DEMODULATOR PROBES [abc's of ELECTRONIC TEST PROBES (1966)]

The usefulness of the present-day oscilloscope is undisputed. Its value as a research tool and service instrument is further enhanced by several types of demodulator probes which permit the oscilloscope to be used at frequencies it would otherwise not be suitable for-such as those in television i-f, r-f, and video stages, which are too high to be observed directly on an oscilloscope . Television stations transmit a composite television video signal consisting of an amplitude-modulated r-f carrier, together with blanking and sync pulses. These carrier frequencies are close to 900 megacycles if we go as far as Channel 83 in the ultrahigh-frequency band.

Although the vertical-amplifier sensitivity of oscilloscopes is quite high, their limited frequency response causes us difficulty. Direct observation of signals above several megacycles becomes unreliable, and often impossible, with a general-purpose oscilloscope. Wide-band oscilloscopes, with a frequency response extending to perhaps five megacycles or more, are not only more expensive, but their deflection sensitivity also is usually less than that of a high-gain, relatively narrow-band oscilloscope. The latter somewhat limits their suitability for observation in low-level circuits . When we talk about the sensitivity and frequency response of an oscilloscope, we will concern ourselves with the vertical amplifier only.

With many oscilloscopes, the signal we wish to observe can be connected directly to the vertical-deflection plates. Thus, we circumvent the frequency-limiting characteristics of the vertical amplifier; but, in doing so, we sacrifice the deflection sensitivity gained by using the amplifier. Therefore, this method of operating is suitable only for signals whose amplitudes are such that we get satisfactory deflection directly, without the need for additional amplification.

Because the vertical-amplifier circuits of oscilloscopes do not respond to the high frequencies in the i-f and r-f circuits of radio and television receivers, and since we are interested in the modulation envelope only, we can use a demodulator probe to demodulate our signal and thus get the desired indication on the oscilloscope . The performance requirements for a demodulator probe are more stringent than for a rectifier probe . The main difference between a demodulator and a rectifier probe is that the former rectifies and removes the carrier frequency before passing the modulation envelope of the r-f signal on to the vertical amplifier of the oscilloscope. The rectifier probe, on the other hand, rectifies and filters both the carrier and the modulation component, giving an output proportional to the peak value of the carrier signal (whether modulated or not) . Accordingly, the filter characteristics of an oscilloscope demodulator probe are determined by the service applications it is designed to meet.

The peak value of our modulated signal varies at the modulation rate. We must therefore design our probe so its output voltage will rise and fall with the envelope of the r-f signal. In other words, in video circuits our probe must completely rectify and filter video frequencies from 100 khz to 4.5 mhz, and must also pass a 60- hz square wave undistorted. For i-f signal tracing, the probe should have a relatively high input impedance from 25 up to 45 megacycles. The demodulator probe thus gives a low-frequency vertical-deflection voltage proportional to the instantaneous amplitude of the r-f signal. Since the voltage levels at which such signals exist are not too low, we do not need a very sensitive probe . So, we can direct our efforts toward designing a wide-band probe. The gain of the oscilloscope is also on our side.

It usually does not take much input signal to get a substantial deflection on a relatively sensitive oscilloscope . Fortunately, the input to the oscilloscope is the demodulated signal; so, we can concentrate on the probe for fidelity and on the oscilloscope for sensitivity . The input characteristics of the probe also are important. Must the impedance always be high? Or should it sometimes below? Does the frequency response always have to be wide? Or is a relatively narrow frequency response sometimes sufficient? Such questions enter into the choice of a probe . Any characteristic can be made predominant by proper design . The highest possible fidelity of reproduction is usually achieved at the expense of input impedance. Therefore, a probe which faithfully reproduces signals over a wide frequency range usually has a low input impedance. Conversely, a high-input impedance probe usually has rather poor frequency characteristics. Demodulator probes are rather good compromise . Some of the more desirable characteristics of a typical demodulator probe include the highest practical input impedance, the greatest possible sensitivity, high fidelity of output, good mechanical construction, and 60- hz hum rejection. Of course, these qualities cannot all be successfully combined in a practical probe-but at least this is our design goal.

Fig. 5-1 . Tuned circuit changes f-m signal to a-m signal; demodulator probe extracts information from a-m signal; oscilloscope displays signal.

The r-f signal we observe will be either frequency- or amplitude-modulated. Nevertheless, we must first demodulate the signal, and then apply the demodulated signal to the vertical amplifier of our oscilloscope. Let us first consider a frequency-modulated signal, which we can use to test the frequency response of an i-f amplifier. Fig. 5-1 shows such a signal applied to a tuned circuit. The output signal from the tuned circuit varies in amplitude, in conformance with the frequency response of the tuned circuit.

(A) Amplitude modulated signal (input). (c) After proper demodulation.

