Circuit Analysis of Reference Oscillator Circuits:
Fig. 1 shows the portion of the RCA chassis CTC25 chroma section which reconstitutes the 3.58 MHz reference signal. These circuits are the burst amplifier, Chroma.
sync phase detector, chroma reference oscillator control, and the Chroma oscillator. Two signals are applied to the control grid of V19, a positive gate pulse and the output of the first video amplifier. Because of the filters which are incorporated in the first video amplifier and the very small value of the coupling capacitor, C25, the low-frequency components of the video signal have been removed and only the chroma information remains.
Fig. 1. Reference signal circuits of RCA Chassis CTC25.
Between pulses from the horizontal output transformer, which are fed to its grid, V19 is cut off by a positive pulse from the horizontal output transformer applied through a divider to the cathode.
This pulse is integrated by C116, and the average cathode potential is maintained at about 35 volts. The positive gate pulse applied to the grid of V19 is sufficient to overcome the cathode bias and the tube amplifies the color burst which is applied at this same instant.
The output of V19 is developed across the primary of L31 which is tuned to 3.58 MHz. Transformer coupling is used between the burst amplifier and the phase detector to block the enabling pulse from the circuits which follow. Thus, the input to the phase detector, which comes from the burst amplifier, is eight or nine cycles of the 3.58 MHz signal which originated at the transmitter, and nothing else.
In the absence of a signal from the reference oscillator, the output of the phase detector at the junction of R173 and R174 is zero.
Since the signals at the extreme ends of these resistors are equal and opposite, and the center of the secondary of L31 is grounded, the junction of R173 and R174 remains at ground potential during the half-cycle when the diodes are conducting-as well as when they are cut off.
Since the reference oscillator signal is fed to the cathode of X14 and the anode of X15. one diode is forward biased at the same instant that the other is reverse biased. During the half-cycle when the burst signal causes the diodes to conduct, it is possible for X14 to be reverse biased or forward biased depending on the phase of the signal from the reference oscillator. At the same instant. X15 will be biased in the opposite direction. Unless the reference oscillator is operating at the correct phase, the amounts of conduction in X14 and X 15 are no longer equal and the voltages at the extremes of R 173 and R174 arc no longer equal and opposite. Therefore, the voltage at the junction of R173 and R174 can swing either positive or negative depending on the phase relationship of the reference signal and the color burst.
The network consisting of R 176, L33, and R1 along with C110, C111, L29, and L30 determines the phase shift of the feedback signal from the reference oscillator to the phase detector. By changing the setting of R1, this phase shift may be varied. This generates an error signal in the phase detector which eventually changes the phase of the reference oscillator. The range of this control is adequate to shift the reference oscillator about 30° in either direction. This will shift a keyed-rainbow pattern one complete color bar from normal in either direction.
The output of the phase detector is integrated. or filtered, by C122, R 177, and C123 and fed to the grid of the chroma reference oscillator control tube. This is essentially an AFC tube and the operation is very similar to that of the horizontal oscillator AFC tube discussed in Part 3 of this series. The principal differences are the frequency of operation and the fact that the DC component of the plate current flows through the oscillator tank circuit.
The reference oscillator is a crystal-controlled electron-coupled oscillator. The tuned circuit, L34 and C127, is tuned to the exact frequency of the color burst and the crystal stabilizes the frequency.
Grid-leak bias is developed by C128 and R181. The output is developed across the primary of L35 which is tuned to the oscillator frequency. The output of the X demodulator is taken directly from the top of the secondary of L35.
The reference signal for the Z demodulation is shifted in phase by L37, C132 and R185.
Color Killer
Fig. 2 is the schematic of the color killer circuit of the RCA CTC25 chassis. This circuit is immune to noise because it is actuated by the phase difference between the reference oscillator and the color burst rather than by the mere presence of the burst signal.
First, consider the operation of the circuit when no color burst is present. Notice that its operation is quite similar to a keyed AGC circuit. A positive pulse from the horizontal output transformer causes current flow through V14 to charge the right side of C103. Between pulses, C103 must discharge through R172 to ground, developing about-15 volts at the top of R172. The long time constant of C108 and R172 maintains this voltage at a steady level. This filtered voltage is used to hold the chroma bandpass amplifier in cutoff.
Fig. 2. Color Killer circuit of RCA Chassis CTC25.
The actual amount of conduction of V14. and hence the amount of bias at the chroma bandpass amplifier, is determined by the amplitude of the positive pulse, also from the horizontal output transformer, applied to the grid of V14. The magnitude of this grid pulse is set by the color killer adjustment.
In the absence of the color burst, only the reference oscillator signal is applied to X12 and X13. Both diodes conduct the same amount and the output at the junction of R 158 and R l 59 is zero. When a color burst is present, the conduction of X12 and X13 is unequal and a negative output appears at the junction of R158 and R159.
This negative voltage opposes the positive pulse voltage at the grid of V14 and the tube remains in cutoff. As a result, no bias is developed in the plate circuit and the chroma bandpass amplifier is allowed to amplify the chroma signal.
Unlike the chroma sync phase detector whose output can be either positive or negative, the output of the color killer detector is always negative. This is true because the phase relationship between the reference oscillator output and the color burst is a constant in the color killer detector but a variable in the chroma sync phase detector. Automatic Chroma Control Another refinement of the chroma system has been incorporated in some sets. Called "automatic chroma contra I" (ACC), this circuit is essentially an AGC circuit for the chroma bandpass amplifier. A means of automatically controlling the gain of the chroma bandpass amplifier is desirable for two reasons: I. The receiver AGC circuit is controlled by the amplitude of the horizontal sync pulses, not the color burst level. In theory. the relation between these two levels is constant, but this may not be true under all circumstances This condition is especially noticeable in fringe areas. Accordingly, the level of the chroma signal may vary even though the sync pulse level is maintained fairly constant by the AGC. No AGC circuit, or any other closed-loop correcting system for that matter. is 100% efficient. (If it were, no error signal could be developed.) Therefore, an "auxiliary" AGC circuit will maintain a more nearly constant output with varying inputs.
The ACC detector circuit used in RCA Chassis CTC21, CTC28, and CTC30 has about the same configuration as the color killer detector circuit shown in Fig. 2.
