CHAPTER 10

image

Amplitude Modulation Signals and Circuits

In this chapter, I will try to keep the math to a minimum, but it is inevitable that it will creep in to describe exactly what’s going on with amplitude-modulated (AM) signals. This chapter will cover some basics of amplitude modulation (AM). Actually, there are many types of amplitude modulated signals. In the standard broadcast or medium wave band from about 530 kHz to about 1.7 MHz, a carrier signal within that band is varied in strength. This variation in carrier strength makes the resulting signal an amplitude-modulated signal.

Other forms of amplitude modulation can be found in cell phones and digital television transmission. In cell phones, a pair of amplitude modulated signals is used in presenting the voice or data information in a signal that possesses both amplitude modulation and phase modulation. In digital television signals (DTV), there are two types. One is similar to standard broadcast AM signals (vestigial sideband AM), and the other is very much the same as cell phone amplitude-modulated (QAM or quadrature amplitude modulation) signals.

This chapter will cover various AM signals and some of the circuits used in AM radios, such as mixers, amplifiers, and AM demodulators. Instead of separately describing single circuits such as oscillators, mixers, or demodulators, they will be explained in the context of an actual radio system. Again, it is recommended that an oscilloscope and an RF (radio-frequency) signal generator (or a function generator with AM capability) be at hand for the projects and experiments.


AM Signals—What Are They? “Hey, Can You Please Turn Down the Volume … Now Can You Turn It Back Up?”

In the case of standard AM radio signals, the best analogy is just volume variation. When we turn up the volume or loudness of music on our radio or stereo, we are in fact modulating the amplitude of the music. And when we turn it down, we are again modulating the music in terms of reducing the amplitude of the music signal (see Figures 10-1 and 10-2).

image


FIGURE 10-1 A music signal turned to low volume.

image


FIGURE 10-2 A music signal turned up in volume.

For standard AM signals, we “replace” the music with a single tone. Although we can use a volume control to vary the intensity of the single tone, that is limiting in the sense that it’s hard to turn a knob that varies the resistance of the potentiometer fast enough to mimic voice or music. The limits of the volume control allow us to turn the music completely off or to a very loud volume, which depends on the amplifier’s output voltage rating.

In the earlier days of telephone and radio, carbon microphones served the purpose of a variable resistor that could vary the intensity of an electric signal for voice and music. Later on, it was found that a voltage or current controlled amplifier yielded the same results in terms of varying the signal strength (see Figure 10-3).

image


FIGURE 10-3 A block diagram of methods to modulate a signal via a carbon microphone and voltage-controlled amplifier.

In the top diagram in Figure 10-3, the carbon microphone acts as a variable resistor in series with the antenna. Depending on the sound pressure (e.g., via voice or music) to the carbon microphone, the RF signal is varied. In the bottom diagram, the carbon microphone forms a varying direct-current (DC) signal with battery BT1 into the primary winding of T1. The secondary winding of T1 provides an alternating-current (AC) signal to vary the grid bias voltage to a power oscillator to vary the oscillator’s output signal. Figure 10-4 illustrates what an AM signal looks like for a single high-frequency carrier tone and a low-frequency modulating tone.

image


FIGURE 10-4 (a) An unmodulated AM signal or an AM signal at 0% modulation. (b) An AM signal at 50% modulation. (c) An AM signal at 100% modulation.

Zero percent modulation of an AM signal represents the carrier amplitude when silence is being transmitted. Normally, when we tune an AM station between music or voice passages, we can still sense that the carrier is there via a drop in the hiss level. When there is no station on the air or the radio is tuned to a frequency where no RF is being transmitted, usually hiss, or random noise, is heard.

At 50 percent modulation, the peak carrier level is one and a half times the unmodulated carrier level, and its minimum is 50 percent of the unmodulated carrier amplitude. At 100 percent modulation of the carrier, the carrier level peaks at twice the carrier level, and the minimum point is where the carrier is “pinched” off.

