Our setup is in fact HVD, going to a Tape Library. Tape libraries have up
until last year most commonly used the high voltage interface because they
were quite often physically separated by many feet from the hosts that used
them.
To summarize the elements: The Adaptec AHA-3944 is an HVD controller with
two VHCDI connectors on the back (yes, I do realize that this is quite a
rare connector type for HVD controllers, but it is what is there). The
VHCDI are cabled by a VHCDI and 68-pin Universal SCSI cable. In theory,
Universal SCSI should work with HVD, but I'm like you skeptical of the cable
at this point. We have thought about buying a VHCDI to 68 pin adapter and
attaching that to an old but thick 28 gauge wire HVD cable, and seeing if we
notice any difference.
The tape drives are in fact HVD SCSI, not LVD. I'm well aware that you can
damage device and/or controller by mixing HVD on either end with non-HVD.
The tape drives set their SCSI ID by a selector on the back of the tape
drive.
We tried a variation on your experiment already: we cabled directly to the
tape drive, and terminated with HVD terminators at the device. We then set
different SCSI IDs on the tape drive. As soon as we hit 8 or above the
device disappears from the chain. The Adaptec BIOS seems to suggest that
it can support 15 devices. The Adaptec BIOS does not see the device either
when you scan devices, if the tape drive is set to 8 or higher. So it is
not an OS issue.
--
Will
NOTE: To reply, CHANGE the username to westes AT earthbroadcast.com
"Overlord" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> Hi Will,
> I'm going to make some assumptions here.
> I assume you are jumpering Wide SCSI drives on a 68 or 80 pin cable
> and know what you are doing by jumpering at least the 4th set of ID
> jumper pins (ID8) in addition to whatever other jumpers are set.
> I assume that each drive has a unique ID which none of the other
> drives has.
> I assume that you have the bus segment terminated correctly.
> I assume you have gone into the SCSI card bios (ctrl A at the banner)
> and used the controller bios to try and detect the drives.
>
> You haven't really posted your configuration but.... try a single
> drive with all the ID jumpers jumped (ID15) on your Wide cable.
> (that is the 68pin cable - the 50pin cable is physically wider than
> the 68pin cable but is not the Wide cable) and see if the 3944 bios
> can detect it. Also, the card will not load the bios if there are no
> bootable drives.
>
> All things being equal.... I would suspect the cable.
>
> However...you haven't given the entire model Adaptec SCSI controller.
> The specs I bring up seem to point to the 3944 being a Differential
> SCSI controller. Differential = HVD. HVD = High Voltage Differential
> as opposed to the standard LVD drives.
> HVD drives are used in electrically noisy industrial settings and a
> standard plain vanilla LVD or SE or 50pin narrow SCSI drive will NOT
> work on the HVD controller. And quite possibly never work again on
> anything if it's been hooked to a HVD controller.
> Is the actual model number of the card something like: AHA-3944AUWD?
>
>
> On Thu, 4 Sep 2003 17:40:49 -0700, "CHANGE username to westes"
> <(E-Mail Removed)> wrote:
>
> >We have an Adaptec 3944 host adapter in a Compaq 1850R, and that adapter
> >doesn't seem to be able to deal with any SCSI ID greater than 7, even
though
> >the BIOS suggests it can see 15 IDs per channel. What would cause the
> >adapter to not see SCSI IDs 8 or higher?
> >
> >--
> >Will
> ~~~~~~
> Bait for spammers:
> root@localhost
> postmaster@localhost
> admin@localhost
> abuse@localhost
> postmaster@[127.0.0.1]
> ~~~~~~
> Remove "spamless" to email me.
|
Similar Threads
