[Info-vax] DEC Multia (UDB) issues

[glen herrmannsfeldt]

Richard B. Gilbert <rgilbert88 at comcast.net> wrote:
(snip)
>>> Generally speaking: Are SCSI failures common? I recently tried to

>> (snip)

> And remember that BOTH ENDS of the bus must be terminated!

> If the controller is at one end of the bus, it supplies termination; a 
> switch or jumper enables termination at the controller.  If not, be 
> certain that BOTH ends are terminated.

> Also be certain that the bus is terminated exactly twice; once at each 
> end!  Zero will not work!  One will not work.  Three will not
> work. . . !

Reasonably often one or three will work.  Years ago with a VME
Sun 3/260 the terminator was so far in I couldn't get to it.
There was a connector to add on, though, and it worked fine.
That is, three terminators.  That was SCSI I, though.

With only one, any reflections off the unterminated end will
be absorbed by the terminated end.  Reasonably chance with
SCSI I, less with later ones.

Some Apple computers drive the bus without any buffers, unlike
the way they are supposed to be done.  The drivers aren't strong
enough for an overterminated bus.  With strong enough drivers,
it works fine.

-- glen