@/bbroadcast sound A. M. Modulation Monitor @ The audio which can be recovered in the home receiver is highly depen- dent upon the modulation that is tak- ing place in the transmitter. It is im- portant, therefore, that this modulation process be continuously monitored at the broadcasting station with an FCC approved modulation monitor. The monitor, essentially an operational test instrument that provides accurate data on the modulation taking place, per- mits adjustments to modulation levels during programming. The monitor contains a number of important func- tions and circuits, but we only have space to discuss a few of the most im- portant ones, THE SIGNAL When modulation is taking place in the transmitter, the output signal is a modulated rf wave that contains three components: the rf carrier, and two equal sidebands that are above and below the carrier. This is a full, dou- ble sideband signal, Each one of the sidebands will carry the entire pro- gram modulation intelligence. But even though the complete information is contained in each sideband and tech- nically we can recover the full pro- gram audio from either one, because of the demodulation process. both side- bands should be identical in every re- spect or the recovered audio will suf- fer in some manner. The sidebands are not audio signals—they are rf wave components that vary at the audio modulation rate. DEMODULATION Demodulation is the reverse process of modulation, The demodulator re- moves the modulation intelligence from the modulator rf wave and con- verts this back into the original audio signal, All other functions and meter- ing circuits work with the audio signal that appears at the output of the de- modulator, so consequently the de- modulator is the most important cir- cuit in the monitor. The accuracy of its performance will determine the accu- racy of the entire monitor. Even if every other circuit and function of the monitor performed perfectly, none could yield any more accurate data than the accuracy of the signal pre- sented to them. THE DETECTOR The circuit most commonly used for demodulation in an a.m. monitor is the diode detector. The diode does have some shortcomings, but when the circuit is well designed and oper- ated within its parameters, it will pro- duce the necessary accuracy in results. The diode, its filter and load make up the detector circuit. The diode re- moves the sidebands from the modu- lated wave by rectification, the filter Temoves the rf components from the Tectified carrier, and the output voltage is developed across the load resistor. This is a peak detector whose output is a dic, voltage that varies at the audio modulation rate. The variation in amplitude of this d.c. voltage is a replica of the original audio signal. The filter design is important to the detector’s operation. The value se- lected for the capacitor must be such that it will remove only the rf com- ponent, leaving the audio envelope component intact. If the value is too large, it will remove the higher audio Figure 1. The basic monitor circuits. % METER OC = MoDULATION eet} RECT AMP weTER oa PL} Lfoete aupiof = -___ | TRIGGER | oe CIRCUIT TO FLASHER 4 FLASHER clnatr CONTROL METER Auoio foo AMP MONITOR Hi-Z {______- pisTORTION METER; 4 : t : : ) > ? broadcast sound (cont.) + oe = i Allah a: a L WwooULATEO AF Lie oeTecToRS iH aypio Le ° oa core Figure 2. Detection process. (Rf wavetorms are shown exaggerated, frequency components and produce a poor audio bandpass for the detector. And if the value is made very large, then the circuit becomes a d.c. power supply instead of a detector. The load must be high in resistance value or ‘it will affect the detector’s efficiency and make it work at somewhat less than a peak detector. DETECTION Two separate detector circuits are used in monitors to provide two inde- pendent output signals. These detec- tors are placed in opposite polarity to the rf wave so that one detector will rectify the positive going half of the wave, and the other detector will rec- tify the negative going half of the wave. Each of these detectors is re- covering the same modulation informa- tion but the separate outputs are out of phase with each other. The output of each detector can be selected by a front panel (+) or (—) peak modu- lation switch. Whichever detector is selected by this switch will then feed all the other circuits of the monitor. In a.m. modulation, it is important that the positive and negative audio modulation peaks are observed and measured, so this is why they are made selectable. The audio peaks could be split by a phase splitter but the com- mon practice is the use of two detec- tors, This makes things less compli- cated, gives a better review of the modulated wave, and makes the fol- lowing circuits more simple in design. The output of the detector which rectifies the positive half of the rf wave is a varying d.c, voltage that is posi- tive with respect to ground. When the d.c. component is removed by a cou- pling capacitor, the resulting audio sig- nal positive peak is in phase with the original audio—that is, both positive ENVELOPE at RECTIFIED, s, + RE PULSES. ENVELOPE Pree AUDIO PEAK AUDIO PEAK aC audio peaks are in the positive direc- tion. The output of the detector which rectifies the negative half of the rf wave is a varying dic. voltage that is negative with respect to ground, and the recovered audio signal is out of phase with the original audio. There- fore, the original positive audio peak is now going negative while the origi- nal negative peak is now in a positive direction. As far as the monitor cir- cuits are concerned, regardless of which audio peak is selected for ob- servation, they see only positive going voltages at their inputs, SENSITIVITY The diode detector is somewhat in- sensitive and requires a relatively large input signal or it will become non- linear and produce a distorted output. This lack of sensitivity