White Paper 10 Dumb Things Smart People Do When Testing Network Cabling Systems

White Paper 10 dumb things smart people do
when testing network cabling systems
9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 1

White Paper: 10 dumb things smart people
do when testing network cabling systems
While anyone who installs, tests and certifies a network cabling plant knows the importance of meeting
standards performance parameters and ensuring application support, even the best of us can make mistakes
that can adversely impact the bottom line and customer satisfaction.
1. Fail to specify the type of link testing you want.

2. Neglect to agree upon marginal test results upfront in writing.
3. Fail to specify all test parameters to be tested.
4. Wing it on Alien Crosstalk compliance.
5. Forget to enable plot data.
6. Go with a non-EF compliant tester for testing multimode fiber.
7. Choose the two-cord reference for Tier 1 optical loss testing.
8. Use patch cords to certify a fiber cabling system.
9. Rely on a duplex tester for certifying MPO trunks.
10. Skip out on proper fiber inspection.
1. Fail to specify the type of link testing you want
The channel link is the complete link over which active equipment communicates and
therefore includes equipment/user patch cords at both ends. While the network ultimately relies on
the performance of the channel, channel testing has less stringent limits and can allow problems
with the fixed portion of the network to go undetected. Permanent link testing assures that a
passing link can reliably be configured into a passing channel by adding certified patch cords and is
required for a warranty application

Cabling Mistakes #1: Specify Channel Testing While Installing

Permanent Links
While the smart people who install, test and certify

network cabling plants know the importance of meeting
standards performance parameters and ensuring application
support, even the best of us can make mistakes that adversely
impact the bottom line and customer satisfaction.
Lets take a look at the #1 Dumb Thing that smart people do
when testing network cabling systemsSpecifying channel
testing when installing permanent links.

Channel vs. Permanent Link Refresher
In a network, the channel is the connection from one active device to another, including any patch
cords and equipment cords. In the data center, this could be from an access switch to a server and would
include any patch cords used for cross-connects or interconnects between the two. In the LAN, the
channel could be the connection from a distribution switch in a telecommunications room all the way to
the network card in a laptop.
The overall length of the channel, including the cabling and any cords should not exceed 100
meters per industry standard guidelines. There are also recommendations on the length of equipment
cords--standards recommend 5 meter patch cords.
In contrast, the permanent links is the fixed portion of the channel, typically patch panel to patch
panel in the data center or patch panel in the telecommunications room to work area outlet or
consolidation point in the LAN. The maximum permitted length of the permanent link is 90 meters.

Not Exactly Intuitive
As the complete end-to-end link over which active equipment on the network communicates,
proper data transmission obviously relies on the performance of the channel. It therefore seems
intuitive that channel testing would certainly be the way to go, right? Wrong.
Some contractors may prefer channel testing for the simple fact that it's easier to pass, but that's
not exactly a good thing. Per TIA standards, combining standards compliant patch cords with a standards
compliant permanent link will always result in a passing channel. This design has enabled the plug-and-
play application of Ethernet devices. For channel testing, this also means if very high. Quality patch cords
are used, the channel can pass even if the permanent link has failed. Why is this a problem?

As the fixed portion of the channel, the permanent links is really considered to be the true
foundation of the network. If you test the channel, problems with that foundation may go undetected.
Besides, let's face it. Patch cords and equipment move around. And once your patch cords are removed
from the channel, you might as well throw you channel testing results right out the window. That's why
it's important to specify permanent link testing upfront.

Permanent link testing also offers your customers the benefit of

being able to retest the permanent link so that the performance at any
time in the future can be directly correlated to the performance at the
time of installation. Later down the road if there's a problem with the
data transmission in a channel, retesting the permanent link can rule out
any problems in the fixed portion. Subsequent channel testing can then
be done to limit the problem to the patch cords. And this is a problem
more often than you think.

The Weakest Link
While patch cords are an integral part of the network, they often are
also considered commodity and quality may be overlooked. They are
handled and manipulated more than any other component, which also
makes them more subject to damage. Compromising on the quality of patch
cords can compromise the performance of the network.
Thankfully, the permanent link adapters that come with the
DSX-5000 CableAnalyzer are of the highest quality cable and contain a
reference plug to ensure that they do not contribute any disturbance to the
permanent link under test and to ensure that the end of the permanent link
provides interoperability with compliant patch cords.
With permanent link testing, a passing link can be reliably configured
into a passing channel just by adding quality compliant patch cords.

