In Comparing Op-Amp & Video Amp Topologies, Part 1, we mentioned that in a video amplifier, such as seen in Figure 1, a DC servo circuit is generally required to ensure that the average voltage in a back-porch region of the video signal is set to zero volt. In Figure 1, with discrete devices for Q1 and Q2, the input offset voltages are generally in the range between plus and minus 150mV. More expensively matched transistors can be used. However, resistor RL or REE must be chosen or adjusted for 0 volt offset at Vout to counter the offset voltages due to power supply voltage tolerances. Also the temperature dependent base-emitter turn on voltage of the transistors affects the offset voltage.
So it is just easier to somehow servo Vout to zero volts with a feedback system. This is usually done with the addition of a low bandwidth op amp.
For many video processing circuits such as analog-to-digital converters or video display circuits, the blanking level voltage of a video signal is established as a reference voltage. This blanking level in the video signal must not vary when the average picture level changes. Often external video signals contain an offset voltage at blanking level, or the signal is AC coupled.
In applications where two or more video sources are mixed in a manner such as a split screen, picture-in-picture, fade-in/fade-out, or dissolve, it is essential to have the blanking level of each video signal source defined to a common voltage. Otherwise in a split screen, for example, one source may look fine in terms or black levels, but the other source may look washed out. Note that when mixing with multiple video sources, the video signals must be synchronized to each other, or synchronized to a reference generator and then mixed together.
If a video signal is AC coupled, in a dark scene, the blanking level is nearly at 0 volt, but if the scene is bright such as a daylight shot, the blanking level will go below the 0 volt level.
See Figures 2 and 3.
As can be seen in Figures 2 and 3, depending on the brightness of the scene of the video program, an AC coupled video signal produces different voltage levels for the blanking level.
In Figure 2, a black level signal for a 525 TV line system is shown to have the blanking level as nearly 0 volt when AC coupled.
However, for an AC coupled video signal that has a substantial brightness level of about 90% peak white, the average picture level is now high. When this signal is AC coupled, the blanking level voltage is pushed down to -0.445 volt as shown in Figure 3.
Before mentioning methods used to ensure the blanking level of a video signal is maintained to a 0 volt level, I should mention something about DC servo circuits for audio amplifiers. In audio amplifiers that do not have an output coupling capacitor, the output signal of the audio amplifier is low pass filtered via an integrator circuit to provide the average DC offset voltage and this average DC offset signal is then compared with a reference voltage such as 0 volt. The integrator circuit's output signal is then generally fed to the (-) input of the audio amplifier via a large value resistor.
Thus, the output of the audio amplifier is “sampled” continuously by the DC servo circuit based on the audio signals on the average yield 0 volt, and any other average voltage at the output of the audio amplifier must be a DC offset voltage. See references  and .
Video signals however, are not symmetrical signals that have an average voltage of 0. Instead, they often have a net positive voltage such as a video signal related to a video program that shows mainly well lit outdoor scenes. Thus a DC servo circuit that is used in an audio amplifier is not applicable to video amplifiers.
There are many ways to establish a stable blanking voltage level. Examples are a DC restoration circuit using a diode and capacitor, a sync tip or back-porch clamping circuit using an FET or bipolar transistor and a coupling capacitor, or a sampled back-porch feedback circuit using a sample-and-hold capacitor with an op-amp. For this discussion, a sampled back-porch feedback circuit will be presented.
Figure 4 shows the amplifier of Figure 1 with a feedback clamp circuit.
In Figure 4, the input video signal is buffered with an emitter follower Q4 to drive a sync separator circuit U1. From the trailing edge of the horizontal sync pulses, a back-porch clamp pulse is provided via a timing generator within U1. The back-porch clamp pulse from U1 (the LM1881) is then a negative going pulse that is coincident with a portion of the back-porch region of the input video signal. See Figure 5.
An output of the amplifier via the emitter of Q3 is fed to a series R-L low-pass circuit (R2 and L1) to remove the high-frequency color burst signal when sampled during the back-porch region into the 0.1μF sampling capacitor connected to the (+) input of U3A (e.g. TL082). The voltage at the (+) input of U3A is thus indicative of the average voltage at a back-porch region of the video signal. The (-) input of U3A is connected to a 100kΩ resistor with a ground reference at a 0 volt level to servo the back-porch voltage to a 0 volt level.
The output of U3A provides an integral and proportional control signal to the resistor labeled “R DC Clamp,” which then provides a negative feedback circuit with the base of Q2 for clamping the back-porch level of the video signal to 0 volt.
Clamping circuits such as the one just described are used to condition the video signal to have a well-defined blanking voltage level.
In the beginning days of black-and-white television sets from the 1940s and even to some color televisions from the 1970s, AC coupling of the video signal was often used for driving the cathode or grid of a cathode ray tube (CRT). The result was that when a well lit scene was adjusted for optimal viewing via the brightness and contrast controls, a dark scene or a fade to black would end up being a washed-out gray or a fade-to-gray. A simple DC restoration circuit or a clamp circuit would have fixed this problem.
I was asked once to repair a color TV set. Once I fixed the major problem, I could see that the video to the CRT was AC coupled and produced washed-out grays instead of true black levels. I modified the set by inserting a DC restoration circuit at the CRT. I replaced the DC bias resistor (i.e., >100kΩ) that was connected to the brightness pot to provide a variable voltage with a 1N4007 diode. The other end of the diode went to the AC coupling capacitor, and a bleeder resistor (e.g., ~500 kΩ) was tied to a voltage reference to slowly discharge the AC coupling capacitor. The result was a TV set that then displayed true black levels.
In the next article, we will look at other video amplifiers. For a sneak preview, see Figures 6 and 7 that show two other video amplifiers.
In the next article, these two video amplifiers  will be examined and compared with a standard op-amp.
- “Audio Power Amplifiers,” Bob Cordell, McGraw Hill, 2011
- “Audio Power Amplifier Design,” Douglas Self, Focal Press, 2013, Sixth Edition
- “BVE-3000 Editing System,” Sony Corporation, 1983, 3rd Edition
- “AVR-3 Videotape Recorder/Reproducer,” Ampex Corporation, 1976