BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a side view of one combiner segment according to an embodiment of the invention.
FIG. 2 is a perspective view of a complete combiner according to an embodiment of the invention.
FIG. 3 is a block diagram of the combiner of FIG. 2.
FIG. 4 is a block diagram of a combiner employing predecessor apparatus.
FIG. 5 is a block diagram of an alternate combiner configuration according to an embodiment of the invention.
FIG. 6 is a block diagram of another alternate combiner configuration according to an embodiment of the invention.
FIG. 7 is a block diagram of a branch combiner employing predecessor apparatus.
FIG. 8 is another block diagram of a branch combiner configuration employing predecessor apparatus.
FIG. 9 is a block diagram of a combiner configuration according to an embodiment of the invention, wherein an additional channel is added.
FIG. 10 is a block diagram of another alternate combiner configuration according to an embodiment of the invention, wherein an additional channel is added.
FIG. 11 is a perspective view of an alternative junction section for a combiner according to an embodiment of the invention.
FIG. 12 is a plan view of another alternative junction section for a combiner according to an embodiment of the invention.
FIG. 13 is a simplified configuration of a combiner according to an embodiment of the invention.
FIG. 14 is a configuration providing an additional tuning component.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION
The invention will now be described with reference to the drawing figures, in which like reference numerals refer to like parts throughout. The present invention provides an apparatus and method that in some embodiments provides a combiner that places a plurality of broadcast-level signals on a single transmission line.
The present invention uses a short-circuited manifold into which a number of signals are fed. The number of signals to be combined determines the number of input ports required. Each input to the manifold requires one bandpass filter tuned for the input frequency. All of the signal power is directed to one manifold output.
The figures herein illustrate embodiments employing rigid coaxial lines for interconnection functions, variously showing cut-off cable ends and flanges that may show center conductors. While rigid coax is a useful form of transmission line in high-power applications such as entertainment broadcasting, it is to be understood that flexible coaxial line (sometimes characterized by helically-grooved outside conductor wall) may be preferred in some embodiments, while other forms of transmission line such as waveguide and open lines may be suitable for some applications. Considerations such as power levels and physical size for a particular frequency regime may be relevant when selecting suitable materials.
It is to be understood that the use of the terms “tee”, “tee section”, and “tee junction” herein in place of “junction section” is a simplification referring to a common and useful physical form of a three-port junction, well suited both to drawings and to physical realization. Other forms of three-port junction sections, including but not limited to what can be termed a “flat Y” section and an “XYZ” section, to be shown in later figures, are equally realizable and may be satisfactory in some embodiments.
FIG. 1 shows in plan view a single segment 10 of a combiner according to the inventive apparatus, having as the component parts of the segment 10 a single filter 12 and a junction section 14 component of the manifold. FIG. 1 further shows a short-circuited stub 16. The filter 12 may in some embodiments be a known type, internally coupling signals within an allowed frequency range from filter input port 18 to filter output port 20 with low loss and substantially invariant phase across the allowed frequency range. The filter 12 preferably rejects out-of-range signals back to their sources, both before the input 18 and after the output 20, with low leakage of the out-of-range signals through the filter 12. The filter 12 appears as a short circuit to out-of-range signals. An in-range signal fed to the input 18 travels through the filter 12 and out the output 20, and arrives at the tee junction 14, its entry point to the manifold, after passing through a connecting section 22.
Under conditions such as termination of both tee outputs in nonreactive loads, rather than in the stub 16 and output 28 shown in FIG. 1, the tee 14 would effectively split the signal at the filter output 20 into two signals. If the signal paths were well matched, the two signals would exhibit equal phase and magnitude, with half of the signal power exiting the tee junction 14 at the left port 24 and half exiting at the right port 26, and with the out-of-band short-circuit locus of the filter 12 (located slightly inboard from the filter output 20) serving in general as a common spatial reference point for phase of the two signals. In the embodiment shown, the signals behave instead as described below.
The stub 16 has an internal short circuit placed at an appropriate electrical distance from the tee junction 14, with L, the physical length 28 of the stub 16, chosen so that the resultant reactive circuit, viewed from the electrical center of the tee junction 14, appears as an infinite impedance. With no effective conduction path in that direction, all of the signal power is directed to the other tee output port 26. It is to be understood that the electrical length (in wavelengths) of the stub 16 is a function of frequency, so that an added filter, in order to see an infinite impedance in the stub 16, requires specific transmission line dimensions, as further addressed below.