Note in Figure 10-4 that the average carrier level of all three examples is the same regardless of the modulation. This fact becomes important later when it is applied as a basis for automatic volume control (AVC) for AM radios. The AVC system in standard AM radios measures the average carrier level to even out the loudness of received strong and weak signals.

A general formula for broadcast AM signals is

image

image

For example, radio station KCBS in the San Francisco Bay Area has a frequency of 740,000 Hz = fc. And the modulating signal m(t) is generally voice signals such that the modulation of the AM transmitter does not exceed 100 percent in the negative peaks.

There is a little quirk in United States AM radio broadcasting, though. Because music or voice is not a symmetrical waveform, the Federal Communications Commission (FCC) in the United States allows the positive peaks to go as high as 125 percent modulation.

In a specific case of a sinusoidal modulation signal, the AM signal can be described as:

image

where the modulation index is m, such as the preceding examples, where m = 100%, 50%, and 0%. The modulation frequency is = fmod. For example, if a radio station at fc = 1 MHz is modulating at m = 75% at a modulating frequency of 440 Hz= fmod, the AM signal would be expressed as:

AMDSBC(t) = [1 + 0.75(cos(2π(440Hz)t)] cos[2π(1MHz)t]

If you will notice, producing a standard AM signal is akin to multiplication of two signals, [1 + m(t)] and sin(2πfct). In the first signal, [1 + m(t)], the added DC offset signal “1” is essential to ensure that the carrier signal never goes lower than 0. This is important because if the carrier signal is multiplied by a negative number, the phase of the carrier flips 180 degrees. [That is, the carrier signal will flip in polarity since m(t) is a signal that is AC in nature and has positive and negative values.]

The reason for calling this signal AMDSBC(t), a double sideband carrier signal, is because this signal contains three signals: a lower sideband (LSB), a carrier, and an upper sideband signal (USB). When a spectrum analyzer is connected to an AM signal, the three signals appear (see Figure 10-5).

image


FIGURE 10-5 A spectrum of an AM signal modulated at 100%, (from left to right) LSB, carrier, and USB.

Figure 10-5 shows for convenience the spectrum of the modulating signal, which does not appear at the output of the AM RF signal. The only signals at the output are the LSB, carrier, and USB.

The frequency spacing is equal from the LSB and carrier and the carrier to the USB. For example, taking the preceding example for:

AMDSBC(t) = [1 + 0.75(cos(2π(440Hz)t)] [cos(2π(1MHz)t]

the frequencies for the LSB, carrier, and USB are, respectively:

1 MHz – 440 Hz

1 MHz

1 MHz + 440 Hz

NOTE To transmit a 440-Hz signal for standard AM, the spectrum occupied is (1MHz + 440 Hz) – (1MHz – 440 Hz) = 880 Hz, which is twice the bandwidth of the modulating signal at 440 Hz.

Thus, the sideband signals are comprised of difference and sum frequencies of the carrier frequency. The more common term is that the frequencies of the upper and lower sideband signals are the sum and difference frequencies from the carrier frequency. Thus, multiplication of two signals having two different frequencies f1 and f2 results in two signals that have sum and difference frequencies of (f1 + f2) and (f1 – f2).

A major advantage of transmitting a standard AM signal to a radio is that the demodulator or detector to recover the modulated signal m(t) is typically a simple rectifier or nonlinear amplifier circuit. Initially, standard AM signals were transmitted in this manner to simplify and economize the manufacture of radio receivers. As we will see later, there are more “efficient” types of AM signals that also require more sophisticated signal demodulators to recover m(t), the modulating signal.


Other Types of AM Signals

Before we leave AM signals that include the carrier signals, there is one other class of standard AM signals with a slight modification. This type of signal is known as the vestigial sideband (VSB) AM signal, and it was and commonly is used in transmitting analog television signals. It is really a standard AM signal that has either the lower or the upper sideband filtered off, but not completely, thus leaving a vestige of that sideband (see Figure 10-6).

image


FIGURE 10-6 Spectra of a standard AM signal and a vestigial sideband AM signal.

