|
|
(source: Electronic Technician/Dealer, Jan. 1980)
By Bernard B. Daien
Where have all the diodes gone?
The proliferation of ICs has brought into vogue another kind of detector circuit that helps eliminate many of the troublesome problems associated with alignment and trap settings. For a rundown on this "new" way of doing things, read on.
In the past, the word "detector" in electronic literature was considered synonymous with "diode." Other detectors (demodulators) have been used occasionally, each offering advantages for certain applications, but they were all more complicated, used more components, and cost more, than the simple diode.
With the advent of low cost integrated circuits, other detectors became more prevalent. For example ... single sideband has been used in commercial and amateur transmission, but lately it got a big boost when Citizens Band (CB) transmitters adopted it too. Since single sideband (SSB) transmissions have no carrier, it is necessary to reinsert a carrier at the receiving end. In this application, the "product detector" offers significant advantages over a simple diode, and integrated circuit product detectors are now in common use.
More recently, the "synchronous detector" has enabled the use of simpler, more economical, TV intermediate frequency amplifier strips.
In most cases, many of the troublesome trap circuits have been eliminated! The result is increased use of the synchronous detector as a trend. If you wish to know more about the synchronous detector read on! The non linear diode
You know that the theory of operation of the diode detector is that it functions as a high frequency rectifier. Ideally any rectifier should conduct in one direction only, have zero reverse conduction, and zero voltage drop in the forward direction. Also, you already know that the silicon diode has a forward voltage drop of about six tenths of a volt ... and when you consider that the diode is not going into full conduction until it gets over half a volt input, you suddenly realize that this is putting quite a burden on the I.F. amplifier.
If we assume a signal of 100 microvolts is useable at the antenna terminals, then the I.F. amplifier would have to deliver almost 80 db of gain in order to deliver enough signal to just turn on the silicon diode detector! To get around this, many circuits employ germanium diodes as detectors, since germanium goes into conduction at about two tenths of a volt (one third the forward threshold voltage of silicon).
Furthermore, in the region of low level operation, the diode is very non-linear, i.e., the voltage drop across the diode is not proportional to the current through the diode. Instead the diode has a logarithmic response. Nonlinear devices are good frequency mixers (which is exactly why the diode is used as the mixer in UHF TV tuners). The result is that all frequencies fed into the diode detector mix with all the other frequencies to produce sum and difference frequencies. In the case of the video detector in a TV set we have several frequencies that are present, including the picture carrier, the sound carrier, the adjacent picture carrier, adjacent sound carrier, and the color signals. These different frequencies, beating together, produce interference frequencies which are visible in the picture because the video amplifiers are broad band circuits and pass the spurious frequencies.
In order to prevent this, traps are incorporated in both the video IF and the video amplifiers to reduce the unwanted signals. In the video IF, by trapping out unwanted signals before they reach the detector, the mixing action is prevented.
In the video amplifier, traps reduce the effects of signals which have already been mixed in the detector. But ... if the detector were linear, there would be no mixing action, and the traps would not be required.
Fig. 1 A basic differential amplifier circuit, the backbone of the synchronous
detector.
Fig. 2 A dual differential amplifier with outputs in parallel and inputs
in push-pull arrangement.
Fig. 3 A double balanced demodulator for use in synchronous detection
circuits.
Fewer stages
Also, if the detector could be designed to function more like an amplifier, the input signal to the detector could be reduced, thus less gain would be required from the IF strip. Putting all of this together, a high class TV set could be built with one less IF stage, and fewer traps. That is exactly what the synchronous detector does. Stated another way, a diode detector has a loss between input and output, while the synchronous detector circuit has substantial gain. There is no mixing action in the synchronous detector.
Since the synchronous detector does not mix, there are no sum and difference frequencies produced. With the diode the difference frequencies are the ones that appear on the TV screen ... since they are low frequencies. The sum frequencies are quite high, and they can radiate back into the front end and the antenna of the set, or nearby sets. This is particularly annoying since it entails mechanical shielding, electronic filtering, and critical parts placement ... which means extra size and cost in set design. Federal regulations concerning radiation are very strict, and must be complied with ... so you can see that the synchronous detector overcomes some very severe problems inherent in the good old diode detector. Sorry to disillusion you, but the old diode detector that you have become so familiar with over the years, is no longer a friend! Here's the bad news ...! At this point you are probably saying, "I'm sold. Let's use the synchronous detector. I'm ready ..!" No, you're not! While it cannot be disputed that the synchronous detector is a technological improvement, there are "growing pains" associated with it. As you will learn later in this article, the synchronous detector requires two inputs, just as the product detector requires a substitute carrier for SSB reception, in addition to the input signal ... so does the synchronous detector require a carrier generator input.
The concept of two inputs to a demodulator is nothing new in color TV.
The color demodulator requires an input signal (chroma), and a restored carrier (color reference signal) derived from the color burst. Further, you will recall that there are two popular methods of generating the phase synchronized color reference signal ... by using a free running 3.58 MHz oscillator locked to the color burst by means of an AFPC circuit, ... or, by using the color burst to ring a crystal, then limit the amplified ringing to produce a constant amplitude signal.