FIG. 2 shows in perspective view a combiner 30 having a plurality of filters 32, 34, 36, and 38, and a complete manifold 40, including the shorted stub 16 of FIG. 1. Each filter 32, 34, 36, and 38 presents a short circuit to signals outside its respective passband at specific physical locations at the input and output side of the respective filter. Each filter thus serves, among other functions, to block signals present in the manifold 40 and originating from the other filters. This ensures that minimal levels of off-frequency signals make their way to the other transmitters. The bandpass filters 32, 34, 36, and 38 are substantially equally effective in both directions in some embodiments.
High-level RF signals enter, past respective out-of-band-signal-short-circuit input-side loci 42, 44, 46, and 48, through the respective filters 32, 34, 36, and 38. The RF signals then pass through respective filter connection sections 50, 52, 54, and 56 and exit, past respective out-of-band-signal-short-circuit output-side loci 72, 74, 76, and 78, into respective tee junctions 58, 60, 62, and 64. The respective tee junctions 58, 60, 62, and 64 are joined by manifold connection sections 66, 68, and 70. It is to be understood that the configuration of FIG. 2 is one of any number of alternatives, wherein the number of filters is a function of the number of channels to be combined, and can be increased or decreased as appropriate. Likewise, the use of a straight-line manifold 40 arrangement is a user option, shown for simplicity of presentation, as will be further addressed below. Details of joints between sections are not shown in the drawings, since many known and future methods are likely to be adaptable to the instant invention. The generic flanges shown are representative, and are familiar to those proficient in the art.
FIG. 3 shows the functional elements of the combiner of FIG. 2 in block diagram 80 form. All input locus, filter, output locus, output segment, tee junction, shorting stub, and interconnecting section components of FIG. 2 are shown in general relation to their respective dimensions and positions in the perspective view. As noted, the use of four filters rather than another quantity is illustrative only.
The combination of the bandpass filters 32, 34, 36, and 38 that block out-of-band signals and the arrangement of the manifold 40 causes the signal at each filter's effective output port out-of-band short circuit locus 72, 74, 76, and 78 to see only one conductive path out of the combiner, with all other paths appearing as open circuits.
In manufacturing a combiner according to the instant invention, appropriate dimensioning of variable components, namely the short circuit stub 16 and the connection sections 50, 52, 54, 56, 66, 68, and 70, is required to ensure that overall combiner losses are minimized. One method for establishing the dimensions begins by identifying precisely the fixed dimensions, such as the output loci 72, 74, 76, and 78 with respect to the respective mounting flanges 82, 84, 86, and 88, computing the desired electrical distances between the respective filters, then assigning dimensions for the above-identified variable components. The variable components may then be fabricated, leaving sufficient unattached parts, such as flanges and the remainder of their associated connection sections, to permit adjustment, followed by clamping the unattached parts together to form complete electrical paths. Performance of this clamped-together combiner may then be measured by injecting, for example, low-power signals into all of the inputs and verifying function. Precise signal measurement, fine tuning of dimensions, and, ultimately, final assembly (such as by welding of clamped joints) and high-power test can confirm that the manifold combiner for the specific set of channels is correctly implemented.
FIG. 4 shows a block diagram of the functional elements of a predecessor apparatus 100 using a plurality of CIFs 102. Capability of this design can be used as a baseline to evaluate a manifold combiner according to the inventive apparatus, with tradeoffs as addressed below. In a predecessor apparatus 100, CIF groups 102 are made up of pairs of filters 104 supported by respective input 3 dB hybrids 106 and output 3 dB hybrids 108. The individual CIFs 102 are manufactured to operate at specified frequencies, with the overall apparatus able to form a combined output signal. Unused ports can be terminated in nonreactive channel loads 110 to suppress reflections. Given that the components, such as the filters 104, transmission lines 112, and the like, are generally equivalent in size to corresponding components of a manifold combiner as shown in FIG. 3, it can be appreciated that overall apparatus size and complexity are substantially greater for the predecessor apparatus 100. It is to be understood that each of the filters 104, at least, is either uniquely sized or precisely adjusted to operate at a specific channel frequency. Thus, while interchange of components and reassignment of frequencies is possible for the predecessor apparatus 100, such modifications are not trivial where each filter 104 and hybrid 106, 108 component in each CIF 102 is on the order of 2 meters by 1 meter by 0.7 meters in size for typical FM radio broadcast applications.
In manufacturing a combiner according to the predecessor apparatus 100, the CIFs 102 are, at least nominally, capable of being combined with any number of like devices to form a combined output signal. In the predecessor apparatus 100, the filter assemblies 104 are, to a significant extent, noninteroperating. As a result, mutual dependency between CIFs 102 is slight, and additional CIFs 102 may be added and removed with relatively little impact in some embodiments. The principal tradeoff in the predecessor apparatus 100, compared to a manifold combiner according to the instant invention, is that the number of large, expensive components is increased on the order of threefold or fourfold with the predecessor, as is the physical size of the total assembly.
