[802.3ae_Serial] From Serial PMD 14 Aug: issues in characterising stressed eye
Updated link models are available from
http://www.ieee802.org/3/ae/public/adhoc/serial_pmd/documents/index.html :
see separate email.
Steve Buchheit reported some experimental difficulties in doing a stressed
sensitivity/jitter measurement.
A test setup:
Pattern gen -> modulator E/O -> pin -> TIA -> BERT showed apparent total
jitter of 37 ps out of 100 with 2^23-1 PRBS. It is important to find out
(how?) whether this is a feature of the transmitter or the receiver (e.g.
setup and hold time), as this will make a big difference to a stressed eye.
Then he used a long cable followed by a driver and laser to create a signal
with 81 ps apparent jitter (anywhere between 44 and 81 ps actual jitter),
and ~2.7 dB ISI, measured with DCA in wider bandwidth mode (i.e. without 7.5
GHz filter), and using a 111...101000...010111... pattern as defined in one
of the drafts.
Sensitivity measurements of a transponder at 10^-12 BER using "clean" eye
and jittered signal gave -20 and -7 dBm respectively!
Tom Lindsay was concerned that a short simple pattern may be needed for the
stressed eye vertical eye closure penalty (ISI) determination. The pattern
described above has, we think, been removed from the draft standard. Also,
in his experience, simple patterns show significantly less apparent jitter
than ones on the order of 2000 bits long.
Peter Öhlén had compared the 10^-5 point on a jitter bathtub with a scope
trace with 10^5 samples and found good agreement.
We should check out whether we get consistent results both for jitter and
ISI measurements with and without the 7.5 GHz filters. Steve is well
positioned to try this, but others should too.
More generally, it would be great if Steve could report his findings with
visible diagrams, even if informally, at the Copenhagen meeting.
Vipul Bhatt, representing the 1310 serial PMD feasibility ad hoc, asked for
advice on using the link model as an aid in their experimental work. To be
progressed off line and included in next week's agenda.
Next phone conference
---------------------
Usual coordinates:
15:15 GMT = 4:15 pm BST = 17:15 CET = 11:15 am EDT = 8:15 am PDT,
Tuesday
+1(816)650-0631 Access code 39209
Piers