In a standard AM signal, the upper and lower sidebands occupy a spectrum of twice the modulation signal. To save on spectrum space, a portion of the upper or lower sidebands is chopped off. This portion relates to the higher-frequency components of the modulating signal. The lower-frequency components of the modulating signal are not affected sideband-wise, and thus there is a standard double sideband carrier AM signal at the lower modulation frequencies in a VSB signal. The lower-frequency signals in a sense bias the AM carrier to a minimum level such that the negative modulation never goes below 100 percent.

One may ask how can we always guarantee that there is a low-frequency signal to bias the AM envelope? The answer is that a video signal always has synchronizing signals that are low frequency in nature. Even when a TV program or show fades to black, the synchronizing signals are still being transmitted.

The question is, then, what other feature do we have in VSB? The answer is that the VSB signal is also compatible with a simple detector circuit such as a half-wave rectifier that is commonly used in demodulating standard AM signals. This makes the receiver circuits simpler and more cost-effective. To illustrate how a TV signal is modulated with VSB for a fade-to-black signal, see Figure 10-7, which is an oscilloscope trace of an analog cable TV signal.

image


FIGURE 10-7 A “black level” television signal (top). The resulting VSB RF modulated signal (bottom).

The top trace in Figure 10-7 shows an inverted video signal, which is normal for modulating an RF carrier for the 525-line (480i) TV analog system. The VSB RF signal on the bottom trace is then detected or demodulated with a simple diode detector. When analog television transmission became commercialized in the 1940s in the western hemisphere, vacuum tubes were primarily used. There were no inexpensive phase lock loop or synchronous detector circuits. So the VSB signal was invented such that a small signal rectifier (e.g., a 6AL5 vacuum tube or a 1N60 germanium diode) could be used for demodulating the VSB signal back into a baseband video signal. This baseband video signal, when amplified, was generally sent to the TV’s picture tube’s cathode for displaying the television image.


Suppressed Carrier AM Signals

There are quite a few AM signals where the AM signal’s carrier is suppressed. These types of suppressed carrier signals are found everywhere today. For example, they are used in cell phone signals, digital TV signals, single sideband signals, and analog color TV signals, just to name a few.

First, what does a suppressed carrier AM signal mean? Let’s look again at Equation (10-1), which states

image

We know that the carrier signal is always on regardless of m(t), the modulating signal, which can go to zero to have

AMDSBC(t) = [1 + 0 ] cos(2πfct) = cos(2πfct)

where sin(2πfct) is the carrier signal at frequency fc.

An AM signal with suppressed carrier is a signal where there is no carrier at all by eliminating the DC offset “1,” and it is expressed as:

image

Note that the new subscript in Equation (10-3) is DSBSC (double sideband suppressed carrier) instead of DSBC (double sideband carrier) in Equation (10-1). A typical double sideband suppressed carrier signal is illustrated in Figure 10-8.

image


FIGURE 10-8 A double sideband suppressed carrier signal (bottom trace) with sinusoidal modulation (top trace) of the reference carrier signal (middle trace).

In a DSBSC signal, if one looks carefully, during the positive cycle of the modulating signal, the DSBSC signal is in phase with the reference carrier signal. However, during the negative cycle of the modulating signal, the DSBSC signal is 180 degrees out of phase with the reference carrier signal. With a double sideband suppressed carrier signal, we cannot use a simple diode detector to demodulate the signal. A comparison between a standard AM signal and a double sideband suppressed signal when a simple half-wave rectifier is used for demodulation is shown in Figures 10-9 and 10-10.

image


FIGURE 10-9 A standard AM signal with half-wave rectification.

image


FIGURE 10-10 A double sideband suppressed carrier signal that has been half-wave rectified.

From Figure 10-9, we can reproduce the modulating signal fine with a standard AM signal. But with the double sideband suppressed carrier signal as shown in Figure 10-10, we get a demodulated signal that will look like a full-wave rectified version of the modulating signal itself (also see Figure 10-11).

image


FIGURE 10-11 A modulating signal and a full-wave rectified version of it.

