Data Center 10 Gigabit Copper Cable Performance Test
In the next few years, more and more data centers, storage area network (SAN), etc. will deploy 10 Gigabit Ethernet in large quantities. Among the many systems in the data center, the cost of a cabling system is less than the total IT budget. 0.4%, but close to 50% of the network failures are caused by the wiring system. It can be seen that the construction and maintenance of the physical medium is a link that cannot be ignored. Although optical fiber is the protagonist of the physical medium of the 10Gbps network in the data center, TIA 942, ISO/IEC 24764, IEEE 802.3an and other standards still put forward requirements for the copper cable medium that supports 10GBASE-T applications. This article focuses on 10 Gigabit copper cables. Media testing issues.
Channel/Link Routine Certification Testing
This test process is the same as the commonly known Category 6 cabling system test, and the stored test report includes the following parameters:
(1) Connection diagram, including continuity test of shielding layer.
(3) Insertion loss.
(4) Near-end crosstalk.
(5) Near-end crosstalk power.
(6) Proximal attenuation.
(7) Remote attenuation.
(8) Return loss.
(9) Propagation delay.
(10) Propagation delay deviation.
(11) DC loop resistance.
Alien Crosstalk Test
The biggest difference between the testing of Category 6A cable and Category 6 cable is the alien crosstalk. This is when running 10GBase-T because the upper limit of the signal spectrum reaches 500MHz, it is more likely to cause interference between adjacent twisted pairs or adjacent distribution frame sockets, which is collectively referred to as alien crosstalk (AXT).
Alien crosstalk will superimpose with normal signals, distort the signal, increase the bit error rate of the link, and even cause the device to fail to connect. Therefore, on the basis of cable certification, a set of important interference measurement parameters must be added. This new set of test parameters is called alien crosstalk. If the alien crosstalk of the cable is tested 100%, the number of tests is (n^2 +n)/2 times, according to this formula (where n is the number of wiring links), the project of testing 500 links requires 125250 A combination of tests, which would take several hundred hours. Therefore, the ISO/IEC 61935-1 standard proposes a method of sampling alien crosstalk, including alien near-end crosstalk (ANEXT), alien far-end crosstalk (AFEXT), integrated alien near-end crosstalk (PSANEXT), and integrated alien far-end crosstalk. (PSAFEXT).
Test Equipment Requirements
(1) Instrument test frequency
For Category 6A class Ea and below cables, the test frequency of the instrument is not less than 500MHz (LANTEKII 500 from IDEAL); for class F and FA cables, the test frequency of the instrument is not less than 1000MHz (LANTEKII 1000 from IDEAL).
(2) Accuracy meets class IIIA or IV
(3) Adapter requirements
Two dual-port AXT adapters (one for the host and one for the remote), and the test kit also includes 12 AXT terminal modules (supporting the “6-pack-1” test model) and 14 test jumpers.
Build a “6-pack-1” permanent link model in strict accordance with the requirements, and complete the pre-assessment of cables.
(1) Press respectively on 6 types of systems: “6 pack 1” model with a length of 37m (121ft); “6 pack 1” model with a length of 37m ~ 55m (180ft); 55m ~ 100m (328ft) after using mitigation technology “6 in 1” model.
(2) In Category 6A and above systems, the 90m permanent link “6 pack 1” model.
The field test of 10 Gigabit copper cables is divided into two stages. The first stage is similar to Category 6, and also uses the instrument to select the corresponding standard (such as TIA CAT 6A or ISO Class Ea) for basic testing. The second stage is to test the alien crosstalk and finally get the six-pack-one comprehensive alien crosstalk result. Obviously, the latter stage requires the necessary instrumentation and time preparation.
There is a saying in the market that shielding products do not need alien crosstalk testing, which is irresponsible to the user. Because shielding is only a means of product realization, and there will be many complex factors in engineering construction, testing without an acceptance stage is dangerous for end-users, so shielding and non-shielding should pay attention to the problem of alien crosstalk.
To save time for CAT 6A testing, the international standard supports sample testing with a sampling rate of 5%, or 10 worst links. Therefore, alien crosstalk is not terrible. As long as there is a relevant tester and a test adapter for 10 Gigabit copper cables, the task can be easily completed.