Question on SJ frequency in SAS-1.1r6 Table 32

Bill Ham bill_ham at ix.netcom.com
Wed Oct 27 03:41:55 PDT 2004


* From the T10 Reflector (t10 at t10.org), posted by:
* Bill Ham <bill_ham at ix.netcom.com>
*
This is a multi-part message in MIME format.
--------------1101ADA10B367D3ACB8A81A5
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi,

Following is my understanding of these requirements.

This requirement refers the responses of the receiver to applied SJ during a
signal tolerance measurement.  This measurement reports the detected bit
error ratio out of the link receiver.  The assumption for the upper limit is
that the receiver will not be able to track applied SJ above the upper limit
frequency.  So whatever the receiver does at the upper limit should pretty
well represent what it does at higher frequencies.

If you have a part that behaves differently above the upper limit then that
part is probably able to track SJ above the upper frequency and is better
than required.  You get no credit if this happens but the part probably has
extra margin.

Note that this requirement only applies to the APPLIED SJ and not to the
frequency content of the rest of the jitter spectrum in the signal used for
the signal tolerance measurement.

Either interpretation that you state should yield the same result in terms
of passing or failing the requirement.

Cheers, Bill

Bill Lye wrote:

> * From the T10 Reflector (t10 at t10.org), posted by:
> * Bill Lye <lye at pmc-sierra.com>
> *
> Working from sas1r06.pdf, in table 32 (page 109) there are two notes,
> (c) and (d) specifying sinusoidal jitter frequency.  One of my fellow
> designers is reading these notes and is finding them ambiguous and would
> like some clarification as to the intent.
>
> Currently the notes read:
>
>   (c) Sinusoidal swept frequency: 900 kHz to >5MHz
>   (d) Sinusoidal swept frequency: 1800 kHz to >5MHz
>
> Now, this can be read in one of two ways, one of which implies that the
> sinusoidal swept frequency range can be terminated at 5MHz (or 5MHz plus
> epsilon), while the other implies it could go all the way to infinity.
> I believe the first interpretation is the correct one, however I was
> not involved at the time these notes were created, so I would like
> confirmation if at all possible.
>
> Thanks in advance...
>
> -Bill
> --
> Bill Lye                                         PMC-Sierra, Inc.
> Technical Advisor, Mixed-Signal Development      8555 Baxter Place
> lye at pmc-sierra.com                               Burnaby, BC
> (604) 415-6000 x2522                             Canada V5A 4V7
> *
> * For T10 Reflector information, send a message with
> * 'info t10' (no quotes) in the message body to majordomo at t10.org


--------------1101ADA10B367D3ACB8A81A5
Content-Type: text/x-vcard; charset=us-ascii;
 name="bill_ham.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Bill Ham
Content-Disposition: attachment;
 filename="bill_ham.vcf"

begin:vcard 
n:Ham;Bill
tel;pager:978 828-9102
tel;cell:978 828-9102
tel;fax:978 470-0321
tel;home:978 475-6850
tel;work:978 828-9102
x-mozilla-html:FALSE
adr:;;92 Wildwood Road;Andover;MA;01810;
version:2.1
email;internet:bill_ham at ix.netcom.com
note;quoted-printable:Please use 978 475-6850 (home/home office)=0D=0A if cell does not work
x-mozilla-cpt:;-5168
fn:Bill Ham
end:vcard

--------------1101ADA10B367D3ACB8A81A5--




More information about the T10 mailing list