This document contains an excerpt from the Report and
Order in Docket 19692, Amendment of Part 73 of the Commission's Rules
and Regulations to Establish Standards for the Design and Installation of Sampling
Systems for Antenna Monitors in Standard Broadcast Stations with Directional
Antennas, released February 12, 1976, whose contents remain relevant today.
The entire text is located at 57 FCC 2d 1085.
Maintenance and Sampling System Performance, and
Internal links: Sampling Lines, Coupling Elements, Maintenance and Sampling System Performance, and Other Considerations
21. In this section, we will review the engineering material submitted in this proceeding [Docket 19692], and, in some instances, draw tentative conclusions as a result of our study of this material. Under the heading Sampling Lines we consider as the first of two major components of a sampling system, the design and disposition of the coaxial cable which delivers the sample currents from each tower of the array to the antenna monitor, and under Coupling Elements, the apparatus used to extract samples of the tower currents to be fed into the sampling lines. In the section Maintenance of Sampling System Performance we discuss measurement procedures which may be used to detect any deterioration in system performance, and under Other Considerations outline other minor details of system design which appear of sufficient importance to record for general guidance, but, in many instances, seem inappropriate for inclusion in broadly drawn rules.
22. A point in which there is almost entire agreement is that sampling lines should have a solid outer conductor -- braided or foil wrapped lines afford insufficient shielding against unwanted fields, are subject to moisture contamination, and are probably the weakest component in many older sampling systems. Also, there is near unanimity as to the necessity for all lines in the system to have identical physical and electrical characteristics. Footnote 1. While most comments assume that the outer conductor will be of copper, aluminum cable is also available. While its electrical characteristics are apparently satisfactory, some of the comments cite difficulties in achieving satisfactory bonding and terminations when this material us used as an outer conductor. In addition to being less expensive than copper line, aluminum appears to have one possible long-term advantage -- the outer conductor is usually seamless, while the copper version has a welded seam. However, there was no indication that rupture of this seam is a frequent occurrence.
23. The question of the amount and kind of dielectric, and the initial processing of the cable, is rather intimately related to the question of whether equal length sampling lines should be required in any or all systems. When the ambient temperature of a coaxial line is raised, the metal in the cable expands, causing an increase in the physical length of the cable. For any given length of cable, this effect, alone, results in an increase in the phase delay in the cable. However, a temperature change in the same direction affects the constant of the dielectric in the cable so as to produce a counter effect -- tending to reduce the phase delay when temperature increases. If a solid dielectric is employed, the latter effect is predominant, and lines with such dielectric exhibit a rather large negative phase / temperature characteristic. On the other hand, lines in which the dielectric is principally air or other inert gas exhibit a much lower coefficient, generally in the positive direction. Cable with foamed polyethylene dielectric has a temperature / phase characteristic which, while inferior to air line, is much better than that exhibited by cable with solid dielectric.
24. The phase of the current sample presented to the antenna monitor is delayed by its transmission over the sampling line. If this delay varies with temperature, obviously it is a possible source of error in the monitor indication. If all sampling lines are of equal length, and exposed to the same environmental conditions, the error is cancelled, since we are concerned with the relative phases between towers, and a comparable change in the electrical length of lines leaves the relative phase indication unchanged. While, therefore, the employment of equal length sampling lines insures that the temperature changes will have the lease effect on the stability of the system, this expedient is quite expensive, necessitating, in many instances, the addition of hundreds of feet of sampling line only to achieve this end. It may also be an unnecessary refinement for those directional antenna systems which are not required to be held to very close operating tolerances. However, even in such cases, the effect of temperature variations on cable characteristics cannot be ignored. If unequal length lines are installed, good quality sampling line with a reasonably low temperature / phase coefficient should be employed. Footnote 2.
25. Generally, for stations not required by their authorizations to hold phase and current variations within restricted and specified limits, we believe that the decision as to whether to limit the differences in the relative length of sampling lines may be approached on the following basis. For the average array, the relative phases should be held within ± 3 degrees (approximately equivalent to a ± 5% variation in current ratio, which is a tolerance specified in our rules (§ 73.52(b) ). Assume that, to maintain the relative phases within this tolerance, a monitoring accuracy of one-half this tolerance, or ±1.5 degrees represents good engineering practice. The repeatability of the type approved monitor is ± 1 degree (§ 73.53(c)(13)(ii) ). The permissible variation in samples presented to the monitor, caused by the temperature variations, should therefore not exceed 0.5 degree. Accordingly, we believe, a reasonable basis for determining the tolerable difference in the length of lines is to ascertain, considering the phase / temperature characteristic of the line employed, and the temperature variation to which the line would be subject on a diurnal and seasonal basis, that, as between the longest and shortest sampling lines proposed, a temperature phase differential in excess of 0.5 degrees will not occur.