The phasing is set so that an increase in the level of the color burst produces a negative-going 2 output and a decrease in the level of the color burst produces a positive-going output. This output voltage is used to raise or lower the bias of the chroma amplifier to maintain a more nearly constant level of chroma signal. This helps to prevent changes in color intensity and precludes the owner having to readjust the color control every time the input signal level changes.
RCA Closed-Loop ACC
The previously described ACC (Automatic Chroma Control) used in RCA chassis CTC21, 28, and 30 is an open loop system. That is, the output of the ACC circuit is not used to control the gain of the amplifier which feeds it. By contrast, the ACC circuit used in the RCA CTC31 chassis is a closed loop system. The loop is from the grid of the first chroma amplifier, through the burst amplifier, through the ACC amplifier, and back to the grid of the first chroma amplifier.
Fig. 3 is a simplified schematic of this circuit.
The color burst as well as the chrominance information are amplified by the first chroma bandpass amplifier. The plate load is the primary of the double-tuned transformer and one of the outputs front the secondary is fed to the burst amplifier. The burst amplifier amplifies the color burst and injects it into the reference oscillator circuit to control its phase.
First. consider the circuit with no burst signal present. The oscillator operates at its natural frequency and develops approximately 3.5 volts of negative bias at its grid. This voltage is applied to the emitter of the ACC amplifier and a positive potential of about 35 volts is present at the collector. Because of the voltage drop across R739, the DC potential at the grid of the first chroma bandpass amplifier is about +5 volts. (This bias voltage may vary considerably front set to set.) During color operation. the color burst from the first chroma bandpass amplifier is fed through the burst amplifier, which is gated on during horizontal retrace, to the grid of the oscillator. This signal increases the drive and causes the bias to increase to about-8 volts. This 5-volt change in voltage at the grid of the reference oscillator is amplified by the ACC amplifier transistor and causes a 3I-volt swing at its collector. The normal collector voltage is about 4 volts when a nominal 80 volt burst signal is applied to the oscillator grid. The bias voltage at the grid of the first chroma amplifier is approximately-5 volts under these conditions. If the amplified color burst signal increases in amplitude. the grid of the reference oscillator becomes more negative and the emitter current increases. This, in turn, causes the collector current to Increase and the collector potential to swing in a negative direction. Finally, this negative-going voltage is used to increase the bias of the first chroma amplifier and reduce its gain.
Conversely, if the amplified color burst decreases in amplitude for any reason. the emitter and collector voltages of the ACC amplifier become less negative, decreasing the bias on the first chroma amplifier and increasing its gain. Thus, the burst amplitude at the grid of the oscillator is maintained at a constant 80 volts. This is the optimum level to properly phase the reference oscillator. Since the first chroma bandpass amplifier also amplifies the chrominance signal, it, too, is maintained at its optimum level, This, of course, is the more important function of the ACC circuit.
Fig. 3. Simplified ACC circuit of the RCA CTC31 chassis.
The principal advantage of the closed-loop ACC circuit is its ability to maintain a more nearly constant level of chrominance signal. The curves shown in Fig. 4 demonstrate the characteristics of the two types of control, open-loop and closed loop. Bear in mind that these curves illustrate the characteristics of the two basic systems and do not apply to any specific circuits.
Fig. 4. Characteristics of open-loop and closed-loop control systems.
Notice that R732, R777, and R735 are quite critical as to value, drift characteristics, and temperature coefficient. For this reason, glass resistors having close tolerance and low temperature coefficients are used.
RCA Color Killer
Incorporation of closed-loop ACC made it necessary to revise several other circuits in the RCA chroma system. Since the color burst is amplified by the first chroma amplifier, color-killer bias had to be fed to a different stage; the demodulators were chosen. The use of transistors in the ACC and color killer is also a significant departure from earlier RCA designs.
A simplified schematic of the color killer circuit used in the RCA CTC31 chassis is shown in Fig. 5. The base voltage of the killer transistor is established by the setting of the killer control and the potential at the grid of the reference oscillator.
Under no-color conditions, the base potential is about .5 volt positive with respect to the emitter and the transistor is cut off. Since the transistor is cut off, the collector voltage is determined by the voltage divider, R737 and R749, connected between the blanker grid and ground. Since the blanker grid is about 100 volts, the collector voltage is about-25 volts. This voltage is also present on the screen grids of the demodulators and keeps these tubes below cutoff.
When a color burst is applied to the reference . oscillator, the grid swings negative and the killer transistor is driven into saturation. This clamps the collector voltage to the emitter potential, raising the screen voltage of the demodulators to about 2 volts, well above cutoff. Notice that the color-killer transistor operates either at saturation or cutoff.
Thus, variations in color-burst amplitude have no effect on the bias supplied to the demodulators.
Fig. 5. Simplified color-killer circuit of the RCA CTC31 chassis.
Fig. 6. Simplified reference-oscillator circuit of the RCA CTC31 chassis.
RCA Reference Oscillator
The RCA CTC31 chassis uses an injection type reference oscillator which is similar to the one used in the CTC 18, CTC20 and CTC24 chassis. Fig. 6 is a simplified schematic of the oscillator used in the CTC31. Basically, the oscillator is of the tuned-plate, tuned-grid, electron-coupled type. The frequency is determined by the crystal in conjunction with the small trimmer capacitor shunted across it.
As with any TPTG oscillator, the oscillator plate tank (L704) is tuned slightly above the oscillator frequency. In this circuit, it is adjusted so that the self-bias developed at the oscillator grid is-3.5 volts with no burst signal applied. In earlier models using the injection oscillator, the counterpart of L704 was not adjustable. In the absence of a color burst, the oscillator runs at the reference frequency, but with a random phase.
When the burst is injected through T702, this signal pulls the oscillator into phase with it. The oscillator is stable enough to remain properly phased until the arrival of the next burst.
