Twisted pair

Twisted Pair Wiring Schemes

The twist of the pairs in Category 5e and 6 cables is necessary to maintain high
performance. All connectorization techniques require that the twist be maintained up to the
point where the individual wires enter the connector or a piece of equipment.
While twisted pair connectors are interchangeable (one manufacturers 8P8C style jack fits
into anothers outlet), they do vary in termination techniques. See each manufacturers
instructions for specific.
Twisted pair wiring is designed so that the same wires continuously connect from one end
of the system to the other (i.e. green to green) just like electrical wiring. This differs from
fiber optics (see below).
Twisted Pair Termination
Twisted pair cables typically terminate in one of two TIAs recognized standards; T568A
and T568B. The US government and residential installations employ T568A; commercial
installations typically employ T568B. Either method is acceptable. However, it is important
that only ONE method be used consistently throughout the entire network.
General Practices
U/UTP data connectors are of the Insulation Displacement Connector (IDC) type in an 8P8C
size (eight pin). As the wires are crimped or inserted into place, the connector automatically
displaces the insulation to permit clean conductor contact and a gas-tight seal.
Maintaining conductor twist is essential for top performance especially at termination. Other
proprietary tools and methods exist; always refer to the connector manufacturers specifications.
Use a ring tool to remove about 7.5 cm (3 inches) of jacketing. This will expose four twisted
pairs color-coded as pair 1 (blue with white/blue), pair 2 (orange with white/orange), pair
3 (green with white/green) and 4 (brown with white/brown). Separate the pairs but DO
NOT UNTWIST the conductors while preparing them for connectorization.
Place the conductors in the appropriate slots in the jack or the outlet module (striped
conductors in the odd numbered slots, solid in the even) and crimp or insert them into place
with the appropriate tool. Rack termination (i.e. punch-down blocks) are usually colorcoded to aid in placing the pairs.
Follow the same untwist rule as connectors. Refer to the manufacturers instructions for the
actual connection.
Optical Fiber
Fiber optic schemes
Optical fibers can be either spliced together by fusion, mechanical methods or terminated
with a connector.
Optical signals travel over transmit/receive pairs. The integrity of the tx/rx signals are
maintained by a system of polarity where connector orientation reverses at each end of the
pair. The diagrams in chapter 9 show a typical duplex link with a transmit/receive fiber pair
as well as the more complicated systems of today, with duplex solutions over an array cable
or parallel optics.
Due to the different styles and manufacturers of fiber optic connectors, this section covers
only general practices. Refer to the connector manufacturers specific instructions for
detailed procedures.
Loose tube preparation
Prepare the cable end for termination. If the cable is loose tube buffered, furcate the fibers.
Furcation is not required for tight buffered fibers unless the connectivity occurs outside of a
protected enclosure.

96

www.commscope.com

Fiber stripping
Use a fiber stripping tool to cleanly remove the acrylate coating. The stripped length will be
determined by the connector being used; consult the manufacturers instructions. Remove the
coating residue from the stripped fiber with a lint-free cloth soaked in a solution of 97%
isopropyl alcohol. Avoid handling the bare fibers as much as possible.