To demodulate a double sideband suppressed carrier AM signal, we have to use a product detector (see Figure 10-12). The product detector provides a multiplying function. We multiply the incoming double sideband suppressed carrier signal by another signal of the same frequency and the correct phase φ. In general, the detection of this signal is characterized as:

image


FIGURE 10-12 A product detector to demodulate a double sideband suppressed carrier signal.

AMDET_DSBSC = AMDSBSC(t) × cos(2πfct + φ) = [m(t)]cos(2πfct) × cos(2πfct + φ)

where φ is the phase angle or a phase shift that is different from the phase of the original transmitted carrier signal.

After filtering out the high-frequency signals at 2fc, we are left with the following:

image

Figures 10-13 through 10-15 illustrate demodulation with a product detector for phase angles φ of 0 degrees, 45 degrees, and 90 degrees.

image


FIGURE 10-13 Demodulated output of a DSBSC signal with a 0-degree phase shift leads to the correct and full output from the detector.

image


FIGURE 10-14 Demodulated output of a DSBSC signal with a 45-degree phase shift leads to reduced output at the demodulator.

image


FIGURE 10-15 Demodulated output of a DSBSC signal with a 90-degree phase shift leads to zero output from the detector.

As one can see if the demodulating phase angle is incorrect (at 90 degrees), the detected DSBSC AM signal can drop to zero.

On another note, a “perfect” product detector does not leak through any of the input signals to the output. So when two signals such as an RF signal and an oscillator signal are multiplied via a perfect product detector, the output contains neither the RF signal nor the oscillator signal.

In terms of bandwidth requirements, the double sideband suppressed carrier signal takes up the same amount of spectrum or bandwidth as a standard AM signal—only that the carrier signal is missing. So, for example, if the carrier frequency is 1,000 kHz and the modulating frequency is 5 kHz, the upper and lower sidebands will be 1,005 kHz and 995 kHz. The bandwidth (BW) is just (1,005 kHz – 995 kHz) = 10 kHz = BW. Again, the bandwidth of the DSBSC AM signal is twice the bandwidth of the modulating signal.

It should be noted that in practice, all product detectors or multiplying circuits (even ones that are implemented via digital signal processing) produce some other leak-through and/or distortion signals along with the intended sum and difference frequency signals. Also, the receiver’s demodulation signal cos(2πfct + φ) usually has to be regenerated or recovered somehow from the transmitter. Because the carrier is not sent in the usual manner like standard AM signals, generally the receivers that demodulate suppressed carrier AM signals require more circuitry.

Moving on, what else can we do with a double sideband suppressed carrier signal?


Knowing Your I’s and Q’s

A combination of double sideband suppressed carrier signals can lend itself to transmitting two channels within the same bandwidth. For example, if the modulating signals of each channel have a spectrum of 5 kHz, two independent channels of 5 kHz bandwidth will be provided. The bandwidth of the RF signal is still 10 kHz, and two 5 kHz channels amount to 10 kHz of information.

By taking advantage of the fact that Equation (10-4) shows if the demodulating signal is 90 degrees out of phase from the original transmitting signal, the output is zero, that is:

image

We can then send two signals that occupy the same bandwidth but are 90 degrees out of phase from each other. At first, the combining of the two signals at the transmission end makes them seem inseparable. But at the receiving end, we can use two product detectors with demodulating signals that are 90 degrees out of phase from each other to demodulate two separate channels of information (see Figure 10-16).

image


FIGURE 10-16 A two-channel information transmission and receiving system.

In Figure 10-16, the first channel is arbitrarily defined as the I channel, with the cosine waveform defined as 0 degrees or in phase (I). The second channel, a sine waveform, is then phase shifted 90 degrees from the cosine waveform and is called the quadrature (Q) channel. Quadrature in this case means a “quarter cycle” or 90 degrees shifted from the reference signal, a cosine waveform. V1(t) and V2(t) are usually two separate channels of information to be transmitted. For example, V1(t) and V2(t) can be the red color difference (R – Y) = Pr and the blue color difference (B – Y) = Pb video signals such as the video signals you may notice at the output of a DVD or Blu-ray player, Pr and Pb. Figure 10-17 shows for cosine and sine waves.

image


FIGURE 10-17 Cosine and sine waveforms on the top and bottom, respectively.