Admiral Burst Amplifiers and Reference Oscillators
Fig. 7 shows the burst amplifier and reference oscillator circuit of the Admiral 1G1155 and other chassis. The more recent chassis, 3H10, 4H10, 5H10, and 4H12, use essentially the same burst amp and oscillator circuitry except for the tube type. The chroma signal from the plate of the first chroma bandpass amplifier and a positive enabling pulse from the horizontal output transformer are both fed to the burst-amplifier grid. The enabling pulse, having an amplitude of 50 volts, brings the tube out of cutoff and the color burst is amplified to a peak-to-peak amplitude of about 170 volts. The large value of cathode resistance, 39K ohms, prevents saturation. During the time that V5 is conducting, the drop across R166 charges C120. Between pulses, current flows upwards through R166 to discharge C120, maintaining an average cathode bias of about + 45 volts. This prevents the chrominance signal from appearing in the plate circuit of the burst amplifier.
The reference oscillator is an injection type, electron-coupled oscillator and its operation is similar to that of the RCA circuit discussed above. The self-bias under free running conditions (no color) is -.3 volt. When a burst signal is injected, this bias swings negative and the negative excursion is used to operate the color killer and the ACC circuit. The output of the reference oscillator is coupled through the plate transformer, L35, to the phase shifting circuits, L36, C129, and L37, which establish the correct phase of the reference signal for R-Y and B-Y demodulation. C128 and R6, the tint control, are used to shift the phase of the reference signal without disturbing the phase displacement between the R-Y and B-Y axes.
Admiral ACC and ASC Circuits
As shown in Fig. 7, the ACC circuit used in the chassis series 1G1155-1, 2G1156-2, 2G1157-1, 3G1155-2, and 3G1155-3 has two variations. In the solid-line drawing, the ACC control voltage is taken from the grid of the reference oscillator and, after filtering, is used as bias voltage for the first chroma amplifier. Since the oscillator grid swings more negative as the color burst increases in amplitude, the chroma-amplifier gain is reduced as the level of the composite chrominance signal (chroma and color burst) increases. The operation of the dashed-line circuit is much the same, although a diode detector has been added. Since this addition increases the amount of the control voltage, a divider network, 180K ohms and 270K ohms, is also added to the circuit.
The ASC (Automatic Saturation Control) circuit used in the 3 H 1 ONC57-1 chassis is shown in Fig. 6. This circuit, as well as the one described above, is a closed-loop system. A portion of the output from the first chroma amplifier is rectified by XII and the negative voltage which is produced is used, after filtering, to control the gain of the first chroma amplifier. Thus, as the chrominance level increases, the 'bias increases to reduce the amplifier gain and vice versa. If this were the only control voltage, the dynamic range of saturation would be seriously limited (the degree of color saturation of the picture would remain constant regardless of the degree of saturation of the scene being televised). To prevent this, a second voltage is combined with the bias derived from X11. The negative voltage at the grid of the reference oscillator, which is proportional to the color-burst amplitude, is also fed to the grid of the first chroma amplifier. Thus, sufficient bias is always available to maintain the desired dynamic range of saturation.
Fig. 7. Reference-signal circuits of the Admiral 1G1155 chassis.
Fig. 8. Reference-signal circuits of the Admiral 3H10NC57-1 chassis.
Admiral Color Killer
The color-killer circuit shown in Fig. 8 is typical of many of the circuits used in late-model Admirals.
The positive cathode bias of the burst amplifier is divided across the threshold control, R9, and negative voltage is obtained from the grid of the reference oscillator. In the absence of a color burst, this negative voltage is slight and the color-killer tube will conduct if plate voltage is supplied. The source of plate voltage is the positive pulse from the horizontal-output transformer which is fed to the left side of C128. Current flows through V 15A, charging the right side of C128 to a negative potential. Between pulses, C128 partially discharges through R170 and the negative voltage which is developed holds the second chroma amplifier below cutoff. When a color burst is received, the negative voltage at the grid of the reference oscillator increases and cuts off the color-killer tube. This allows C128 to completely discharge and the cutoff bias is removed from the second chroma amplifier.
Notice that the setting of R9 affects the bias of the first chroma amplifier and, if R9 is misadjusted, the operation of the ASC circuit will be impaired. To properly set R9, adjust all front-panel controls for proper operation and set the color control at mid-range. Turn to an unused channel, set the color-killer control fully clockwise, and then adjust it until the color in the snow almost disappears.
Zenith Burst Amplifier and Reference Oscillator Circuit
The chroma-reference circuits of Zenith's 20X1C36 and 20X1C38 are shown in Fig. 9. In many respects, they are similar to the circuits of the RCA CTC25 chassis discussed under the heading "Circuit Analysis of Reference Oscillator Circuits" in Part 5. The composite chrominance signal from the plate of the first chroma amplifier and a positive enabling pulse from the horizontal-output transformer are fed to the grid of the burst amplifier.
Since the color burst is coincident with the enabling pulse, the burst is separated from the remainder of the chrominance signal and amplified. The positive cathode bias of about 45 volts is developed by conduction through R176 while the tube is gated on, and this voltage is sustained between pulses by the charge stored in C127.
The output of the burst amplifier is fed to the chroma-sync phase detector and to the ACC and color killer detector. The hue control, R3, in conjunction with C131 allows the viewer to vary the phase of the amplified color burst.
The chroma-sync phase detector compares the relative phases of the reference oscillator signal from L30 and the color burst from the burst amplifier. Any phase error is converted to a voltage error which is used to change the conductance of the chroma reference-oscillator control tube. The operation of this type of circuit was explained in Part 3 of this series.
The reference oscillator is typical of the type of oscillator used in conjunction with an AFC tube. Since the system of chroma demodulation used by Zenith requires four reference signals in quadrature. a special output transformer is used instead of the usual RLC phase-splitter network.
Fig. 9. Reference-signal circuits of the Zenith 20XIC36 chassis.
Zenith Color-Killer and ACC Circuits
The color-killer and ACC circuit of the Zenith 20X I C36 chassis is also shown in Fig. 9. The color killer and ACC detector is a conventional phase detector, hut, since the phase relationship of two inputs is constant, the amplitude of the output becomes a function of the amplitude of the color burst. When no burst is present, the output is-.7 volt, but, during normal color reception, this potential increases to approximately-6 volts. If the amplitude of the color burst decreases from its normal value for any reason, the detector output also decreases.
