RE: modal noise, PMD power penalty
Vipul,
To question 1:
This response may be just a clarification, not a disagreement. I believe
that strictly speaking, the two phrases are related but not the same. Modal
noise is caused by the time variance of the channel (which can be observed
as changes to the speckle pattern at the end of the fiber) passing thru a
mode-selective loss. Without the loss, there is no modal noise generated
even though there is time variance.
Paul
----------
From: Vipul Bhatt [SMTP:vipul.bhatt@xxxxxxxxxxx]
Sent: Thursday, November 02, 2000 12:58 PM
To: Equalization Ad-Hoc Reflector
Subject: modal noise, PMD power penalty
Importance: High
For those who won't be able to attend the Sunday meeting, but feel
qualified to answer the following two questions... please attempt a
reply before Saturday evening. These two questions are unrelated to
each other.
1. I don't see any difference between the phrases "time variance of
a
multimode fiber as a channel" and "modal noise". Do you agree? I am
assuming that the effect of laser is implicitly accounted for. If
you
agree, isn't the wheel already invented - don't we already know that
modal noise is a slow variation phenomenon? [If there is no mode
partition noise - and this is where the answer may be different for
10G VCSEL and DFB lasers - refractive index changes due to
mechanical
stress and local temperature variations will occur on a time scale
representative of the strain mechanisms, i.e., on the order of
milliseconds or longer. Paraphrased from Lasky, Osterberg and
Stigliani, Optoelectronics for Data Communication. ISBN
0-12-437160-4.]
2. Is it true that PMD Power Penalty, if accounted for in the link
power budget, will make it unnecessary to account for horizontal eye
closure? The converse may not be true - compared to a zero-PMD link,
a
link with PMD will see the receiver eye suffer from both vertical
and
horizontal eye closure. But if more power is thrown at the link, the
eye will open up both vertically and horizontally. Do you agree? [I
don't want to recommend something to P802.3ae that falsely includes
a
double penalty.]
Thanks,
Vipul
vipul.bhatt@xxxxxxxxxxx
(408)542-4113