As a corollary, since all of the components in both predecessor and instant approaches are finitely efficient, power dissipation due to insertion loss per component may be expected to increase with combiner complexity, so the predecessor apparatus 100 may have increased waste power, assuming comparable workmanship. This can translate to either requiring larger or more highly stressed transmitters to achieve a specific antenna power level, or losing some broadcasting range if the transmitters' output levels are constrained. The waste power from added apparatus and from the channel loads 110 can likewise add to cooling requirements for the broadcast operator.
A number of alternative embodiments for manifold combiners according to the instant invention are realizable. The layout shown in FIGS. 2 and 3 uses a straight-line approach, with all feeds from one side and the output at right angles thereto. As shown in FIG. 5, addition of elbows 122 at desired angles in place of or as partial substitutes for one or more of the straight sections, for example, can position the inputs substantially radially. Similarly, as shown in FIG. 6, alternating the orientation of the tees 130 can alter positioning of the respective transmitters 132, 134, 136 and 138 without appreciably affecting performance. Position order of channels according to frequency, e.g., with the lowest channel frequency proximal to the stub 16 in FIG. 2 and highest farthest therefrom, or another arrangement, optional in some embodiments, may be dictated by the requirements for particular channel assignments in other embodiments.
FIG. 7 shows a two-way branch combiner 140 according to known practice. The branch combiner shown is capable of combining two transmitter output signals of different frequencies. Since each of the filters 142 and 144, respectively, appears as a short circuit to out-of-band signals, it is possible to realize a combiner by ensuring that the lengths of the signal paths from the respective effective short-circuit loci 146 and 148 to the common tie point at the center of the tee 150 are each a quarter wavelength of the signal applied to the other filter. For example, if a first signal F1, applied to the first filter 142, is centered at 95.7 MHz, where a quarter wavelength (in large 50 ohm air-filled coax) is about 60 cm, and a second signal F2, applied to the second filter 144, is centered at 103.5 MHz, where a quarter wavelength is about 55 cm, then the effective path from the output short-circuit locus of the first filter 142 to the center of the tee 150 should be 55 cm, so that the short circuit in the first filter 142 appears as an open circuit to the F2 signal, and the only available path for the F2 signal leads out through the third port of the tee 152. An equivalent process defines the length of the second leg of the branch combiner. It is to be understood that any odd multiple of one-quarter wavelength, that is, ((2n+1)*λ)/4 for n a counting number, will realize the required port impedance translation, but it is noted that bandwidth tends to drop off with increased line length, so that a minimal value of n may be preferred.
FIG. 8 shows a three-way branch combiner 160 substantially analogous to the two-way combiner of FIG. 7. In this application, the three signal line lengths must each generally satisfy the requirement of translating both of the other short circuits to opens. This is realizable to an acceptable approximation, provided the channel frequencies are fairly close together and fairly uniformly separated. Branch combiners approach a practical limit according to known practice with the three-way apparatus shown. The requirement that each filter appear as a short circuit for all of the other frequencies an odd number of quarter wavelengths back from a single, common combining node becomes marginally realizable with the three-way branch combiner. Both the computations for the dimensions and ordinary spatial limitations for adding still more devices make branch combiners with four or more inputs increasingly complex without providing offsetting benefits in likely applications.
By contrast with the branch combiner, the manifold combiner of the instant invention is capable of providing a more satisfactory approximation of the impedance translation function, wherein the limitations of branch combiners are overcome to a significant extent while the cost and size penalties of CIF-based combiners are avoided. The use of a plurality of tee junctions, a plurality of connecting sections, and at least one stub provide an increased number of variables, so that the individual sections can be kept short enough to reduce bandwidth loss while allowing all of the rejection requirements to be realized.
Lengths of all connecting sections and the shorting stub may be calculated for each allowable configuration permutation, with excess variables assigned values that maximize performance terms such as bandwidth. For example, in a simplified case 240 shown in FIG. 13, using only two filters, F1 242 of wavelength λ1 and F2 244 of wavelength λ2, wherein F1 242 is closer to the stub 246, the distance from the tee 250 proximal to F2 244 to the short circuit (end) of the stub 246 and from the F2 tee 250 to the F1 output short circuit point 252 are each ((2ni+1)*λ2)/4 to a good approximation, with the values of the ni allowed to differ, and the distance from the tee 256 proximal to F1 242 to the short circuit (end) of the stub 246 and from the F1 tee 256 to the F2 output short circuit point 258 are each ((2ni+1)*λ1)/4, with the ni again allowed to differ. This case 240 has four adjustable elements 246, 260 and four lengths to be set, and thus may have only a single solution for minimum element lengths. Similarly, the case shown in FIG. 3 has eight adjustable elements, 16, 50, 52, 54, 56, 66, 68, and 70, for four frequencies, which should ordinarily be sufficient to develop at least one solution. Where too many variables in the form of connecting section and stub lengths exist, as shown in FIG. 14 for a four-filter variation 280 on the embodiment of FIG. 3, providing ten adjustable elements 282, preferred values such as standard product dimensions can be assigned. If multiple solutions occur, selection may be made based on extrinsic considerations such as fit of filters adjacent to each other, room size, and the like.
