In broadband communications systems such as CATV equipment, system performance may critically rely on gain or attenuation flatness. In particular, CATV systems are often plagued by issues resulting from the frequency-dependent attenuation of very long cables (increasing with frequency) as well as the negative gain slope of certain amplifiers. This negative gain slope exhibited by CATV system components can cause a variety of headaches for system designers.
For example, suppose the system’s operating bandwidth is split into a number of channels. Channels at lower frequencies are subject to much less attenuation than those at higher frequencies, so the “louder” low-frequency channels may saturate an amplifier exhibiting sufficient gain to amplify the “quieter” high-frequency channels. Obviously an amplifier driven to saturation in a spectrally rich multi-channel system would result in a whole mess of impossible-to-filter-out intermodulation products, so this situation is highly undesirable.
Out-of-band spurious spectral content (e.g. intermodulation products, harmonics) from the louder lower-frequency signals may also degrade the quality of the quieter, higher-frequency signals (be it in terms of MER, SNR, CNR, or any other Figure-Of-Merit flavors) – and so on.
Because both the amplifiers and the cables exhibit a negative gain/attenuation slope, in many cases a designer may wish to flatten this slope at the expense of overall gain. This can be done with a device that has a positive attenuation slope (over the desired operating frequency range), called an equalizer. Equalizers may be realized, for example, as resistive pi or tee networks with the series resistive elements “bypassed” by parallel capacitors and the shunt resistive elements “bypassed” by parallel inductors, effectively forming a lossy low-order high-pass structure. Figure 1 illustrates this concept.