June 22, 2009
Testing should be performed in the field when deploying 10GBase-T
By Dan Payerle
Alien crosstalk describes the general phenomenon where energy is coupled between cables in a common bundle or installation. It was brought to light by the development of active network hardware that could provide 10G Ethernet over twisted-pair cabling (10GBase-T). Alien crosstalk becomes worse as the operational frequency increases.
Since the frequency range required to support 10GBase-T is higher than
that of 1000Base-T, crosstalk—both internal and alien—has become a
difficult obstacle to overcome when designing and installing LAN
cabling to support the latest technologies.
Mitigation techniques
1. Proximity is the key contributor to alien crosstalk. When 10GBase-T
is being selectively deployed, avoid using adjacent ports in the patch
panel. There may be no alternative, however, when deploying 10GBase-T
to workstations located near each other.
2. When deploying 10GBase-T in adjacent ports of a patch panel, alien crosstalk testing should be performed in the field.
3. In the event the alien crosstalk test fails, take the following actions to reduce the level of alien crosstalk.
b) If the cords cannot be separated, use either Cat 6A or Cat 6 screened twisted-pair (ScTP) cords.
c) Reconfigure any cross-connect as an interconnect.
d) Replace Cat 6 connecting hardware with Cat 6A.
e) Replace the Cat 6 horizontal cable with Cat 6A.
4. Retest the channel after performing any mitigation techniques to be
sure that the techniques have brought the alien crosstalk margins back
to acceptable levels.
{mospagebreak}
Field testing
Alien crosstalk testing involves testing various combinations of links
that are identified as ‘victims’ and ‘disturbers’. Performing a test on
a single victim cable involves at least six different testing
configurations, and as many as 12—depending on the manufacturer of the
field test equipment.
Given that multiple tests are required for each victim link, requiring
100% testing of every link as a victim is not practical. During
testing, when the first three victim/disturber combinations reveal a
condition known as ‘insignificant alien crosstalk’, the test can be
stopped without finishing the 1% or minimum five victim links.
Insignificant alien crosstalk is a condition where the measurement is
below a certain level and may not be detectable by some field test
instruments.
The selection of disturber links has to be done individually for every
victim link. Select all links that run in the same cable bundle or are
most consistently positioned relative to the victim cable. These
bundles may be found in the patch panel, crossconnect or conduit. Add
any additional links that occupy adjacent positions in the patch panel
or outlet.
When selecting links to test, the routing of links should also be
considered in addition to the location of the links in the patch panel.
The disturber links should be run in the same pathway as the victim
link to have the most impact on alien crosstalk measurements.
The proper selection of links for alien crosstalk testing is critical
and requires a certain degree of knowledge about the topology of the
cabling plant. Without knowing where the various links are routed to
within the building, the process of testing can be inaccurate, since
the chosen disturber links may not be close enough to the victim to
provide any significant data.
After deciding on the victim and disturber links to check, the field
tester needs to be connected to the cabling according to the
manufacturer’s directions. Some field testers require a personal
computer to be attached to the field tester during the measurement
process to gather the data and compute the alien crosstalk results.
Additionally, the tester and computer may need to be moved to the
opposite side of the link for the second half of the alien crosstalk
testing process.
Because the number of links to test and the time to test each
victim/disturber combination can be significant, choosing the right
field tester can save time and trouble. A field tester that does not
require a computer for data acquisition or double-testing of each
victim/disturber combination can cut 75% of the total alien crosstalk
testing time while eliminating the need to bring a computer into the
field.
Dan Payerle is business unit manager with Ideal Industries Inc., and
has been actively involved in the LAN cabling business, providing
network design, testing, troubleshooting, consulting and training
services for a variety of companies over the last decade. Working with
several national training companies, Dan developed training programs
for copper and fiber optic installation courses, and created curricula
for trade schools to use in the process of becoming nationally
accredited.
CLICK HERE for Ideal Industries Inc
Print this page
