Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
SECTION 5 PROTECTIVE DEVICES TARIFFS AND PROTECTIVE DEVICES Unrestricted interconnection of user-owned communications devices or of privately owned unregulated communications systems to.the public telephone network, as discussed in detail in Section 3, introduces the possibility of harm to the users of the networks in the form of degraded performance or an increase in the hazards of exposure of carrier personnel to dangerous voltages and currents. As a safeguard against these potentially harmful effects, AT&T has incorporated in FCC tariffs 259, 260, and 263 not only protective criteria relating to levels, bandwidth, and signaling frequencies, but, in some cases, a requirement for carrier-furnished and installed protective and coupling arrangements to be placed between the telephone network and customer-owned and customer-maintained equipment and systems. Private-line customers obtaining service under FCC tariff 260 are not, in all cases, required to obtain protective devices. This Section discusses this concept of protection along with alternative arrangements. At the present time, the selection of devices and priority of design and manufacture rests with the carrier. The number of different types of coupling devices available is limited and are intended to fill immediately-known requirements. They are to be followed by additional types as needs are identified, economics are justified, and as development is completed. Systems innovation and development of user-owned devices may be influenced by the willingness of the carrier to produce specialized interface units. This approach will be discussed in depth in later portions of this section. PROTECTION AFFORDED BY PRESENT CARRIER-FURNISHED DEVICES It is not intended here to provide a detailed description of every available coupler. Each is described in detail in a Bell System Technical Reference. The couplers are similar in their basic functions, which are: (a) To isolate the line from hazardous voltages (b) To limit signal levels (c) To preserve longitudinal balance - 40 -
- 41 - (d) To protect the network control and signaling functions In its simplest form, the coupler is designed around an isolation transformer that interfaces directly, via a jack, with the user-owned equipment. This transformer serves three functions: (a) It ensures longitudinal balance on the loop regardless of any unbalance in the customer's equipment (b) It isolates DC currents in the customer's equipment from the loop (c) It prevents hazardous A.C. voltages from being impressed on the loop by virtue of its saturation capability Varistors, shunted across the line side of the transformer, limit peak signal voltages. A capacitor in one side of the line blocks line current from saturating the transformer core. Some of the more complex forms of coupler include a more sophisticated signal limiter designed to reduce distortion of data signals that exceed the allowable limits. Others include arrangements for signaling and supervision, either manual or automatic, answering only, or answering and calling. Coupler for interfacing customer-owned PBX equipment are much more complex units. Degree of Protection 1. Hazardous Voltages. The major hazard is that involving personnel and the protection provided here is excellent in the carrier-furnished units. A saturable transformer is an effective method of protection. Fuses and circuit breakers rated for equipment protection do not provide personnel protection. 2. Signal Amplitude. The protection provided here also is excellent. The various types of limiter all ensure that proper levels are not exceeded. 3. Spectrum Limitations. No attempt is made in any of the couplers to limit signal spectrum. The couplers provide no protection against unwanted frequencies. 1Section 3
- 42 - Longitudinal Unba Lance ,". The isolation transformer provides 4., excellent protection against any defects in the customer's equipment or installation that could cause unbalance on the user loop and consequent hazard of cross talk and noise. 4. Improper Network Control Signaling. The subject of network- control signaling and the consequences of improper control are dealt with 'in some detail in Section 4. In this section, conclusions are reached as to the effectiveness of the current carrier-provided interface arrangements in preventing improper network-control signaling. The degree of protection afforded to customer-generated network-control signals is minimal. DC isolation is indeed provided between the customer's equipment and the loop, but since signals are usually merely repeated, there is no protection against dial-pulse speed variation, make- break ratio (in most cases), or repetitive dialing from a malfunctioning auto-calling device. In certain cases, particularly with relays that repeat dial pulses, the coupling device can, in fact, degrade the dial pulses by inferior timing characteristics of the relay. In another instance, the dial repeating function in one of the protective devices was less tolerant to dial pulse variation than if no protective device were used. In this latter case, AT&T is redesigning the unit. IDENTIFIED ISSUES Reliability The protective arrangement or coupler introduce another electronic box into the system. What are the chances of failure occurring in a coupler with an attendant reduction in reliability? The answer, of course, depends on the complexity and soundness of design of the coupler. In the very simplest type of voice coupler, several solid-state diodes and an isolation transformer are all that is involved. Since all elements are solid-state, life under normal operating conditions is indefinitely long. Transformer- insulation failure at telephone-line voltage is extremely rare unless the quality of the insulation is initially poor. At the next higher level of coupler complexity, the diodes are replaced by amplifiers and an AGC circuit with power supply. Additional resistors, capacitors, transistors, and diodes are introduced. Under normal conditions, the life of this sort of coupler should be comparable to the life of the attachment. Certain of these couplers use relays for