(B) Output from half-wave rectifier.

(D) Demodulation with negative peak clipping.

Fig. 5-2. Successive steps involved in demodulating an amplitude-modulated signal.

A frequency-modulated test signal has a center frequency from which it deviates, usually at a 60- or 120- hz rate. For example, if the signal has a center frequency of 44 megacycles and a deviation of ±4 megacycles, it will contain all frequencies from 40 to 48 megacycles. The tuned circuit will pass only those frequencies within its passband, and the magnitude of each frequency at its output will be directly proportional to the characteristics of the tuned circuit. The signals here are of such high frequencies that they cannot be displayed directly on an oscilloscope, but must first be demodulated. This is accomplished either with a demodulator probe, or by the demodulator in the equipment under test. If we wish to observe the characteristics of successive tuned stages as we progress toward the earlier stages of a receiver, we apply the output of a demodulator probe to our oscilloscope . We start out by applying a constant-amplitude, frequency-modulated signal to our circuit under test. Then we take the output signal , which now has amplitude characteristics corresponding to the frequency response of the circuit under test, and demodulate it. (This is shown in Fig. 5-1.) We thus come up with a modulation envelope representing the characteristics of the tuned circuit. Now we have a means of observing the characteristics of a circuit that operates at frequencies far beyond the capabilities of an oscilloscope.

At first glance, a demodulator probe looks much like the rectifier probe discussed previously. That is true-the probe circuits do look alike. However, the values of the components in the two probes differ.

With the rectifier probe, we charge a capacitor to the peak value of the applied signal, and keep it charged to that value. Then we measure this charge with a DC measuring device. A different situation prevails with the demodulator probe. Here, we are interested not so much in the exact amplitude of the signal, but in its shape. The probe must therefore faithfully demodulate an amplitude-modulated signal and present on the oscilloscope an exact reproduction of the modulation envelope. We are now faced with a time-constant problem. Whereas before we wanted a long time constant, now we want a relatively short one, compared with that of the signal under test. Fig. 5-2 shows the action of a half-wave demodulator probe. When a signal is demodulated, either the positive or the negative half of the modulated signal is rectified to provide either the positive- or the negative modulation envelope. Therefore, from the probe is obtained a unidirectional output signal which changes in step with the modulated signal. If the time constant of the probe is too long, a sharp drop (fall time) cannot be followed because the charge on the capacitor will not have a chance to leak off. This will cause negative-peak clipping. On the other hand, if the time constant is too short, our signal will be ragged or fuzzy, and it will tend to follow the carrier frequency rather than the modulation.

The magnitude of the signal is generally such that semiconductor diodes can be used in place of vacuum tubes. This is an advantage, for semiconductor diodes require no heater voltage ; nor is there the hum problem associated with AC filament voltage, to cause trouble- some hum modulation on an oscilloscope. In addition, the semiconductor diode makes possible a much smaller (and thus more easily handled) probe than does a vacuum tube.

The front-to-back resistance ratio of the diodes should be as high as possible. In a balanced probe, this ratio (as well as the values of both the front and back resistances) should be matched for both diodes.

The characteristic curve of a germanium diode becomes nonlinear below about 0.5 volt. This causes difficulties at signal levels of 0.5 volt or lower. The curvature, which distorts and magnifies changes in signal voltage, becomes troublesome when we measure voltage ratios.

For example, a 5-to-1 change in actual signal level from 0.5 to 0. 1 volt may show up as a change of perhaps 6- or 8-to-1 because of the nonlinear characteristics of the diode. Another time this characteristic may become bothersome is in observing the response curve of a tuned circuit. At the extremes of the curve, where there is very little signal, the probe may show that the curve falls off more rapidly than it actually does. This is an important point to consider when measurements are made at low levels.

The response of a probe to complex signals is determined by the relationship of the resistors and capacitors following the rectifier. The r-f high-frequency limits are influenced by the circuit preceding the rectifier probe. A high-impedance diode and a short lead will greatly improve the high-frequency characteristics of a demodulator probe and proper shielding is most important. Available are series- and shunt type demodulator probes which use series resistors to isolate the oscilloscope and cable shunt capacitances from the probe.

As the frequency increases, the impedance of our demodulator probe becomes very low. Therefore, we must be sure to measure from a low-impedance point if possible, so the low impedance our probe presents at frequencies above 100 megacycles does not disturb the circuit response. At 1 mhz the equivalent input resistance of a typical semiconductor demodulator probe is approximately 25,000 ohms. It drops to about 5000 ohms at 100 mhz, continuing to drop as the frequency rises. The input capacitance of a demodulator probe should be kept as low as possible ; the average is somewhere between 3 and 10 pf.

