2 m UTP CAT 5 w. 2^7-1 PRBS
Hi all
Just another example. I don't know how to post these results, in order not
to over-fill the reflector. If you think it should be put to another
relevant forum, or placed on the server - please comment.
This was just to show how rude one can be and still get signals transmitted
to the CDR. It is only a result from my investigations on making a
sufficiently distorted test signal (ISI - DJ, and loss) for GE compliance
far-end eye template test of our XAUI CDR characterisation.
I guess, that a decent PCB layouter on FR-4, can do a lot better job than
what is expected from the worst case compliance channel (I think 2 meter
cable is worse than 50 cm on FR-4 + PCB backplane connectors). It seems to
be a huge over-specification in the 10GE compliance channel. Alright, I have
not included connectors or cross talk impairments, but I will continue my
investigations.
I use our 12.5 Gb/s HP test-set running at 3.125 Gb/s. Using differential
signaling with 2.5 GHz SDH filters (4'th order Bessel Thomson) driving a
pair of the 2 meter UTP 24 AGW * 4 pair. (CMG E1644469 - ISO/IEC 11801 & EN
50173, TIA/EIA 568A 3P verified it says on the jacket). I cut off the
connectors and soldered the wire pair to SMA connectors. This cable has a
fairly nice 50 ohm impedance (between 45 and 55 - I measured using
differential TDR measurements. 78 ohms for single ended TDR).Also no
cross-talk is included.
I deliberately took this eye diagram with a 2*150 mVpp and 2^7-1 PRBS in
order for you to see and resolve the pattern effects (ISI DJ just about 100
ps pp).
The differential inner eye opening is about 85 mVpp.
The GIGA CDR will run error free even with just 20 mVpp differential inner
eye opening (2*40 mVpp driving amplitude). When the pattern is increased to
2^31-1 it still needs about 20 mVpp, hovever the driver amplitude has to be
increased about 20% due to the spectral frequency components emerging at
much lower frequencies.
Stay tuned. I will do some additional SI jitter tolerance measurements at
different eye opening amplitudes.
Benny
--------------------------------------------
llllllll ii llllllll llllll
ll ll ll ll
ll llll ll ll llll llllllllll
ll ll ll ll ll ll ll
llllllll ll llllllll ll ll
GIGA, an Intel company
Benny Christensen, M.Sc.E.E, Ph.D.
Mileparken 22, DK-2740 Skovlunde, Denmark
Tel: +45 7010 1062, Fax: +45 7010 1063
e-mail: benny.christensen@xxxxxxxxx, http://www.giga.dk
Sdhutp7.ZIP