- 43 - dial-pulse repeating. Relays are notably poorer in reliability than solid-state devices and can, therefore, be expected to have a somewhat higher, but still acceptable, failure rate. Redundancy Redundancy, for purposes of this discussion, means that essential functions are duplicated in the coupler and in the devices attached to it and the requirement for protection, in many cases, can cause such a redundant condition. For example, redundancy occurs in some of the couplers provided for use with PBX's; In these cases ,all functions of the coupler are repeated within the PBX itself from transformer isolation to regeneration of subset dial pulses which themselves may meet the dial criteria. One approach would be to delete the redundant features from the user equipment designed for interconnection to the common carrier network. A manufacturer, on the other hand, would then be required to supply two types of equipment -- one to interface with the carrier provided coupling unit and another where a coupler is not required. Another approach would be to allow interconnection under the provisions of a Certification Program. 2 Transparency Ideally, the protective device should be "transparent"; that is, its presence should have no effect upon normal system functions. In this connection, the present coupling units are not transparent in that they do not pass DC due to the transformer prOVided for line balance. "Transparency" has another, and somewhat different, meaning to the designer of equipment attached to the telephone network. The ideal protective device to him is one that does not require that he make design changes in his equipment. For example, the AT&T CDR coupler for PBX's presents a manufacturer of PBX's with a ten-terminal interface, whereas a PBX is designed for a two-terminal connection direct to the carrier's line. Certainly, the greater the transparency of the protective device, the fewer the problems presented to the designer and manufacturer of terminal equipment. As with the redundancy case, transparency can be improved by cooperative action by the carrier and the supplier of attachments to produce improved couplers or by incorporating the protection into a unit built under an enforced certification program. 2section 6
- 44 - Availability Ce.ta~n types of p.otective devices a.e said not to be available. AT&T states that the most f.equent1y .equi.ed types a.e available and the cauie. is p.oc~edingwiththedeve1opmentof othe.vadeties. They funhe. state that the sUddenness of the tariff filing created problems with regard to the supply of protective devices. A m~imum number of types were ready for dist.ibution at the time of, and shortly after, the filing. Nevertheless, a number of users have complained about lack of availability of announced units. Some have complained that, due to the difficulty in defining all protective requirements in advance, design and production of devices by the car.ier could unduly delay installation of systems. There is also concern on the part of manufacturers that their desi.e and ability to innovate will he limited by the decisions of the carriers. At this time, availability is further complicated by a lack of a firm interpretation of tariff language. A lack of uniform interpretation among the many telephone companies and the various state Public Service Commissions is also a factor. Power Supply Dependence P.otective arrangements (above the simplest level) requi.e a source of power and typically commercial AC power is used. In the event of a power-line failure, therefore, the protective arrangement becomes inoperative. Communications within the customer's site can continue if the customer has provided emergency power for his own equipment, but communications with the outside world, where it is most needed, is cut off. This problem can be resolved and, fortunately, many solutions exist. Automatic means for bypassing the coupler in the event of an emergency is one possib1ity. The problem disappears, of course, if the protection is incorporated into the design of the user's equipment. Glare "Glare" is a condition that occurs on trunks or lines when the circuit is seized at both ends at, or nearly at, the same time (or during what is called the "unguarded interval"). When this happens, the switching machines at each end of the circuit are confused, each fruitlessly waiting for an answer from the other end. Early-type protective couplers were designed to a 1.5 second unguarded interval. The addition of this coupler introduced a three-fold increase in potential glare with customer-provided PBX's over normal operation. However, a field change order for all CDH units, which reduces their unguarded interval from 1.5 to .5 seconds, has been issued. The risks of glare with this change are no different with the protective coupler and user-provided PBX equipment than that with carrier-provided PBX' s. The increase in glare incurred by the addition of the protective arrangement would, therefore, appear to be a minimal problem at this stage.