If using a polish-type connector

Follow the manufacturers instructions for connector preparation (adhesive placement, etc.)
and place the fiber in the connector. If an adhesive is being used to hold the fiber in place,
allow it to cure; again, refer to the instructions. Once the fiber is secure in the assembled
connector, scribe the fiber and remove the excess fiber from the face of the connector.
Polish the connector face per instructions. Clean the fiber with the isopropyl alcohol.
Deviating from the manufactures procedure, using non-recommended materials or using out
of date adhesives are not recommended. This may cause performance issues either during
installation or during the lifetime of product.
If using a no epoxy/no polish connector
These connectors typically have a short factory-polished fiber stub installed inside the
connector. Therefore, it is important to verify that the fiber type of the connector matches
that of the cable. Cleave the fiber end of the cable using a quality fiber cleaver. The cleave
should be clean, without chips and within 1 of perpendicular. Non-polish connectors use
mechanical methods to hold the fiber in place. Insert the fiber into the connector and
activate the holding mechanism per manufacturers instructions.
Again, following the manufacturers installation process is critical to success. As many no
epoxy/no polish connectors are difficult to evaluate with a microscope, frequent testing is
recommended so errors in process do not propagate throughout the installation. Some of
these connectors recommend the use of a VFL (Visual Fault Locator) to provide a visual
guide of success during termination.
If splicing on a pigtail
Remember that a pigtail is a length of fiber that has one end already terminated from the
factory and the other end bare, waiting to be spliced (joined) to an incoming cable.
Optical fibers are spliced in two ways. Fusion splicing uses a machine that precisely aligns and
melts together two prepared fiber ends with an electric arc; the splice is then reinforced with a
splice protector. Mechanical splicing holds two prepared fiber ends together in a sleeve filled
with index matching gel. Indoors, spliced fibers are placed in splice trays and secured in a
rack. (Outdoors, spliced fibers are placed in splice trays that are usually sealed in a waterproof
splice enclosure.)
The splicing environment should be as free as possible from dirt and humidity. By splicing
indoors, the harsh conditions are usually avoided but there may be dust and debris still
within the computer room. Regardless of your splicing location, make sure to follow all
appropriate OSHA procedures.
Before exposing cable components and working within the enclosure, the installer should
consider how the cable and pigtail (or pigtail module) will lay when the process is finished.
The fiber can be test-routed to make sure that it can fit into the location and that the fiber
bend radius can be maintained.
Cable preparation/jacket removal for splice
Prior to splicing, secure the cable to the enclosure. The cable end can then be prepared for
splicing. The instructions for the facility/enclosure tell how much of the jacket to strip away.
Measure that distance from the end of the cable. Carefully make a ring cut through the
jacket at the choke point using the appropriate cable prep tool. DO NOT cut or nick the
fibers within the cable jacket. Make a second cut about 15 cm (6 inches) from the cable
end and remove that part of the jacket to access the ripcord.
Removing non-fiber elements and securing
Trim the aramid yarns and any strength element flush with the cable jacket. Review the
manufacturers instructions as to how the cable should be secured. If the pigtail is integrated
within a module, unwrap several loops of fiber in order to have enough length to work with
while splicing.
Connector loss
TIA standards set connector power loss to be no greater than .75 dB per connector set.
However, experienced craft personnel can prepare connector pairs to deliver loss of 0.5
dB or less. Pre-terminated single-fiber connectors typically provide loss of 0.5 dB or less.

www.commscope.com

97

Cleaning
Finally, clean each connector before inserting into the patch panel. Keep dustcaps on any
connectors reserved for future use.
Coaxial wiring
Coaxial cable has a center conductor and an outer conductor which acts as a shield.
Using BNC-style connectors is a popular method of termination for DS3/4 cable.
Specialized connectors exist for other coaxial cable types; contact CommScope Technical
Support at support@commscope.com for more information.
Connectorizing Braid-Shielded Coax Cable
These instructions refer to general practices. Exact instructions may vary with the maker of
the tools and connectors being used. Always refer to the manufacturers instructions.
If using a boot to protect the connection at the tap or if using a compression fitting with an
independent sleeve, slide it over the cable end first. Strip the end of the coax cable by
clamping a cable prep tool around the cable and making sure that the end of the cable is
flush against the stop. Operate the tool per manufacturers instructions until a clean cut is
achieved.
A clean cut will have a square end and the cable will have a round cross-section. The
dimensions in the drawing are approximate; always check with the connector manufacturer
for exact dimensions.
Carefully bend the braid back over the jacket. On tri- and quad-shielded cables, carefully trim
outer layer of tape. If using a crimp-style or one piece compression connector, slide it over the
cable end and use the crimping tool to firmly attach the connector. A good connection will
have solid contact between the neck of the connector and the braid under it. The conductor
should extend no more than 3 mm (1/8 inch) beyond the front edge of the connector.
If a two-piece compression connector is being used, slide the main boot over the cable
prior to cable preparation. Then slide the compression sleeve to the back of the connector.
Use the proper compression tool to compress the connector per the manufacturers
specifications. If a non-crimpable connector is being used, follow the manufacturers
specific instructions.
Labeling termination
TIA standard 606-A calls for machine-generated labels to be used for circuit identification
at both the desktop and the telecommunications and equipment rooms. This applies to all
cable media.
13. Test cabling infrastructure permanent link
The details of testing copper and optical links is discussed in Chapter 15. In general
though, every link should be tested as installed. Up-front testing and fixing of troubled links
will create a much smoother process when the system is brought online and will have
additional benefits throughout the life of the system.
14. Install electronics
It is common practice to activate and run, or burn-in, electronics before placing into
production. Review manufacturers recommendations on the timing, but one week burn in
should be sufficient to discover so called infant mortality, or early failure of electronic
gear. If there is no failure after that initial time, the device will likely run for years. Burn-in is
typically done in a different room in order to keep separate room in order to keep separate
from operational devices.