SHUNT-TYPE DEMODULATOR PROBES

The performance characteristics of the shunt-type demodulator probe are similar to those of the equivalent shunt-type rectifier probe in Section 4. The schematics of both probes are similar, except the capacitance and resistance of the shunt-type demodulator probe are smaller in order to provide a shorter time constant. In some demodulator probes, the semiconductors are inserted in a direction which will give a positive-going output voltage. Therefore, the anode of the diode is at ground potential, and the cathode is connected to the input signal whose positive modulation envelope we will then display. If this output is applied to an oscilloscope designed to give an upward deflection when a positive-going voltage is applied to its vertical-input circuit (most oscilloscopes are designed this way), maximum output from the probe will result in maximum upward excursion on the response curve. The response curve will then be of a polarity usually considered normal or upright. If the oscilloscope gave an upward deflection for a negative-going voltage, the response curve would be inverted.

Fig. 5-3. Basic schematic of shunt-type demodulator probe.

Although not necessarily incorrect, inverted response curves may be misleading (or inconvenient) if they are to be compared with certain illustrated examples. A reversal of the diode would re-invert the curve, bringing it back to the accepted, normal position. For convenience, both types of probes could be kept on hand-one giving a positive and the other a negative output. The desired polarity of the response curve could then be obtained by using the proper probe.

Fig. 5-3 shows the schematic of a shunt-type demodulator probe.

Like the rectifier probe, it has a capacitor which charges to the instantaneous peak value of the modulation envelope. This is a most important requirement; so we try to achieve it as closely as we can. The series resistor here is not used for calibration, but rather to isolate the cable capacitance from the input circuit. Unlike the rectifier probe (which develops a constant DC voltage), this one develops a varying DC voltage. If the r-f filtering is not sufficient, the r-f pulses may become troublesome if permitted to travel the length of the cable.

A demodulator probe for video-amplifier display must be designed to demodulate a 60- hz, square-wave-modulated r-f signal without introducing noticeable distortion. The ability of the probe to do this depends on the resistance and capacitance values within the probe, as well as on the prevailing distributed capacitances. We want these components to be small-yet, if they are too small, they will not do their job. For instance, too small a series resistor will not isolate sufficiently, and too small a shunt resistance (if used) will short out the signal voltage applied to the probe. We must therefore arrive at a compromise. The time constant of the probe is made up of the cable and input capacitances of the probe, as well as any additional shunt filter capacitance, plus the combined resistance of the shunt and series resistors . The probe is less susceptible to hum because fortunately we now want a small value of charging capacitor (usually not more than a few hundred pf). We can improve high-frequency characteristics of our demodulator probe by inserting a small inductance in the cable . The situation here, however, is somewhat more delicate because we have to be concerned with the waveshape of the output signal. Therefore, if we have a high frequency square wave modulating a high-frequency signal, and want to display the demodulated square wave on our oscilloscope, we can insert a small peaking coil to reduce the rounding of the leading edge.

However, we must make sure that our inductance does not cause ringing. The coil must therefore be selected very carefully. It may even have to be damped by a shunting resistor. To further reduce the loading effect of a demodulator probe, we can also use an isolation resistor of a few thousand ohms ahead of our blocking capacitor.

Probes are often used to display AC waveforms in the presence of relatively high DC voltages. If so, suitably rated blocking capacitors must be used. The semiconductor diodes must not only have a high front-to-back ratio, but must also accommodate reasonably high AC signal voltages without loss of sensitivity or burn-out.

At the video-amplifier output we may find a rather large signal-in fact, one greater than the voltage-handling capability of our diodes.

If so, we can put two or more semiconductor diodes in series to increase our signal voltage-handling capabilities, as we did for the rectifier probe . In low-level circuits, however, the output of our probe may not be sufficiently great to provide a usable deflection on our oscilloscope. So we may have to use additional amplification. This must be a high-quality amplifier which introduces very little, if any, hum at all on its own ; its frequency response should be essentially flat from about 20 to at least 500,000 cycles. A resistance-coupled amplifier must therefore have a gain of at least 20-and if possible, higher.

SERIES DEMODULATOR PROBES

Here again we have a circuit like the one discussed at great length in the section on rectifier probes. This is still about the simplest probe we can make-which we do by placing a semiconductor diode in series with our shielded cable, as shown in Fig. 5-4. Between successive peaks of the modulated r-f signal, the capacitor is discharged some- what, and then is of course, charged again by the following r-f pulse.

If the cable acts as the capacitor, it will carry a series of pulses at the carrier frequency rate. These pulses may have serious consequences at those frequencies where the cable is a multiple of a quarter wavelength. Depending on the frequency, a condition of resonance or anti-resonance may exist which greatly changes the sensitivity of the probe.

Furthermore, since this cable capacitance is quite large, we would experience the negative-peak clipping mentioned previously-horizontal-sync pulses would be greatly attenuated and severely distorted.

Fig. 5-4. Basic schematic of a series demodulator probe.