A marginal pass suggests you are close to the limit, something which many customers will
question and possibly reject. The problem is that ANSI/TIA-1152 and IEC 61935-1 state that a marginal
pass simply results in an overall pass condition. It has met the requirements of the standard. If you
want to reject marginal results, you will need to specifically state that in your test specification

Cabling Mistakes #2 and #3: Neglecting Marginal Test Results and Parameters
While the smart people who install, test and certify network cabling plants know the importance
of meeting standards performance parameters and ensuring application support, even the best of us
can make mistakes that adversely impact the bottom line and customer satisfaction.
Lets take a look at the #2 and #3 Dumb Things that smart people do when testing network cabling
systems Neglecting to agree upon marginal test results upfront in writing and failing to specify all
test parameters to be tested.

A Marginal Pass is Still a Pass
Back in the days of category 5, marginal passes were rarely seen due to the fact that the
components were well ahead of their time in terms of performance. But those days are long gone.
Today's category 6A components specify a much higher level of performance that is more difficult to
meet. Everything from marginal performance of the components themselves, to installation practices
and even the quality of the test equipment can be enough to push the results into the marginal pass
region, which on today's standard-compliant test equipment is indicated with an asterisk (*).
Despite the appearance of an asterisk next to your test result, TIA and ISO/IEC standards dictate
that any marginal pass is still considered a pass and therefore compliant. The standards specifically
state that a parameter shall be marked with an asterisk when the result is closer to the test limit than
the measurement accuracy published by the field tester manufacturer. That clearly indicates that tester
accuracy plays a role.

By selecting an industry-compliant tester like the

DSX 5000 CableAnalyzer that offers repeatable, exceptional
accuracy, you can minimize marginal test results. And beware of
any tester than allows you to disable the asterisk and hide marginal
results. This makes the tester non compliant; not to mention that it
can put your reputation and business at risk. Also, make sure to
keep your tester in tip top shape--keep up with the latest firmware,
service your tester as needed, and make sure your permanent link
adapters are not worn out.

Despite all efforts, marginal passes will happen--especially if

the link contains a consolidation point that adds another
connection. Some customers may refuse to accept a marginal
pass, wondering what happened to their supposed "premium"
cabling system. This is when you want to make sure you're
covered. Remember what the standards say. Unless it was
specified and agreed upon in writing that marginal test results
are not acceptable, your marginal passes still pass. Agreeing
upon marginal test results upfront is a smart move.

Know Your Parameters
When it comes to copper certification, it's not just the margins that have changed. With the
introduction of category 6A and a better understanding of applications, data transmission and what
impacts performance, we now have new parameters specified as a minimum in field testing. For
example, when it comes to certifying a category 6A copper system for 10GBASE-T operation, Alien
Crosstalk testing is required.
There are also other newer parameters that if not passed, can cause an application to have issues.
Mode conversion parameters like Transverse Conversion Loss (TCL) and Equal Level Transverse Conversion
Transfer Loss (ELTCTL) are used to measure the common mode signal within a pair, indicating proper
balance for high frequency or noisy environments. And we now have DC resistance unbalance (discussed
in our last blog), which can distort Ethernet signals in a PoE connection.

What you don't want is the time and expense involved in going back and retesting for parameters
you didn't test for the first time. Make sure you know which parameters are required and specify them
upfront so your technicians are prepared. And make sure your tester has the capability to test for every
parameter specified, especially the newer ones like TCL, ELTCTL and DC resistance unbalance--which by
the way, can all be tested with the DSX 5000 CableAnalyzer.

3. Fail to specify all test parameters to be tested
When it comes to copper certification, ANSI/TIA-1152 and ISO/IEC 11801:2011 specify which
parameters should be tested as a minimum in field testing, suggesting there are other parameters
which there are. These include DC resistance unbalance within a pair and between pairs for fully
supporting PoE applications. Additional parameters such as TCL and ELTCTL are also specified, but are
relegated to laboratory testing only. The latest field testers are capable of making these additional
important measurements, but you will need to specify which parameters you want tested.

Cabling Mistakes #2 and #3: Neglecting Marginal Test Results and Parameters
While the smart people who install, test and certify network cabling plants know the importance of
meeting standards performance parameters and ensuring application support, even the best of us can
make mistakes that adversely impact the bottom line and customer satisfaction.
Lets take a look at the #2 and #3 Dumb Things that smart people do when testing network cabling
systems Neglecting to agree upon marginal test results upfront in writing and failing to specify all
test parameters to be tested.

