[Tfug] SCSI disks

John Karns johnkarns at gmail.com
Tue Aug 21 12:04:07 MST 2007


On 8/20/07, Bexley Hall <bexley401 at yahoo.com> wrote:
>
> --- John Karns <johnkarns at gmail.com> wrote:
>
> > On 8/16/07, Bexley Hall <bexley401 at yahoo.com> wrote:
> > > However, the LX uses "50 pin (SE) SCSI".  I've
> > > never seen a 50 pin drive larger than 4G (but
> > > I suspect they exist -- or existED!).  In any
> > > case, I don't *have* such a device.
> > >
> > > *But*, I can probably find a *wide* SCSI drive
> > > (non-SCA) of a larger size.
> >
> > > But, is this as "simple" as getting a cable
> > > with a 50 pin berg connector on one end and
> > > a 68 pin wide scsi connector on the other?
> >
> > Yes, I've done exactly that, connecting a 68-pin
> > wide drive to an
> > Adaptec 50-pin 1460 slim SCSI pcmcia card to read /
> > write the drive.
>
> Ah, OK.  But, how did you handle the termination on
> the drive end of the cable?  ("Termination??  *What*
> termination???  :>  )

In the case I mentioned with the 50-pin pcmcia card connected to a
68-pin hd, I probably had the hd jumper set for termination on the
drive.  In any event, it was just a temporary connection, so I didn't
consider it critical to have proper termination.  Of course it's
always a good idea to verify the data after copying.

FWIW, the shorter the cable, the better.  In cases where the cable is
short, and there is only one device connected to the cable,
termination can often be left off, particularly in situations where
the setup is temporary.


> I have lots of IDC (e.g., Berg-style) connectors.
> But, *my* problem is the cables that I have are
>
> 68W --------------- 50SE ----- terminator
>
> I.e., they are expecting the "host" to be located
> at the 68 pin end (with its on-board termination)
> then, the narrow, SE drive on the 50 pin connector.
> Finally, the terminator dangling off the end to
> terminate the *drive* end of the cable.
>
> I would cut off the terminator in my application
> since that would be the "host" end of the cable
> (and the host is already terminated).  *But*,
> there is no way to "extend" the 68 pin end to tack
> a terminator on there for the *drive* end of the
> cable!  So, I would have to count on the drive
> having termiation onboard, etc.  (I haven't looked
> into this, yet)

Since you said that the Sparc uses 50-pin SCSI, I assumed that:

1) the controller was 50-pin
2) the cable was also 50-pin
3) the drive was non-SCA

I'm not sure I'm clear about what you stated about the termination is
on your setup.  AFAIK, there should be termination on both ends of the
cable (SCSI host and hd end, as last device connected).

For clarity, I'll add that neither end should have multiple
termination - e.g., if the drive is jumpered for termination, there
should be no other termination device at that end of the cable.

In short, if you have a 50-pin, terminated controller with a 50-pin
cable attached, and want to attach a 68-pin hd, all you should need to
do is connect it with an adaptor and set the jumper on the drive for
termination.  If the hd has no such jumper, then attaching a 50-pin
terminator to the end of the cable should do the trick.

-- 
John




More information about the tfug mailing list