98

www.commscope.com

The electronics can be installed into the rack or cabinet frame using the flanges and screws
that are supplied with the units. Electronics, especially heavier units, are typically installed
from the bottom up for ease of handling and stability during the process.
15. Install patch cords
A patch cord is a relatively short length of cable connectorized at both ends and can be
used to connect electronics to other electronics, electronics to a patch panel or connect two
passive links together.
Unlike backbone or behind-the-wall cabling, patch cords are expected to be mated and
unplugged frequently. Therefore a robust design is strongly desired, and CommScope
recommends the use of factory terminated patch cords for both copper and fiber cabling to
provide a robust, durable solution along with a consistent and low loss.
Note that patch cords are part of the cabling link and that their quality is of great
importance, just like that of the backbone cabling. A system utilizing all Category 6A
components must utilize Category 6A patch cords in order to achieve the expected
performance.
Similarly, for fiber cabling, the cord fiber type needs to match the cabling in the backbone.
For fiber cables trunk cable with 50 m core fiber must be connected to patch cords that
also have a 50 m core size. Similarly, singlemode cabling must be connected with
singlemode patch cords.
Matching fiber core sizes is standard practice. Today the challenge is making sure that the
fiber patch cords meet or exceed the bandwidth of the cabling. Laser-optimized OM3 or
OM4 fiber cabling must be connected to the electronics through laser-optimized fiber as
well. For any 50 m fiber application, CommScope recommends using patch cords that
only contain OM4 50 m fiber. These will match up well with OM4, OM3 or any lower
grade 50 m cabling. Having one 50 m cabling type for your patch cords also limits
excessive inventory or accidentally mixing fiber types within the system.
A good exercise for checking fiber consistency in active networks is to walk through the
data center and run a patch cord color check. Much of todays backbone and trunking
cable within the data center will be aqua in color, to denote the high bandwidth laser
optimized fibers. Any patch cords that are orange in color, which is the traditional color for
standard multimode fiber, then there was likely no patch cords that match the bandwidth of
your high bandwidth backbone cabling. Orange patch cords are likely standard 50 m
grade and may even be 62.5 m fiber, which would be the wrong core size.
Fiber patch cords are typically cleaned in the factory after polishing is completed, but this
does not mean that they arrive to the job site free of debris. It is important to clean the
connector endfaces of the patch cord before inserting into the patching field. Connectors
that are not patched should be left with their duct caps on to limit endface contamination.
Similarly, optical ports should have their dust caps reinstallled immediately after patch cords
are removed.
16. System test
Lastly, the system should be tested. Although individual links were evaluated earlier in the
process, it is important to test the whole system for several reasons.

to verify the cable connections are routed to the proper location

to verify proper polarity is maintained (transmit goes to receive)

to verify connections are properly mated into the adapters and free of debris

Testing is described in detail in the following chapter.