The two separate channels are demodulated with the correct phases at 0 and 90 degrees referenced to the original transmitting signal. We also get “demodulation” of the “opposing” channel in each case, but the demodulation signal referenced to the transmitting channel is 90 degrees out of phase, which results in zero output. That is, the I and Q channels transmitted are decoded without cross talk as long as the demodulating signals are perfectly aligned and 90 degrees from each other. If channels are amplitude matched and there is an imperfect 90-degree separation, such as when the I channel is 0 degrees and the Q channel is 89 degrees, then there is an error of 1 degree. And the resulting cross talk or leakage of “contamination” into each channel will be on the order of 1 percent. The amount of leakage for small angles (<10 degrees) is about 1 percent per 1 degree of error.

And now for something related to single sideband signals (SSBs)—we can use a special form of I and Q signal modulation to generate SSB signals. If instead of independent signal sources V1(t) and V2(t) at the input of the transmitter V2(t) is phase shifted 90 degrees from V1(t) and combined with the two mixers, then we get a single sideband signal. In essence, a single sideband signal is a frequency-translated signal. Figure 10-18 shows an input signal Vin(t) that is split out to a 0 degree and a 90 degree signal into two multipliers, I Mult Tx and Q Mult Tx. The carrier signals are Cos (Bt) and Sin (Bt)

image


FIGURE 10-18 A single sideband transmitter and receiver.

When the adder or subtractor is set for addition of the two signals from the multipliers, the result is a lower sideband signal, which frequency translates the input signal’s spectrum but causes an inversion in frequency spectrum as well. For example, if the input signal is from music and has a spectrum of 20 Hz to 20 kHz and we set the adder/subtractor for addition with a 1 MHz RF signal, the resulting spectrum will be a lower sideband signal from (1 MHz – 20 Hz) to (1 MHz – 20 kHz). Alternatively, while using an adder, we can generate an upper sideband signal by flipping the phase by 180 degrees of just one of the carrier signals to either mixer.

If the adder/subtractor is set for subtraction, we instead get an upper sideband signal that is then a straight frequency translation of the input signal’s spectrum. There is thus no spectrum inversion. For example, if we have a voice signal whose spectrum spans 100 Hz to 4 kHz, and if we want to frequency translate that voice spectrum upward by 100 kHz, the resulting spectrum will span from 100.100 kHz to 104 kHz. In this example, we have produced an upper sideband signal at 100 kHz. Alternatively, while using a subtractor, we can generate a lower sideband signal by flipping the phase by 180 degrees of just one of the carrier signals to either mixer.

Demodulating the SSB signal is simpler in that almost any type of mixer circuit can be used. Fortunately, we do not require two mixers in the receiver. That is, a multiplier circuit is not the only circuit that can be used to recover the modulation signal. A simple diode and an oscillator signal with the same frequency as the transmitting frequency are combined and rectified. Another advantage to demodulating an SSB signal is that it is phase independent. The reason is that we are just “beating” or frequency translating the (RF) SSB signal downward back to the baseband signal, such as a voice signal. If the demodulating frequency at the receiver end is incorrect, the pitch of the recovered signal will be shifted high (e.g., “Donald Duck–sounding”) or shifted low (e.g., a deep voice).


Basic Radio Circuits

Radio receivers can be categorized into two types—the ones with an RF mixer (e.g., multiplicative function or multiplier circuit) and those without an RF mixer. The simplest standard broadcast AM radio circuit would include a tuned circuit, variable capacitor, and antenna coil with a detector diode connected to a crystal earphone. This is known as a TRF (tuned radio-frequency) circuit. Figure 10-19 shows variable capacitors.

image


FIGURE 10-19 Air dielectric variable capacitors.