In order to avoid these difficulties, we must remove the cable capacitance from the circuit. We can do this with the rectifier probe by inserting an isolation resistor between the cable input and the demodulator output. The value of this resistor must not be too large. Otherwise, it will seriously distort the waveshape of the demodulated signal , as well as shift the marker position on the steep side of the response curve. The isolation resistor reduces the sensitivity of the probe some

what. It is therefore best to make the resistor small (or even leave it out if we can) and compensate by other means-such as using a capacitor as a charging capacitor in lieu of the cable capacitance.

We can rapidly discharge the cable capacitance by shunting it with a resistor, so our oscilloscope can follow the modulation envelope of high frequencies and steep pulses. The back resistance of the semiconductor diode determines the performance of the probe. This is a variable value ; in fact, if the diode has a low back resistance, the shunting resistor may not even be needed because the cable capacitance can discharge through the diode.

Also to be considered is the effect of measuring an AC voltage when a DC biasing voltage is present. This bias voltage may even exceed our signal voltage ; if it does, we will get no indication at all. If the signal peaks exceed the biasing voltage, we will get an indication only while the diode receives a signal in the conducting direction. This difficulty can occur in a series demodulator probe if a good blocking capacitor is not used and DC is present. (It can also be caused by a leaky oscilloscope input capacitor, even though a good blocking capacitor is used.) Ideally, the blocking capacitor is an open circuit. However, it does have some leakage resistance which, although high, is in series with our rectifier and input circuit. As such, it forms part of a voltage divider across any DC circuit we apply our probe to. It takes only a minute leakage, resulting in a small biasing voltage, to disable our probe.

Therefore, if a probe is not working properly, the blocking capacitor should be one of the first items checked.

The DC blocking capacitor in the oscilloscope also becomes important when we use a series demodulator probe. Its leakage resistance is in series with the vertical attenuator. Thus, we do have a DC resistance path (it may be a great many megohms) between the vertical input terminals. It and the back resistance of the semiconductor diode now form a voltage divider. If the probe is applied to the plate of a vacuum tube (which may be several hundred volts above ground), DC current will flow through the semiconductor . diode, the leakage resistance of the capacitor, and the vertical attenuator, causing a certain amount of DC voltage to appear across our diode. This is undesirable because (1) the current, if sufficiently high, will damage the diode, and (2) the voltage drop across the diode may bias the diode into an operating region where its full sensitivity will not be realized.

Because these circuits are powered from a rectifier, hum voltages may be present. Although not noticeable on the picture tube of a television receiver, these hum voltages would be displayed, along with the signal under test, on our oscilloscopes. This is so because up to now we have had no means of 60- or 120-hz rejection. We can correct this by adding a small capacitor in series with our semiconductor diode. That is the first job of this capacitor. The second one is based on the fact that, being small (usually mica or ceramic), its insulation resistance is almost infinite. We therefore will not have the DC current and diode biasing problems we would have without this capacitor. Of course, the problem is aggravated if our oscilloscope has no blocking capacitor. We could add a resistance, of a few megohms at most, from the input terminals to ground. However, large-and probably destructive-currents would develop through the diodes if we measured at the plate of vacuum tubes without using any DC blocking capacitors.

The input capacitance of the probe may be sufficiently high to detune the circuit under test. This effect not only is detrimental in some circuits, but may also cause regeneration (oscillation which may damage the diode in the probe) . This difficulty is overcome by connecting a probe with a low input impedance across the stage following the one we are observing. The intervening stage, which acts as a sort of buffer, may also be tuned, but the low impedance of the probe will effectively swamp the tuned response of the circuit and thus prevent regeneration. Because it is now a low-impedance probe, its output is much lower than if it were a high-impedance one. We can usually take care of that by advancing the gain control on the oscilloscope.

Although some output is lost, we have prevented our circuit from breaking into oscillation, which would make any observation impossible.

The probe in Fig. 5-5 is of conventional design, with a blocking capacitor, a series rectifier, and a network to filter out any r-f impulses.

The blocking capacitor should be as close to the rectifier as possible ; its other end, designated the tip, should be short so it can be brought out directly to the point under measurement. The blocking and charging capacitors, if of a good quality, should allay our worries that the leakage resistance of the oscilloscope input capacitor will harm our rectifier. (This type of low-impedance probe is often shown in the service manuals for television receivers .) If a low-impedance probe is not available, we can still use a high-impedance demodulator probe for measuring in a tuned circuit. This we do by adding a swamping network consisting of a DC blocking capacitor, plus a resistor of several hundred ohms, in series across our tuned circuit. The leads between the probe tip and the ground lead must be very short for both components.

Fig. 5-5. Series demodulator probe (low impedance).

