CPD Troubleshooting

1
CPD Troubleshooting
2
The CPD Problem

• Most frustrating, time-consuming network issue to
  troubleshoot
• Comes and goes unpredictably
• Disappears when you think youve found it
• Reappears after the problem has been fixed
• No cable service provider is immune

3
Why treat CPD?

• CPD greatly diminishes network performance and
  reliability
• Services with higher modulation formats are
  easily distorted by CPD
• digital video
• High-speed Internet
• Digital phone services

4
What is CPD?

• An intermodulation distortion of downstream
  signals due to nonlinearities found at metallic
  junctions.

5
The Common Path
6
Where is CPD Found?

• Every connector interfaces in coaxial plant is
  potential source
• Terminators
• Splitters
• Amplifiers
• Junctions
• Splices
• Taps
• Filters

7
Physics of CPD

• Contaminated interfaces oxidation
• Bimetallic Corrosion - Electrochemical corrosion
  between dissimilar metals
• Single Metal Corrosion - Corrosion from air and
  moisture in unprotected metals
• Crevice Corrosion - Connectors
• Stress Corrosion - Stress corrosion at bend,
  shear, and clamp points
• Temperature, vibration, humidity
• Lubricants, cleaning solvents, skin oils
• Diode Effect

8
Causes of CPD Loose Connections

• At microscopic level, contacts are comprised of
  many contact surfaces separated by cavities
• Surface area and pressure govern conductivity
  characteristics
• As connectors age, adequate contact surface area
  and pressure may be lost, creating CPD
• Insertion/installation may create stress points
  that are catalysts for corrosion formation
  leading to CPD

9
Nonlinear Intermodulation Diode Effect

• Nonlinear - A signal that passes through a device
  and the shape of the signal from the input to the
  output is changed.
• Beats
• F1 - F2
• F1 F2

Channel Beat
Channel Beat
Channel 7 175.25 MHz Channel 8 181.25
MHz Channel 9 187.25 MHz Channel 10 193.25
MHz Channel 11 199.25 MHz Channel 12 205.25
MHz Channel 13 211.25 MHz
8 minus 7 6 MHz 9 minus 7
12 MHz 10 minus 7 18 MHz 11
minus 7 24 MHz 12 minus 7 30
MHz 13 minus 7 36 MHz
8 plus 7 356.50 MHz 9 plus
7 362.50 MHz 10 plus 7
368.50 MHz 11 plus 7 374.50 MHz
12 plus 7 380.50 MHz 13 plus 7
386.50 MHz
10
Return path without CPD
Signs of CPD

• 6 MHz spikes in return path 6, 12, 18, 24 MHz
  ..
• Raised return path noise floor
• Elevated CSO/CTB in the downstream

Return path with CPD
11
Why treat CPD? - cont

• Non-linear Intermodulation
• Ingress
• Impedance mismatch / suck outs
• BER/MER
• Reflections

12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
Severity, Unpredictability and Prevention

• Degree of corrosion in the formation of a diode
• Distance of corroded connector and forward output
  of an amplifier
• Magnitude of the forward and reverse signals at
  the CPD source
• Day-Night Effect and Temperature
• Vibrations
• Clean connectors
• Dry connectors
• Proper connection

29
Conventional CPD Remediation

• Wait until the problem is severe
• Use two or more techs
• One to monitor spectrum analyzer
• One or more techs in the field to disconnect
  devices until problem is isolated
• Labor intensive and disruptive
• Unreliable with no positive confirmation of fix

30
Conventional CPD Remediation

• Last amp in
• First amp out
• Halves
• From known feeder leg termination back to amp
• Contractors to change out equipment

31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
(No Transcript)
42
(No Transcript)
43
(No Transcript)
44
Go Interactive
45
Procedures and Results
46
Background

• The following are results from the first week of
  the live trial performed from 12/3/07 through
  12/7/07.
• Network troubleshooting using Hunter was
  performed 12/4 12/6.
• Two teams of technicians with Arcom
  representatives worked in 10 of the 32 designated
  trial nodes fixing 13 impairments using Hunter.
• The trial effectively demonstrated that CPD
  sources could be easily identified and
  permanently repaired.
• The following is a description of the work that
  was performed, procedures taken and ending
  results.