The output of the detector is filtered by C64 and used as bias for the grid of the first chroma amplifier, V4B. Under no-color conditions, V4B is near saturation and the screen potential is about 75 volts.
This voltage is at one end of a series network consisting of R165, R17, and R164. The opposite end of R164 is connected to the grid of the horizontal discharge tube which is 65-volts negative. When R 17 is properly adjusted, the voltage at its junction with R164 is about-28 volts. This voltage is used to bias the second chroma amplifier below cutoff. C118 is a bias filter which integrates the horizontal pulses from the horizontal discharge tube.
During color reception, the grid bias of V4B increases to-6 volts and the screen and plate voltages rise to 225 volts. This would cause the voltage at the grid of the second chroma amplifier to rise to a positive potential if it were not for the clamper diode connected across C118. The actual bias of the second chroma amplifier is 0 volt.
As stated before, the output of the color-killer and ACC detector is-6 volts under conditions of normal color reception. If the chrominance level varies from its normal value, the detector output will also change. Thus, a decrease in chroma level reduces the negative bias on V4B, increasing its gain. Conversely, an increase in the chroma level increases the bias on V4B to reduce its gain. Notice that small variations in bias on V4B do not affect the bias of the second chroma amplifier because of the action of the clamper diode. This, too, is a closed-loop system.
General Electric Reference Circuits
Fig. 10 shows the burst gate and subcarrier amplifier circuits of the General Electric HC chassis. The output of the chroma-bandpass amplifier is fed to the cathode of the burst gate tube, V5B, and the 100-volt enabling pulse from the horizontal-output transformer is fed to the grid. This allows the color burst to be separated from the composite chroma signal and amplified.
The positive enabling pulse causes the grid of V5B to draw grid cur7 4Mc OSC rent, charging C72 and C73. Between pulses. these capacitors discharge through R87, developing about 85 volts of bias. This bias, of course, holds the tube below cutoff" between pulses.
The output from the burst-gate tube is coupled through L20 and excites the 3.58-MHz crystal. causing it to ring. Because of the high Q of the crystal, this ringing continues throughout the interval between color bursts. Each successive color burst rephases the crystal if there has been any drift. The amplitude of the ringing signal at the grid of V5C is large enough to overdrive the tube, and thus the output remains constant throughout the interval between bursts.
C86, connected between the plate of V5C and ground, shifts the phase of the reference signal to provide tint control. Quadrature reference signals are required, so a transformer having two secondaries is used as the plate load of V5C. No color-killer circuit, as such, is used in this chassis. Since the demodulators have no output unless there is a reference-signal input, and since the 3.58-MHz crystal "rings out" if there is no color burst, the modulators are, in effect, cut off during b-w operation.
Fig. 10. Burst gate and subcarrier amplifier of the General Electric HC chassis.
Fig. 11. Reference-signal circuits of the Motorola A22TS-918A chassis.
Motorola Reference Circuits
The burst amplifier, chroma sync amplifier, color killer, and demodulator of the Motorola A22TS-918A are depicted in Fig. 11. The chroma cathode follower (not shown) drives both the chroma bandpass amplifier and the burst amplifier. An enabling pulse from the horizontal-output transformer turns on the burst amplifier, V 16A, during the horizontal retrace interval, allowing the color burst to be separated from the composite chrominance signal.
The interstage transformer between VI6A and V 16B is tuned to the burst frequency. The network consisting of R4, L31, and C149 is a phase-shifting network which allows the phase of the burst to be adjusted for correct hue. V 16B further amplifies the color burst and feeds it, via the 3.58-MHz crystal, to the chroma-demodulator tube.
Notice that the positive pulse applied to the screen grid of V 16B gates this tube on during the horizontal retrace interval only.
The chroma-demodulator tube not only demodulates the chroma signal, but serves as the reference oscillator as well. This combination of both functions in a single tube was not noted in any of the other makes of sets examined. Since this particular portion of the color training series is limited to reference signal and associated circuits, the method of demodulation will be discussed at a later time. At present, we will consider only the functions of the cathode, control grid, and screen of V15.
Consider V15 as an electron coupled Hartley oscillator. In the absence of color bursts, oscillations are sustained by virtue of the split inductance tank typical of a Hartley oscillator. During color reception, the amplified color burst is injected through the crystal to the grid of V15 and re-phases the tank circuit to synchronize it with the burst signal. In this respect, the oscillator is similar to the injection-locked oscillator used in a number of other sets.
Since the oscillator is an integral part of the demodulator circuit, there is no way to split the oscillator phase prior to demodulation. As we shall see later, chroma demodulation may be achieved so long as the phase of either the reference signal or the chroma signal is split. Motorola's decision to split the phase of the latter instead of the former is unique but equally acceptable.
The color-killer circuit is similar to the ones used in many of the sets discussed in the preceding pages. In the absence of color, a pulse from the horizontal-output transformer causes V5B to conduct, charging C131 and producing cutoff bias for the chroma amplifier. During color reception, the cathode current of V15 increases (because of the increased oscillator activity) and this drives the cathode of V5B positive into cutoff. Since C131 cannot charge when V5B is cut off, the bias is removed from the chroma-bandpass amplifier.
Fig. 12. Simplified phase-sensitive detector using diodes.
Since the function of the chroma bandpass amplifier (or amplifiers) is simply one of increasing the level of the chrominance signal, there is very little to be said about it. The bandpass considerations were discussed in Part 4 of this series. The means by which it is gated by the color killer and controlled by the ACC circuit were covered in Parts 4 and 5.
Malfunction in the chroma bandpass amplifiers will normally result in insufficient color or no color at all. If the gain of the amplifier is reduced, color saturation will be decreased; if the amplifier fails completely, there will be a complete loss of color. Improper alignment of the amplifier will cause color smearing or "grainy" color, but alignment problems are more likely to be the result of tampering than drift. Realignment should be attempted only if the necessary test equipment is available. "Eyeball" alignment will usually result in further degradation of picture quality.
Diode Chroma Demodulators
In essence, a chroma demodulator is simply a phase-sensitive detector- no more, no less. Although phase sensitive detectors (PSD) have been used in the majority of b-w receivers built in the past 20 years, the use of a pair of phase detectors to extract the color-difference signals from the chroma sidebands seems to excite a great deal of interest.