It is of prime importance that the detuning effect of the probe be reduced. One way is to apply the output from a cathode-follower probe to the demodulator probe. Being a low-capacitance device, the cathode-follower probe is designed to do nothing more than pick off a signal with the least disturbance to the circuit. Its output can then be directly connected to the demodulator probe and, in turn, fed to the oscilloscope through a shielded cable. With signals of up to several megacycles, a demodulator probe is often unnecessary if the vertical amplifier response of the oscilloscope is wide enough to accommodate them. We simply pick off our modulated signal with the cathode follower probe, applying its output directly to the vertical-input circuit of the oscilloscope. This is a good arrangement for us to keep in mind when observing video-amplifier or sync circuits.

The series-demodulator probe in Fig. 5-5 is used most frequently for television receiver servicing, where it is moved from stage to stage during alignment or signal tracing. Since it is a demodulator (detector) probe and travels from stage to stage, it is often referred to as a "traveling detector." Fig. 5-6 shows the circuit and construction in

formation of another simple probe suitable for alignment of video-i-f stages. This probe is designed to provide the proper loading for correct adjustment of over-coupled i-f stages.

Fig. 5-6. Series demodulator probe suitable for alignment of overcoupled video i-f stages.

Fig. 5-7. Series demodulator probe in kit form (EICO Model PSD).

Fig. 5-8. Probe recommended for television receiver alignment.

Fig. 5-9. High-impedance series demodulator probe.

Fig. 5-7 shows the circuit of a series demodulator probe available in kit form; Fig. 5-8, a semiconductor-diode detector probe recommended by a television manufacturer for aligning his television receiver. A high-impedance demodulator probe is shown in Fig. 5-9.

Fig. 5-10. Balanced demodulator probe having 300-ohm input.

Fig. 5-11. Two identical half-wave probes used to measure a balanced line.

BALANCED DEMODULATOR PROBES

A balanced demodulator probe will often be useful, just as the rectifier probe has been, for observations in such balanced circuits as a twin-lead transmission line or the input of television receivers, boosters, or converters. The probe in Fig. 5-10 has a balanced 300-ohm input and a single-ended, or unbalanced, output. Two rectifiers, connected in the same direction, are employed. This arrangement is satisfactory without a DC blocking capacitor because there is usually no DC voltage where balanced measurements are made. The charging capacitor receives alternate pulses from either diode. When one diode is conducting at its maximum, the other has a signal in the opposite direction, also of maximum magnitude. These polarities alternate at the frequency of the signal under measurement. The output signal from the probe is the modulation envelope of the amplitude-modulated signal under test. The ground lead from our probe should be kept as short as possible and connected to the nearest ground. If a balanced demodulator probe is not available and measurements under balanced conditions are desirable, we can use two identical demodulator probes in parallel, and also connect their outputs in parallel to our oscilloscope. This method is shown in Fig. 5-11. Be sure the diodes in both probes are connected in the same direction. A balanced high-impedance probe is diagrammed in Fig. 5-12.

Fig. 5-12. Balanced high-impedance probe.

Fig. 5-13. A 300-ohm balanced demodulator probe using only one semiconductor diode.

If we use only one unbalanced probe to measure on a balanced line, we will upset the balance and thus get an erroneous indication. It is therefore advisable, when making measurements in a balanced circuit, to construct a balanced probe, or else use two probes as outlined in the previous paragraph. Fig. 5-13 shows the schematic of another balanced probe which uses only one diode, but still offers a balanced input of 300 ohms and an unbalanced output to our oscilloscope. This probe can be constructed on terminal strips. It does not have to be shielded because the low-value resistors make it relatively insensitive to extraneous pickup.

PEAK-TO-PEAK OR VOLTAGE-DOUBLER DEMODULATOR PROBES

We can get a more sensitive probe by making a voltage-doubler or peak-to-peak reading probe. The probe is referred to as one or the other, depending on the type of signal measured. With a symmetrical wave, we will have a voltage-doubler action; but with an unsymmetrical wave, the output voltage will equal the peak-to-peak value of that waveform. A typical voltage-doubler demodulator probe is shown in Fig. 5-14. With a symmetrical signal, a voltage-doubler probe will produce, on our oscilloscope, twice the deflection a half-wave probe would. Such a voltage-doubler probe is therefore useful at low signal levels. However, because it has a lower input impedance than the half-wave probe, it is suitable only for measurements where impedance levels are relatively low. This characteristic may turn out to be the limiting factor in some applications. The frequency response of the voltage-doubler demodulator probe i , therefore, not nearly as good as that of a half-wave probe.