www.commscope.com

99

15. Testing & Documentation

When the design and installation of a data center is complete, testing of the installed cabling is
the next important step.
A permanent link is considered to be all of the cabling and terminations within a cable run
except for the patch cords, which are expected to be interchangeable over time. Cabling
should be tested AFTER installation into the panels in order to evaluate the quality of the fully
installed system. Found issues should be investigated and solved before commission of the
system can be considered complete.
Testing of the passive system, whether copper or optical fiber, is typically completed on the
passive links before connection to electronics. This is very important, because once a system is
operational; end users typically do not want to disturb any of the cabling.
Consider, for example, an optical fiber run utilizing a 24-fiber MPO trunk cable connecting to
a 24-fiber LC-MPO module on each end. A request comes in to connect one user quickly and
therefore testing is by-passed in order to meet the request; the new user is patched in through
fibers 1-2. Later it is determined that the trunk cable was damaged during installation and fibers
13&14 are dark. Although the active connection is only utilizing one pair of fibers, it is now
impossible to replace or repair the link without taking down that active pair.
CommScope goes to extra lengths to ensure that our cables perform as promised.
CommScopes unique WebTrak identifier printed on the cable allows customers and installers to
enter this information online and receive the test report for that specific reel of cable, whenever
and wherever it is required.
Testing installed cable is critical to establishing network performance and integrity. It reveals
problems such as cable bent tighter than its recommended minimum bend radius, or a poorly
installed connector.
Documenting the test results is equally essential as it provides a baseline for performance prior
to actual network operation and helps trouble-shooting efforts should there be problems during
turn-up and operation. CommScope recommends end-to-end testing for each installed (or
permanent) link as outlined in TIA/EIA 568 C.0 Annex E.

Twisted Pair Cable Testing

All twisted pair cable permanent links should be tested and the results documented. Some
warranty programs may require additional testing - see the warranty for details.
Permanent links and channels should be tested for wiremap, length, insertion loss (or
attenuation), NEXT loss, power sum NEXT, ELFEXT loss, power sum ELFEXT, return loss,
propagation delay and delay skew. These are the same parameters used in testing cable
master reels at the factory. A master test report (CommScope provides these for Category 6
and 5e box/reels) is a good benchmark for the expected performance of a link. Test reports
are available at www.commscope.com.
Test equipment
Test twisted pair channels with a test set at one end and a remote at the other. For Category 6A and
6 testing, use Level III meters; for Category 5e, use Level III or IIe testers. Exact testing methods will
differ because of the wide variety of equipment and features, but these general rules apply:
Prior to testing, field-calibrate the test set. Make sure the equipment is set up for the proper
network and cable type. Connect the test set and remote to the ends of the permanent link or
channel either through directly plugging the patch cords into the tester or by using the
appropriate adapter modules.

100 www.commscope.com

The link must be tested at several frequencies from 1 MHz up to 100 MHz for Category 5e
and 250 MHz for Category 6A and 6. Worst-case values for link and channels are shown in
the following tables.

Performance Standards

TABLE 16: CATEGORY 6A U/UTP PERFORMANCE STANDARDS (TIA 568 C.2)