Thus, a thorough discussion of each of the four popular types of chroma demodulators is included here. Since the circuit which utilizes diodes is perhaps the most easily explained, we will begin this discussion with it.
Fig. 12 shows a simplified phase sensitive detector and associated waveforms. The input from the reference transformer is constant, while the phase of the information input varies. Cases 1 and 2 show the information signal at two possible phase angles. In case 1, diode X1 cannot conduct since the instantaneous cathode and anode voltages are equal throughout the cycle.
Consequently, the voltage at point A is a simple sine wave whose first excursion is positive.
During the first half of the cycle, X2 conducts because its cathode is negative with respect to its anode, and the instantaneous voltage at point B is equal to the sum of the two applied voltages. These are equal in amplitude, but opposite in polarity; the potential at point B is 0 during the time that X2 is conducting. During the second half cycle, X2 is cut off because the cathode is positive with respect to the anode. The voltage at point B is the positive excursion of a sine wave.
Fig. 13. Typical output curve for a phase-sensitive detector.
The voltage at point C, the output, is the sum of the instantaneous voltages at points A and B. During the first half of the reference sine wave, the voltage at point A completes a positive half-cycle while the voltage at point B is clamped to zero. Therefore, the voltage at point C is a positive half-cycle having an amplitude equal to one-half the amplitude of the half-cycle at point A. During the second half of the sine wave, the voltages at points A and B are equal in amplitude but of opposite polarity, and the voltage at point C is zero. Thus, if the input signals arc in phase, the circuit functions as a half-wave rectifier with a positive output.
Case 2 shows the instantaneous voltages that arc present when the information signal is 180° out of phase with the reference signal. Now it is X2 which never conducts, and X I acts as a half-wave rectifier.
But, X1 is connected in the opposite polarity from X2, so its output is negative instead of positive. Thus, if the input signals are out of phase, the circuit functions as a half-wave rectifier having a negative output.
Additional waveforms to show the output of the PSD for intermediate phase relationships could be included, but they add little to the discussion. Instead, Fig. 13, showing the output for various phase relationships, is presented. Notice that the curve has the form of the familiar sine wave.
The polarity of the output from a PSD may be reversed by two means: reversing the diodes or reversing either of the transformers that supply the signals. Another characteristic which is of particular interest is this: With the exception of the positive and negative maximums, any output amplitude (including zero) may be the result of two different phase relationships.
These statements become meaningful when the PSD we have been discussing is renamed a chroma demodulator and placed in a TV set.
Since the phosphors used in a color CRT are red, blue, and green, a minimum amount of circuitry will be used if one demodulator produces a maximum output when a red chroma signal is received. If we arrange the circuits preceding the demodulator so that a chroma signal representing red reaches the demodulator in phase with the reference signal, we have a red (R-Y) demodulator. (The red axis is at 76.6° and the R-Y axis is at 90°. This discrepancy is more apparent than real since the color difference amplifiers shift the axis slightly and also because the various red phosphors in use have slightly different colors. Depending on these variables and the position of the "tint" control, the true axis of operation of the R-Y demodulator may be any angle from, perhaps, 60° to 120°). In actual practice, the R-Y demodulator of a TV receiver may be followed by an amplifier. This inverts the signal so that the R-Y demodulator output must be maximum negative, instead of maximum positive, to produce red on the CRT. To be absolutely correct, we must refer to an R-Y demodulator followed by a difference amplifier as a (R Y) demodulator.
A second demodulator might well be connected so that its maximum outputs occur when the chroma signal is "all blue" and "no blue." Notice that we may cause this demodulator to have either a positive or negative output (for an "all-blue" signal) merely by reversing the phase of the reference signal.
Fig. 14 shows the chroma demodulators and difference amplifiers of the General Electric HC chassis.
Consider the R-Y demodulator. The phases of the two inputs are such that a "red" (R Y) chroma signal produces a negative output.
This output is fed to the R-Y color-difference amplifier where it is amplified and inverted. The output from the R-Y difference amplifier is the R-Y signal which is fed to the red control grid of the CRT gun.
The B-Y demodulator is identical to the R-Y demodulator but the phase of the reference signal has been changed. In the B-Y demodulator, a "blue" (B Y) chroma signal produces the maximum negative output. This is amplified and inverted in the B-Y color-difference amplifier and finally appears as a positive signal at the blue grid of the CRT, turning on the blue gun and causing a blue field.
Fig. 14. Chrome demodulators and color-difference amplifiers of the General
Electric HC Chassis.
It was stated previously that a specific amplitude and polarity of output from a demodulator may be the result of either of two phase relationships. For example, observe from Fig. 13 that phase relationships of I60° and 200° each produce a negative output which is 94% of the 180° output. However, since the reference signal applied to the B-Y demodulator is shifted 90° from the reference signal at the R-Y demodulator, the 160° signal at the R-Y demodulator becomes a 70° signal at the B-Y demodulator.
Again referring to Fig. 13, this signal will produce an output from the 13Y demodulator which is positive with an amplitude that is 34% of maximum. By the same token, the 200 signal at the R-Y demodulator becomes a 110° signal at the B-Y demodulator, and a negative output with an amplitude which is 34% of maximum is produced.
From the above, it is apparent that even though two chroma signals having different phase angles can produce the same output from a single demodulator; when these same chroma signals are fed to the second demodulator, they cause radically different outputs. Therefore, any phase of chroma signal may be described in terms of the outputs it produces from the two demodulators. Stated another way, two demodulators are sufficient to extract all of the color information from the chroma signal.
In spite of this, three color difference signals are necessary to operate the color CRT because the colors from three phosphors are required to produce all the visible hues. There are two methods of producing control voltage for the excitation of the third phosphor. A third demodulator operating on the color axis of the third phosphor (green) may be used, or this voltage may be derived from the outputs of the R-Y and B-Y demodulators.
This latter method is more popular although the former method is often used.
