).
>
> Thanks a lot for the info, I'll try to get all I need that way.


To get the official Ethernet electrical specifications,
you have to buy the standard from IEEE. It is way over
1000 pages - not too cheap and tedious to read - have
a nice time with it.

The decoding of Ethernet signals from scope images is
very tedious, nearly impossible. Been there - done that.

The thing I'm using to look at Ethernet traffic is called
Ethereal - Google for it.

--

Tauno Voipio
tauno voipio (at) iki fi

On 22 Apr 2006, in the Usenet newsgroup comp.os.linux.networking, in article
<(E-Mail Removed) .com>, (E-Mail Removed)
wrote:

>The reason for my "Ethernet oscilloscope" is not to gather specific
>information about some network (or at least not yet).

An oscilloscope isn't going to provide useful information relating to the
network. The information is carried on a 10 (or 100 or 1000) MHz square
wave, using NRZ - meaning it's extremely difficult to read "by eye". In
fact, about the only usefulness of a scope on Ethernet is looking at
rising/falling edges for transition times, and time domain reflections.
I can say that with nearly fifty years of experience with scopes, and
twentythree years with Ethernet.

>I'm more interested in learning electronics related to networking.
>That's my main goal.

That covers every thing from RS-232 at 75 bits/second on up through
multi-gigabit fiber. There are LOTS of different things going on.

>I may also be interested in building applications with electronic
>adapters to have some internet connection work from a printer port or
>anything as funky as this (or as useless as it may appear to you).

-rw-rw-r-- 1 gferg ldp 46049 Apr 26 2001 PLIP

That's actually a mini-howto - IP over the printer port.

Old guy

Tauno Voipio wrote:

> To get the official Ethernet electrical specifications,
> you have to buy the standard from IEEE. It is way over
> 1000 pages - not too cheap and tedious to read - have
> a nice time with it.

It may be free now - not 100% sure on that.

Andy.