The input capacitor and output charging capacitors are again made rather small in order to make the probe relatively insensitive to 60- or 120- hz hum. The signal-handling capabilities of our probe are once more limited by the voltage rating of the diode . The DC level at which measurements are made should not exceed the voltage rating of the

Fig. 5-14. A voltage-doubler demodulator probe.

probe input capacitor. Alternating voltages in excess of 50 volts peak will tend to produce pattern distortion, whereas inputs in excess of 60 volts peak can impair the sensitivity of the semiconductor diodes, or even burn them out completely. Because of the 60- hz rejection capabilities of our probe, we can make tests for r-f signals in filament, automatic gain control, and B+ circuits, and thus check to see whether or not the filtering, bypass, and decoupling capacitors are doing their jobs.

The peak-to-peak probe may provide more or less than twice the output of a half-wave probe. The reason is that the signals may not always be symmetrical in their positive and negative excursions.

Hence, if we measure the larger excursion with our half-wave probe, the peak-to-peak probe will give us less than twice the output of the half-wave probe . Conversely, if we measure the smaller excursion of the modulated signal, the voltage-doubler probe will then give us more than twice the reading than that of the half-wave probe.

In order to check the half-wave and voltage-doubler probes against each other, obtain a symmetrical signal, preferably one you can control (from a signal generator, for example). First, observe this signal on an oscilloscope, to see whether both halves of the modulation envelope are equal. Then take measurements with both probes to see whether the voltage-doubler probe gives exactly twice the output of the halfwave probe.

A peak-to-peak indicating circuit can be modified to give only peak voltage readings. This is done by means of a switch, which eliminates one diode and capacitor from the circuit. An example of this type of circuit is shown in Fig. 5- 15. When the switch is in the "peak" position, only one diode is in use. In the "peak-to-peak" position, both diodes are active.

Fig. 5-15. Combination peak and peak-to-peak reading demodulator probe.

DEMODULATOR PROBE SELECTION AND USE

The r-f and i-f frequencies in television receivers are too high to be displayed directly on an oscilloscope. Before these signals can be observed, they must be demodulated so a modulation envelope can be obtained. In other words, a demodulator probe extracts the signal from the r-f carrier. This signal will have a DC component and, on a DC oscilloscope, will be displayed vertically by an amount equal to the DC component . On an AC oscilloscope, the signal will be centered about the zero axis.

The demodulator probe adds a certain amount of capacitance. This we must take into account when measuring the stage gain and characteristic of tuned circuits, because the output voltage of our probe may not be the same as the actual voltage at the point of measurement (depending on which direction our probe detunes the circuit) . Sometimes the probe may supply the necessary additional capacitance to properly tune our circuit; then we would get a better indication than without the probe. On the other hand, if our probe detunes the circuit, the indication would be lower than normal. The circuit may also break into oscillation when the probe is applied, in which event any indication would be completely meaningless.

Fig. 5-16. Inductance of single wire at high frequencies.

Demodulator probes should be shielded so they will not pick up voltages, other than those at the point under observation. If a probe is simply held near a field-producing element (such as the horizontal output transformer of a television receiver), there will of course be an indication, which should disappear as soon as the probe is connected to the point under test. To check the shielding of our probe connect a small resistor of perhaps 10,000 ohms between the tip and ground.

With the resistor connected in this manner, move the probe around the television chassis. If the probe is properly shielded, there will be no indication on the oscilloscope.

Briefly comparing the characteristics of series, shunt, and voltage doubler probes, we find that the series-demodulator probe is somewhat more sensitive, but does not attenuate hum as much as the shunt type demodulator probe does. Furthermore, the series-type probe, although more sensitive, causes more distortion--which becomes objectionable if faithful reproduction of the modulation envelope is important. The shunt-type demodulator probe is less sensitive, but provides a somewhat greater rejection of hum and higher fidelity of demodulation. The voltage-doubler probe has a higher sensitivity, but its frequency response is more limited.

The ambitious person who wants to build his own probe will do well to pay close attention to the following. The probe should be carefully shielded. If a balanced probe, it must be balanced both mechanically and electrically. Connections within the probe should be short, and components must be suitable for high frequencies. The resistors should be small, noninductive carbon or carbon-film; and capacitors must be disc ceramic or small mica. Furthermore, two or more diodes used together, especially in a balanced arrangement, should have matching forward and back resistances. Keep signal and ground leads short to minimize the inductive effect. (See Fig. 5-16.)

A tuned circuit can be used with a demodulator probe-either a coil with a capacitor across it, or one resonating with its own distributed capacitance. The tuned circuit is connected between the tip and the ground connection of the probe, and is then placed near the circuit whose waveform we wish to observe. The resonant circuit of the probe should be tuned to the same frequency as the circuit under test. The advantage of such an arrangement is that the coupling to the tuned circuit can be made quite loose. Thus, the circuit under test will not be loaded and thus disturbed as much as it would if a direct connection were made with a demodulator probe . For this purpose, we can use an i-f coil if we wish to make observations in the i-f circuit, or a video-peaking coil in a video circuit. Minimum additional capacitance is almost always desirable because the distributed capacitance of a probe will add enough to assure resonance at the desired frequency.