The generation of the G-Y signal by combining portions of the R-Y and B-Y signals is the method used in the General Electric HC chassis shown in Fig. 14. Since the signals at the plates (pins 5 and 2) of V 11 are the RY and BY color-difference signals, respectively, the grid signals of these two triodes are the (R-Y) and (B-Y) signals. These signals also appear at the cathodes of the respective triodes since C96 and C93 have significant impedance at the frequencies contained in the color difference signals.
(At .5 MHz, C93 has an impedance of 145 ohms.) In part 1 of this series it was stated that G-Y =-.51(R-Y) .19(B-Y) Thus, by combing suitable portions of the cathode signals of VII B and V IIC, a GY signal is fed to the cathode of VII A. Since the VII A is operated as a grounded-grid amplifier, there is no inversion and the G-Y signal at its plate may be connected directly to the green grid of the CRT. Since the cathode-to-ground resistance of each section of V 11 is different, the bias present on each cathode is also different. To develop the desired bias on each section of V 11, the grids are returned to a bias-bleeder network, consisting of R89, R88, R98, and R99, which is connected between ground and the B + supply.
R8A and R8B are the blue and green brightness controls, respectively. Since the red phosphor is the least brilliant, the red grid is operated at the maximum positive potential and no adjustment is required.
R8A and B are used to set the blue and green conduction to produce reference white.
Blanking of the CRT during horizontal and vertical retrace is accomplished by the signals fed through C90 to the cathodes of VII. These signals are negative pulses taken from the horizontal and vertical output transformers. Since there is no signal inversion in a driven-cathode amplifier, the amplified pulses at the plates of VII are negative, and cut off the three guns of the CRT.
Low-Level Triode Demodulators
Any attempt to prove that the use of the diode phase-sensitive detector is inferior or superior to the use of amplifying devices in a phase-sensitive detection system is inconclusive. Each system has inherent advantages and limitations.
The use of amplifying devices (tube or transistor) has the advantage that some gain is contributed to the system, but increased complexity and lowered gain-stability is the price of this amplification. Similar arguments apply in choosing from among the various types of amplifying detectors: triode, pentode, sheet-beam, and twin-pentode. The use of transistorized demodulators offers some interesting possibilities, although no such circuits were used in the 1967 and early 1968 product lines that we examined.
The demodulator and difference amplifier circuits used in the Philco 17QT85A chassis are shown in Fig. 15. Although pentodes are used as demodulators, they are connected as triodes, the plates and screens being tied together. Fig. 16 is a plot of the combined plate and screen currents of the pentode section of a 6BL8 when it is connected as a triode. The circuit used in developing the curve was quite similar to the Z demodulator of Fig. 15 although bias was applied to the grid instead of the cathode. This has the effect of increasing the plate potential 8 volts, which effectively increases the steepness of the curve a slight amount.
In the Z demodulator (V 14A of Fig. 15) the reference signal which drives the cathode has a p-p amplitude of 4.5 volts. In the absence of any other signal, the conduction of V 14A and V 15A produces a positive bias of 8 volts on each tube.
Since this is the approximate cutoff bias of the tubes, V 14A conducts only during the negative half-cycles of the cathode signal. At the instant when the reference voltage is peak negative, the tube bias is reduced to 3.5 volts and the tube current is about 8.2 ma. The unfiltered plate waveform would consist of negative going half-cycles. However, the pi filter (C145. L35, and CI46) integrates the pulses, and the voltage at the output of the pi filter is at a DC level of 235 volts.
Fig. 15. Chroma demodulators and color-difference amplifiers of the Philco
17QT85A chassis.
Fig. 16. Characteristics of the 6BL8.
When a chroma signal is fed to the grid of V 14A. the instantaneous current is determined by the instantaneous grid-cathode voltage. This voltage is the sum of three potentials: (1) the cathode bias. (2) the instantaneous reference-signal voltage, and (3) the instantaneous chroma-signal voltage. For example, if the chroma and reference signals are 180° out of phase, the control grid is nominally 1.25 volts positive at the same instant that the cathode potential is + 3.5 volts (bias less the peak reference signal of 4.5 voltage is 2.25 volts. From Fig. volts.) Thus, the total grid-cathode 5, the tube current is about 12.7 ma at this instant. This is an increase of about 4.5 ma from the peak current with no chroma signal applied. After a number of cycles of voltage, the plate voltage will stabilize at a new voltage which is less positive than it was when no chroma signal was applied.
In the example just cited, the instantaneous peak tube current reaches its maximum value. The minimum instantaneous peak current may be approximated by considering the effect on tube conduction of in-phase chroma and reference signals. In this case, the grid is maximum negative at the same instant that the cathode is maximum negative (minimum positive) and the instantaneous grid-cathode potential is -4.75 volts. Again referring to Fig. 16, the current is approximately 4.5 ma. A third condition which is easily described is the 90° or 270° phase relationship.
In this case, the chroma signal is passing through zero when the reference signal is maximum negative, and the peak tube current is the same as if no chroma signal were applied; 8.2 ma.
From the above examples, we find that the maximum and minimum instantaneous peak tube currents are 12.7 and 4.5 ma, respectively. Since the plate load resistor is 3.9K ohms in the actual circuit of Fig. 15, the swing in plate voltage may be predicted to be 32 volts. Since the chroma reference signal amplitude at the grid of V 14A s 2.5 volts, p-p, during normal operation with a keyed-rainbow generator as a signal source, the predicted "gain" is 12.8. Actual observations showed this "gain" to be 10.
Bear in mind that the above examples are an oversimplification of the operation of the circuit since only the instantaneous peak currents were considered. In actual practice, the tube conducts throughout approximately 180° of the reference signal sine wave, and the magnitude of the current may be represented by the positive half of a sine wave.
Nevertheless, we can conclude that be tube conduction is proportional o the sum of the instantaneous amplitudes of the reference signal and chroma signal during the intervals when the tube is out of cutoff. In the circuit of Fig. 15, these intervals correspond to the negative half cycles of the reference signal. The operation of the X demodulator of Fig. 15 is similar to that of the Z demodulator just described.
