Biblio5

Figure 15.6 A simplified layout of a CATV system showing its basic elements. The objective is to provide +10.5 dBmV signal level at the drops (tap outputs). LEA c line extender amplifier.

measured in dB at the highest frequency of interest. Grant (Ref. 1) describes a 13-dB equalizer for a 300-MHz system, which is a corrective unit for a length of coaxial cable having 13-dB loss at 300 MHz. This would be equivalent to approximately 1000 ft of 12 -in. coaxial cable. Such a length of cable would have 5.45-dB loss at 54 MHz and 13-dB loss at 300 MHz. The equalizer would probably present a loss of 0.5 dB at 300 MHz and 8.1 dB at 54 MHz.

15.3.8 Taps

A tap is similar to a directional coupler. It is a device inserted into a coaxial cable which diverts a predetermined amount of its RF energy to one or more tap outputs for the purpose of feeding a TV signal into subscriber drop cables. The remaining balance of the signal energy is passed on down the distribution system to the next tap or distribution amplifier. The concept of the tap and its related distribution system is shown in Figure 15.6.

Taps are available to feed 2, 4, or 8 service drops from any one unit. Many different types of taps are available to serve different signal levels that appear along a CATV cable system. Commonly, taps are available in 3-dB increments. For two-port taps, the following tap losses may be encountered: 4, 8, 11, 14, 17, 20, and 23 dB. The insertion loss for the lower value tap loss may be on the order of 2.8 dB and, once the tap loss exceeds 26 dB, the insertion is 0.4 dB and remains so as tap values increase. Another important tap parameter is isolation. Generally, the higher the tap loss, the better the isolation. With an 8-dB tap loss, the isolation may only be 23 dB, but with 29-dB tap loss (two-port taps), the isolation can be as high as 44 dB. Isolation in this context is the isolation between the two tap ports to minimize undesired interference from a TV set on one tap to the TV set on the other tap.

For example, a line voltage signal level is +34.8 dBmV entering a tap. The tap insertion loss is 0.4 dB so the level of the signal leaving the tap to the next tap or extender amplifier is +34.4 dBmV. The tap is two-port. We know we want at least a +10.5 dBmV at the port output. Calculate +34.8 dBmV − X dB c +10.5 dBmV. Then X c 24.3 dB, which would be the ideal tap loss value. Taps are not available off-the-shelf at that loss value, the nearest value being 23 dB. Thus the output at each tap port will be +34.8 dBmV − 23 dB c 11.8 dBmV.

there are no more than three amplifiers in tandem to reach any subscriber tap. Also note that with this system layout there cannot be a catastrophic failure. For the loss of an amplifier, only one-sixteenth of the system is affected in the worst case scenario; with the loss of a fiber link, the worst case would be one-sixth of the system.

15.4.1 Design of the Fiber Optic Portion of an HFC System

Before proceeding with this section, it is recommended that the reader turn back to Chapter 9 for a review of the principles of fiber optic transmission.

There are two approaches to fiber optic transmission of analog CATV signals. Both approaches take advantage of the intensity modulation characteristics of the fiber optic source. Instead of digital modulation of the source, amplitude modulation (analog) is employed. The most common method takes the entire CATV spectrum as it would appear on a coaxial cable and uses that as the modulating signal. The second method also uses analog amplitude modulation, but the modulating signal is a grouping of subcarriers that are each frequency modulated. One off-the-shelf system multiplexes in a broad FDM configuration, eight television channels, each on a separate subcarrier. Thus a 48-channel CATV system would require six fibers, each with eight subcarriers (plus 8 or 16 audio subcarriers).

15.4.1.1 Link Budget for an AM System. We will assume a model using a distributed feedback laser (DFB) with an output of +5 dBm coupled to a pigtail. The receiver is a PINFET with a threshold of −5 dBm. This threshold will derive approximately 52-dB S/ N in a video channel. The C/ N required is about 49.3 dB (see formulas 15.4 and 15.5). This is a very large C/ N value and leaves only 10 dB to be allocated to fiber loss, splices, and link margin. If we assign 2 dB for the link margin, only 8 dB is left for the fiber/ splices loss. At 1550-nm operation, assuming a conservative 0.4- dB/ km fiber/ splice loss, the maximum distance from the headend to the coax hub or first fiber optic repeater is only 8/ 0.4 or 20 km. Of course, if we employ an EDFA (erbium-doped fiber amplifier) with, say, 20 dB gain, the distance can be extended by 20/ 0.4 or 50 km.

Typical design goals for the video/ TV output of the fiber optic trunk are:

C/ N (carrier-to-noise ratio) c 58 dB,

Composite second-order (CSO) products c −62 dBc (dB down from the carrier level);

CTB c −65 dBc.

