International Telecommunication Union   ITU
عربي  |  中文  |  Español  |  Français  |  Русский
 
 Advanced Search Advanced Search Site Map Contact us Print Version
 
Home : ITU-T Home
   
ITU Telecommunication Standardization Sector (ITU-T)
Handbook Contact: tsbedh@itu.int

Mitigation measures for telecommunication installations

PART 2: CASE STUDIES


Case study #

2.7

Download PDF ()

Title

Lightning damage to a telephone

Type of trouble

Damage.

Source of trouble

Lightning.

System affected

Customer equipment.

Location

Customer premises.

Keywords

Damage, lightning surge, inherent resistibility, GDT

Version date

2004-01-01

 

System configuration
The problem equipment is a telephone powered from the telecommunication line, in a customer premises. An integrated circuit (IC) in the phone was being physically damaged (impulse type damage). It was assumed that lightning was causing this damage and loss of function. Of the 50 customer premises investigated, half had primary protection and the other half did not. Most customers remembered a very close lightning strike.



Figure 2.7-1 – System configuration

 

Measurement/Searching techniques/Experiment

Identifying failed components

The first part of the investigation was to identify the damaged components using normal service techniques. The component most damaged was a transmission IC. The plastic coating on top of the IC was then removed using chemical etching to expose the die. The die was then inspected using an optical microscope and photographed. Typical field damage can be seen in Figure 2.7-2.



Figure 2.7-2 – Damage to the transmission IC by lightning strikes
 

This damage is due to a breakdown between a conductor track and a ground as a result of an impulse overvoltage.

Replicating field damage in the laboratory

In an attempt to replicate this damage, different types of lightning surges, as recommended in Rec. ITU-T K.21, were applied to the telephone. No damage could be caused using up to 4 kV 10/700 µs waveshape, without external surge protective devices (SPDs), i.e., an inherent test voltage of 4 kV. When the voltage was further increased, damage could be caused but it was a different type of damage.

The next type of test was to investigate whether the operation of a GDT or sparkgap could cause a very high dV/dt and replicate damage to the IC. This test was performed with a GDT or sparkgap across the line or line to earth. Voltages as high as 6 kV could not damage the IC.

The next experiment was to check if high voltages (up to 100 kV) applied longitudinally could cause high charging currents and damage the IC due to the capacitance of the phone to earth. The charging currents did not cause damage but when one side of the line discharged to earth, the damage was replicated.

The current waveform of the current conducted transversely through the phone, when one side of the line breaks down, is as shown in Figure 2.7-3. This waveshape caused similar damage to the audio IC compared with field damaged ICs.



Figure 2.7-3 – Current waveshape entering the phone to replicate field damage
 

Increasing the resistibility of the phone

To "harden" the phone against damage, the current path from the line into the IC was identified from knowing the entry pins on the IC. The circuit was redesigned to minimize inductance of SPD lead lengths and components added to minimize the amount of current which could enter the IC. The inherent resistibility of the phone was increased fourfold from approximately 200 A to 800 A. Hardening the phone to this level eliminated damage to the phone, saving millions of dollars in maintenance per annum.

Unfortunately, while damage was no longer occurring, customers began complaining about stored number loss. Stored number loss was replicated by using the waveshape above but with a very high frequency oscillation added to the first half cycle.

The test generator used for the hardware damage test and the memory loss test is given in Figure 2.7‑4.



Figure 2.7-4 – Test generator for hardware damage and memory loss test
 

Lead lengths

The lead resistances and inductances are critical factors in determining the maximum current amplitude obtainable for a given generator source voltage and hence, the shorter the lengths (phone & earth) involved, the less source voltage required.

Test method

  • Hardware component test: Spark gap 2 shall be closed. Spark gap 1 shall be opened to 1 mm. The voltage source for the generator shall be slowly increased until breakdown just occurs across Spark gap 1. The breakdown across the spark gap will be repeated as the capacitor recharges and discharges. The peak-to-peak current, of the first cycle, flowing to ground shall be measured and noted. The waveform shall be the same as the ringwave. The phone shall be subjected to 10 ringwave impulses before the telephone is tested for correct operation. Spark gap 1 shall then be opened in 1-mm steps and the process repeated until a peak current of 800 A is achieved. This is the current that the telephone is expected to resist without damage or misoperation. A new phone sample may be tested at this point to remove any deleterious combined effects of previous testing.
  • Memory retention test: The opening of Spark gap 1 shall be adjusted to achieve a peak current of 400 A. It will be necessary to reduce the source voltage until a controlled discharge repetition rate is achieved. Once this current has been achieved, Spark gap 2 shall be opened to 2 mm. A minimum of 10 ring-wave impulses shall be performed on the telephone after which correct operation and memory retention shall be confirmed.

NOTE – The first half cycle of this ringwave will have a very high frequency ring wave superimposed on it. No attempt should be made to characterize this very high frequency ring wave as it would be exceeding the oscilloscope and current transformers operational performance parameters. Furthermore, it would require specialist measurement techniques, which are beyond the scope of this handbook.

Surge coupling

Because half of the services had GDTs installed, it is assumed that the coupling must have taken place between the phone and the GDT. Two methods have been thought of:

  • Induction into the internal cabling: This is thought unlikely as the surge is a transverse current requiring breakdown to ground of one side of the cable. This would require an induced voltage of tens of kilovolts.
  • Breakdown from earth to one side of the cable due to a high earth potential rise (EPR) due to a close ground strike: In Australia, indoor cabling is often tacked under the floor of the building and hence breakdown is possible.

 

Mitigation method/Results/Conclusion

The damage was solved by developing a test to replicate the field damage and by increasing the inherent resistibility of the phone fourfold. It should be noted that the original phone had an inherent resistibility of 4 kV 10/700 µs. Mitigation was very successful as the damage bill was dramatically reduced.

 

References

Recs ITU-T K.21 and ITU-T K.44.

 

 
Table of content

 

Top - Feedback - Contact Us -  Copyright © ITU 2010 All Rights Reserved
Contact for this page : TSB EDH
Updated : 2010-06-29