The above process can be repeated with F1 242 and F2 244 in FIG. 13 swapped, so that a second set of solutions may be compared for desirable attributes. For situations more likely in practice than that of FIG. 13, such as with three or more transmitters and associated filters, the increased number of permutations may produce several near-optimal solutions. Preference criteria such as cost, expandability, and/or one or more additional factors may be used to rank the configurations. Moreover, in applications wherein it may be anticipated that additional transmitters of known frequencies are likely be added subsequently to an original installation, solutions compatible with the later configuration can be provided for using such extra components. In all cases, it is to be understood that engineering experience may be the final determinant.
The result of employing a manifold combiner according to the instant invention is to combine several isolated signals onto a single transmission line, with the signals maintaining their isolation until presented to an antenna for broadcast or otherwise employed. In typical applications, the signals are of similar power and individual bandwidth but differ from one another in center frequency.
Where a user has a working manifold combiner according to the instant invention but subsequently requires an additional broadcast channel, at least two realization methods for adding a channel are feasible.
FIG. 9 shows one method 200 for adding a channel, wherein the assembly process above, illustrated in FIGS. 2 and 3 and/or FIG. 5 or 6, is repeated, adding a channel (T5) 202 to the set 204 previously in use. Because all of the variable connection sections 206 and the stub 16 tend to interact, it can be necessary to deactivate the manifold combiner long enough to alter or replace an appreciable number of the variable sections 206 and/or the stub 16, as well as to perform low-level and high-level testing after the alteration. In an extreme case, all variable sections 206 and the stub 16 may be reusable without alteration, albeit possibly shifting positions of the variable sections 206; at the other extreme, all variable sections 206 and the stub 16 may require alteration or replacement. The particular channel frequencies combined, or a capability within an installation to reorder the filters themselves, may affect the required extent of alteration.
FIG. 10 shows a second method 210, wherein the combined output signal from the combiner 212 can be joined with the new signal T5 214 using a constant impedance filter module 216, which can be of the type shown in FIG. 4, to perform the extra combining function. It should be noted, as is evident in the drawing, that the cost in added materials and physical space within a transmitter building for this method may be comparable to that of the entire original installation of the combiner 212, in contrast with the method of FIG. 9, wherein size and material impact are generally less. The advantage of potentially avoiding acquiring, calibrating, or reworking numerous small components in the manifold combiner is thus traded off against size and material cost for the large components.
The significance of adopting manifold combiners according to the instant invention, when compared to some established methods (such as CIFs) for performing like functions, includes at least allowing a much smaller combiner to be built at a greatly reduced material cost with essentially no sacrifice in electrical performance. Indeed, as noted, energy cost, signal reach, and/or facility reliability may be improved in some embodiments due to reduced overall insertion loss. In addition, the number of channels that can be combined, as well as flexibility in frequency spacing of channels, may exceed practical limitations of other established methods (such as branch combiners).
All of the embodiments thus far presented, for both prior art and the instant invention, are coplanar—i.e., place all of the filters and other apparatus on a common, generally horizontal surface—and employ various orientations to facilitate feed from transmitters. While this may be appropriate in general because of the large size and massive weight of individual components, it is to be understood that some embodiments may place components on more than one level, with elbows and the like used as needed. The effective lengths of interconnecting lines must nonetheless satisfy the requirement for the manifold that only a single output node from the combiner exists for all signals, with the out-of-band short circuits of all other nodes so spaced as to appear as open circuits. Where appropriate, three-axis (orthogonal) tee junctions 220, as shown in FIG. 11, or non-orthogonal junctions 230, as shown in FIG. 12, may be included in coplanar or non-coplanar embodiments.
The many features and advantages of the invention are apparent from the detailed specification, and, thus, it is intended by the appended claims to cover all such features and advantages of the invention which fall within the true spirit and scope of the invention. Further, since numerous modifications and variations will readily occur to those skilled in the art, it is not desired to limit the invention to the exact construction and operation illustrated and described, and, accordingly, all suitable modifications and equivalents may be resorted to that fall within the scope of the invention.
Patent Application
20070252661
