Reminder: SAS PHY teleconference, 6/21, 10 am CDT

Alvin.Cox at Alvin.Cox at
Wed Jun 20 14:19:19 PDT 2007

Formatted message: <A HREF="r0706203_f.htm">HTML-formatted message</A>

Toll Free Dial in Number: (877)810-9442 
International Access/Caller Paid Dial In Number: (636)651-3190 
Webex information: 
Topic: SAS-2 PHY WG 
Date: Thursday 
Time: 10:00 am, Central Daylight Time (GMT -05:00, Chicago) 
Meeting number: 826 515 680 
Meeting password: 6gbpsSAS 
1.  CJTPAT versus JTPAT
Should we stick with CJTPAT as the required pattern and not allow JTPAT as 
an equivalent?
Discussion leaned toward CJTPAT being a more stringent test since it 
includes a wider spectral content.
Question to all:
Do the header and CRC need to be valid? If these are generated by a tester 
that just makes up fake data, is that acceptable for the test?
2. Review of Annex B since de-embedding has a significant impact at 6Gbps.
3. JTF
We accepted the concept but have not discussed the implementation. No 
responses have been voiced concerning the Synthesis Research proposal to 
help close the measurement accuracy.
Does the use of this filtering give an equivalant value to using the 
single pole f/1667 if SSC is not applied? This question is in regards to 
the text in the eye diagram section. I have been reviewing how to 
incorporate 6G and think it may be best to integrate rather than make a 
totally new section.
Test equipment question:
Comments from Bent Hessen-Schmidt, SyntheSys Research, Inc.: 
I suggest that we use text equivalent to: 
The Reference Clock characteristics are controlled by the resulting JTF 
(Jitter Transfer Function) characteristics obtained by taking the time 
difference between the PLL output (the Reference Clock) and the data 
stream sourced to the PLL. The PLL CLTF -3 dB corner frequency, and other 
adjustable CLTF parameters such a peaking, are determined by the value 
required to meet the requirements of the JTF. 
The JTF shall have the following characteristics for an encoded D24.3 
pattern (1100110011 0011001100). This is the MFTP which is a test pattern 
that has clock-like characteristics and a transition density of 0.5. 
1) The -3 dB corner frequency of the JTF shall be 3 MHz +/-1 MHz. 
2) The magnitude peaking of the JTF shall be 3.5 dB maximum. 
3) The attenuation at 30 KHz +/-1% shall be 75 dB +/-3 dB. 
The JTF -3dB corner frequency and the magnitude peaking requirements shall 
be measured with sinusoidal PJ applied, with a peak-to-peak amplitude of 
0.3 UI +/-10%. The relative attenuation at 30 KHz shall be measured with 
sinusoidal phase (time) modulation applied, with a peak-to-peak amplitude 
of 20.8 ns +/-10%. 
You will see that we have changed to from 72 db to 75 dB and from 2.1 MHz 
to 3 MHz and added the word “relative” to the last sentence. Relative 
should indicate that the 75 dB are with respect to the actual magnitude of 
jitter on the 30 kHz stimulus. The +/-10% therefore merely sets the 
starting point and still allows the other vendor. All uncertainties are 
then included in the +/-3 dB term. Effective tightening of the tolerances 
can be seen on the residual of the 30 kHz being confined to less than 5.2 
ps instead of (7.4 x 1.1 x 1.1 ps = 8.95 ps). 
4.  SAS-2 6Gbps PHY Electrical Specification
Rev 8 to be posted late 6/20 to just before the call. Will have text from 
SAS-2 r10 with comments concerning incorporation of 6G in several areas.
Alvin Cox
Seagate Technology, LLC
Office 405-381-8067
Cell 405-206-4809
E-Mail	alvin.cox at

More information about the T10 mailing list