On Mar 19 2005, 23:43, Jules Richardson wrote:
We couldn't get it to work with any transceiver we
tried (SQE test
was
off, FWIW), whether hooked up to a hub or not - it
still gave a
complaint about heartbeat at startup and then subsequent "no carrier"
errors (and the "Ethernet jammed" errors).
The complaint about "heartbeat" is because the interface is expecting
to see the SQE test appear a few bit times after the end of the frame.
"Ethernet jammed" might mean it's seeing a continuous jam signal, or at
least sees a jam signal when it listens to the wire to see if it can
transmit. A jam signal is what a repeater puts on the wire when it
detects a collision (and it would see the SQE test signal as a
collision). It's an alternating set of 1s and 0s that lasts 96 bit
times, as I recall (an ordinary station can also generate a jam signal
but not such a long one), and the object of that is to force everything
to see a collision and back off. Now it so happens that 96 bit times
is also the interframe gap length, so if you have a repeater with a
transceiver set to do SQE tests, it sends an SQE test starting a few
bit times into the IFG, then generates a jam which will still be there
slightly after the end of the IFG, just when the station might be
checking for carrier again. Have you got another repeater somewhere
with a transceiver on it? Or are you missing a terminator so that the
voltage is out of spec and might look like a collision?
As said in a private mail just now though, we'd
had this problem with
our PDP 11/84 - on the advice of an ex-DEC chap, it'd only talk to
one
of these DEC units rather than any kind of
AUI-equipped hub (or a
transceiver). On a whim we tried the same with the Sun, and it seems
to
have improved matters.
I've had various DEC machines, like my 11/83, attached to both
thickwire and thinwire with non-DEC transceivers and never had a
problem. I do check the SQE test, though :-)
--
Pete Peter Turnbull
Network Manager
University of York