Comments on Steve Finch's "Comments on Ultra SCSI"

Michael Smith msmith at dalsemi.com
Mon Feb 14 13:23:02 PST 1994


  Steve Finch recently wrote several comments on Ultra-SCSI, but the 
section that caught my eye was the following:


>One solution is to scale the 5ns slew rate requirements down to 2.5 
ns.  
>We should be able to control the maximum slew to around 7.5ns.  But 
if 
>we do this, what happens to the signal quality?  Wasn't this the 
reason 
>for 5ns slew originally?
>

  I agree that signal quality would definately suffer as the slew rate 
is increased. As a terminator manufacturer, I have been able to mostly 
ignore special precautions for large di/dt because of the 5nsec slew 
limit (and 24mA max current). However, recently some of the controller 
silicon has exceeded the 5nsec slew limit, causing unexpected drooping 
and ringing in the signal, due in part to parasitic inductance effects 
that can be ignored in a spec-compliant situation. Most of these 
manufacturers are reworking their drivers to slow them down.

  A "fast" slew rate signal can be terminated properly by special 
design precautions. For example, Backplane Transceiver Logic (BTL) 
slews 41mA in 2nsec; in a 128 line Futurebus+ system, we're talking 
more than 5 Amps! And BTL noise margins are much tighter than TTL. The 
transciever and terminator designs are tough (and expensive), but can 
be designed.

  The upshot is that faster slew rates are possible, but very painful 
to implement.

  By the way, Steve was not advocating faster slew, but I wanted to 
get my $0.02 in before somebody does.

Michael Smith
Dallas Semiconductor
msmith at dalsemi.com





More information about the T10 mailing list