CATV Modulator Return Loss Effects On Headend Combining Isolation

Size: px

Start display at page:

Download "CATV Modulator Return Loss Effects On Headend Combining Isolation"

Jerome Bradford
6 years ago
Views:

Loss Effects On Headend Combining Isolation Ron D

1 Bi-Level Technologies the SelectedWorks of Ron D. Katznelson January 18, 2005 CATV Modulator Return Loss Effects On Headend Combining Isolation Ron D Katznelson Available at:

2 CATV Modulator Return Loss Effects on Headend Combining Isolation Dr. Ron D. Katznelson, CTO, Broadband Innovations, Inc., San Diego CA. 1 Introduction This note addresses the impact of broadband return loss of frequency agile modulators when combined through headend combining networks. In particular, the effect of less than ideal output return loss values are examined and it is shown that even under the worst case conditions, a broadband return loss of 10 db produces small losses in isolation values that still permit cable operators to meet their headend combining specifications. There are two possible isolation concerns in the headend. The first is the effect of port-to-port isolation of the common combiner, potentially changing the modulator-to-modulator isolation values. The second is the composite effect of all return loss values from all modulator ports as it affects the output return loss of such a combiner, which would further affect the port-to-port isolation of the Broadcast Splitter feeding multiple zones, thereby reducing the -to isolation values. As analyzed below, we conclude that reduced output return loss of modulators from 14 db to 10 db provides no significant degradation for the isolation values in the first case and causes measurable differences for the second case but that such differences are well within the specification range of MSO s headend combining networks. r n I' C I C (a) (b) Figure 1. Port-to-port isolation at the input of an RF combiner. 2 Modulator Port-to-Port Isolation Figure 1(a) shows a terminated 8-way combiner with well-terminated input ports. Under such conditions, an ideal combiner will have no leakage from port to port. This means that no voltage can develop at one terminal due to the imposition of a signal voltage on another terminal. It is a wellknown fact of circuit theory that an impedance across which no voltage develops can be replaced by any other impedance without any effect on the circuit variables. Thus, replacing the termination with any other impedance having any return loss would not change the fact that there is zero voltage 1

3 across that impedance, meaning that isolation is not affected by input port source impedance variation in ideal combiners. Practical combiners, however, still exhibit port-to-port parasitic coupling as shown by I C in Figure 1(a). The question arises as to how the isolation is affected in this case if the input port is not well terminated and is driven by source impedance causing a returned signal r n as shown in Figure 1(b). Here we use r n as a complex return phasor of magnitude less than 1 indicating the voltage fraction and phase of a returned signal due to impedance mismatch. We similarly designate the parasitic insertion loss I C (isolation) of the combiner and note that typically its magnitude corresponds to 30 db. The effective isolation for the case we are interested in as shown in Figure 1(b) is designated by I C and it can be approximated by using the following observation: A voltage signal reaching port n from another port will be attenuated by a factor I C, which for simplicity we shall assume is identical to all port pairs. The reflected signal from port n back into the combiner will have a voltage level of I C r n and thus other ports will receive a further I C attenuated version of this signal, namely they will receive a total voltage of I C I C + I C r n. Exact calculation would include infinite number of reflections (and powers of I C ) which diminish in value rapidly. It can be seen that adding a term that is 40 db (r n of 10 db) instead of 44 db (r n of 14 db) to a base 30 db does not change appreciably the effective isolation in this situation. The change in port-to-port isolation is but a small fraction of a db. 3 Broadcast to Splitter Isolation Typical architectures for headend and hub combining/splitting networks for broadcast, local and narrowcast services are described in a paper that forms a specification guideline for AT&T Broadband [1]. A relevant portion of the narrowcast combining is shown here in Figure 2. As can be seen, signals on a given channel that is narrowcast to each zone may produce unwanted channel interference for the same channel on other zones with a relative level that is equal to the narrowcastto-narrowcast co-channel isolation. Co-channel isolation of narrowcast channel X serving Zone 1 to any others depends, among other things, on the output port-to-port isolation of the Broadcast tier splitter shown in the Sublevel 1 combiner network of this figure. The sublevel terminology in Reference [1] is adopted and thus we associate the combining network designated by Sublevel 1 in our figure with that shown in Figure 4 of Reference [1]. It is reflections from the Broadcast combiner section that we are interested in quantifying in order to see their effect on effective isolation of the splitter, as its presented to the narrowcast signals. These reflections will be added to the nominal parasitic isolation term R S, resulting in higher possible leakage. For a good approximation, it can be shown that to a first reflection order, the normalized leakage R T between output port 1 and output port 2 of the Broadcast splitter (effective output port-to-port isolation) is given by N 2 2 jωτ n (1) RT = RS + A LS1LS 2 RC + LCn rn e n= 1 where (refer to Figure 2) A is a voltage fraction indicating the attenuation of the pad in the figure, R S is the nominal splitter isolation vector, R C is the nominal broadcast combiner s output return loss vector, L S1 and L S2 are the splitting (combining) loss vectors for port 1 and port 2 of the Broadcast Splitter, L Cn is the combining (splitting) loss vectors for port n of the Broadcast, r n is the return loss vector of the modulator connected to port n of the Broadcast, τ n is the roundtrip delay associated with the cable connecting port n to its respective modulator and ω is the RF frequency on which the value of this leakage vector is obtained. In general, all the variables (except 2

