[SunRescue] sun3/80 scsi

[Greg A. Woods]

[ On Friday, December 29, 2000 at 11:02:23 (-0500), jeff borisch wrote: ]
> Subject: Re: [SunRescue] sun3/80 scsi
>
> When you say "system" I think of the mainboard or the scsi controller.
> I have only seen jumpers on drives to supply TERMPWR to the bus. But from
> reading the SCSI FAQ, I learned the initiator (the mainboard SCSI controller
> in the 3/80s case) also is supposed to apply voltage to this line.

Yes, I did intend to imply the SCSI host adapter/controller on the
mainboard when I used the word "system".

Since the host adapter is (almost) always an initiator, and so according
to the standard must always supply TERMPWR, there's no jumper to turn it
off (though usually there's a fuse to prevent it from supplying too much
current, and if the fuse is easily removable it's effectively a jumper
too! :-)
 
Drives usually have an option to supply TERMPWR (and almost always an
option for internal termination too) because they might be at/near the
end of the bus where a passive terminator may end up pulling TERMPWR
down too far, and thus out of spec, if it's only being supplied from the
other end of a long bus.

> Is it common to have devices that supply mismatched TERMPWR?

I don't think so.  In my experience it was only very early Sun systems
that on occasion made the mistake of connecting the TERMPWR line to
ground instead of to +5 Vdc.  Some systems that straddled the SASI to
SCSI transition might have had issues too.  Of course I didn't really
start working with SCSI much until SCSI-2 was well entrenched, and given
its near "plug&play" success I've never paid too much attention to
specific implementations unless I've had problems.  I have seen several
ribbon cables with cond#26 burned to a crisp though, and at least one
was explicitly due to a Sun 3/60 (the others were always due to
fumble-fingered people playing with "hot" un-mounted equipment and
accidentally grounding a part or jumper-pin on the drive controller
board).  I learned about the 3/60 issue before I saw the results from
someone else who hadn't learned it though....

> I have never
> seen the voltage or polarity for TERMPWR listed on any drive specsheet.

The X3T9.2-v10l (7-SEP-93) draft of the SCSI-2 standard says:

    Single-ended SCSI devices providing terminator power on cable B shall have
    the following characteristics:
    VTerm        = 4,5 V d.c. to 5,25 V d.c.
                   1500 mA minimum source drive capability

The implication is that since they didn't prefix those values with a '-'
that they mean a positive (and indeed it is essentially standard
practice these days to consider ground to be equivalent of the negative
supply).

-- 
							Greg A. Woods

+1 416 218-0098      VE3TCP      <gwoods at acm.org>      <robohack!woods>
Planix, Inc. <woods at planix.com>; Secrets of the Weird <woods at weird.com>