[802.3ae_Serial] noise penalty in stressed receiver sensitivity
Hello,
Happy New Year. I hope all of you had great holidays. I have a question
regarding the definitions of stressed receiver sensitivity and TP3
jitter compliance in D4.0:
Receiver jitter tolerance is specified at TP3 using the bathtub mask
(52.8.2.) That mask specifies the maximum jitter (including sinusoidal
jitter) that the receiver (CDR) has to be able to tolerate with better
than "design" BER. The center of the eye OMA that is required at the
input of such a stressed receiver is the "stressed receiver sensitivity"
(52.9.11).
Meanwhile, in the 3.1.16a version of the link model, stressed receiver
sensitivity is defined as the sum: Nominal_RX_Sens + ISI + DJ +
baseline_wander penalties (source noise is not included). If the above
is correct then I am confused about the following:
The random jitter component measured in the bathtub plot at TP3 (52.8.2)
originates from (a) input random jitter, and (b) the contributed
source/fiber noises (RIN, MPN, MN). The presence of (b) means that the
measured receiver sensitivity at the center of a TP3 compliant signal
WILL be penalized by the presence of RIN, MPN, MN. This is not
consistent with the definition of stressed receiver sensitivity used in
the link model that does not include the laser/fiber noise. Am I
missing something?
Thanks
Dubravko Babic
Alvesta Corporation
500 Oakmead Parkway
Sunnyvale, CA 94085
tel: 408-331-4858
fax: 408-732-4942