One technique used on an HFC system is to employ optical couplers (a form of power splitter), where one fiber trunk feeds several hubs. A hub is a location where the optical signal is converted back to an electrical signal for transmission on coaxial cable. Two applications of optical couplers are illustrated in Figure 15.9. Keep in mind that a signal split not only includes splitting the power but also the insertion loss of the coupler.5 The values shown in parentheses in the figure give the loss in the split branches (e.g., 5.7 dB, 2 dB).

5Insertion loss (IEEE, Ref. 10) is the total optical power loss caused by the insertion of an optical component such as a connector, splice, or coupler.

Figure 15.11 FM system model block diagram for the video transmission subsystem. (Courtesy of ADC video systems.)

transmit block diagram for the video portion of the system. Figure 15.12 illustrates a typical FM/ fiber hub. Figure 15.13 shows the link performance of an FM system and how we can achieve an S/ N of 67 dB and better.

As illustrated in Figure 15.11, at the headend, each video and audio channel must be broken out separately. Each of these channels must FM modulate its own subcarrier (see Figure 15.10). It should be noted that there is a similar but separate system for the associated aural (audio) channels with 30 MHz spacing starting at 70 MHz. These audio channels may be multiplexed before transmission. Each video carrier occupies a 40-MHz slot. These RF carriers, audio and video, are combined in a passive network.

where K c a constant (23.7 dB) made of weighting network, deemphasis, and rms to p-p conversion factors;

CNR c carrier-to-noise ratio in the IF bandwidth; B1F c IF bandwidth;

458 COMMUNITY ANTENNA TELEVISION (CABLE TELEVISION)

Figure 15.12 A typical FM/ fiber hub.

BF c baseband filter bandwidth; and

DF c sync tip-to-peak white (STPW) deviation.

With DF c 4 MHz, B1F c 30 MHz, and BF c 5 MHz, the SNRw is improved by approximately 34 dB above CNR.

Figure 15.13 Link performance of an FM system. (Courtesy of ADC Video Systems.)

Example. If the C/ N on an FM fiber link is 32 dB, what is the S/ N for a TV video channel using the given values. Use Eq. (15.14):

S/ N c 23.7 dB + 32 dB + 10 log(30/ 5) + 20 log(1.6 × 4/ 5)

c23.7 + 32 + 7.78 + 2.14

c65.62 dB.

Figure 15.13 illustrates the link performance of an FM fiber-optic system for video channels. Table 15.1 shows typical link budgets for an HFC AM system.

Table 15.1 Link Budgets for AM Fiber Links

460 COMMUNITY ANTENNA TELEVISION (CABLE TELEVISION)

Figure 15.14 Typical frame structure on a single fiber in a CATV trunk. (Courtesy of ADC Video Systems.)

15.5 DIGITAL TRANSMISSION OF CATV SIGNALS

15.5.1 Approaches

There are two approaches to digitally transmitting both audio and video TV signals: transport either raw, uncompressed video or compressed video. Each method has advantages and disadvantages of which some are application-driven. For example, if the objective is digital to the residence/ office, compressed TV may be the most advantageous.

15.5.2 Transmission of Uncompressed Video on CATV Trunks

Video, as discussed in Chapter 14, is an analog signal. It is converted to a digital format using techniques similar to the 8-bit PCM covered in Section 6.2. A major difference is in the sampling rate. Broadcast quality TV is generally oversampled. Here we mean that the sampling rate is greater than the Nyquist rate. The Nyquist rate, as we remember requires that the sampling rate be two times the highest frequency of interest. In this case, for the video the highest frequency of interest is 4.2 MHz, the video bandwidth. Using the Nyquist rate, the sampling rate would be 4.2 × 106 × 2 or 4.2 million samples per second.

One example is the ADC Video Systems scheme, which uses a sampling rate of 13.524 × 106 samples per second. One option is an 8-bit system; another is a 10-bit system. The resulting equivalent bit rates are 108.192 Mbps and 135.24 Mbps, respectively. The 20-kHz audio channel is sampled at 41,880 samples per second and uses 16-bit PCM. The resulting bit rate is 2.68 Mbps for four audio channels (quadraphonic).

ADC Video Systems, of Meriden, CT, multiplexes and frames a 16-channel TV configuration for transmission over a fiber optic trunk in an HFC system. The bit rate on each system is 2.38 Gbps. The frame structure is illustrated in Figure 15.14, and Figure 15.15 is an equipment block diagram for a 16-channel link.

A major advantage of digital transmission is the regeneration capability just as it is in PSTN 8-bit PCM. As a result, there is no noise accumulation on the digital portion of the network. These digital trunks can be extended hundreds of miles or more. The complexity is only marginally greater than an FM system. The 10-bit system can easily provide an S/ N ratio at the conversion hub of 67 dB in a video channel and an S/ N value of 63 dB with an 8-bit system. With uncompressed video, BER requirements are not very stringent because video contains highly redundant information.

15.5.3 Compressed Video

MPEG compression is widely used today. A common line bit rate for MPEG is 1.544 Mbps. Allowing 1 bit per hertz of bandwidth, BPSK modulation, and a cosine roll-off