By Robert M. Blumenkranz, Jerrold/Century III
It is well known that a cable system's performance depends to a large extent on the peak-to-valley response maintained along the cascade. As the peak-to-valley response beings to degrade, it becomes increasingly difficult to keep proper operating levels through the cascade. In the past, minimum cascade performance specifications have allow a maximum peak-to-valley of (N/10)+1 at the end of the cascade. Other formulas have been used as well depending on the length of the cascade or on system sensitivity to changes over temperature. This performance specification could only be met after the module had been aligned in the system.
Prior amplifier designs allowed each amplifier to have a minimum peak-to-valley variation, which if allowed to propagate through the cascade, could result in additional alignment time required on the part of the field service technician. If the number of "mop-up" controls within the amplifier were too large, then the number of variables under the technician's control would increase. This resulted in a great degree of control over the cascade peak-to-valley, but with more variables the chance of detuning in the field was increased. In order to solve this problem a "superflat" amplifier would have to be designed which would allow the field technician to plug it directly into the cascade without requiring any additional field alignment.
This amplifier would require a peak-maximum peak-to-valley of less than 0.2 dB over the 50 to 550 Mhz bandwidth, in order to maintain a maximum of 2.6 dB peak-to-valley response at the end of a 16 amplifier cascade. If the amplifier could meet this specification on the bench it would then be installed into the cascade without any additional mop-up. The time required to obtain satisfactory system performance would be decreased. If an amplifier should fail then replacing it would not require any system realignment. This paper will describe the basic design concepts and performance results of such an amplifier.