A Marginal Pass is Still a Pass
Back in the days of category 5, marginal passes were rarely

seen due to the fact that the components were well ahead of
their time in terms of performance. But those days are long gone.
Today's category 6A components specify a much higher level of
performance that is more difficult to meet. Everything from
marginal performance of the components themselves, to
installation practices and even the quality of the test equipment
can be enough to push the results into the marginal pass region,
which on today's standard-compliant test equipment is indicated
with an asterisk (*).

Despite the appearance of an asterisk next to your test result, TIA and ISO/IEC standards dictate
that any marginal pass is still considered a pass and therefore compliant. The standards specifically state
that a parameter shall be marked with an asterisk when the result is closer to the test limit than the
measurement accuracy published by the field tester manufacturer. That clearly indicates that tester
accuracy plays a role.
By selecting an industry-compliant tester like the DSX 5000 CableAnalyzer that offers repeatable,
exceptional accuracy, you can minimize marginal test results. And beware of any tester than allows you
to disable the asterisk and hide marginal results. This makes the tester non compliant; not to mention
that it can put your reputation and business at risk. Also, make sure to keep your tester in tip top shape--
keep up with the latest firmware, service your tester as needed, and make sure your permanent link
adapters are not worn out.

Despite all efforts, marginal passes will

happen--especially if the link contains a
consolidation point that adds another connection.
Some customers may refuse to accept a marginal
pass, wondering what happened to their supposed
"premium" cabling system. This is when you want to
make sure you're covered. Remember what the
standards say. Unless it was specified and agreed
upon in writing that marginal test results are not
acceptable, your marginal passes still pass. Agreeing
upon marginal test results upfront is a smart move.

When it comes to copper certification, it's not just the margins that have changed. With the
introduction of category 6A and a better understanding of applications, data transmission and what
impacts performance, we now have new parameters specified as a minimum in field testing. For
example, when it comes to certifying a category 6A copper system for 10GBASE-T operation, Alien
Crosstalk testing is required.
There are also other newer parameters that if not passed, can cause an application to have
issues. Mode conversion parameters like Transverse Conversion Loss (TCL) and Equal Level Transverse
Conversion Transfer Loss (ELTCTL) are used to measure the common mode signal within a pair,
indicating proper balance for high frequency or noisy environments. And we now have DC resistance
unbalance (discussed in our last blog), which can distort Ethernet signals in a PoE connection.

What you don't want is the time and expense involved in going back and retesting for parameters
you didn't test for the first time. Make sure you know which parameters are required and specify them
upfront so your technicians are prepared. And make sure your tester has the capability to test for every
parameter specified, especially the newer ones like TCL, ELTCTL and DC resistance unbalance--which by
the way, can all be tested with the DSX 5000 CableAnalyzer.

4. Wing it on Alien Crosstalk compliance
The cabling vendor offering the warranty may not require alien crosstalk testing. Often
considered optional by many, its not optional in either ANSI/TIA or ISO/IEC standards. Unless the test
specification states no alien crosstalk testing required, you, the installer could be asked to do it by the
end user, regardless of what the cabling vendor may say. Making the situation worse, this is a sample
test. With no agreement in place, you could be looking at 100% testing of the installation for PS NEXT
and PS AACR-F, at your cost.

Cabling Mistakes #4: Winging it on Alien Crosstalk Compliance
Lets take a look at the #4 Dumb Thing that smart people do when testing network cabling systems
Winging it on Alien Crosstalk compliance.

On the Hook of Off?
Alien Crosstalk is the key performance parameter that can
impact the ability of 10GBASE-T to operate, representing more than
90% of the total allocated noise budget. To certify a copper system
for 10GBASE-T, Alien Crosstalk testing is therefore required. But in
reality, all testing is optional. If everyonethe customer, installer,
consultant and cabling vendoris in agreement to not test for Alien
Crosstalk, then technically you are off the hook.
But to truly demonstrate compliance for 10GBASE-T
applications, you are still on the hook. Besides, given the
importance of the Alien Crosstalk parameter, most cabling vendors
will not warranty a system without testing it. But its not
economically feasible or practical to test all of my links for Alien
Crosstalk compliance, you say? Thats why standards recommend
specifying a sample size for Alien Crosstalk testing.
What do I test?
When specifying a sample size for Alien Crosstalk testing,
the ISO/IEC 61935-1 standard recommends testing a sampling
size as shown in the table. The standard also recommends testing
an equal number of short, medium and long disturbed links and
also states that if three of each of these disturbed links exceeds 5
dB in margin, its okay to stop testing.
To ensure that the connectors and cables will meet

requirements for PS ANEXT and PS AACR-F, you will need to
specify a sample size for Alien Crosstalk testing per standards
recommendations.