26. The following table, based on a study of material supplied by manufacturers of cable for sampling systems, gives a general idea of the magnitude of changes in phase delay occurring in two different types of cable, for different lengths of cable.
------------------Cable Length-------------------- Frequency (kHz) 100 feet 500 feet 1000 feet Dielectric 550 0.016 0.08 0.16 Primarily 1000 0.03 0.15 0.29 Air 1600 0.05 0.24 0.47 Foamed 550 0.09 0.45 0.89 Polyethylene 1000 0.17 0.83 1.7 Dielectric 1600 0.27 1.32 2.65
These figures are representative for phase-stabilized cable of the types shown. Solid dielectric cable may have a phase / temperature coefficient as much as three times as great as cable with foamed polyethylene dielectric, and its use obviously would present problems in many cases.
27. While it is obvious that air dielectric line has the much better temperature / phase characteristic, such line is considerably more expensive than line with a dielectric of foamed polyethylene. Also, because air dielectric line mist be pressurized with dry air or other inert gas, auxiliary apparatus is required for this purpose. Finally, if the performance of such line is not to deteriorate, it must be subject to careful maintenance procedures. Several consulting engineers who comment discourage the use of such line for the latter reason -- they have found a number of instances where maintenance has been neglected, and antenna monitoring systems have become inaccurate or inoperative primarily because moisture has penetrated improperly scavenged and pressurized lines. Footnote 3.
28. The majority of engineers appear to favor line with foamed polyethylene dielectric. It has fewer maintenance problems, and has a reasonably good phase / temperature characteristic. It seems entirely satisfactory for the majority of installations and is probably adequate for even the most critical uses, if measures are taken to minimize phase / temperature differentials to a degree commensurate with the monitoring accuracy required for the particular array.
29. As we have pointed out previously, the employment of lines of precisely equal length seems unnecessary, in most cases, to insure adequate monitoring accuracy. It actually may be undesirable if the result of such a procedure is to present to the monitor, as between pairs of towers, phase differences at or near zero or 180 degrees. Monitors are usually less accurate at these extremes, and near these phase points it may be difficult to resolve ambiguities in the sign of the phase difference. To avoid such situations, sufficient extra cable may need to be added to one or more lines to displace the presented phase difference by several degrees.
30. Potomac [Instruments], a manufacturer of type approved antenna monitors, points out that it is also desirable to avoid line lengths approximating 1/4 wavelength (and presumably odd multiples thereof), since a line of this length, driven by a low impedance high Q loop may produce a high voltage when unterminated. Should an open terminating resistor occur in the monitor input, severe damage may be caused to the monitor.
31. Phase-stabilized line is coaxial line which, on order by a customer, and, of course, subject to an additional charge, has been heat cycled by the manufacturer to reduce semi-permanent stresses produced in line manufacture. Such stabilized line has an additional phase / temperature coefficient much lower than line not so treated. Footnote 4. Unstabilized line, subject to normal temperature cycling after installation, eventually becomes stabilized. We had raised the question in our Notice as to the circumstances in which phase-stabilized line should be required. Generally, this would appear to depend on design considerations. If sampling lines are to be substantially equal in length, and so disposed as to be subject to equivalent environmental conditions, pre-stabilization would appear to be unnecessary -- since all lines presumably will change equally to the same final condition. However, if lines have lengths that differ greatly, the possibility exists that as the lines approach a stabilized condition, phase indications may drift from those established in the initial adjustment of the array. For this reason, we believe phase stabilized cable should be required for any sampling system in which the sampling lines are not of approximately equal length. Footnote 5.
32. It is apparent that if lengths of cable between each tower base and the transmitter building are buried, not only are these portions of the sampling system subject to less extremes in temperature, but are better shielded from troublesome ambient fields, better protected from damage, and less subject to deterioration from weather. Buried line should, of course, be jacketed. While it is therefore apparent that line should preferably be buried, it is emphasized by several parties that soil conditions in particular areas, or other factors, may make this disposition of the line undesirable, or perhaps, impossible. In such cases, each cable must be run above the ground between the tower and the transmitter house. It is evident that, where this is done, adequate support and protection of the cable is necessary. In addition, to avoid the buildup of troublesome currents in the outer conductor caused by the high fields to which it is subject, the cable must be tied to the station ground system at periodic intervals throughout its horizontal run.