Now, consider the operation of the two demodulators if the signal source is an un-keyed-rainbow generator. Using an un-keyed-rainbow signal, there is a constant rate of change of phase difference between the chroma signal and the reference signal. Since the chroma signal is fed to both demodulators at the same phase angle, the outputs of the two demodulators differ in phase by the same amount as the phase difference of the reference signals applied to the two demodulators. In a system using X and Z demodulation, the phase of the reference oscillator signal applied to the two demodulators differs by 63.9°. Observations of the outputs of the two demodulators when an un-keyed rainbow generator is used as a signal source confirm that the sinusoidal output of the Z demodulator passes through zero about 64° later than the output of the X demodulator passes through zero.
The operation of the color-difference amplifier shown in Fig. 15 is typical of most color-difference amplifiers used in conjunction with low-level, vacuum-tube demodulators.
The difference signals from the demodulators are coupled through C147 and C151 to the B-Y and R-Y amplifiers, respectively. These signals are amplified and fed to the respective control grids of the CRT. Notice that there are two apparent discrepancies in the foregoing discussion. First, since the signals at the red and blue control grids of the CRT are R-Y and B-Y respectively, the signals at the inputs to the difference amplifiers must bear the negative sign, -(R-Y) and-(B-Y) hence, the demodulators are actually-X and-Z demodulators. However, it is common practice to ignore this reversal of polarity.
Also observe that while the demodulators operate on the X and Z axes, the difference amplifiers operate on the R-Y and B-Y axes. Although this seems unlikely, it is actually the case. Since the difference amplifiers have a common cathode circuit, a portion of the-Z signal which is fed to the B-Y amplifier appears at the cathode of the R-Y amplifier and vice versa. Thus, a small portion of the-Z signal is added vectorially to the-X signal, shifting the true axis of the difference amplifier to R Y. Also, a portion of the -X signal is added vectorially to the-Z signal in the B-Y amplifier to establish its true axis.
The means of deriving the G-Y signal is similar to the method described in the explanation of the General Electric circuit. In the circuit of Fig. 15, an additional portion of the R-Y signal is fed from the plate of V 15B to the grid of V 13B where it is added to the (R Y) signal being fed to the cathode of V 13B. A sample of the (B Y) signal is also fed to the cathode of V 13B. The total of these three signals produces a voltage at the plate of V13B which is-.51 (R-Y)-.19(B-Y), equal to G-Y.
Low-Level Pentode Demodulators
The operation of a low-level pentode demodulator is quite similar to the triode demodulator just discussed. The demodulator circuit used in Packard-Bell Chassis 98C15 is shown in Fig. 17. The 6GY6 tubes used as demodulators belong to the family of tubes having two independent control grids. That is, both the suppressor grid and the control grid have considerable effect on the plate current.
The plate current of V18 and V19 are controlled by the instantaneous values of chroma and reference signals and detection takes place at the plate of the tube. The operation of the color-difference amplifiers is essentially the same as the ones discussed previously.
Sheet-Beam Demodulator
The sheet-beam demodulator used in Zenith receivers is actually a variation of the pentode demodulator. However, there are two significant departures from the conventional design. Since the amplitude of the output from sheet-beam demodulators is sufficient to drive the CRT directly, they are known as high-level demodulators. The more important difference is the means by which detection of the chroma signal is accomplished. In the triode and pentode circuits just discussed, the amount of cathode current is controlled jointly by the chroma and reference signals. In the sheet beam demodulator, cathode current is controlled solely by the chroma signal, but the selection of which plate receives the current is controlled by the reference signal.
Fly. 17. Chroma demodulators and color-difference amplifiers of Packard-Bell
Chassis 98C15.
Fig. 18. Sheet-beam demodulator of Zenith Chassis 24MC32.
Fig. 19. Chroma demodulator of Admiral Chassis 1G1155-1.
Fig. 18 shows the demodulators used in the Zenith 24MC32 and 42 chassis. Consider the B-Y demodulator under no-signal conditions. The reference signal is fed to the two deflection plates in opposite phases and so the cathode current is directed alternately to each of the plates. Since the amplitude of the reference signal is considerably greater than the minimum required to deflect all of the current to one plate, the current at each plate consists of a series of square-wave pulses of constant amplitude.
During color reception, the chroma signal is fed to the control grid and determines the amount of instantaneous current that is available to the plates. Assume, for example, that the left deflection plate, pin I, is maximum positive at the same instant that the control grid is maximum positive. In this case, the current of the left plate is maximum and the plate voltage is minimum. Since the right deflection plate is positive one-half cycle later, the control-grid voltage is negative during the time that the right plate is gated on. Thus the current of the right plate is minimum and the voltage is maximum.
A 90° phase difference between reference and chroma signals causes one plate to be gated on while the chroma signal swings from peak positive to peak negative. The other plate is gated on while the chroma signal swings from peak negative to peak positive. Therefore, the average of all the values of instantaneous grid voltage present while one plate is gated on equals zero and the average value of plate current is equal to the amount of current under no-color conditions. For intermediate phase relations of the reference and chroma signals, the current of a specific plate varies in be same manner as in any of the other demodulators described. The curve shown in Fig. 13 applies to this circuit except that the phase relationship is shifted 180°. The phase of the reference signal s established so that the left side of V21 demodulates on the R-Y axis and the output is fed to the red grid of the CRT. The right side of L3I is a filter which attenuates the 3.58-MHz ripple. Since the right deflection plate of V21 is driven 180° out of phase with the left deflection plate, the right side of V21 s operating on the (RY) axis, another way of saying that its output polarity is reversed. In the same fashion, the right side of V20 operates on the B-Y axis and drives the blue grid of the CRT while the left side produces the negative or (B Y) output.
As we have pointed out previously, the G-Y signal is composed of -.51(R-Y) and-.19(B-Y).
Fig. 20. Chroma demodulator of Motorola Chassis A22TS-918A.
The signals from the left side of V20 and the right side of V21 are combined in the proper proportions and the GY signal which results is fed to the green grid of the CRT. Twin-Pentode Demodulators The twin-pentode demodulator used in Admiral and Motorola receivers is still another significant variation of the conventional pentode demodulator. The circuit shown in Fig. 19 is used in the Admiral 1G1155-1 chassis.
Insofar as demodulation of the R-Y and B-Y signals is concerned, the operation of this circuit is not greatly different from the operation of a circuit employing a pair of pentode demodulators. The reference signal is fed to the suppressor grids at the correct phase to cause the left and right halves of the tube to demodulate on the B-Y and R-Y axes, respectively.