When choosing your disturbed links, it doesnt make sense to choose the link that terminates at
the end of a row of connectors as this is not the worst case scenariothe disturbed link should be
surrounded by connectors above and below. But make sure the disturbers are in the same bundle
Alien Crosstalk across cable bundles is not considered significant.

Shielded from Testing?
Shielded cabling offers far superior noise immunity compared to unshielded, and if installed
correctly, you should see hardly any Alien Crosstalk on a shielded cabling system. But if its not installed
correctly, even shielded cabling can fail. And as with any cabling, what seems to be theoretically
impossible seems to happen in the field.
For example, in a data center application where you are running from one grounded patch panel
to another grounded patch panel, an open shield (not connected) on a cable can result in failed Alien
Crosstalk testing even though Wire Map will still pass. This can be caused by not installing the shield
correctly, such as clamping down on the non-conductive side of the foil in the cable.

While most testers look for simple d.c. continuity between the shield on the main unit and the shield
on the remote unit, that d.c. signal will look for anyway it can to get to the remote unitincluding
through the common building ground to which the patch panels and racks are connected. That means
that the tester will show a connected shield even when it isnt. Thankfully this can be avoided with the
DSX-5000 CableAnalyzer, which reports distance to shield integrity issues using a patented a.c.
measurement technique.
Still think you can wing it on Alien Crosstalk testing? Think again. To certify a system, you must
specify a sample size per standards recommendationswhether the cabling is shielded or not. If you
dont, you risk having to test every link for ANEXTan extremely costly oversight.

5. Forget to enable plot data
A test report without plot data is an empty test report. While there is nothing in either ANSI/TIA-
1152 or IEC 61935-1 that states plot data must be recorded, plot data is the only way to really tell what
might be causing an issue with your link under test. And besides, alien crosstalk testing requires plot
data from in-channel tests. Furthermore, if you need Fluke Networks troubleshooting support and you
dont provide plot data, we will need it to help you, forcing you to retest with it enabled and more than
doubling your test time. There is nothing in either ANSI/TIA-1152 or IEC 61935-1 that states plot data
must be recorded.

Cabling Mistakes #5: Forgetting to Enable Plot Data
Lets take a look at the #5 Dumb Thing that smart people do when testing network cabling systems
forgetting to enable plot data

What is Plot Data?
Plot data is the full color graphical depiction of measured
test parameters required for certification testing. For copper
certification, plot data includes graphs for key parameters,
including Insertion Loss (IL), Return Loss (RL), Near End
Crosstalk (NEXT), Power Sum NEXT, Attenuation to Crosstalk
Ratio (ACR-N), Power Sum ACR-N, Far End ACR-F, Power Sum
ACR-F, Power Sum Alien Near End Crosstalk (PS ANEXT) and
Power Sum Alien Crosstalk Ratio Far End (PS AACR-F).
For each graph, the smooth red line shows the testing
limits of the standard youre testing to (i.e., TIA, ISO/IEC, etc.).
The jagged color lines show the results for individual pairs and
pair combinations. These lines are jagged since actual results
tend to have peaks and valleys, which is not a problem as long
as they stay above the limit.
Each of these graphs shows the values in decibels (dB) as they vary by frequencythe
horizontal axis is frequency and the vertical axis is dB. The horizontal axis will vary depending on the
type of cabling being tested. It will run to 100 MHz for Category 5e, 250 MHz for Category 6, 500 MHz
for Category 6A and so on. (You can also extend the frequency that you want to test to, which can be
handy for recertifying a cable to a higher standard.)

Better Than a Thousand Words

In the case of plot data, a picture is better than a thousand
words. It shows valuable visual information about the performance
of individual pairs within the cable with respect to the test limits.
Your customer will appreciate being able to visually see the
headroomreports without plot data will look empty.
Plot data is also the only way to really tell what might be
causing an issue with your link under test, and it is required for
relating test results to distance. Measurements made based on
frequency are great for certification, but are of little use for
diagnostic purposes. Its important to see where the crosstalk is
happening so the failing link can be fixed or replaces. Thats where
time domain data comes into play, which is achieved based on the
knowledge of the speed at which electrical signals travel over copper
cabling.
For example, Fluke Networks DTX 5000 CableAnalyzer use a patented digital signal processing
technique to report the distance to a location on the link under test where crosstalk or return loss is
excessive. The two parameters that show the time domain information are HDTDX (High Definition
Time Domain Crosstalk) and HDTDR (High Definition Time Domain Reflectometry). Plot data must be
enabled for these two parameters to be saved in the results.

