D. McKinnon, ast technology Labs, Inc. (To be ansi/tia-571-B) Telecommunications Telephone Terminal Equipment Environmental Considerations


C2. Traditional telecom surge specification



Download 202.91 Kb.
Page7/7
Date14.08.2017
Size202.91 Kb.
#32409
1   2   3   4   5   6   7

C2. Traditional telecom surge specification


When the 500-type set was being designed, a field study of induced lightning surges was conducted to provide design information for the insulation. Direct lightning surges were not studied. Surge volt-ages were measured behind a primary protector (using 3-mil carbon blocks) and found to have a peak value of 600 volts. The voltage waveshape (needed to determine the energy of the surge) had an envelope of a 10 ms rise time to peak voltage, and a 1000 ms fall time to half of peak voltage (referred to as a 10x1000 ms waveshape). The maximum induced voltage on the line side of the primary protector was about 1000 volts. Therefore, a protector functions mainly to arrest voltages from direct lightning hits and power crosses, which can be much higher than 1000 volts.

The peak induced current for cable runs is limited by cable impedance. The longer the cable run the higher the induced voltage but the higher the wire resistance. This results in induction looking like a constant current source of about 5 amperes but with a wide variation possible.

For metallic surges, current is limited by the impedance of the telephone terminating the line. The traditional 500-type telephone has an impedance of 600 ohms at lightning frequencies, but surge cur-rents saturate the hybrid voice transformer and cause its impedance to drop to the dc resistance of the windings, about 40 ohms. If a 1000 volt surge with no cable impedance is applied to such a telephone, the surge current would be 25 amperes.

For longitudinal surges applied to 500-type telephones, only insulation is stressed so that very little current actually flows unless there is a breakdown. The voltage and current waveshapes are identical.

To test 500-type telephones, the current should be limited by the equipment under test, not the surge generator impedance. Having a surge generator capable of delivering 100 amperes satisfies that aim and was a design parameter for the Bell System surge simulator. It was not based on field studies.

C3. Surge types

L-type (Longitudinal)


Lightning surges induce voltages onto both tip and ring conductors. When a power line fault causes arcing to cable pairs, the arcing usually occurs to both conductors. A ground potential rise has the same effect as a line fault but in the reverse direction. All of these result in a longitudinal voltage, which is a common mode voltage.

When several telephone line protectors are connected to the same protector ground, a discharge through some protectors can cause a ground potential rise. For telephone lines whose protector is not firing the ground potential rise adds to the voltage induced onto the telephone lines.


M-type (Metallic)


Metallic voltages (between tip and ring) are created when the primary protector grounds only one conductor of the tip-ring pair. The longitudinal voltage on the other conductor then becomes a metallic voltage, which is a differential mode voltage between tip and ring. A ground potential rise is common to both tip and ring, so it does not affect the metallic voltage.

P- type (Power)


Power line surges commonly appear on a grounding conductor (causing a ground potential rise for the telephone line as well as the power line) but the phase and neutral conductor can also be hit by lightning between the power company transformer and the building being served. A high voltage on the building’s grounding system can also arc over to phase and neutral, and the surge is then transmitted through the power system as a common mode voltage.

T-type (Transverse)


Equipment with a 2-wire power cord usually has no ground reference but a P-type common mode surge can appear on phase and neutral. The telephone line then becomes the ground reference for the surge and arcing can occur between the power line and the telephone line. This is known as a transverse surge. For equipment with a 3-wire power cord, transverse surges are still possible if the insulation between phase/neutral and the grounding conductor is better than the insulation between phase/neutral and the telephone line.

G-type (Ground)


A surge along ground leads or through building steel is common. Some equipment may have a protector ground, a power cord grounding conductor, and a permanent grounding conductor to building steel. Lightning could then produce voltage on one ground connection and not the other, producing a surge through the equipment ground.