The top air dielectric variable capacitor has four gangs, two for the AM band and two for the FM band. The variable capacitor section with the most densely populated plates will be for the RF section for AM radios. The other AM section is similar but has smaller plate areas or a smaller number of plates. The two FM sections are identical and have three rotor plates (rotor plates are connected to the shaft and usually are grounded or AC grounded) and two stator plates. The middle two-gang variable capacitor is designed just for AM radios. Again, the section with higher capacitance is for the RF section, and the other section is for the oscillator circuit.

The variable capacitor shown at the bottom of Figure 10-19 is a single-gang 365 pF version. This variable capacitor is commonly found on eBay or at electronics supply stores (e.g., www.tubesandmore.com).

For TRF radios, one or more sections (e.g., two sections connected together in parallel) of the four- or two-gang variable capacitors may be used to match the antenna coil or RF inductor. Note that the frame of an air dielectric variable capacitor is normally grounded or AC grounded.

In Figure 10-20, the variable capacitor on the left has two dissimilar sections and is generally for superheterodyne radios. The middle lead is the ground or common lead. The top lead is for the RF section and is usually connected to an antenna coil, and the bottom lead is a smaller-capacity variable capacitor that is used for the oscillator section of the radio and is connected to an oscillator coil.

image


FIGURE 10-20 Polyvaricon variable capacitors.

The polyvaricon variable capacitor on the right is normally used for AM and SW (shortwave) superheterodyne radios. One section is used with the antenna coil for the RF filter, and the other is used with the oscillator coil. For AM bands, there is usually a series capacitor with one of the twin sections of the polyvaricon variable capacitor. The series capacitor is usually about 110 percent of the maximum capacitance of a twin variable capacitor. For example, if you are using a twin 270 pF variable capacitor, the series capacitor for the oscillator coil is about 300 pF.

Generally, antenna coils are available with ferrite rods or bars, as seen in Figure 10-21. The top and middle antenna coils have separate primary and secondary windings for a total of four leads. The bottom antenna coil has three leads, which is a tapped inductor or autotransformer. The low-side tap has about 5 to 10 turns of wire. A TRF crystal radio circuit is shown in Figure 10-22.

image


FIGURE 10-21 Antenna coils.

image


FIGURE 10-22 A crystal radio circuit.

A crystal radio typically requires a long wire antenna, typically at least 20 feet, strung up above the ground. The antenna should be coupled into the tuned circuit via a small capacitance value, typically about 5 pF, to isolate the low source resistance of the antenna. A low source resistance will load the tuned circuit (L1 and VC1) down and reduce the radio’s ability to separate the various radio stations. For example, if the antenna is connected directly to L1/VC1 by shorting out the 5 pF capacitor C0, then the radio will receive many stations at once, and tuning VC1 will have little effect. With C0 in series with the antenna, the low source resistance of the antenna is buffered or isolated via C0, and VC1 and L1 are essentially allowed to form a narrow band-pass filter that allows tuning to be more selective.

One question arises: why do we need an antenna coil (in the first place) when a small inductor can be used instead? The answer is that most small inductors of the same inductance value have resistive and other types of losses that antenna coils generally do not have. One can replace the antenna coil with an inductor, but even with a 5 pF capacitor connected to the antenna, the user will notice that the tuning is less sharp with the inductor versus the antenna coil.

Also, the antenna coil on its own without the long wire antenna is capable of receiving radio stations better than an inductor. For example, having a 6- or 8-inch ferrite-rod antenna with Litz wire provides very high sensitivity, and without a long wire antenna, usually one or two local stations of strong signal strength can be heard.

To improve on the crystal radio’s sensitivity, one can add an audio amplifier. However, due to the turn-on voltage of the diode detector, the improvement in sensitivity is limited. The audio amplifier must also have a high input resistance (e.g., >100 kΩ) so as to not load down the tank circuit (VC1 and L1).