Simply Required
Another reason for enabling plot data is that is it required for Alien Crosstalk certification testing
of Category 6A cabling. Furthermore, if you need Fluke Networks troubleshooting support and you
dont have plot data, we will ask you to retest with it enabled, more than doubling your test time. Our
experts cannot analyze testing results without it.
The default under instrument settings for the DTX 5000 CableAnalyzer sets plot data to
standard for displaying and saving plot data for frequency-based tests required by the selected test
limit. As a minimum, we recommend keeping it that way. Choosing extended will save data beyond
the frequency range required by the selected test limit.

6. Go with a non-EF compliant tester for testing multimode fiber
If your specification states ANSI/TIA-568-C, ISO/IEC 11801 or ISO/IEC 14763-3, youre

required to use an encircled flux (EF) compliant launch for multimode fiber testing. Measurement
uncertainty with EF-compliant launch conditions are significantly reduced, which can save you
from over-optimistic results that can leave your customers wondering why their application is not
performing. Some cabling vendors are requiring EF testing for warranties, while others will not
provide onsite support without seeing EF compliant results first .

Cabling Mistakes #6 - #7: Using Non-EF Compliant Cable Tester and Two-Cord
While the smart people who install, test and certify network cabling plants know the
importance of meeting standards performance parameters and ensuring application support,
even the best of us can make mistakes that adversely impact the bottom line and customer
satisfaction.
Lets take a look at the #6 and #7 Dumb Things that smart people do when testing network
cabling systems Using a non-EF compliant tester for testing multimode fiber and choosing the
two-cord reference for Tier 1 optical loss testing.

If It's in the Spec
Approved in October 2010 with the publication of ANSI/TIA-526-

14-B, Optical Power Loss Measurements of Installed Multimode Fiber
Cable Plant, encircled flux (EF) testing is now required by the ANSI/TIA-
568 series of standards. So if your specification states ANSI/TIA-568-C.0,
you must use EF compliant launches for multimode fiber testing.
If you think you can stick with the old method of using a mandrel
to stabilize a launch, think again. Decades ago when loss budgets were
relatively high, simple stability via a mandrel was all that was needed to
strip out modes more susceptible to loss and ensure accurate results.
But that's no longer the case with the tighter loss budgets of high-speed
10, 40 and 100 gigabit Ethernet over multimode fiber.

EF testing uses a more restricted template that more closely matches the tightly controlled
launch conditions of today's high speed optical fiber transceivers. This significantly reduces
measurement uncertainty that can save you from over-optimistic results and leaving your
customers wondering why their application is not performing.
So if you're testing today's multimode fiber, going with an EF compliant tester like the
CertiFiber Pro is a smart move. Besides, some cabling vendors are now requiring EF testing for
warranties, while others will not provide onsite support without first seeing EF compliant results.

Keep it at One
While a two-cord reference method while testing might appear to make sense, there is a
reason why it is not recommended. When both cords are referenced, one of the end connections
is improperly referenced out, providing overly optimistic or negative loss results.
In contrast, the one-cord reference method recommended by both TIA and IEC standards
assesses the condition of the channel end faces against the high quality multimode connector of
the test reference jumper and includes the loss of the connections at both ends of the channel.
Since most of the loss in a channel is caused by these connections, the one-cord reference
method delivers the highest accuracy. Thankfully a negative loss caused by a two-cord reference
will result in both a warning and a FAIL on the CertiFiber Pro. Besides, many cabling vendors will
reject results carried out with a two-cord reference, and that could prevent you from acquiring a
warranty.

7. Choose the two-cord reference for Tier 1 optical loss testing
While a 2-cord reference may seem easier, referencing both test cords results in optimistic
results and may provide negative loss results. For this reason, many cabling vendors reject results
carried out with the 2-cord reference, which could prevent you from acquiring a warranty.
ANSI/TIA and ISO/IEC only recommend a 1-cord reference. The test specification must call out the
reference method, and only the 1-cord reference includes the loss of the connections at both
ends of the link for the highest accuracy.

Cabling Mistakes #6 - #7: Using Non-EF Compliant Cable Tester and Two-Cord
satisfaction.
Lets take a look at the #6 and #7 Dumb Things that smart people do when testing network
cabling systems Using a non-EF compliant tester for testing multimode fiber and choosing the
two-cord reference for Tier 1 optical loss testing.