Sometimes we may wish to pick up a signal by placing a floating shield over the tube and connecting our demodulator probe to it. We will indeed pick up a signal-plus extraneous signals which may be picked up from the horizontal- and vertical-sweep circuits. In many instances, the interfering signals will override the one which is under test. To overcome this difficulty, therefore, it is best to disable the sweep circuits.

In r-f measurements, keep the ground lead very short, and make the ground connection as close as possible to the ground return of the point where the signal is taken off. This fact cannot be over emphasized. More often than not, beginners mistakenly use just any point as a ground, and then simply move the probe clip around. More than anything else, the ground connection may, at frequencies above 100 megacycles, determine the correctness of our response curve.

The length and position of the ground lead are also important. The ground return lead is part of the r-f carrying circuit and, as such, the demodulator circuit. In a probe used for measurements up to the video range, the position of the ground clip may not be overly critical.

On the other hand, with television carrier frequencies, it must be short and direct . If the ground return is too far from the probe, deceiving displays will be obtained because of grounDCurrent loops in the chassis. A simple DC ground connection made to arbitrary point somewhere on the chassis, is not always satisfactory.

Because of its loading effect, the demodulator probe should always be applied to low-impedance points. If we have no such point and do not want to swamp the circuit, we can insert a resistor of several ohms in the cathode circuit of a tube . The resistor will often develop across the tube sufficient signal to give a usable display on our oscilloscope, and yet not load the circuit if picked up by the demodulator probe.

The blocking capacitor, of course, will eliminate any DC voltage developed from this connection.

A detailed treatise on the use of demodulator probes in servicing television receivers is beyond the scope of this guide. We will there

fore touch only on some of the highlights, to whet the reader's interest.

The response of a video amplifier can be checked in two ways. Both require that a swept video signal be applied at the video-amplifier input. The output signal can be applied directly to the vertical-input circuit of a wide-band oscilloscope, or it can be demodulated with a demodulator probe and its output displayed on an oscilloscope.

A much simpler way to display the video-amplifier characteristics on our oscilloscope is to use a demodulator probe. What we do is apply the video-amplifier output to the demodulator probe, and then observe the modulation envelope on our oscilloscope. However, the oscilloscope must definitely be able to display a 60-hz square wave with no distortion, because the sweep generator usually operates at a 60-hz sweep rate. If the video-amplifier characteristics were ideal, amplification would be equal for frequencies up to 4.5 megacycles . As a result, the response curve would increase sharply at the low-frequency end, stay flat all the way up to the highest frequency, and then drop off sharply. This, of course, is a square wave ; and since each sweep is completed in 1/60 of a second, a 60- hz square wave is thus displayed. When a demodulator probe is used, the picture-tube socket should be removed from the picture tube and the probe inserted at the point where the video signal is obtained, because video circuits are critical as far as shunt capacitance is concerned. What we have done is remove the effective capacitance of the picture tube and substitute an equal amount of capacitance presented by our demodulator probe.

The swept signal, if used directly, should be connected to the oscilloscope through a low-capacitance probe. We apply this signal to the video-amplifier input, and apply the output from our demodulator probe to the vertical-input circuit of the oscilloscope. A marker signal should also be applied so the frequency characteristics of the video amplifier can be determined. The value of the series filter resistor in the demodulator probe now becomes of great importance. The time constant of the resistor and the filter capacitance of the probe must be long enough to give the probe a good response for a signal as low as 60-hz square wave. In this way, the display near zero frequency will be correct, not fuzzy, because of the inability of the probe to respond to those low frequencies.

Video amplifiers can also be tested with square waves at several frequencies. The demodulator probe is not used. Instead, we apply the output from our video amplifier through a low-capacitance probe, to the vertical-input circuit of a wide-band oscilloscope.

Fig. 5-17. Equipment setup for testing flatness and linearity of sweep-generator output.

We can test our probe by using the output from a 60-cycle squarewave generator to modulate an r-f signal and then apply the modulated r-f signal to our demodulator. The output of the probe should again be a 60- hz square wave, and the fidelity of the square wave will be a direct indication of the demodulation capabilities of our probe.

Fig. 5-18. How to compare the frequency of one r-f signal with that of another.

Fig. 5-19. How to check the characteristic impedance of a balanced line.

Fig. 5-20. How to check the 60- hz sync buzz voltage in a receiver.

Demodulator-probe characteristics at low frequencies (in the audio frequency range) are of interest. With most probes, if we apply the signals of various frequencies-starting from perhaps 20 to 30 cycles or so and going up to several hundred thousand cycles--we will experience feed-through over a certain broad frequency range . This will manifest itself as an output voltage equal to the input voltage, without being modified in any way. This feed through characteristic SCOPE TRANSFORMER will drop off at high and at low frequencies for two reasons. (1) At the high-frequency end, the filter characteristics of our r-f network will attenuate ; and (2) at the low-frequency end, the reactance of the input capacitor will become very high. This behavior, which is normal for demodulator probes, should be remembered.