I-type (Intrabuilding)


An intrabuilding surge occurs when the steel structure of a building conducts a lightning discharge which in turn induces a longitudinal voltage in telephone cables running parallel with the steel. This is a source of longitudinal voltages for cables that do not connect to the outside plant.

C4. Open circuit voltage and voltage waveshape


Induction voltages are usually less than 1000 volts peak, but have elongated (10x1000 ms) wave-shapes. Direct voltages are often 4000 volts or more, with short (1.2x50 ms) waveshapes.

Surges cause components to fail by different mechanisms, depending on the component’s weakness, and one surge parameter cannot account for all failures. The voltage parameters are:



  • Peak voltage: The peak voltage can cause carbon tracking in insulation and is a common source of breakdown. A peak voltage of 1000 V covers nearly all induced surge voltages.

  • Voltage heating: Leakage current through insulation can cause V2 dt heating of the insulation. This is not considered significant for insulators but played a role in deriving the FCC surges.

C5. Short circuit current and current waveshape


Peak available current and current waveshape are very important for CPE that use voltage and cur-rent limiting devices. In many older surge definitions the voltage waveshape and current waveshape were considered the same since the surges were applied to insulation and only leakage current resulted. In modern CPE, a surge protection element (like an MOV) becomes a low impedance during a surge and significant surge current flows. The current waveshape under these conditions is much shorter than the open circuit voltage waveshape, and 10x300 ms is a typical current waveshape.

The current parameters are:



  • Peak current: The peak current can cause heating of diode junctions that have a constant voltage drop. This is usually insignificant.

  • Current heating: The V2 dt heating of resistance elements can cause operation of fuses used to provide power line fault protection. This is a service affecting fault that should be avoided.

C6. Surge studies and data

Telephone line monitoring


The Bell System11 collected detailed lightning surge data at several locations in the 1970’s. The data did not distinguish between induced and direct lightning surges. The I2t plot followed a normal distribution with a maximum value of 0.6 A2-s, except for one surge that had an I2t of 1.2 A2-s.

Simultaneous voltage-current waveforms were also recorded, which showed that the events with the maximum voltage and current had a voltage and current decay time of less than 300 ms, and tended to be ringing waveforms (rather than simple exponential decays). While unipolar test waveshapes cannot capture the variety of actual waveshapes, the energy content can be replicated.


Survey data


The Bell Labs studies were detailed records taken at a few sites. Other studies surveyed many sites but with limited data. The two methods largely support each other. For example, in order to qualify solid-state protectors (SSP) for Central Offices, BellSouth Services conducted a surge survey12 that showed a maximum energy of 0.55 A2-s. Since the telecom installations around Central Offices is well controlled, it is likely that only induction surges were observed. The SSPs withstood these induced surges.

GTE Telephone Operations conducted a survey at the station end of the telecom loop13,, where the Bell System studies were made. Much higher energies were observed. Damaged SSPs were sectioned and the energy required to accomplish the observed damage was estimated. Also, damage from direct lightning was distinguished from damage from power cross (which was more severe). The maximum surge energy was equivalent to a 500 A, 10x1000 ms waveform, which has an energy of 175 A2-s. This is 3 orders of magnitude greater than the energy from induced surges.


C7. Standards on surges

FCC Rules, Part 68


The 600 V, 10x1000 ms waveshape used in the Bell System was not judged to be adequate when the FCC instituted a registration program for telephones. Industry wide, protector let-through voltages were more like 800 V peak, and that was the voltage selected for the FCC surge. However, the energy of the 600 V, 10x1000 ms surge as measured by V2 dt (=360), was kept constant for the 800 V surge by adjusting its waveshape to 10x560 ms (V2 dt =358).

For longitudinal voltages, a peak voltage of 1500 V was selected to represent a 1000 V surge and a 500 V ground potential rise. To maintain the same energy as used for metallic surges, the waveshape was adjusted to 10x160 ms which gives V2 dt =360.