MHz

Insertion
Loss (dB)
Channel/Link

NEXT (dB)
Channel/
Link

PSum
NEXT (dB)
Channel/Link

ACRF
(dB)
Channel/Link

PSum ACRF
(dB)
Channel/Link

Return
Loss (dB)
Channel/Link

2.3/1.9

65.0/65.0

62.0/62.0

63.3/64.2

60.3/61.2

19.0/19.1

4.2/3.5

63.0/64.1

60.5/61.8

51.2/52.1

48.2/49.1

19.0/21.0

5.8/5.0

58.2/59.4

55.6/57.0

45.2/46.1

42.2/43.1

19.0/21.0

10

6.5/5.5

56.6/57.8

54.0/55.5

43.3/44.2

40.3/41.2

19.0/21.0

16

8.2/7.0

53.2/54.6

50.6/52.2

39.2/40.1

36.2/37.1

18.0/20.0

20

9.2/7.8

51.6/53.1

49.0/50.7

37.2/38.2

34.2/35.2

17.5/19.5

25

10.2/8.8

50.0/51.5

47.3/49.1

35.3/36.2

32.3/33.2

17.0/19.0

31.25

11.5/9.8

48.4/50.0

45.7/47.5

33.4/34.3

30.4/31.3

16.5/18.5

62.5

16.4/14.1

43.4/45.1

40.6/42.7

27.3/28.3

24.3/25.3

14.0/16.0

100

20.9/18.0

39.9/41.8

37.1/39.3

200

30.1/26.1

34.8/36.9

31.9/34.3

250

33.9/29.5

33.1/35.3

300

37.4/32.7

31.7/34.0

400

43.7/38.5

500

49.3/43.8

23.3/24.2

20.3/21.2

12.0/14.0

17.2/18.2

14.2/15.2

9.0/11.0

30.2/32.7

15.3/16.2

12.3/13.2

8.0/10.0

28.8/31.4

13.7/14.6

10.7/11.6

7.2/9.2

28.7/29.9

25.8/27.1

11.2/12.1

8.2/9.1

6.0/8.0

26.1/26.7

23.2/23.8

9.3/10.2

6.3/7.2

6.0/8.0

NOTE: Propagation Delay is 555 nanoseconds for channel/498 nanoseconds for link tested at 10 MHz.
NOTE: Delay Skew is 50 nanoseconds for channel/44 nanoseconds for link tested at 10 MHz.

TABLE 17: CATEGORY 6 U/UTP PERFORMANCE STANDARDS (TIA 568 C.2)

MHz

Insertion
Loss (dB)
Channel/Link

NEXT (dB)
Channel/
Link

PSum
NEXT (dB)
Channel/Link

ACRF
(dB)
Channel/Link

PSum
ACRF (dB)
Channel/Link

Return
Loss (dB)
Channel/Link

2.1/1.9

65.0/65.0

62.0/62.0

63.3/64.2

60.3/61.2

19.0/19.1

4.0/3.5

63.0/64.1

60.5/61.8

51.2/52.1

48.2/49.1

19.0/21.0

5.7/5.0

58.2/59.4

55.6/57.0

45.2/46.1

42.2/43.1

19.0/21.0

10

6.3/5.5

56.6/57.8

54.0/55.5

43.3/44.2

40.3/41.2

19.0/21.0

16

8.0/7.0

53.2/54.6

50.6/52.2

39.2/40.1

36.2/37.1

18.0/20.0

20

9.0/7.9

51.6/53.1

49.0/50.7

37.2/38.2

34.2/35.2

17.5/19.5

25

10.1/8.9

50.0/51.5

47.3/49.1

35.3/36.2

32.3/33.2

17.0/19.0

31.25

11.4/10.0

48.4/50.0

45.7/47.5

33.4/34.3

30.4/31.3

16.5/18.5

62.5

16.5/14.4

43.4/45.1

40.6/42.7

27.3/28.3

24.3/25.3

14.0/16.0

100

21.3/18.6

39.9/41.8

37.1/39.3

23.3/24.2

20.3/21.2

12.0/14.0

200

31.5/27.4

34.8/36.9

31.9/34.3

17.2/18.2

14.2/15.2

9.0/11.0

250

35.9/31.1

33.1/35.3

30.2/32.7

15.3/16.2

12.3/13.2

8.0/10.0

NOTE: Propagation Delay is 555 nanoseconds for channel/498 nanoseconds for link at 10 MHz.
NOTE: Delay Skew is 50 nanoseconds for channel/44 nanoseconds for link for all frequencies.

www.commscope.com 101

TABLE 18: CATEGORY 5E U/UTP PERFORMANCE STANDARDS (TIA 568 C.2)