If It's in the Spec
Approved in October 2010 with the publication of ANSI/TIA-526-

14-B, Optical Power Loss Measurements of Installed Multimode Fiber
Cable Plant, encircled flux (EF) testing is now required by the
ANSI/TIA-568 series of standards. So if your specification states
ANSI/TIA-568-C.0, you must use EF compliant launches for multimode
fiber testing.
If you think you can stick with the old method of using a
mandrel to stabilize a launch, think again. Decades ago when loss
budgets were relatively high, simple stability via a mandrel was all
that was needed to strip out modes more susceptible to loss and
ensure accurate results. But that's no longer the case with the tighter
loss budgets of high-speed 10, 40 and 100 gigabit Ethernet over
multimode fiber.
EF testing uses a more restricted template that more closely matches the tightly controlled
launch conditions of today's high speed optical fiber transceivers. This significantly reduces
measurement uncertainty that can save you from over-optimistic results and leaving your
customers wondering why their application is not performing.
So if you're testing today's multimode fiber, going with an EF compliant tester like the
CertiFiber Pro is a smart move. Besides, some cabling vendors are now requiring EF testing for
warranties, while others will not provide onsite support without first seeing EF compliant results

Keep it at One
While a two-cord reference method while testing might appear to make sense, there is a
reason why it is not recommended. When both cords are referenced, one of the end connections
is improperly referenced out, providing overly optimistic or negative loss results.
In contrast, the one-cord reference method recommended by both TIA and IEC standards
assesses the condition of the channel end faces against the high quality multimode connector of
the test reference jumper and includes the loss of the connections at both ends of the channel.
Since most of the loss in a channel is caused by these connections, the one-cord reference
method delivers the highest accuracy. Thankfully a negative loss caused by a two-cord reference
will result in both a warning and a FAIL on the CertiFiber Pro. Besides, many cabling vendors will
reject results carried out with a two-cord reference, and that could prevent you from acquiring a
warranty.

8. Use patch cords to certify a fiber cabling system
Fiber patch cords typically have a maximum mated loss of 0.5 dB. That can and does result
in inconsistent readings, possibly failing perfectly good links. ANSI/TIA and ISO/IEC specify the use
of reference grade connectors for test cords, 0.1 dB for multimode and 0.2 dB for singlemode.
Such cords are often referred to as Test Reference Cords or TRCs.

Cabling Mistakes #8: Using BIMMF Test Cords and Common Mandrel
satisfaction.
Lets take a look at the #8 Dumb Thing that smart people do when testing network cabling
systems Using BIMMF (Bend Insensitive Multimode Fiber) test cords and a common mandrel.

Eliminating Unwanted Modes
Multimode fiber can support hundreds of modes. These modes are
categorized into lower order modes that travel a shorter distance closer to
the fiber core and higher order modes that bounce back and form from
between the core-cladding boundary and therefore travel a longer distance.
As you can imagine, these unstable higher order modes are much more
susceptible to bending losses.
Unlike lasers used in today's high speed fiber network that concentrate
light in the center of the fiber core, the recommended LED light source used
in testing tends to overfill a fiber, exciting both the lower order and the
higher order modes. So why are LED recommended for testing? The
underfilled launch of a laser can overlook high loss events such as a
misaligned connection, resulting in an overly optimistic lower loss
measurement and the potential for a bad link to PASS. Standards bodies
therefore long ago determined that an overfilled launch (i.e., LED) with
pessimistic results is the better option. www.flukenetworks.com| 2006-2017 Fluke Corporation 50
9/21/17
Unfortunately, exciting all the modes with an LED source also ends up
with wide variations in the measurement--as much as 0.1dB on every
connection. While that variation wasn't as much of a concern when
dealing with 1 and 10 Gig links, a 0.1dB variance per connection could
mean the difference between PASS and FAIL in today's advanced 40 and
100 Gig fiber links with much more stringent loss requirements.
Controlling launch conditions so that the light source excites
primarily the relatively stable lower order modes and removes the
higher order modes improves accuracy. One of the early and simplest
methods to do this was the use of a common mandrel. When a launch
cable is tightly bent around the mandrel, the higher order modes that
are more susceptible to this bending are removed so that only low
order modes emanate from the launch cable. Sounds like a good idea,
right? Well not so much anymore--especially when using BIMMF test
cords!
Goodbye Mandrel, Hello EF

While the mandrel does a good job of removing unwanted modes,
it still is not accurate enough for testing today's high speed fiber links.
That's why we now have the Encircled Flux (EF) method that more
precisely launches the light to better match the laser launch condition
of the transceiver (Click here to learn more about encircled flux).
If you do still use a mandrel, beware of your test reference cords.
Because test reference cords are not cheap, some may choose to use
normal cords. But normal cords are often made from bend insensitive
multimode fiber (BIMMF). And if you use BIMMF test cords in
conjunction with a mandrel, you are certainly asking for trouble. Why?

