Challenges for testing 10GBASE-T

Transcription

1 Challenges for testing 10GBASE-T Adrian Young Customer Support Manager

2 Check list Category 6A component compliant connector Category 6A compliant cable Perfect terminations Method of protecting cable from other contractors Bundles of 12 or 24 cables Well designed path ways avoiding sharp turns Level IIIe (500 MHz) Accuracy field tester DSP-4x00, OMNIScanner2, WireScope 350, LANTEK 6 Permanent link test lead with Category 6A compliant RJ45 plug Alien Crosstalk Adapters compliant to IEC & TIA/EIA-568-B.2-10 Batteries charged? Technician trained on how to test 2

3 Channel Length Limitation for Cat 6 / Class E TIA: TSB-155 (Telecommunications Systems Bulletin) Support for 10GBASE-T over limited channel length should operate for channel length 37 m should operate for channel length of 37 to 55 m, depending on AXTalk test results may operate over channels > 55 m after mitigation ISO/IEC: TR24750 (Technical Report) Similar limitation for Class E as outlined in TSB-155 for Cat 6 Note: To test is to know: The Alien Crosstalk performance should be (sample-) tested for all cabling links 3

4 Permanent Link test lead Ever swapped your test leads around and got a different result? It can t be the tester, it s the same tester you are using May be it s the test lead.. well kind off.. It s that RJ45 plug on the end of the lead that is causing it to change. The standard states:? The plug is terminated by hand The variation of performance can be wide spread. So the TIA and IEC define an acceptable range for RJ45 plugs 4

5 Importance of the RJ45 test plug- If the performance of your plug changes, so does your NEXT margin It s the placing of the twisted pairs that affects it s performance It is a real struggle to create a test plug One solution remove the twisted pairs in the RJ45 plug 5

6 Importance of the RJ45 test plug- If the performance of your plug changes, so does your NEXT margin Removing the human element increases accuracy/consistency Also allows the plug to be tested without destroying it Typically yields a better Return Loss margin as well 6

7 Links less than 15 m Before installing links less than 15 m, check with your cabling vendor to see if they are able to support those links TIA does not mention a minimum distance directly (well kind of <10 m) ISO/IEC PDAM 1.1 does mention a minimum distance 7

8 Step 2 Alien Crosstalk Testing Not considered mandatory in the standard Methods are well established and defined in TIA and IEC Selecting which links to test is the biggest challenge Victim cable - select 1 % or 5 links which ever is the greatest (longest links) Disturbers - choose the cables in the same bundle as the victim cable ANEXT comes mainly from the panel, select links above and below 8

9 PS ANEXT Power Sum Alien Near End Crosstalk 9

10 Example PS ANEXT measurement 10

11 PS AACR-F Power Sum Alien Attenuation Crosstalk Ratio Far end 11

12 Why do it this way Voltages are for quantative purposes only Why not just energize all the disturbers? Normalization Let s look at the PS AACR-F measurement Standard requires an adjustment of the disturber voltage Instrument Rx V Tx Instrument 1.63 V The Rx of the signal generators clearly overwhelm the receiver. In the real world, this would cause the PS AACR-F measurement to FAIL. Solution Make the signal generators adjust their voltage depending on the length of the link they are connected to, also known as Normalization. Signal Generators 12

13 Why do it this way TIA/EIA 568-B.2-10 Adjustment is only applied when the disturbed permanent link or channel is longer than the disturbing permanent link or channel. This adjustment is referred to as AFEXT loss normalization. Instrument Rx 1.63 V Tx Instrument 0.3 V How do the signal generators know the length of the disturbed cable? Signal Generators 13

14 14

15 15

16 Troubleshooting Alien Crosstalk If you are unfortunate to see this: 16

17 Disturber Ranking Patch panel or cable issue? It s grouped around our victim link This is a Patch Panel issue 17

18 Disturber Ranking Patch panel or cable issue? It s spread out around our victim link This is a Cable issue 18

19 Mitigation for worst Between-Channel Performance 1. Reduce the alien crosstalk coupling by separating the equipment cords and the patch cords and un-bundling the horizontal cabling 2. Alternative to (1): utilize equipment cords specified to mitigate the alien crosstalk coupling such as Category 6 ScTP or Augmented Category 6 3. Reconfigure the cross-connect as an interconnect 4. Replace connectors with Augmented Category 6 connectors 5. Replace the horizontal cable with augmented Category 6 After applying one of the mitigations, re-test See whether a noticeable improvement has been achieved in the worst case margin of the offending test parameter 19

20 10GBASE-T is not Category 6A TSB-155/TR24750 Ch = 10GBASE-T TSB-155 PL = 10GBASE-T 20

21 PS AACR-F fails the Pair-to-Pair and the Avg-Pair test The Overall Test results shows Pass based on the outcome of the Alien Crosstalk Margin Calculation (ACMC) the final criterion for 10GBASE-T 21

22 The Result Detail Note: PS AACR-F fails Alien Crosstalk Average Margin for all wire pairs and the average is positive, the victim link PASSes the 10GBASE-T requirements 22

23 Summary 10GBASE-T Done. June 2006 TIA TSB-155 Done Feb 2007 TR24750 Done Aug 2007 TIA/EIA-568-B.2-10 Category 6A Draft 9.0 Ballot (no technical changes) Test plug performance defined Channel and Permanent Link limits set (no changes for over a year) Alien Crosstalk methods defined You can feel safe testing to this standard TODAY ISO/IEC PDAM 1.1 Channel values only No component performance specified ISO/IEC PDAM 1.2 Component values appear to contradict PDAM 1.1 No general agreement on this 23

24 Thank you