Like the Bell System surge, the surge generator was specified to provide 100 A to make sure the generator did not limit the surge current, and no distinction was made between the open circuit voltage and the short circuit current waveshapes.

CCITT Recommendation K.17


The CCITT established a surge circuit for telecom purposes in Recommendation K.17. The open circuit voltage waveshape is about 10x700 ms, while the short circuit current is about 10x310 ms, in good agreement with field surveys. By specifying a surge circuit, a linear response is guaranteed for any load impedance. The FCC surge specifies only open circuit and short circuit waveshapes, and permits undesirable non-linear responses for loads other than at the specification points.

The surge voltage is not specified in K.17, since that is dependent on the installation. A value of 1000 volts metallic and 1500 volts longitudinal is appropriate. The output impedance of the CCITT circuit is high, 40 ohms, which limits available current at 1000 volts to 25 A. This results in an I2t of 0.136 A2-s for the metallic surge, which represents about 98% of all surges but is low compared to the maximum values seen in the surveys.

The energy from the longitudinal surge through each conductor should be the same as for a metallic surge, since a metallic surge is a longitudinal surge with one conductor earthed. This is accomplished for the level A surge with the CCITT generator. For level B surges, the power line generator is a better model for achieving the energy desired.

ANSI/IEEE C62.45


IEEE 587 (adopted as ANSI C62.45 with some updates) established a surge for power lines that has an open circuit voltage waveshape of 1.2x50 ms, a short circuit current waveshape of 8x20 ms, and an output impedance of 2 ohms. This is typical of direct lightning surges. The peak voltage is specified as 6 kV for location category B, limited by the arc over characteristic of power receptacles.

IEC 1000-4-5


This IEC surge standard has two circuits, the CCITT circuit for telephone line surges and the ANSI waveshapes for power line surges (giving a schematic but without values). The CCITT circuit with the 25  resistor bypassed is also shown. Thus, the CCITT and ANSI circuits have become the worldwide norms for surges. The standard gives severity levels for the power line surge of 1, 2, and 4 kV, de-pendent on the building installation category.

The open circuit voltage is specified at the output terminals of the surge circuit. Some standards specify the voltage the capacitor is charged to. To achieve 5000 volts at the output terminals, the capacitor must be charged to 6000 volts because of the voltage divider in the output.


C8. Surge likelihood

Level A and level B


Induced surges are described as level A or level B. Level A does not mean average, but that level of stress that equipment must withstand to have a reasonable life. Level B surges represent the envelope of surge energies that equipment also needs to withstand. However, the equipment sees many more level A surges than level B surges.

Level C


Terminal equipment is normally protected against direct lightning strikes by protectors, but occasional poor grounding is unavoidable. Therefore, it is desirable for terminal equipment to withstand direct strikes, and must at least be safe under such conditions. The level C lightning condition is taken from the Bellcore generic requirement GR-1089.

C9. Surges for telecommunications equipment

Metallic


Surge M-1 uses the CCITT generator. With the 25  resistor at 1000 volts, the I2t is 0.136. Surge M-2 uses the FCC 800 V, 10x560 µs generator which produces an I2t of 4. The M-3 surge uses the GR-1089 1000 V, 10x1000 µs generator which produces an I2t of 7.

Rational: These changes are required to make this section correspond with the changed requirements in SP 3283-A-2.


Longitudinal


The level A longitudinal surge uses the CCITT generator at 1500 volts, with an I2t of 0.129 for each leg. The level B longitudinal surge actually represents a ground surge and has an I2t of 1.01 for each leg. The level C surge has an I2t of 7 for each leg.

Power


The power line surges use the IEEE waveshape and an exposure likelihood based on IEC 1000-4-5. That is, surges on phase and neutral rarely exceed 2500 volts, the value used for the FCC power line surge. Ground surges are more likely to reach 5000 volts, the value used in TIA-571. The I2t at 2500 volts is 17.5, and at 5000 volts is 70.