BIMMF was introduced to withstand tighter bends with substantially less signal loss, which is
critical in high speed fiber applications where the strict loss budgets leave little margin for bend-
induced loss due to installation. With BIMMF, a common 25mm mandrel therefore does not strip out
the higher order modes at the 850nm wavelength. While a smaller, specially designed mandrel might
do the trick with a BIMMF test reference cord, you can't use that same mandrel at the 1300nm
wavelength--you will likely find that your link passes at 1300nm but fails at 850.

Problem Solved
The answer is an easy one. Don't use BIMMF for test cords at all--even when testing a BIMMF link.
To find out more about the ins and outs of testing BIMMF, download the white paper here.
And it's time to put away that mandrel and use the latest EF method of testing. Not only is EF a
more accurate method of testing, it also required by industry standards for testing of today's fiber links.
An even easier way to ensure accuracy? Fluke Network's CertiFiber Pro Optical Loss Test Set
which is encircled flux compliant out of the box and comes with EF non-BIMMF test reference cords.
Problem solved.

9. Rely on a duplex tester for certifying MPO trunks
Testing 12-fiber MPO fiber trunks with a duplex tester requires 15 steps in the setup process, and
with the time pressure put on technicians in the field, the chance that they will follow all 15 steps is
slim to none. A tester with MPO testing capabilities, such as the Fluke Networks MultiFiber Pro,
requires just 5 setup steps and groups all 12 fibers of the MPO onto one test report. It also certifies 8
and 10 fiber count MPO links for the support of 40GBASE-SR4 and 100GBASE-SR10. Testing MPO trunks
with a duplex tester can make it virtually impossible to stay on schedule.

Cabling Mistakes #9: Rely on Duplex Tester for MPO Cable Certification
systemsrelying on a duplex tester for certifying MPO trunks.

Now the Challenging Norm
With the growth of big data, virtualization and higher bandwidth speeds in
data center fiber backbone links to 40 and even 100 Gbps, most data centers today
are standardizing on multi-fiber push-on (MPO) connector solutions. While required
testing of these 12-fiber MPO connectors may have been limited in the past, it is
now the norm in the majority of data center installations. Field testing is the only
way to ensure that MPO links meet the application performance requirements.
Despite the fact that pre-terminated MPO fiber cables are manufactured and
tested by vendors to comply with ANSI/TIA and international standards, there are
many factors that can potentially impact performance. First of all, MPO connectors
are harder to clean than duplex connectors. The 12-fiber MPO interface features an
array with a much larger surface area, which unfortunately makes it easier to move
contaminants from one fiber to another within the same array during the cleaning
process. 40 and 100 Gbps MPO fiber applications also have much lower loss budgets
so its important to ensure the highest testing accuracy as possible.
The Problem with Duplex Testers

It seems only logical that testing a 12-fiber MPO fiber link with a duplex tester would be time
consuming. In fact, it requires about 15 steps with a 3 jumper method and the use of MPO to LC fan-
out cords that separate the 12-fiber cable into single fiber channels (or LC-MPO cassettes can be used if
they have no more than a total loss of 0.35dB).
To test MPOs with a duplex tester, technicians first need to verify and set the SC-LC and LC-LC test
reference cords. A third LC test reference cord then needs to be verified, set and removed before even
connecting to the first pair of fibers on the MPOsand there are 5 more pairs to be tested on both
ends.
Not only is there a good chance that technicians pressed for time in the field will not follow all 15
steps, but the complex, time-consuming process inherently lends itself to the potential for greater
inconsistenciesand it makes it more difficult to keep all the connectors clean during the process.

A Better Way
Testers with an on-board MPO connector take the complexity, time and inconsistency out of MPO
testingeliminating the need for the fan-out cord or cassette. Fluke Networks MultiFiber Pro is the
first tester to offer this functionality with the ability to scan all 12 fibers simultaneously and display test
results in an easy-to-read bar graph. The MultiFiber Pro even has the ability to identify each fiber of the
MPO, providing more precise, per-fiber data and reporting to improve MPO connector verification and
troubleshooting.
MultiFiber Pro even has the ability to test for correct polarity, which ensures a continuous
connection from the transmitter to the receiver and requires patch cords with different polarity types
depending on the chosen polarity method. MultiFiber Pro allows the user to test individual patch cords,
permanent links, and channels for correct polaritywithout needing prior knowledge of the links
polarity.
The next time you think a duplex tester will suffice for your MPO testing, keep in mind that the
innovative features of the MultiFiber Pro will eliminate the complexity and allow you to test 90% faster.
And if you only have one MPO job for now, you can still save all that time by renting one call us and
well set you up! 9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 59
10. Skip out on proper fiber inspection
Contaminated connections remain the number one cause of fiber related problems, which is why
fiber end faces should always be inspected before connecting. Unfortunately, relying on subjective
human inspection produces inconsistent results. If you have inspection equipment, use it. And consider
using the IEC 61300-3-35 standards cleanliness grading criteria to avoid disputes. The latest generation
of field testers can automatically grade an end face in around 1 second.

