CROSSTALK CANCELLATION
The electronics in the chroma recording section had two separate influences on the phase of the colour-under signal recorded on tape. The purpose of locking the new low-frequency colour subcarrier to line sync has been fulfilled in the replay de-jittering loop just described. Considerable trouble was taken to rotate the phase of head B’s subcarrier signals ‘backwards’ during record, and ‘forwards’ during replay in order to establish a specific pattern of chroma phases on tape. Its purpose is to facilitate a method of cancelling inter- track crosstalk.
In standard-play VHS format, the start points of successive video tape tracks are offset by 1^ TV lines, but because of the shallow angle which they make to the tape ribbon itself, recorded TV lines lie side by side on tape, with all line sync pulses adjacent. The effect is that crosstalk picked up by either head consists of signals from the same line in the alternate field, and as such their signal content will be similar to that of the main signal read out by the head in question. The cancellation of crosstalk interference depends on this correlation of recorded TV lines on tape, and on the carefully contrived pattern of chroma phases on tape described earlier, and shown in Fig. 14.7. The third row of that diagram shows the burst phasors picked up by head A during replay – the long arrows represent the main signal (from the top row), and the short arrows the crosstalk interference from the adjacent track on either side written by head B.
The fourth row of Fig. 14.7 shows the final effect of head B’s replay. In the up-conversion process its chroma phase is advanced by 90° per line to restore normality to the chroma signal; in doing so, the crosstalk signal picked up from adjacent ‘A’ tracks will also be advanced by 90° per line. Again, the large arrows represent the main signal and the small ones the crosstalk components. A two-line (128 μs) delay is now introduced. The direct and delayed signals made available represent a time-coincidence between line n and n + 2 as shown by the fifth and sixth rows of the diagram. The content of lines n and n + 2 are highlighted by the boxes drawn in rows three and five of the diagram. It can be seen that the main signals are phase- coincident so that they will reinforce when added, whereas the spurious crosstalk signals are in antiphase – adding these components will cause them to cancel to zero. The arrangement of the two-line delay and add-matrix is depicted in Fig. 14.9, along with a vectorial representation of the crosstalk cancellation process.
That the process works on every line is illustrated by the boxed examples in Fig. 14.7’s rows four and six. These head B lines n + 4
and n + 6 are also brought into time-coincidence by the delay line, and their addition in the matrix will again double the amplitude of the wanted main signal while cancelling out the crosstalk components. For perfect cancellation the time delay must be exactly 128 μs with in-phase arrival of both direct and delayed signals at the adder; both signals must also have equal amplitude. Adjustment of phase and amplitude is made with an inductive and a resistive preset respectively. In practice they are trimmed for minimum crosstalk on replay of a colour-bar signal from a special alignment tape.