Apple SCSI TermPWR; Was Re: This is funny (ebay)

From: Doc <doc_at_mdrconsult.com>
Date: Sun Feb 24 09:09:32 2002

On Sun, 24 Feb 2002, Richard Erlacher wrote:

> Is it possible you're mixing two things up? Most of the later SCSI drives
> from SEAGATE permitted TERMPWR to come from any device. The standard suggests
> that any/all devices can be providers of Termpwr, since only one will have the
> highest voltage, hence be the source, as it supposed to be diode-isolated from
> the SCSI bus. To implement this option, they provided two jumpers, typically,
> which allowed for four options. (1) Termpwr provided by the drive, and to its
> terminators from its own power supply, (2) Termpwr provided by the drive, and
> to the bus, but its terminators powered fom the SCSI bus, (3) Termpwr provided
> by the bus to its internal terminators, and (4) isolation from Termpwr, but
> its terminations powered from its own supply. This did not affect whether or
> not internal termination was enabled or disabled. That was a separate choice,
> handled by a separate option.

  If I understand correctly, you're asking if I know (and implemented)
the diffference between Term Power and Termination. Yes. Typically, I
supply termPWR from the terminating device, if possible. Easier for a
feeble mind to track.

> Since the Apple SCSI doesn''t provide Termpwr, and since Termpwr is supposed
> to be diode-isolated (source can drive the bus, but bus can't drive the
> source) from the bus, any or all devices are permitted to drive Termpwr. Only
> one will actually drive it until the voltage is pulled down to where another
> might do so. It's not risky to have more than one device drive the Termpwr
> line, though it's safest to know which is driving it and make sure it's
> enabled there. If your system consists of a computer and a varying set of
> external devices, it's easiest just to enable Termpwr at the host adapter and
> be done with it. Since that's not an option with the Apple devices, you must
> have it supplied by some device that's usually attached to the SCSI bus. Just
> pick one that's normally there, and be prepared to enable it on some other
> device if it isn't provided otherwise.

  Right. If you look at the list of computers I've had trouble with,
they're all Apples with a single internal device. I don't know if it
makes any difference, but all the Hawk drives I've got came from Sun.
All have internal terminators, not R-packs. I have three or four
around. The one that's in the IIci is an ST31200N. It works just fine
if I power termination from the drive and use a terminated cable. The
Mac can't see it at all if I power termination from the drive and enable
the drive's terminators.

> In general, it's easy simply to stick an external terminator on the ends of
> the SCSI chain. Termination has no effect on addressing or on Termpwr, so the
> only consideration is its location. If you seldom use more than one external
> device, or if you ALWAYS use a specific one, it's reasonable to install/enable
> the terminations in that device and install other SCSI devices between the
> host adapter and that device, which then saves you the
> installation/moving/easter-egg-hunt-for the external terminator.

  I'm quite familiar with the easter-egg-hunt.... Whenever I spend more
than 5 minutes looking for a terminator, external or internal, I buy one
next time I see one. At GoodWill, if possible. That means I have a
bunch of spares, stashed in suitably surprising locations.
  I also have to admit I'm real lazy about this. If I use cable
termination. I don't have to remember the jumper layout of all the SCSI
drive models I have. I just use a terminated cable if possible and plug
stuff in.

        Doc
Received on Sun Feb 24 2002 - 09:09:32 GMT