33. The great majority of those filing suggestions on this point favor a single turn, unshielded loop as the coupling element, rigidly constructed, and mounted on a tower leg at a point near the current maximum, but not less than about 10 feet above ground level. The usual alternative to this kind of tower coupling element is the shielded loop, in which the shield imparts rigidity, and encloses the conductor. Such a loop is usually mounted so that it may be rotated on a vertical axis (with, of course, provision for locking in any particular orientation), to adjust the degree of coupling to the tower. While the shielded loop has theoretical advantages, in that the coupling to the tower field takes place only electromagnetically, the advantage is generally held to be of no practical significance, and it is outweighed by the fact that shielded loops too often have been found to accumulate internal moisture, with a consequent deterioration in performance. The comparative ease with which such loops may be rotated is seen as a liability, rather than an asses, the general opinion seeming to be that if the loop can be rotated, in the course of time it will be, either inadvertently by high winds, by workmen painting the tower or servicing its lighting, or perhaps by a misguided operator seeking to adjust his antenna monitor readings to values he considers more suitable.
34. While the great majority of engineers believe that the orientation of the loop with respect to the tower should be rigidly fixed, it is suggested that our proposed requirement that the plane of the loop in all cases include the vertical center line of the tower is too restrictive. An alternative orientation, with the plane of the loop including a tower face, should also be permitted.
35. While we had proposed that coupling loops on all towers be of equal size and shape, many engineers believe that such a requirement leaves too little flexibility for adjusting the degree of coupling to each tower so that sample voltages delivered at the line terminations will be within a range of vales which the antenna monitor can accommodate. The use of loops of equal size would seem to go hand in hand with the employment of sampling lines of equal length, when a primary objective in installing equal lines is to reflect at the antenna monitor, as closely as possible, the phase differences actually obtaining among the fields of the array elements. As we have pointed out previously, we believe sampling lines of comparable length should be required only where it is necessary to hold phase / temperature differentials to extremely low values. Thus, while we believe that the loops should be of the same general construction, we will not preclude such adjustment of the effective size of individual loops as may be necessary to establish proper coupling levels.
36. Those who favor the use of tower loops as coupling elements believe that hey are the best means for extracting samples from towers of any height. However, most of these parties concede that base sampling, usually with a shielded current transformer, is an acceptable alternative for towers of limited height ("limited, in this context, includes towers of electrical heights from less than 90 degrees up to 110 or 120 degrees). Outside this general body of opinion are a few parties who argue either that base sampling should never be employed, or that it is entirely unsuitable for towers of any height.
37. The advantages of base sampling are that the coupling element may be enclosed in the tuning house, protected from the weather, from air contaminants and precipitation which may affect the short and long term performance of an exposed coupling loop. At the base location, furthermore, the coupling unit is readily available for testing and maintenance. Cited as another advantage (but certainly a dubious one) is the rather obvious fact that monitor sample currents, taken at the base, "track" base currents more closely than do currents obtained by tower mounted loops.
38. Those who favor loop sampling point out that a sample taken at or near the point where the tower current is greatest can be expected to be a more accurate reflection of the relative magnitude and phase of the field radiated by the tower than can a sample taken from the current in the antenna feed line at the tower base. The latter sample includes a reactive component primarily representing the current flow through the antenna capacitance to ground. The magnitude of this current, at any time, is affected by ground moisture content and cover and by other variable factors. Since this capacitance effectively shunts the antenna resistance, its effect on the accuracy of samples taken at the antenna base depends upon the relative magnitudes of the tower capacitance and its base resistance. For towers of uniform cross section up to one quarter wavelength in height, and perhaps somewhat higher, it would appear that the antenna base resistance is sufficiently low, and the shunt reactance is sufficiently high, that the error introduced by base sampling is usually small. On the other hand, self supporting towers normally have such a high capacitance to ground that some parties who would approve base sampling for uniform cross section towers of moderate heights believe it should not be employed with self supporting towers of any height. Tower base resistance increases rapidly with increases in electrical height, and, for relatively high antennas, substantial and varying errors may be involved in base sampling. That this effect is real is illustrated by the difficulty frequently experienced in maintaining adequate correspondence between base and loop current ratios in arrays having one or more tall towers.
39. Other objections raised to base sampling is that the sampling element is usually located near other components producing intense fields. The importance for adequate shielding of such transformers is emphasized, particularly the need for an electrostatic shield between the antenna feed line and the transformer secondary. There is one observation that base sampling can produce errors in a tower having a negative resistance, a situation which often occurs in multi-element arrays.
