Glitch Filter

Tak Asami asami at
Fri May 9 17:56:14 PDT 1997

* From the SCSI Reflector (scsi at, posted by:
* Tak Asami <asami at>
Regarding the proposal on glitch filter:

Let me make a statement here...

In terms of (a) delay and (b) duration of reflections, the tranfer rate
has nothing to do with them.  They have everything to do with device to
device distances and stub lengths (a.k.a. configuration rules), as well 
as electrical characeristics of cable itself (whose attenuation 
characteristics are not specified in the standard).

I think the proposed glitch filter settings are primarily focused on 
how precisely you can control the edges give a transfer rate, and really,
very little to do with how effectively remove the glitches.

If that's the case, is it not a crime to specify these filter parameters
giving it a notion that if you do this, you will be immune from this
type of glitches?  Why bother specifying certain class of filtering at
all?  I think there are effective means of glitch filtering which does
not necessarily conform to "so many nanoseconds from the edges" type
specification.  Are we not killing the innovation?

In other words, do we want to design a filter to actually prevent the
glitches from bothering you?  Or is it just to conform to the standard?

Tak Asami ==============================================================
Adaptec / ITC
9701 Jeronimo Rd, Irvine CA 92618
(714) 455-8202 / Fax: (714) 455-8102

asami at
* For SCSI Reflector information, send a message with
* 'info scsi' (no quotes) in the message body to majordomo at

More information about the T10 mailing list