Instead of a diode, one can use the “fuzz” circuit seen in Chapter 8 to act as an AM power detector. See Figure 10-23. Although the transistor amplifier input resistance is moderately high at 45 μA collector current with about 57 kΩ for a current gain β = 100, use of the one-transistor power detector circuit provides adequate selectivity and sensitivity.

image


FIGURE 10-23 Crystal radio with power detector Q1.

The power detector crystal radio works at a minimum of 1.5 volts and better at higher voltages such as 3 volts. Q1 is really a common emitter or AC-grounded emitter bipolar transistor amplifier. We know that for small signals < 5 mV, Q1 amplifies the signal with low distortion. But when the input signal into the base is >5 mV, Q1 starts to distort. We take advantage of the distortion feature because with a large enough amplitude AM signal (see Figure 10-23) into the base of Q1, Q1 distorts and becomes a demodulator by clipping off almost half of the AM signal. The load resistance to the tank circuit L1/VC1 is about 50 kΩ for a current gain of 100 = β. A higher-current-gain transistor such as a 2N5089 (β = 800) may be used for better selectivity and sensitivity by raising the load resistance to about 200 kΩ. To achieve still better selectivity and sensitivity (see Figure 10-24). An emitter-follower Q0 is added to raise the load resistance to >500 kΩ using a transistor such as 2N3906.

image


FIGURE 10-24 Higher-selectivity version with Q1 and emitter-follower Q0.

At some point, amplifying the detected audio signal leads to diminishing returns. The reason is that in order for the power detector to work, we have to drive the base-emitter junction of Q1 into distortion. If the RF signal from the tank circuit VC1 and L1 is too small, Q1 merely acts as an amplifier, and there is no detection of the AM signal. But maybe we should amplify the RF signal and then demodulate the AM signal afterwards (see Figure 10-25).

image


FIGURE 10-25 Converting the power detector into an RF amplifier and adding a power detector afterwards.

Figure 10-25 keeps Q0 as an emitter follower but makes Q1 into an RF amplifier by removing the 0.0033 μF capacitor from the collector to ground, as shown in Figure 10-20. The amplified RF signal via Q1 and R3 is then coupled to power detector circuit Q2, R6, and C7.

An advantage of power detectors is that the demodulated output signal is usually higher than a diode detector for the same input signal. Thus the power detector demodulates and amplifies the demodulated (e.g., audio) signal.


Tuned Radio-Frequency Radio Projects

So far we have described TRF (tuned radio-frequency) receivers and for projects, the following circuits can be built:

     • A low-power TRF radio

         • A TRF radio with an MK484/TA7642 and its improved version using an emitter-follower circuit

A low-power TRF radio using just an AA battery is shown in Figure 10-26.

image


FIGURE 10-26 TRF radio using inductive loading for low-voltage operation.

By using inductors, the signal voltages at the collectors at Q1 and Q2 are allowed to swing above the 1.5 volt supply. The collectors also have a quiescent voltage at 1.5 volts, which allows for reliable biasing of the bases of Q1 and Q2, low-capacitance transistors MPSH10. Note the pin-out for an MPSH10 is base, emitter, collector instead of the emitter, base, collector configuration for a more common transistor such as the 2N3904. Diodes D1 and D2 form a voltage reference for the base biasing. The voltage at the D1 anode maintains a voltage of about 1 volt and thus keeps a reliable voltage at the bases even when the battery voltage drops from 1.5 volts to 1.2 volts. Q1 and Q2 are RF amplifiers, and detection of the AM signal is done via D3. A germanium diode is chosen for AM detection because it has very low turn-on voltage (~0.1 volt) compared with a silicon diode (~0.5 volt).

When AM radios of this type are built, try to keep the wires short (within a couple of inches). See Figure 10-27. The prototype here is built to mimic the flow of the schematic diagram. The reader can build this circuit or others in this book, although they are not designated as official projects. An intermediate to advance skill level of electronics may be required.

image


FIGURE 10-27 TRF radio using inductive loading.


