LINE OUTPUT STAGE
The line output transistor acts purely as a fast switch, clamping an inductor across a d.c. power source for approximately one-half of the line scan period. Consider Fig. 10.9(a), in which a line output stage is reduced to its most basic form. Initially the transistor switch is off, and no current flows in L. Passage of a turn-on current through the transistor base links point A directly to ground, placing the entire supply voltage across inductor L. As a result, a linearly rising saw- tooth current (t1 to t2 in Fig. 10.9(b)) flows in the coil. Some 26 μs
later at t2 the line drive ceases and the transistor switches off. The collapsing magnetic field about L causes an immediate reverse in the direction of magnetic flux and coil-current flow, which now reverses to charge capacitor C; this charging current flows via large capacitor Cres, which effectively links point B to ground for a.c. purposes. At a time t3 determined by the LC time-constant, one half-cycle of oscillation has taken place, and the energy in the capacitor is ready to feed rapidly back into L. Since point B in the circuit is effectively grounded by Cres, this would involve point A going below ground potential: it is prevented from doing so by the action of clamp diode D. The result is that the charge on capacitor C effectively becomes a d.c. voltage source, whose energy is linearly discharged to zero between times t3 and t4. At t4 the circuit is at rest, with no energy left in L or C. This corresponds to the situation at t1, and the transistor is at this point switched on once more to repeat the sequence – at 64 μs intervals. In this way a sawtooth current is built up in L, which represents the line scanning coils themselves. In practice L is a multi- winding transformer (line output transformer, l.o.p.t.) to which the scan coils are coupled.
The voltage across an inductor is proportional to the rate of change of current in it. Since this rate of change is constant during the forward scanning stroke, but fast and varying during the retrace (flyback period), the voltage waveform across the l.o.p.t. and scan coils is a series of pulses about 12 μs wide recurring at 64 μs intervals – see Fig. 10.9(c). Plainly the flyback period is determined purely by the LC time-constant rather than any characteristic of the line drive waveform.
Translating line output theory into practical terms, the transistor switch in Fig. 10.8 is Q401, L is formed between pins 2 and 1 of l.o.p.t. T402, D corresponds to D403, and C to C402/404 which
together amount to 1.19 nF for tuning. The bottom ends of the latter three may be regarded as grounded – D404 and C405 will be discussed shortly. The line scan coils themselves are in effect connected across the l.o.p.t. primary winding, with their return connector (top RHS of diagram) going via a relatively low impedance path to ground. The components which provide this path will be examined next.
Line scan correction
The circuit of Fig. 10.8 is somewhat complicated by the need to introduce various shaping influences on the line scanning waveform; they will now be explained in turn. L401 is the line linearity corrector, acting as a saturable reactance. Its magnetic field embraces a small permanent magnet which at some point in the sawtooth cycle causes the ferrite core to saturate, whereupon the coil’s characteristic changes from an inductive to a resistive one, with marked effect on the scanning current. In some designs the onset of magnetic saturation is adjusted by rotation of the permanent magnet which thus controls horizontal picture linearity.
The flat face of the picture-tube would show a picture somewhat cramped in the centre and stretched at the sides if a truly linear scanning current were used. To compensate, the rate of change of the line (and field, incidentally) scan current is slowed down at the beginning and end of each sweep, giving a characteristic S-shape to the current waveform. For line scan it is easily achieved by a careful choice of yoke-coupling capacitor – in Fig. 10.8 the 0.27 μF capacitor C408.
Some picture-tubes use deflection yokes which cannot themselves compensate for the geometrical distortion of the image which arises from scanning a virtually flat tube face, especially in wide angle types. To correct the resulting cushion-shaped picture, a diode-modulator is used as a controller of picture width. When fed with a field-rate parabolic waveform it provides dynamic correction of cushion distortion: adjustment of the d.c. working point of this E−W (EastWest) control system sets up the picture width. In Fig. 10.8 the action is based on the elements L402 and C302 in the scan coils’ ground return path.
During the flyback time the magnetic field around the line scan- ning coils collapses and energy is transferred to C408, the S-correction capacitor. C408 acquires a charge from this energy, and it is this charge, held across the scan coils via D403, which contributes towards the first half of the scanning stroke. Any variation in the charge modifies the current in the scan coils, and hence picture width. During line flyback the energy lost in the scanning circuit is replenished from that stored in the l.o.p.t. This replacement energy is divided between yoke-series capacitors C408 and C302 during flyback. If C302 were shorted to ground the charge across it would be zero and that across C408 at a maximum: the result is maximum picture width. With the short removed from C302 minimum picture width would result. By varying the impedance of a circuit connected across C302 picture width can be varied without altering the tuning or flyback time of the stage, thus keeping e.h.t. voltage (see later) constant. The ‘variable- impedance circuit’ in Fig. 10.8 is in fact the transistor Q403. Its base is fed – via amplifier Q404 – by (a) a field-rate parabolic waveform obtained from the field timebase; (b) a standing d.c. current to set picture width; (c) a small correction current derived from a beam current sensor – it compensates for picture ‘breathing’ effects due to imperfect e.h.t. regulation, and is applied to the height control circuit for the same reason; and (d) in some sets, a sawtooth waveform at field rate, with which any keystone distortion of the picture can be corrected.
In simple designs and older models these horizontal corrections are trimmed by preset potentiometers. More common now, however, is I2C bus control of these parameters, carried out in the ‘jungle’ chip or in a dedicated bus-decoder IC.
Widescreen displays
In widescreen (16:9 aspect ratio) display tubes the working principles are the same as described above, though the horizontal scanning cur- rent is greater and the amount of correction necessary for S- and pincushion distortion greater. Fig. 10.10 shows the various ways in which older broadcast picture formats can be zoomed under I2C bus control to fill a widescreen display. All of them except Fig. 10.10(e) involve distortion of the picture or loss of part of it. Zoom 1 mode involves changing the linearity of the line scan progressively from the picture centre to its edges, but the displayed image necessarily has geometric distortion. Such are the problems of displaying an image of one shape on a screen of another!
EHT and auxiliary voltage supplies
The l.o.p.t. is a useful source of the many auxiliary voltages and power supplies required elsewhere in the receiver, monitor or viewfinder. Fig. 10.8 shows a secondary winding between l.o.p.t. pins
7 and 9 which provides pulses at 6.3 V r.m.s. to energise the picture- tube heaters. During flyback a pulse voltage (Fig. 10.9c) appears across the l.o.p.t. windings and this can be caught and held by a diode and reservoir capacitor (e.g. D405/C415 in Fig. 10.8) to provide auxiliary supplies. A 205 V supply is thus provided on C415 to operate the RGB amplifiers.
The flyback-rectification system of D405 does not give good regulation of the secondary supply it provides because the flyback pulse is present for less than 20% of the time. For low-current requirements