- 45 - Transmission Degradation Although the ideal protective arrangement should be without loss, coupler losses amounting to 2 db to 3 db are practically achieved. Normal variations of attenuation in the received signal of the line can vary over. 10'sof dB's due to differences of loop length and other circuit variables. Therefore, losses induced by the coupler are small compared to normal circuit variations. There is usually no problem in compensating for this additional loss. Most modems and other attachments have adjustments or taps by which these losses may be fully compensated. Packaging The carrier-supplied protective device now appears as a separate entity in its own cabinet or box. While clean-cut from the carrier's point of view, it represents to the user just another box that has to be put somewhere. Presently, the protective device cannot be physically located in the customer's equipment, although the carrier indicates it is willing to discuss this issue. Integrated Protection Assuming a Certification program3to allow direct connection between carrier and users, the following are some factors involved in the inclusion of the protective arrangement within the equipment cabinet. Redundancies can be removed in various ways; 1. one way is through repackaging. Amanufac- turer, having complete control over both the protective device and his own attach- ment, will tend to eliminate all redundancies in order to get the best cost advantage. 2. There may be small maintenance advantages. An interface of two wires is easier to maintain than the interface of eight or more wires of the more complex couplers. 3. There are fewer hardware variations. Manufacturers of the user's equipment will build the protective arrangement from the same hardware building blocks that are used i~ the rest of this equipment. The number of types of spare assemblies needed for maintenance is consequently reduced. 3Section 6
46 4. The appearance of the installation is enhanced if there is one less box to contend with. The space occupied by the protective arrangement within the user equipment should be considerably less than as a free-standing box. The sharing of common facilities (power supplies, framework, etc.) will contribute to the better packaging efficiency. 5. No conclusions can be drawn with regard to manufacturing-cost advantages. It appears that a large-volume manufacturer would have a manufacturing-cost advantage through elimination of redundancies and the sharing of common facilities (as discussed in land 3). 6. A built-in protective device has greater potential for mobility where that feature is important. Carrier-supplied protective devices would other- wise be required at each point of use of the portable attachment. PROTECTION AT THE TELEPHONE CENTRAL OFFICE This section discusses the feasibility of transferring the protection function to the telephone central office itself. Perhaps the most significant observation to make about providing protection in the central office compared to protection of the customer's station is that no protection can be provided in the central office for certain effects. Protection at the centraL office cannot affect high-level signals that cause cross talk in exchange cables, high voltages that may be hazardous to those working on the loops, or unbalance which destroys the inherent balance of cables. Protection in the central office could, in principle, prevent excessive levels on carrier systems in the trunk plant. Present central-office designs, however, do not provide facilities to limit signals to the levels required to prevent overloading carrier systems or to prevent cross talk in loops or on voice-frequency intertrunks. In any case, such facilities, if provided, would also have to be provided on a per-loop basis or switched into service as required. At this writing, the Panel does not have enough information to make recommendations. OTHER PROPOSED PROTECTION ARRANGEMENTS Although the present coupling arrangements provide an acceptable way of providing protection from the hazards discussed in Sections 3 and 4,
- 47 - there may well be other and better ways of accomplishing it. An approach proposed by one manufacturer· provides partial protection. The exact nature of the protective device, which uses solid-state elements, is not disclosed by the manufacturer. Its virtue is apparently low cost. The device does not use transformer isolation, yet appears to guard against hazardous voltages and out-of-limit signals. The protection, however, is not complete in that capacitive unbalance can still exist. CONCLUSIONS The need for some forms of protection is well established. The questions are: How much? Where? and In what form? Clearly, there must be protection against harmful voltages, excessive signal amplitudes and longitudinal unbalance introduced by attached equipment. We draw the following conclusions: 1. Existing carrier-provided protective devices are indeed effective in providing protection for hazardous voltages, excessive signal amplitudes, and longitudinal unbalance from users. 2. Existing carrier-provided protective devices provide, on the whole, minimal protection against faulty network control and signaling. 3. A protective device obviously introduces another potential point of failure. Reliability of the protective devices under normal operating conditions, however, should be comparable to the attachment and should, therefore, present no great concern. 4. There are redundancies between the functions of the protective devices and those of certain user-provided equipments; e.g., PBX's. 5. Carrier-provided couplers are not inherently transparent. 6. The present dependence of some couplers on commercial power is a significant and probably unnecessary disadvantage. 7. Protective arranggments do not contribute to any significant performance degradation. Increase in glare is minimal. Transmission loss is a small effect. Central~ffice 8. protection cannot provide the same degree of protection as customer-site protection.