MHz

Insertion
Loss (dB)
Channel/Link

NEXT (dB)
Channel/
Link

PSum
NEXT (dB)
Channel/Link

2.2/2.1

>60/>60

>57/>57

57.4/58.6

54.4/55.6

17.0/19.0

4.5/3.9

53.5/54.8

50.5/51.8

45.4/46.6

42.4/43.6

17.0/19.0

6.3/5.5

48.6/50.0

45.6/47.0

39.3/40.6

36.3/37.5

17.0/19.0

10

7.1/6.2

47.0/48.5

44.0/45.5

37.4/38.6

34.4/35.6

17.0/19.0

16

9.1/7.9

43.6/45.2

40.6/42.2

33.3/34.5

30.3/31.5

17.0/19.0

20

10.2/8.9

42.0/43.7

39.0/40.7

31.4/32.6

28.4/29.6

17.0/19.0

25

11.4/10.0

40.3/42.1

37.3/39.1

29.4/30.7

25.4/27.7

16.0/18.0

31.25

12.9/11.2

38.7/40.5

35.7/37.5

27.5/28.7

24.5/25.7

15.1/17.1

62.5

18.6/16.2

33.6/35.7

30.6/32.7

21.5/22.7

18.5/19.7

12.1/14.1

100

24.0/21.0

30.1/32.3

27.1/29.3

17.4/18.6

14.4/15.6

10.0/12.0

ACRF
(dB)
Channel/Link

PSum
ACRF (dB)
Channel/Link

Return
Loss (dB)
Channel/Link

NOTE: Propagation Delay is 555 nanoseconds for channel/498 nanoseconds for link at 10 MHz.
NOTE: Delay Skew is 50 nanoseconds for channel/44 nanoseconds for link for all frequencies.

Documentation
Document each channels performance for the criteria listed above, the test date, the name(s) of
the test personnel and the equipment used (manufacturer, model number and calibration date).
Record (or download if the equipment has that function) the test results and store them with the
as-built drawings. Keep hard copies of the documentation in the telecommunication or
equipment room.
Twisted Pair Troubleshooting
Fail Wiremap
This error is caused by improperly wired connectors and is easily discovered and repaired.
Most test sets will display a graphic representation of the problem (see Figure 52). Fix wiremap
problems by inspecting and correcting miswired termination hardware.

Figure 52: Wiremap Test Set Display

Correct
display

1
2
3
4
5
6
7
8

1
2
3
4
5
6
7
8

Crossed
pairs

1
2
3
4
5
6
7
8

1
2
3
4
5
6
7
8

Pin 1
not connected

1
2
3
4
5
6
7
8

1
2
3
4
5
6
7
8

Miswiring
T568B to
T568A

1
2
3
4
5
6
7
8

1
2
3
4
5
6
7
8

Fail Length (test set 10%)

This occurs when a link exceeds 90 meters; links sometimes grow as cable may not be
placed exactly as planned. Check if the master/meter is set for the correct Nominal Velocity of
Propagation (NVP) of the cable being tested. For instance, a non-plenum cable has a lower
NVP than a plenum cable.
If the test set is correctly set and the length test still fails, the system may have to be redesigned
to eliminate the cable links that are too long. If system redesign is not possible, retest to ensure
the cable passes all other parameters. This link may be limited to slower equipment or services.
Fail Crosstalk or Return Loss (RL)
Fail crosstalk may be caused by several situations. The quickest check is to make sure that the
test set is set up for the correct Category (5e or 6) of cable. Another very common reason is
untwist at the connector.
If the test set displays a distance to the failure, check that location for bends tighter than the
minimum bend radius or for overly-tight cable ties. Check for kinks in conduit; the inside diameter
of conduit decreases as it is bent, and these bends may be crushing the cable. If the distance to
the failure is shown to be less than 3 meters, re-terminate the connection on the failing end.

102 www.commscope.com

If re-termination does not solve the problem, swap the locations of the test set and the remote

