sas21-r03.pdf, Remove Simulation BER-15 Rj Note? Table 33, Note G.
Scot Baumgartner
scotb at us.ibm.com
Thu Aug 27 07:16:35 PDT 2009
Formatted message: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r0908270_f.htm">HTML-formatted message</a>
Hello.
I probably do not understand all the finer details, but the more I think
about this the more I think that the last sentence of Table 33, Note G is
not correct and should probably be removed.
i.e. Remove "For simulations based on a BER of 1E-15, the Rj specified is
17 times the Rj 1 sigma value" on Note G.
The TX device is defined to have a max BER-12 Rj peak-to-peak of 0.15 UI.
That would be 0.15/14.069 = 0.01066 UI Rj RMS.
The SAS StatEye simulation is done to BER-15, but I think should still use
the same TX max Rj. Which is 0.01066 UI Rj RMS.
So of course the BER-15 simulation would have a larger Rj peak-to-peak
value. 0.01066 * 15.883 = 0.1693 UI. As one would expect.
If the user followed Note G, then the TX max Rj would be decreased to
0.15/15.883 = 0.0094 Rj RMS for the simulation.
So I am thinking that the SAS StatEye code is correct with "Rj =
0.15/14.0", and it is the last sentence of Table 33 Note G that is
incorrect.
Scot Baumgartner. IBM Austin, Texas. 512-286-5273. 363-5273
Scot Baumgartner/Austin/IBM
08/03/2009 09:04 AM
To
t10 at t10.org
cc
Subject
Possible Typo in SAS-2.1 Rev02 19May2009. BER-15 Multiplier Should be
15.88 Instead of 17?
I noticed what I think is a typo in sas21-r02.pdf 19May2009.
pg 125. 5.8.4.6.1 Table 29. Note g.
"RJ is 14 times the RJ 1 sigma value, based on a BER of 1e-12. For
simulations based on a BER of 1e-15, the RJ specified is 17 times the RJ 1
sigma value."
I believe it should be "... on a BER of 1e-15, the RJ specified is 15.88
times the RJ 1 sigma value."
I think it should be 15.88 rather than 17.
Here is a link to one document that shows the Multipliers for different
BER values. See pg 3.
http://pdfserv.maxim-ic.com/arpdf/AppNotes/3hfan402.pdf
BER-12 14.069
BER-15 15.883
Scot Baumgartner. IBM Austin, Texas. 512-286-5273. 363-5273
More information about the T10
mailing list