47
Node WI137 step 1

• This node presented a CPD source located at
  120.525uS from the headend server. Once
  determining to work on this this node, the first
  step was to go to the node and isolate which leg
  the source of the impairment is located.

48
Node WI137 step 2

• Using the Quiver handheld radar, we measured CPD
  9.15uS from the node coming from the west leg.
  After inputting the time-distance into Quiver
  Navigator maps, Navigator directed us to the next
  logical split at amp AM0022.

49
Node WI137 step 3

• Now at amp AM0022 we used the Quiver handheld
  radar again and without taking the network down
  we were able to measure a CPD time-distance at
  4.87uS coming from the western leg. After
  inputting this time-distance into Navigator we
  determined to proceed to amp AM0034.

50
Node WI137 step 4

• At amp AM0034 we used the Quiver probe on the
  seizure screw test point of splitter SP0012 and
  measured CPD 2.4uS heading north. Quiver
  Navigator showed the source of the impairment to
  be located at tap TP0197 or terminator TM0026.

51
Node WI137 step 5

• At the tap TP0197 and terminator TM0026 location,
  technicians went through their normal
  troubleshooting of visually inspecting the
  equipment, tightening connections and replaced
  line terminator TM0026. The pin of the
  terminator was corroded and loose. Using headend
  mode in the Quiver we verified that the CPD
  impairment was fixed.
• We were able to locate and fix the source of this
  nodes impairment within 50 minutes and in 3
  determined stops after the node.

52
CPD Stats for WI137

• The graph below represents the CPD statistics for
  Node WI137 from 12-4-07 through 12-7-07. This
  nodes impairment was in high alarm from 12pm to
  11pm on 12-4-07. Even though this impairments
  amplitude dropped to a low level on 12-5-07, we
  were able to track and fix during normal work
  hours using Hunter.

9.15uS repair
53
Node WI122

• This node had 3 impairments fixed over the first
  week of the trial. On 12/4/07 there was CPD at
  4.15uS from the node. In 3 stops we found and
  fixed a cracked cable at the input connector at
  the tap. On 12/6/07 there was CPD 8.4uS and
  10.89uS from the node. Both impairments were
  tracked to line terminators and were fixed within
  1 hour.

54
Node WI122 Graph 12/4 12/7
4.15uS repair
8.4uS repair
10.89uS repair
55
Node VC003

• Using Quiver at the node we measured CPD 13.78uS
  (approx 6063 feet from node). By knowing the
  time-distance and using our Quiver Navigator maps
  we were able to bypass 5 amplifiers, 6 splitters
  and 4 DCs. The impairment was found in 4 stops
  after the node and fixed within an hour and a
  half. Source was a line terminator.

56
Node BA001

• BA001 showed 2 sources of CPD at 2.62uS and
  10.8uS from the node. We were able to track both
  concurrently down the same leg. Sources split
  off of different legs out of amp AM0031. After
  fixing the closer impairment we went back to
  AM0031 and proceeded to track and fix the farther
  impairment. Both sources were line terminators.

2.62uS repair
10.8uS repair
New intermittent CPD source
57
Node WI042

• We measured CPD 4.60uS from the node. In 2 stops
  from the node we located a cracked cable wrapped
  in aluminum foil and tie wraps. I new piece of
  cable was spliced in and fixed the impairment.

58
Node WI043

• Measured CPD 5.64uS from node. In 4 stops we
  tracked source to a terminated tap with a broken
  port and 2 blackened terminators

59
Node WI093

• Distance to the CPD source from the node was
  3.87uS. In two stops after the node source of
  impairment was a bad terminator on a DC-7

60
Node WI048

• CPD source measured at 9.505uS from node.

Node BA006

• CPD source was measured 7.575uS from node.

61
Before/After CPD Statistics

• CPD Alarms 12-4

• CPD Alarms 12-7

62
Before/After CPD Charts

• Nodes with CPD 12/04/07

• Nodes with CPD 12/07/07

63
References
Bharat Patels - Report on Common Path
Distortions http//www.arcomdigital.com/files/CPD_
report.pdf Ronald Mesavage Common Path
Distortion Communication Technology June 1,
2006 http//www.cable360.net/ct/data/15091.html
www.arcomdigital.com