40. Assuming reasonable precautions are taken, and that well designed coupling units are employed, we believe base sampling is an acceptable alternative to tower sampling for uniform cross section towers up to 110 degrees in electrical height.
41. A single turn, unshielded loop may be operated either at tower potential or at ground potential. When operated at tower potential, the inner leg of the loop is electrically bonded to the tower, and the cable which it feeds is maintained at tower potential by electrically connecting the outer conductor to the tower at frequent points as it descends the tower. Since the point at which the cable leaves the tower is above ground potential, and the remainder of the cable run is at ground potential, a transfer device must be employed -- an isolation coil, consisting of many turns of sampling cable wound on a cylindrical form. This coil may have sufficient inductance alone to present a high reactance at the station operating frequency, or it may be tuned to anti-resonance at that frequency by a suitable capacity. A fixed capacitor may be employed, and tuning accomplished, by shorting turns of the coil, or by a variable capacitor equipped with provision for locking its adjustment, once tuning is completed.
42. Alternatively, the loop may be mounted on standoff insulators, and the cable insulated from the tower in similar fashion. When this type of construction is used, no isolation coil is required. Since such is the case, this type of installation is usually less costly than the one just discussed. However, many engineers consider it a less desirable type of installation, since the loop and tower cable structure introduce a capacitive shunt to ground. Generally, opinions as to its employment are similar to those with respect to base sampling -- the insulated loop is tolerable for towers of moderate electrical height -- up to about 130 degrees -- but, for higher towers, all sampling loops should be installed to operate at tower potential.
43. No one commenting on our suggestion that some advantage might be gained by establishing an impedance match between the sampling element and the transmission line considered such a procedure as either necessary or desirable. We will pursue the matter no further.
44. Similarly, a proposed specification requiring that the inner conductor of the sampling line be connected to the inner leg of the loop was pointed out to be contrary to good engineering practice in many cases, and completely infeasible when the loop is connected at tower potential.
45. In our Notice in this proceeding, we requested suggestions as to measurement procedures which might be prescribed to determine whether the electrical performance of the sampling system remains at a satisfactory level. We envisaged the procedure as a comparative one -- an initial set of measurements made and recorded when the system is first installed, would be repeated at periodic intervals, and compared with the original measurements in an attempt to detect incipient or actual deterioration in monitoring system performance.
46. It would appear that many engineers presently do conduct more or less elaborate tests of sampling systems, and we have been furnished, in some cases, with detailed descriptions of the procedures followed. Simple DC measurements of resistance at the monitor line termination, both with the line open and when terminated by the sampling element, are a part of virtually every test program. RF impedance measurements are also common, although the particular procedures employed differ among engineers. In one or two instances, it is indicated that such measurements are supplemented by reflectometer observations.
47. R.F. impedance measurements at the station's operating frequency are the easiest to accomplish, but it is pointed out that such measurements are a more sensitive indication of sampling system performance if made at a frequency at which the RF impedance of the sampling line is high. Footnote 6. However, if such measurements are to be duplicated at a later date, the frequency at which the measurements are made must be accurately known. Thus, a highly stable oscillator, with means for checking the frequency within close limits, must be employed. When the system is separable into various sections (for instance where isolation coils are employed) some engineers measure the various sections separately. In addition to the tests described above, checks of characteristic impedance and the electrical length of the lines may be made.
48. While we believe measurements of the nature discussed, made at the time the sampling system is first installed, and repeated at periodic intervals thereafter, are valuable in detecting the existence or incipience of conditions which may adversely affect the stability of the monitoring system, we have decided not to adopt rules which would require such measurements to be made, since their performance appears to require capability which are beyond the average licensee and his regularly employed personnel. Thus, impedance measurements require the employment of an RF bridge and associated equipment, rather expensive items of equipment which are not found in the typical station's workshop. The performance of the apparently simple DC resistance measurements, aimed at determining the insulation resistance of the sampling line, would require the periodic disconnection of each sampling element from its line, which, in the case of tower mounted loops, necessitates the disruption of carefully waterproofed connections. Moreover, on loops mounted at considerable distances above ground, the operation could be performed only by an experienced tower man, who might have to be specially retained especially for this purpose. Those kinds of considerations generally remove periodic sampling system measurements from the "routine" category. Therefore, while we recommend the an appropriate program of measurements be adopted to establish the initial and continuing integrity of the sampling system, we will not require that such measurements be made.
