| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Bryan - thanks for your effort and thoughts. The reason we went for a test with an injected noise source on 10GBASE-T (and I would recommend something similar
here) was: 1.
We wanted the test to be repeatable, and experience showed that coupling crosstalk from live cables
did not result in repeatable levels from setup to setup (at least not without a lot of work) 2.
We wanted the test to be a receiver quality test, separating out the cabling characteristics. 1000BASE-T1 may be different in that it is likely used as an engineered (cable +
phy) system, which may make the second reason more debatable, but I believe the first reason still stands. PHY designers and system evaluators will want tests to be repeatable from one users configuration to the other. (FYI – I have told Steve that I will be in the 802.3bp room after the break (around 10am) this morning for a
little while) George Zimmerman Principal, CME Consulting Experts in Advanced
PHYsical Communications Technology george@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx 310-920-3860 From: Moffitt,
Bryan [mailto:BMoffitt@xxxxxxxxxxxxx] On George Zimmerman's comment #288 97.5.4.2.1, my expectation of the bit error rate test is that it is performed in a defined alien test configuration so text as follows is suggested:
Differential signals received at the MDI that were transmitted from a remote transmitter and have passed through a link specified in 97.5.5 with alien crosstalk from other bundled links as outlined in Figures
97B-2 and 97B-3 are received with a BER... However it may also need to include something about Table 97–13, and I'm not sure how that would fit in. Bryan Moffitt CommScope Systems Engineering |