After the 3.58-MHz ripple has been filtered, these plate signals are used to drive the blue and red grids of the CRT. Before attempting to explain the method used to develop the G-Y signal, we must digress somewhat to consider the distribution of the cathode current between the screen grid and plate of a pentode. An oversimplified explanation may be developed by considering what happens if the plate voltage is removed from a pentode having a low-impedance screen supply. There is an immediate increase in screen current followed by a gradual decrease as the screen-grid structure melts and flows to the bottom of the envelope. While this demonstration is, perhaps, more dramatic than useful, it does indicate that the screen current increases if the plate current decreases.
Actually, the screen current depends to a great degree on the velocity of the electrons passing between its conductors. If the electrons are travelling at very high velocities, the positive potential of the screen grid deflects them from their path only slightly. Thus, the only electrons which impinge on the screen grid are the ones which happen to be travelling directly towards the wires of the grid. However, if the velocity is decreased, the screen grid causes more deflection of the electrons. Since the screen grid is positive, the electrons are deflected toward its conductors and more of the electrons actually strike the wires. This, of course, causes an increase in screen current.
Now, consider the electron velocity with various suppressor grid potentials. If the voltage on the suppressor is driven negative, electrons in the region between the screen and the suppressor are decelerated to a lower velocity and the screen current increases. Conversely, if the suppressor is driven positive, electrons are decelerated a lesser amount, and the screen current is reduced.
A positive suppressor voltage which decreases the screen current also increases the plate current, and, while the plate swings negative, the screen swings positive. Thus, a signal on the suppressor grid may be amplified in both the plate circuit and the screen circuit. The signal is inverted between suppressor and plate, but, since a decrease in plate current is attended by an increase in screen current, the signal at the screen is in phase with the suppressor grid and out of phase with the plate.
Fig. 21. CRT grid dampers of RCA Chassis CTC27.
Fig. 22. Color-tracking circuit of Hoffman Chassis 913-187486.
Returning to the circuit of Fig. 19, since the signals at the two plates of V16 are BY and R-Y, the common screen grid has elements of the -(B-Y) and -(R-Y) signals on it. However, the suppressor-to-screen transconductance is comparatively low and the GY signal at the screen grid is insufficient to drive the green grid of the CRT. To increase the GY signal at the screen, portions of the R-Y and GY signal from the plates of V16 are coupled back to the control grid. This signal is amplified in the control-grid/screen-grid circuit and the amplified signal adds to the G-Y signal already present at the screen. Since the (R Y) " (BY) signal at the control grid is a video frequency and the chroma-signal frequency is much higher, the interaction between them is too small to be objectionable. The concept of amplifying two dissimilar signals in the same tube (reflex amplification) is by no means a new one. For example, a single tube is used as the second video IF amplifier and first sound IF amplifier in the Muntz J chassis (PHOTOFACT Folder 444-2). The demodulator used in Motorola chassis A22TS-918A (Fig. 20) uses the twin-pentode which serves as the reference oscillator to demodulate all three color axes.
The explanation of the division of current between plate and screen grid which was given above also applies to this circuit.
In this part of this guide, it was pointed out that the cathode, control-grid, and screen-grid circuits of V 15 (Fig. 20) function as a reference oscillator in a Hartley configuration. In considering V15 as a demodulator, we need only remember that a high-level reference oscillator signal is being fed to the control grid.
Since this control grid is common to both pentode sections, the reference signal cannot be split prior to demodulation. Consequently, the phase of the chroma signal is shifted between the two suppressor grids.
Since the axis of a particular de modulator depends on the relative phases of the reference and chroma signals fed to it, it is not important where in the reference or chrominance circuits the phase-shifting network is located. In Fig. 20, the phases of the chroma signal are established by L28 and its associated components so that the left plate of V15. pin 6, demodulates on the B-Y axis and the right plate, pin 1, demodulates on the R-Y axis.
Because the division of current between plate and screen grid is determined by the instantaneous voltage of the suppressor grid, elements of the (RY) and (BY) signal are present on the screen grid of V 15. A portion of the B-Y signal from the junction of R208 and R207 is added to the screen-grid signal, reducing the (B-Y) content. The total of the three signals is the G-Y voltage fed to the green grid of the CRT. C161 , C162, and C163 in conjunction with L36 are used to attenuate the 3.58-MHz ripple which is present on the plates and screen grid of V15.
Miscellaneous Circuits:
Three circuits which are being used more frequently are CRT grid dampers (DC restorers), color tracking circuits, and tint controls.
Fig. 20 shows an example of the latter. R6A is connected between the red and blue grids of the CRT and allows the customer to adjust the relative conduction of these two guns.
Fig. 21 shows the CRT grid dampers used in the RCA CTC27 and CTC31 chassis. By inserting coupling capacitors between the plates of the difference amplifiers and the CRT grids, it is possible to operate these two elements at non related DC potentials. This allows simplification of the circuits and also precludes changes in color temperature caused by changes in the static conduction levels of the difference amplifiers. Thus, the CRT grids are not affected by aging of the difference-amplifier tubes and components.
Although the use of coupling capacitors is desirable for the reasons stated above, a means of maintaining the correct CRT grid voltages during b-w reception had to be devised. Otherwise, the right sides of the coupling capacitors would slowly charge through the 2.2-megohm resistors to 405 volts, saturating the CRT. To prevent this, a negative-going pulse with a negative peak of 180 volts positive above ground (refer to waveform in Fig. 21 is fed through the damper diodes to the CRT grids during horizontal-retrace time. Between pulses, the CRT grid voltages rise from 180 volts to about 181 volts, but this is insufficient to cause a noticeable change in brightness across the screen.
A method of increasing color saturation, as well as contrast, with a single control is illustrated in Fig. 22. As the contrast control is adjusted for greater contrast, the forward bias on Q1 is increased. The increased current through Q1 causes the collector voltage to decrease, reducing the positive cathode bias of the chroma bandpass amplifier and increasing its gain.
Positive blanking pulses are also fed to the cathode of the chroma-bandpass amplifier.