49. The loop should be mounted on the tower at a point well removed from the lighting conduit. The sampling line, after being equipped with an approved waterproof end terminal or cable connector designed for use with the specific cable employed, and attached to the loop. is brought down the tower inside leg, for mechanical protection, and attached to the tower (by standoff insulators in case the loop is operated at ground potential or by clamps electrically bonding the outer conductor to the tower, if the loop is at tower potential), at sufficiently close intervals to provide substantial lateral restraint from "whipping" or displacement.
50. In critical installations, it may be desirable to employ Austin ring transformers to feed tower lighting, rather than chokes, since the ring transformer has a lower shunt capacity to ground.
51. Where feasible, the sampling line should be in a single length, without splices and connectors from the point it leaves the loop to its termination near the monitor. Short lengths of more flexible line (such as RG cable) may be used to connect the sampling line to the monitor. Where isolation coils are employed, of course, the line must be broken for its insertion, with waterproof connectors at both ends.
52. The isolation coils should be constructed of the same kind of cable which makes up the sampling line, unjacketed if tuning is to be accomplished by the inductance adjustment. The form on which it is wound should be of such construction as to provide rigid support for the coil.
53. It has been suggested that, in many cases, a better installation results if the cable manufacturer cuts the cable to length and installs the connectors. It is further suggested that isolation coils be fabricated to specifications by the cable supplier. The coils should be supported with its "hot" end well above ground level.
54. The outer conductor of the buried cable should be tied to the station ground system at the tower, and at its termination within the transmitter building. Care should be taken to insure that the cable sheath and the monitor enclosure are at the same ground potential.
55. Where the sampling lines must be run above ground from the radiating elements to the transmitter house, grounding of the outer conductor of the cable at intermediate points is necessary. Recommended grounding intervals vary from 20 to 50 feet [6 to 15 meters], according to the preferences of individual engineers. Footnote 7. The line must be adequately supported and protected in conduit or raceways. Generally, clamps intended for the purpose should be employed to make electrical connections to the outer conductor, to avoid possible damage to the dielectric by the heat generated by soldering.
Footnote 1: Except for short lengths of more flexible cable needed to connect the sampling lines from their transmitter house termination to the antenna monitor.
Footnote 2: Equal length sampling lines, of course, offer another advantage, which, alone, may recommend their installation in particular situations -- that, when they are used, the antenna monitor can be made to reflect closely the relative phases of the fields actually existing in the elements of the array. This can simplify the initial adjustment of an array, and the subsequent monitoring and maintenance of its performance. The advantage may be sufficient to justify the extra expense of equal length lines, particularly for multielement systems. Many of the engineers who commented in this proceeding [Docket 19692] apparently consider this to be the case, and a number of consulting engineers regularly install lines of uniform length for monitoring directional systems which they design. In some cases, extreme care is exercised by engineers to adjust the various lines to precisely the same electrical length. If a primary aim is to reproduce, at the monitor, the specific phase relationship existing in the array, such precision may be desirable. However, when equal length lines are used solely to minimize temperature / phase differentials, it seems unnecessary -- nominal differences in length would normally introduce negligible errors.
Footnote 3: Older types of semi-flexible air line, such as used in sampling systems, had another frailty -- the inner conductor was supported by solid dielectric beads or washers, spaced at intervals along the conductor. Such lines were prone to "shorts" in handling, during which the inner conductor might be displaced between supporting beads sufficiently to contact the outer conductor. Modern semi-flexible air line has the solid dielectric material disposed in continuous helical fashion along the conductors, and apparently does not suffer this weakness.
Footnote 4: One comment suggests that the coiling and uncoiling of line after stabilization may create new stresses and, therefore, pre-stabilization is useless. However, no evidence is offered on this point.
Footnote 5: Phase-compensated line is air-dielectric cable in which the amount of supporting solid dielectric has been critically proportioned to yield an extremely low phase/temperature coefficient. The engineer may choose to use such line in extremely critical installations. Like all air-dielectric cable, it requires careful maintenance.
Footnote 6: We have one suggestion that RF impedance be made a three separate frequencies: (1) the station's operating frequency; (2) the frequency at which impedance is high; and (3) a frequency at which impedance is low.
Footnote 7: The approach usually employed is the "brute force" method -- to add enough ground connections to preclude the possibility of trouble from ambient fields. We have noted at least one procedure, not described in this proceeding, for determining optimum grounding points -- to search along the line, with a field strength meter with the loop held parallel to the ground, for "hot spots" -- points of standing wave maxima -- where grounds should be made.