David Gillett <[log in to unmask]> wrote:
> On 14 May 98 at 7:02, Russ Poffenberger wrote:
>
> > Putting a narrow device on a wide bus does NOT slow down the other
> > peripherals. This is a common misconception.
> >
> > Thje SCSI controller will negotiate the transfer rate individually
> > with each device.
>
>   Every nice simple rule has its exception, it seems.  If you connect
> a narrow device to a wide bus between the controller and a wide
> device, with a wide-narrow adapter that terminates the wide aprt of
> the bus, then you only have a narrow data path available from the
> controller to the wide device.  [This is not the scenario that Russ
> was responding to, and its fix involves rearranging devices rather
> than installing a second host adapter.]

The narrow device has to go at the physical end of the chain. The bus has to be
terminated anyway. This does NOT preclude the use of wide devices. Obviously, if
you place the narrow device before the wide an terminate the upper data bits,
then you have a problem.

The Adaptec web site has some information about the way to wire up these types
of scenarios.

I also saw a FAQ on the Adaptec site that dispells the other myth that placing a
non-Ultra SCSI device on an Ultra SCSI bus slows them all down.


--
Russ Poffenberger               Engineering Specialist
Schlumberger Technologies ATE   DOMAIN: [log in to unmask]
1601 Technology Drive           CIS:    72401,276
San Jose, Ca. 95110             Voice: (408)437-5254  FAX: (408)437-5246