We do exactly the same thing with a synchronous detector ... we derive the reference signal from the transmitted picture carrier in the video IF, in one of two ways ... the first is by using a phase locked oscillator ... the second is by amplifying the carrier itself, and limiting it to assure constant amplitude. Doesn't that sound familiar? It should be! This requirement for a reference signal is where the pain comes in. Your old sweep generator won't hack it for alignment, because it puts out a single, constantly varying frequency. The synchronous detector needs a fixed, stable, carrier. (An ordinary TV video signal has a carrier with sidebands. The carrier is stable, the sidebands vary in frequency and amplitude, and phase relationships.) You can add the output from a marker generator at the video carrier frequency, and it would work, providing you adjust the amplitude with care. Unfortunately many sweep generators of modern design do not inject the marker generator into the IF along with the sweep. (There are several other ways to generate a marker display on a scope screen, and they are used more frequently than direct IF marker injection.) So you may have to either modify your present generator, or buy a separate crystal controlled generator for the carrier frequency, or buy one of the new generators made for use with synchronous detectors, AND LEARN HOW TO USE IT!
Most of us have never really felt we were masters of alignment, now we have to learn another way of alignment! But that's the price of progress ... some of us simply drop out with each new technological advance, but fortunately, most of us go back to the books.
"Low level" detectors
The synchronous detector has been around for quite some time, even in integrated circuit form. Motorola has been selling it as the MC1330, and the MC1331, labeled "'Low Level Video Detector." They replace one IF stage, the detector, first video amplifier, and first AFC stage. In addition, the MC1331 also incorporates the sync separator and sound detector! So you probably have been looking at the synchronous detector for some time without realizing it.
Motorola also makes the MC1496 and MC1596, "Balanced Modulator and Demodulator," a very flexible IC which can be used in an AM modulator, suppressed carrier modulator, FM detector, phase detector, product detector, or synchronous detector! RCA makes the CA3136, "TV Video IF Phase Locked Loop Synchronous Detector.”
This IC also has an AFT circuit. National Semiconductor makes the LM273 and LM274, "Monolithic AM/FM/SSB IF Strip" which incorporates an IF amplifier, multimode detector, and other features. Signetics makes the N5596 which is similar to Motorola's 1496 and 1596.
And, there are others. The fact that the synchronous detector has been hidden inside an IC, as a part of a multi -function circuit, or has been given another name, has effectively camouflaged it. The synchronous detector is really an oldie in need of a press agent! This is similar to the "Phase Locked Loop," which has been around in TV horizontal phase lock circuits, and in AFC circuits for many years. For those of you who are old timers in the TV business ... think back to the early 1950's ... remember the Philco TV sets (black and white of course), with AFC? That was a very important feature in those days, since the early TV sets did not use the "inter-carrier Sound System," and you had to tune the fine tuning knob ever-so-carefully for the sound. If you mistuned a bit, the sound vanished.
Thus AFC was an effective means to prevent the sound from drifting clean out of hearing! That was an example of a basic form of phase locked loop, although no one called it that at the time.
We do go through these "rediscovery cycles" periodically in electronics. The semiconductor was in use in 1923, in the form of the crystal detector, using silicon diodes (called "carborundum" at that time), lead sulphide (called "galena"), and the copper oxide rectifier. Yet the semiconductor was "rediscovered" with amazing regularity in the form of the selenium rectifier in the 1940's, germanium, also in the 1940's, and silicon in the 1950's. So do not be discouraged by this apparently new development.
How it works
We can explain how the synchronous detector works with the aid of a few simplified basic schematic diagrams.
Figure 1 is a conventional differential amplifier, consisting of two identical transistors with a single emitter resistor.
The input signal is applied between the two inputs (bases), and it only takes about 30 millivolts difference between the bases to turn one transistor fully off, and the other fully on (saturation). Thus, with a signal greater than 30 millivolts, the differential amplifier becomes an electronic switch.
In Figure 2 we have two identical differential amplifiers, with the outputs in parallel, and the inputs connected phase opposing (push-pull inputs, parallel outputs). We have also added a switch in series with the emitters of each differential amplifier. If we now apply an RF signal to the inputs of the differential amplifiers, at the terminals marked, "Carrier," with both switches closed, we will get NO OUTPUT. Remember, the inputs to the two amplifiers were out-of-phase, and therefore the outputs will be out of phase on the two amplifiers, and cancel each other out. (Those of you who are familiar with single side band will recognize this circuit as a "balanced modulator" used to suppress the carrier in SSB.) If we now open one of the emitter switches, we will disable one of the amplifiers, and output will appear since the cancelling action will be defeated.