and test the link from the opposite end. If the failure follows the test equipment (i.e. the failure
was at the remote end and after switching locations is still at the remote end), the problem is
the meter interface cable. If the problem location doesnt move with the equipment, try
replacing the outlet. If that fails, it may be that the installation caused cable damage and the
cable may need to be replaced.
Fail Insertion Loss
Using the wrong category of cable or incorrect category of termination jack are two causes of
this problem. It is corrected by installing the correct category of cable/hardware. Another
common cause is that the cable is too long (see Fail Length).
Post-testing Problems
Some problems appear only when the active equipment is attached. These are a little more
difficult to troubleshoot and repair.
Failure to Link (link light does not come on)
This failure can have numerous causes; the most obvious is that the cabling cannot support the
application. If the correct cable is being used, troubleshooting becomes more complex.
If installation testing rules out wiremap problems, the most likely solution is that improper
patching has produced an incomplete circuit path. Disconnect the equipment at the
telecommunications room and the work area prior to testing the circuit path. Attach a tone
generator to the cable in question and use an inductive probe to check for tone at the far end.
If tone is detected, then the link is continuous. If tone is not detected, the link is not continuous
and each segment of the link must be tested until the broken segment is found.
If the circuit is continuous, use a tester with a noise check test to see if this is the problem.
Note that too much noise is relative; the noise floor varies with the application. If the noise
floor is too great for the application, the only option is to re-route the cable to avoid the noise
source (i.e. fluorescent lights, high-voltage electrical cable, etc.).
Other problems
If the link light is on but the circuit is not working, the cause could be that the horizontal cable is
too short and the receiver is being overpowered by too strong a signal. Solve this by using
excessively long patch cables at either end of the link to add extra insertion loss to the circuit.
Another possibility is that the cable is fine but the active equipment is improperly configured.
Check the active to ensure proper configuration.
Fiber Optic Testing
Testing is especially important when confirming the optical loss for a fiber optic system. The
power loss budget is the allowable system loss between the transmitter and receiver. System
gain, transmitter power and receiver sensitivity all influence the power loss budget.
Take, for example, a link of 500 meters (1640 feet) with three connector pairs (crossconnect,
patch panel and desktop). The TIA maximum loss allowance for a connector pair is .75 dB.
Adding the fiber loss (3.0 dB/m x 0.5 km = 1.5 dB) to the loss from three connector pairs
(.75 dB max x 3 = 2.25 dB) establishes a calculated total loss of 3.75 dB (1.5 + 2.25). It is
possible for a link to deliver more power than the rx end can handle (called saturation). In this
case, a device called an attenuator is installed at the rx end to add loss to the system.
While standards such as TIA/EIA568C.3 provide baseline performance level of components,
there are many options in todays market that provide significantly higher performance levels in
order to achieve high data rate performance with more connector pairs and/or at an extended
distance. Therefore, one cannot simply refer to the standards for what should be expected in the
field.
For example, a manufacturer may specify a 0.5 dB maximum (vs. 0.75 dB allowable by TIA)
for factory terminated LC connectors. Therefore a point-to-point link should see a maximum loss
of 1.0 dB for the two connector pairs, which is less than the 1.5 dB allowable by the
Standards. This issue arises frequently in the data center when examining the expected loss of
an MPO to LC module. Although the module creates a situation with 2 connector pairs, the loss
is typically specified at well below the maximum allowable by the Standards.
Determining the expected loss can become complicated, and CommScope has a link loss
calculator available to provide its business partners with expected loss based upon distance,
fiber type and connectivity. A calculator can take into account statistical data to provide a more
accurate picture of what the system test results should look like.

www.commscope.com 103

Figure 53: www.myCommScope.com, Link Loss Calculation

Test equipment today has become very sophisticated and can often calculate expected loss
values based upon standards provided values. Although valuable, the allowable loss based
upon the standards may be higher then what the system should actually experience based
upon the product specifications. It is recommended to use a link loss calculator based upon
the manufacturers product performance beyond just the standards baseline requirements.
During the testing process, you are very likely to have test results that initially are not passing.
The good news is that an initial test failure does not guarantee that the product is faulty. With
optical fiber testing, simply cleaning both the test cord and the behind-the-wall connector will
ensure that dirt and dust will not adversely affect the test results. If a failing result is obtained,
one should clean both the system connector and test lead before investigating further.
Most issues can be resolved by cleaning the connectors.
There are many devices that can be used to clean connectors and adapters. It is very important
to have the correct cleaning device for the component being cleaned. Dirt can be removed from
the endface of connector utilizing a cleaning tape, often within a cassette that allows for automatic
rotation of that tape to a clean each time it is used. Even connectors that are behind-the-wall
are accessible with cleaning sticks that are sized for the appropriate ferrule diameter. Cleaning
and inspection kits can be purchased that contain all of the cleaning supplies as well as a
microscope to examine the endface of the connector for dirt and scratches.
IMPORTANT: Before looking at the endface of a connector through a microscope always
make sure that the link is unplugged from the transmitter and that no optical power can
be directed towards your eye.

104 www.commscope.com