Transverse


A transverse surge occurs between phase/neutral and tip/ring (acting as the ground path). Because the telephone line has more resistance than the power line, an additional 3 ohms is added to each tip and ring lead based on field experience. The I2t at 2500 volts is 2.8, and at 5000 volts is 11.2.

Ground


A ground surge is the same as a power line surge on the grounding conductor.

ANNEX D (Informative) GROUNDING PRACTICES


The following telecommunication equipment grounding practice is frequently used:

  1. All circuit commons within the equipment enclosure are derived from a single ground concentration point within the cabinet. Each cabinet's ground concentration point derives ground from a single ground concentration point serving all system cabinets and peripherals collocated with the system.

  2. The system cabinets and all associated ducting hardware along with all collocated peripherals are not connected to any ground source other than the system single-point ground, de-scribed in (1).

  3. Service wires bringing commercial power to the cabinets do not share an enclosure or raceway with any other system grounds, dc power distribution wires, or signaling wires. Commercial power terminations not made by means of a connector are enclosed by race-ways and termination boxes, whether these enclosures appear outside or within system cabinets. This is to ensure that ac service wires cannot fault to circuitry within system cabinets or associated ducting hardware.

  4. All system hardware are provided with an ac fault return path to the system single-point ground which, in turn, is provided with a reliable path to the equipment’s grounding conductor. The path from system equipment to single-point ground need not be a direct, dedicated path but can be any reliable path to other system hardware which receives the above grounding path.

  5. All sources of earthing (i.e., system signaling ground to the approved ground source etc.) connect only to the system single-point ground. The intent of providing for a system single point ground is to minimize ground loops and prevent lightning from finding a path through system components.

Other techniques that achieve equivalent results are also used.

1 In Canada, comparable regulations are IC CS-03 (Ref. A9) and IC ICES-003 (Ref. A10)), and comparable safety requirements are in CSA C22.2 No. 225 (Ref. A11) or CSA C22.2 No. 950-95 (Ref. A12), respectively.

2 The FCC Part 68 Rules specify 18 drops in testing for network harms. The Rules are not concerned with proper operation after the drop stresses.

3 The telephone line surges are from CCITT K.17 (Ref. A17), and the power line surges are from IEEE C62.45 (Ref. A18).

4 The FCC surges used to determine network harms are as follows:

Power2,500 V1,250 A2x10 sL-to-N3 surgesMetallic800 V100 A10x560 sT-to-R1 surgeLongitudinal1,500 V100 A/lead10x160 sT/R-to-G1 surge



5 In some extreme environments, e.g. close to broadcast transmitters, field strengths up to 20 V/m at frequencies up to 1 GHz have been observed. Under such extreme environments, equipment compliant with TIA-631 may experience RF interference.

6 The waveshapes and calibration methods are given in IEC 1000-4-2 .

7 A calibration method is given in MIL-STD-883C, Test Methods and Procedures for Microelectronics.

8 Besides ESD currents carried on intentional paths to ground, the free space capacitance of auxiliary equipment may cause an interconnecting cable to carry significant ESD current even though the auxiliary equipment has no path to ground.

9 The correction is for the scanning voltage only, to determine if breakdown will occur at a given point on the equipment. The compliance test voltage is not adjusted for altitude. The correction is from IEC-950.

10 A loop current interruption of 300 ms or more is considered likely to cause a flash that may result in disconnect.

11 R. L. Carroll, “Loop transient measurements in Cleveland, South Carolina,” and other articles, BSTJ, November 1980.

12 Mel Thrasher, “A solid-state solution,” Telephony, June 12, 1989.

13 C. A. Francis II, W. J. McCoy, “The analysis of solid-state overvoltage protection at customer premises locations,” Compliance Engineering, May-June 1996.


Download 202.91 Kb.

Share with your friends:
1   2   3   4   5   6   7




The database is protected by copyright ©ininet.org 2024
send message

    Main page