4 perhaps A) are complex and may depend on the frequency ω, which we have not made explicit in Equation (1). Zone 1 Zone 2 Sublevel 1 Desired Signal Co-Channel Cross-talk on To Zone 1 To Zone 2 r n LC LS A R C Pad R S Splitter s To Zone N-1 Broadcast Channels To Zone N Zone N Zone N-1 Figure 2. Co-Channel isolation of narrowcast channel X serving Zone 1 to any others depends, among other things, on the output port-to-port isolation of the Broadcast tier splitter shown in the Sublevel 1 combiner. As can be seen from Equation (1) the actual effective isolation will depend on relative phasing conditions but a worst case bound can be obtained on the magnitude of the leakage R T by using Equation (1): (2) RT RS + A LS ( RC + N LC r ) where we assumed that all port specific isolation and return loss numbers have identical magnitudes and thereby dropped their respective indices and replaced the sum with N identical terms. A similar analysis can be done in the case of a cascade of Broadcast combiners, wherein more than 8 modulators can be combined. We shall use here the case of two cascades of 8-way combining for a total of 64 modulators. We will skip the expression that is the counterpart to Equation (1) and provide here the result of using upper bounds as in Equation (2). For such a cascade combining we obtain: 3

5 (3) RT RS + A LS RC + 8 LC ( RC + 8 LC r ) The results of Equations (2) and (3) are summarized in Table 1 for the indicated values of the magnitudes of the parameters. Although a larger value than 3 db for A is implied in Reference [1], we conservatively assume minimal attenuation of 3 db. Results for r values corresponding to 10 db and 14 db are shown in the table for either 1 level of combining or 2 levels of combining as shown in the leftmost column. As in Reference [1], we use a 4-way Broadcast splitter for this analysis. With 10 db Modulator Return Loss Number of Combining Levels Item 4-Way Splitter Pad 8-Way Isolation symbol Rs Ls A Rc Lc N r Total 1 db Magnitude db Magnitude With 14 db Modulator Return Loss Number of Combining Levels Item 4-Way Splitter Pad 8-Way Isolation symbol Rs Ls A Rc Lc N r Total 1 db Magnitude db Magnitude Table 1. Isolation impact of modulator return loss in one and two level combining. 4 Summary and Conclusions As can be seen in the table, under the worst case conditions in two level combining, less than 0.6 db degradation of broadcast splitter isolation will be incurred if all modulators exhibit a 10 db return loss rather than 14 db. That decrement goes up to 0.9 db in a single level combining application. It should be emphasized that in all cases the worst case results tabulated above are better than a major MSO s port-to-port isolation guideline of 20 db as presented in Table 4 of Reference [1]. 1. Oleh Sniezko, Multi-Layer Headend Combining Network Design for Broadcast, Local and Targeted Services. NCTA Technical Papers, pp , (1997). 4

Delivering On The 256 QAM Promise

Bi-Level Technologies From the SelectedWorks of Ron D. Katznelson October 22, 2002 Delivering On The 256 QAM Promise Ron D Katznelson Available at: https://works.bepress.com/rkatznelson/5/ DELIVERING ON