Cabling Mistakes #10: Skipping Fiber Inspection Steps

systemsskipping out on proper fiber inspection.

That One Tiny Speck

Despite the constant reiteration and education surrounding the
need to inspect fiber endfaces before connecting any piece of
equipment, contaminated connections remain the number one cause
of fiber related problems and failures in a data center, Enterprise
backbones and other fiber networking environments.
Whether at a fiber cross connect, equipment port or at the end
of a jumper, wherever a fiber endface is exposed there is a risk of
contamination. Even the slightest particle on the core of a fiber can
cause loss and reflections that increase error rates and degrade
performance. And unfortunately, even the smartest fiber technician
who knows how to properly clean a fiber endface can make the
mistake of forgetting to properly inspect before making that final
connection.
And when it comes to inspecting DO NOT think that a manual inspection with a fiber microscope
is the best method of proper inspection. Depending on your experience, ambient lighting, eyesight, or
even how rushed or how tired you are, what you deem as clean with a manual inspection may not be.
There is always a chance that you didnt see that one tiny speck of dirt on the fiber core.
And with todays faster data rates of 40 and 100 gigabit having tighter loss budgets that ever
before, one tiny speck could mean the difference between moving on to your next job with a satisfied
customer and having to spend the extra time (and money) troubleshooting a problem.

Theres a Standard for That

Remember Apples ubiquitous catchphrase, "There's an app for that.? Well the saying in our
industry is, Theres a standard for that. And when it comes to fiber inspection, thats good news.
The IEC 61300-3-35 Basic Test and Measurement Procedures Standard for Fiber Optic
Interconnecting Devices and Passive Components contains specific cleanliness grading criteria to assess
pass or fail certification for inspection of a fiber end-face, removing the human subjectivity factor and
avoiding any disputes.
For various connector types and fiber size, IEC 61300-3-35 certifies the cleanliness of a fiber end
face based on the number and size of scratches and defects found in each region of the endface,
including the core, cladding, adhesive layer and contact zones.
And the even better news? Fluke Networks FI-7000 FiberInspector Pro certifies fiber end-faces to
IEC 61300-3-35 industry standards in just over a second, providing automated PASS/FAIL results that
take the human subjectivity (and hopefully mistakes) out of fiber inspection.

Thank You!

White Paper 10 Dumb Things Smart People Do When Testing Network Cabling Systems

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

White Paper 10 Dumb Things Smart People Do When Testing Network Cabling Systems

Загружено:

Авторское право:

Доступные форматы

White Paper 10 dumb things smart people do

when testing network cabling systems

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 1

1. Fail to specify the type of link testing you want.

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 3

Cabling Mistakes #1: Specify Channel Testing While Installing

While the smart people who install, test and certify

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 4

Channel vs. Permanent Link Refresher

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 5

Not Exactly Intuitive

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 6

Not Exactly Intuitive

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 7

Not Exactly Intuitive

Permanent link testing also offers your customers the benefit of

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 8

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 9

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 10

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 11

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 12

By selecting an industry-compliant tester like the

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 13

Despite all efforts, marginal passes will happen--especially if

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 14

Know Your Parameters

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 15

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 16

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 17

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 18

A Marginal Pass is Still a Pass

Back in the days of category 5, marginal passes were rarely

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 19

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 20

Despite all efforts, marginal passes will

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 21

Know Your Parameters

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 22

Know Your Parameters

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 23

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 24

Cabling Mistakes #4: Winging it on Alien Crosstalk Compliance

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 25

To ensure that the connectors and cables will meet

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 27

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 28

Shielded from Testing?

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 29

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 30

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 31

Cabling Mistakes #5: Forgetting to Enable Plot Data

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 32

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 34

Better Than a Thousand Words

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 36

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 37

If your specification states ANSI/TIA-568-C, ISO/IEC 11801 or ISO/IEC 14763-3, youre

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 38

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 39

Approved in October 2010 with the publication of ANSI/TIA-526-

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 40

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 41

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 42

9/21/17 www.flukenetworks.com| 2006-2017 Fluke Corporation 43