is no real prob- lem since the monitor is designed to couple directly to the transmitter out- put circuit by coaxial cable. In this arrangement, however, there is a dan- ger that too much signal may be ap- plied to the diode. This would over- load the circuit and produce a dis- torted output, and it can damage the diode and circuit components. The semi-conductor diode does not produce a “clean” baseline (a.c. signal zero axis line). Some reverse current will flow on the reverse half cycle of the a.c. signal and this will allow part of that half-cycle to appear in the out- put. The forward conduction curve of the diode exhibits a bend at the portion where the conduction begins. When the input a.c. signal is too small in amplitude, then operation is in this area of the curve and the output will be very non-linear. The input signal must be large enough so that the op- eration is on the straight line portion of the forward conduction curve. Somebroadcast sound (cont.) )DWWYW tr F-ABRHZ Fe FISKE DETECTOR: LOWER UPPER SIDEBAND SIDEBAND. © 15 KHz IDEAL AUDIO. RESPONSE CURVE MISSING MISSING voweR wd 608 “18k Fe DETECTOR Vea SIDEBANDS CLIPPED, O 3k !OkHz AUDIO RESPONSE CURVE Figure 3. Non-symmetrical sidebands affect the amplitude response curve of the recovered audio. monitors forward bias these detectors almost to the point .of conduction in order to improve their linearity; But even with this design technique, the input level must be carefully adjusted. SELECTIVITY The diode is a non-selective device that will rectify any a.c. signal that appears at its input with enough ampli- tude to make the diode conduct. The diode cannot distinguish the various components of the modulated wave, but will rectify the instantaneous volt- ages of all the components present at its input at any instant. In so doing. the diode is recovering both sidebands at the same time and adds them to- gether in its output (in phase). It is be- cause of this detector action that both sidebands should be identical in every respect, When they are not, when they are added together in the detector out- put, the audio will suffer according to the differences of the sidebands. For example, assume that sideband slipping has removed most of the lower sideband because of faulty tun- ing. If we used a sweep signal to modu- late the transmitter, the waveform of the sweep envelope at the output of the detector would appear similar to the FCC's Ideal Detector Curve for tv. picture transmission. In the band- pass area where both sidebands are present and equal in amplitude, these are added together in the detector out- put, But in the area where the lower sideband is missing, only the voltage of the upper sideband is present, so the voltage amplitude is down one-half, or 6 dB. Sideband clipping in this case has produced a very poor audio band- pass for the system. (The t.v. receiver, by the way, corrects for the unequal sideband transmission by having its IF response curve sloped.) CARRIER LEVEL The average value of the carrier does not change during modulation, If any change does occur, this is termed carrier shift. FCC rules require that any carrier shift be kept within a 5 per cent tolerance, A number of conditions can cause carrier shift in the transmit- ter, but a measurement of this factor is also a measure of transmitter per- formance under modulation conditions. The output of the detector is a dic. voltage that is derived from the rf carrier, so the current flowing in the Fig. 4. The modulation meter circuit is a high impedance fet meter with special AUDIO FROM DETECTORS METER SPECIAL” FILTER «= RECTIFIER % MODULATION rae METER IN| oc AMPLIFIERbroadcast sound (cont.) detector output circuit is measured with a front panel meter which is labeled Carrier Level or similar ter- minology. Metering the d.c. current of the de- tector in this manner serves two use- ful purposes. Since this meter measures the output of the detector, then it can be used to set the rf input to the de- tector properly. And since any change in the average carrier value will change the average current in the detector out- put, a proportional amount, we can read the carrier shift directly from this meter. MODULATION METERING The modulation percentage caused by either the positive or negative audio peaks is measured with a high impe- dance voltmeter circuit that is the equivalent of a vtvm or fet meter. This is a quasi-peak indicating circuit and it must contain special damping and meter ballistics, according to the FCC rules. The positive or negative modulation peaks selected by the switch will feed the audio signal to a meter rectifier circuit. The output of this rectifier cir- cuit is a varying d.c, voltage that now has the special damping, and this volt- age is amplified by a d.c. amplifier which drives the front panel modula- tion meter. The r-c filter in the recti- fier circuit provides the special damp- ing characteristics. The meter face has two scales: one in percentage for mod- ulation of program, the other in dB for tone modulation and testing. PEAK FLASHER Since the modulation meter cannot follow peaks because of its character- istics, a true peak indicating circuit is provided. This is an electronic circuit, such as a Schmitt Trigger or similar circuit, which has its threshold sensi- tivity preset by the front panel flasher control. When an audio peak exceeds this threshold, the trigger will fire and turn on a lamp or other visual device to warn that a peak has exceeded its setting. This circuit will catch many peaks that do not show up on the modulation meter. OTHERS The monitor will provide an audio output for aural monitoring and for making distortion measurements. Indi- vidual models also provide many other features. But regardless of the features, the monitor should be kept in calibra- tion, good repair and operated prop- erly. Next month we will discuss some of the problems that can occur in mon- itoring.