22 Apr
2001
22 Apr
'01
2:51 p.m.
IN '85 Intel's 82586 and others were already quite available, though I'm not
sure how long they'd been around. AMD's 7990 was on a board I worked back in
'84, though it was not a production part. I have an AMD 7990 on the floor here
that's copyright 1985. Since the 7990 used a separate modulator, I guess it's
really not a single-chip device, however. By late '89 there was a British
company that made a part, of which I believe I was one of the earliest users,
which was a single-chip device (called ENZO and packaged in an 80-pin (?) pqfp)
tailored for the PC-market, with a host interface and the modulator and
everything in between on board. Only the COAX interface had to be external,
which required a couple of TTL/ECL translation stages and the reverse, in
addition to the usual transformer and coax driver (which always was a costly
component!).
Dick
----- Original Message -----
From: "Tony Duell" <ard(a)p850ug1.demon.co.uk>
To: <classiccmp(a)classiccmp.org>
Sent: Sunday, April 22, 2001 11:52 AM
Subject: Re: 3Mb Ethernet
As we all know, Ethernet was only running at 3 Mbit in the early days. When
was the move to 10 Mbit done and is 10Mb Ethernet backwards compatible?
AFAIK it is not backwards-compatible, at least in the sense that no
10Mbps ethernet controller supports the 3Mbps data rate (at least, I've
never seen one that does).
Also, were there any Ethernet controllers back in
the old days, or what kind
of interfacing did old Ethernet capable equipment such as the SUN 1 or DECNA
use?
There were no single-chip ethernet controllers, if that's what you're
asking. The ethernet circuitry for the classic PERQ (which was generally
10Mbps, but early enough that single-chip controllers didn't really
exist) is a 2910 sequencer, some microcode PROMs, some 9403 FIFOs, and a
lot of TTL glue logic (and some comparator-type parts for the interface
to the AUI connector). It takes up about 1/3rd of the EIO board.
-tony