If we alternately open and close the switches in sequence, the output will appear and disappear ... but remember, the output is an RF signal at the carrier frequency. Now let's go one step further, if we can synchronize the switches with the carrier frequency, as the phase of the output tries to change due to carrier going positive to negative, and vise versa, the switches will also be changing, and the result will be full wave rectified DC. This circuit is called a "double balanced modulator," and depends upon balancing due to identical components in all circuits. This is why the synchronous detector really became practical only after integrated circuits were developed. Since all the components in an IC are made at the same time, of the same materials, in the same way, they can be made very closely alike. To build a double balanced modulator out of discrete parts would require careful matching, plus the generous use of trimmer pots ... very costly in labor.
Fig. 4 Here is one method of obtaining a carrier frequency with the same frequency and phase as the modulating frequency. The carrier frequency becomes much greater in amplitude than the modulating frequency after passing through the narrow band amplifier.
Current regulators
Now let's look at Figure 3. We have now replaced the switches with transistors, which can not only turn on and off, but also act as variable current regulators, limiting the current flow to each differential amplifier in accordance with the input signal labeled "modulating signal" ... because that is exactly what these transistors do, they modulate the current flowing to each of the differential amplifiers in accordance with the modulating input signal. Now here is an important point ... the carrier signal is deliberately maintained at a level which exceeds 30 millivolts, therefore the differential amplifiers function as switches. The bottom transistors are operated in the linear mode, as current amplifiers, and therefore the output from the differential amplifiers is modulated by the modulated signal input. If the carrier and the modulating frequency are identical in frequency and phasing, the output will contain only the sum and difference frequencies between the carrier, and the modulating frequency's sidebands.
Remember, any modulation on the carrier frequency will not appear in the output so long as the carrier level is well over the 30 millivolt level. Thus we try to keep the carrier level around 300 millivolts, and we keep the modulated signal level under 30 millivolts.
At this point you may be asking, "How do we get a carrier that is exactly on frequency, and in phase with the modulating frequency?" (The modulating frequency, as you have probably already figured out, is the frequency we are "detecting.") If you will examine Figure 4, you will see that the carrier frequency is derived from the modulating frequency. We simply run it through another amplifier stage to insure that the signal level is high enough to exceed the previously mentioned 30 millivolt level.
Since we are only interested in the carrier, this can be a typical narrow band tuned amplifier centered on the carrier frequency. Such an amplifier has considerably higher gain than a broadband amplifier, and is much simpler and cheaper to make. Some circuits amplify, then clip this signal to remove most of the amplitude modulation, in a manner similar to the FM IF limiter stages used with the better FM broadcast receivers. Other circuitry often used simply insures that the carrier signal is so much larger than 30 millivolts that the differential amplifiers act as saturated switches, and therefore do not respond to variations in the amplitude of the carrier (modulation), as described previously.
Fewer traps
Stated another way, the carrier frequency supplies only the switching action, while the modulated frequency supplies only the modulation. Any modulation of the carrier frequency does not appear in the output, ... and the carrier of the modulated frequency is rectified to DC. As you can readily see, an interfering signal would not meet the requirements previously set forth for output from the balanced modulator.
More important, the double balanced modulator does not mix "other" signals to produce sum and difference frequencies, as does the ordinary diode detector. Therefore there is no "beating" with undesired carriers, as is usually the case with the adjacent channel signals and the desired signals.
This ability to discriminate against unwanted signals reduces, or eliminates completely, the need for trap circuits in the TV video IF. It also greatly reduces the need for very precise tuning of the tv receiver, since the troublesome beats that used to appear on the screen when the signal moved out of the proper trap frequencies, are no longer so obvious.
Those of you who recall the early days of TV before intercarrier sound was developed, still remember that tuning was accomplished by adjusting the dial for the best SOUND, and it was super critical. With the advent of intercarrier sound IF, tuning became a magnitude less critical. The synchronous detector makes things still easier for tuning purposes. As a matter of fact, some of you may have to become accustomed to it! As previously mentioned, a separate amplifier is used to boost the carrier level, and is often incorporated in the IC, but the tuned circuit for this narrow band amplifier is not, and therefore must be external. Thus some synchronous detector ICs have an extra tuned circuit, often labeled "Limiter." There are some detector ICs which use a different method of doing essentially the same job. Instead of obtaining the carrier, or "reference" via a separate amplifier, a phase locked oscillator is used, in much the same manner as a phase locked horizontal sweep oscillator, or a phase locked color oscillator. Since this is part of a phase locked loop, the oscillator may have a tuned circuit, labeled "Reference," or "Oscillator." This oscillator must be set so that it is close to the desired frequency, in order to insure "lock in," in much the same manner as any phase locked oscillator. The limiter is simply peaked up on the carrier frequency.
Some sets have a "zero" adjustment, which is an ordinary potentiometer that adjusts the DC output level from the detector IC to match the requirements of the following video amplifier chain. In any event, these adjustments, which vary from set to set, are covered in the alignment instructions that apply to the particular set, and are no more complicated than adjusting the sound alignment on any color TV set.
Now, until a new development comes along, you are (temporarily), "up to date" ... again!