Be extremely careful not to overload the probe and thus damage the semiconductor diode. Never apply the probe to the horizontal- or vertical-deflection circuits, because even a momentary contact will immediately burn out the diode.

Semiconductor diodes are rather sensitive to heat. So, if one must be replaced, hold the lead with a pair of pliers, which will serve as a heat sink to conduct the heat from the soldering iron or gun away from the diode. Also, it is advisable, when using a probe in a television receiver, not to lean it against hot vacuum tubes or resistors because enough heat may be conducted through it to ruin the diode.

The accuracy and linearity of the output from a sweep generator can be tested with a demodulator probe and oscilloscope, as shown in Fig. 5-17 . Connect the demodulator probe to the properly terminated output from the sweep-generator cable, and the output from the demodulator probe to the vertical-input terminals of the oscilloscope. Couple in a marker signal through a small coupling capacitor. (One or more markers can be used.)

(A) Low impedance. (B) High impedance.

Fig. 5-21. Demodulator circuits.

A demodulator probe can also be used to check the frequency of one r-f signal against another, as shown in Fig. 5-18 . The r-f signals from the two sources are fed to the input circuit (one end of the 270-pf capacitor), and the output signal from the demodulator is fed to the vertical-input terminals of the oscilloscope. Then the frequency of one of the signals is varied. As we approach and go through the same frequency as the other signal, we will observe a zero-beat pattern on the oscilloscope screen.

Fig. 5-22. How to check the frequency demodulation limit of a probe.

In order to check the characteristic impedance of a line, we connect one end to a sweep generator with the appropriate center frequency . At the other end we connect a balanced probe, as shown in Fig. 5-19 . The load resistor should be equal to the characteristic impedance of the probe; so, we use two 150-ohm resistors in series to give us the required 300 ohms . The center connection between the cable and resistors is grounded to provide a DC return for the probe . If the characteristic impedance of the line is equal to 300 ohms, the display on our oscilloscope will be a straight line . If other than 300 ohms, the display will be curved, the amount being directly proportional to the degree of mismatch. At least twenty feet of transmission line must be used in order for a satisfactory indication to be displayed on the oscilloscope. If the line is shorter, the standing waves may not be strong enough to develop a satisfactory indication.

We can apply a voltage-doubler probe directly to the 4.5-megacycle sound circuit in order to measure the percentage of downward modulation due to 60- hz buzz voltage . For this measurement, we will require a DC oscilloscope. A pickup loop like the one in Fig. 5-20 can be used to check for sync buzz in intermediate frequency and video amplifiers.

Voltage-doubler probes can also be used for making tests in the chroma bandpass amplifier and other video-frequency circuits of color television receivers. The probe should be applied across a low-impedance point, such as the color-intensity control, in order that the probe shunt capacitance will disturb the circuit as little as possible. It should not be connected across the filter coils because shunt capacitance will change the bypass characteristics of the circuit.

In order not to present a noticeable load to the circuit under test, the probe should have an impedance at least ten times as high as the impedance at the circuit point under consideration. For certain alignment procedures, the oscilloscope must be connected to the output of the video-i-f stages-preferably through a low-impedance detector like the one in Fig. 5-2 1A. On the other hand, a high-impedance detector is more desirable for aligning chroma and sound-i-f stages. One such detector is shown in Fig. 5-21B.

Fig. 5-23. Waveshape having two different values of peak voltage.

The low-frequency limit of the demodulating capability of a de-modulator probe can be checked as shown in Fig. 5-22. The output from a video-frequency sweep generator is fed through a marker box to the demodulator probe, the output of which is connected to the input terminals of an oscilloscope . An absorption-type marker is preferred because it does not give confusing beats . As the lower frequency limit of the probe is approached, the probe output falls off and shows evidence of incomplete rectification and filtering. The lower frequency limit can be determined by adjusting the marker to this point.

The peak-to-peak reading with a vtvm rectifier probe will give an indication of the peak-to-peak value of the signal, but will ignore anything between the peaks. Let us look for the moment at the waveshape in Fig. 5-23 . The smaller signal between the peaks might just as well not be there, as far as the vacuum-tube voltmeter is concerned. Its presence or absence will not be indicated on the meter. On the oscilloscope, however, it is clearly shown.

A demodulator probe can be converted to a peak-to-peak or peak reading rectifying probe by adding a relatively large capacitor (0.01 mfd) across its output circuit. The capacitor can be added externally and the voltage across it applied to the input terminal of the vacuum tube voltmeter. The DC voltage indication will then be proportional to the peak value of the voltage under test.