First Project: A TRF Radio

The MK484/TA7642 integrated circuit (IC) that has a multiple-stage amplifier and a peak detector is generally available on the web, including eBay. We present multiple versions of this radio, as shown in Figures 10-28 through 10-31.

image


FIGURE 10-28 MK484/TA7642 TRF radio, version 1.

image


FIGURE 10-29 MK484/TA7642 TRF radio, version 1.

image


FIGURE 10-30 MK484/TA7642 TRF radio, version 2.

image


FIGURE 10-31 MK484/TA7642 TRF radio, version 3, with the emitter follower Q1.

Parts List

     • C1, C3, and C4, 1 μF electrolytic capacitor (10 to 50 volt rating); C4 is optional

     • C2, 0.1 μF film or ceramic capacitor

     • R1, 100 kΩ resistor, any type available

     • R2 and R3, 1 kΩ resistor, any type available

     • L1, 240 μH or 250 μH antenna coil

     • VC1, 330 pF to 365 pF variable capacitor

     • U1, MK484 or TA7642 three-terminal IC (integrated circuit)

     • 1.5 volt supply via a battery or equivalent (do not exceed 2 volts on the supply)

The 365 pF variable capacitor was bought on eBay. However, the single-gang 365 pF variable capacitor shown in Figure 10-19 can be used instead. The antenna coil shown is a 250 μH ferrite-rod type, which is available at the time of this writing at www.tubesandmore.com. The TA7642 IC was available on eBay for less than $1.

This radio pulled in about seven stations. The radio can be improved slightly by coupling a long wire antenna to the antenna coil L1 via a 5 pF to 10 pF capacitor, as shown in Figure 10-30. By coupling a long wire antenna, sensitivity is increased.

However, it was found that with some TA7642 chips, the input resistance was lower than expected, which loaded down the LC tank circuit. An emitter-follower amplifier was added, and sensitivity and selectivity were improved, with the radio picking up more than 10 stations (see Figure 10-31). Note that a different antenna coil and variable capacitor combination is presented in this version. But the reader can just as well use 365 pF for VC1 and 250 μH for L1.

Solid-State Regenerative Radio

The TRF radios we have shown so far can be improved further by adding positive feedback to the tank circuit. By doing this, two things are accomplished. One is that if there is a sufficient RF signal at the tank circuit, a portion of an amplified RF signal is fed back to the inductor-capacitor tuned circuit in a manner to reinforce the RF signal, which then provides very high gain. The second thing is that if the amount of positive feedback is adjusted just below the threshold of making the circuit self-oscillate, the selectivity of the tuned circuit is also increased. The positive-feedback signal “cancels” out some of the resistive and other losses in the antenna coil to allow better separation of tuned radio stations (see Figures 10-32 and 10-33).

image


FIGURE 10-32 Schematic of a low-power regenerative radio.

image


FIGURE 10-33 The regenerative AM radio.

VR1 adjusts the amount of positive feedback into the antenna coil via its bottom lead. The tap of the antenna coil is AC grounded via C1 and has a DC bias voltage via D1. Diodes D2 and D3 provide a “regulated” voltage to the bases of Q2, Q3, and Q4.

Q1 is an emitter-follower amplifier that has a gain of about 1 and allows for a high-resistance load to the tank circuit L1 and VC1. It drives VR1 to Q2, which supplies positive-feedback signal current into the antenna coil. Q1’s emitter also supplies the regenerated RF signal to RF amplifier Q3. The output of Q3 supplies sufficient amplitudes of the AM RF signal to power detector circuit Q4. Most regenerative radios require an outdoor antenna. This design allows for a built-in ferrite antenna coil to receive sufficient RF signal levels such that the regeneration control is useful. For example, if there is insufficient signal level into the antenna coil, the regeneration control would have to be turned up to the point where the circuit turns into an oscillator. Recall that a regenerative circuit is essentially the same as an oscillator circuit, but we will be deliberately lowering the feedback or the gain such that the circuit is just below the threshold of oscillation.

Comparing Regenerative and TRF Radios

The regenerative radio requires the operator to readjust the regeneration control (VR1) when tuning to different radio stations. For example, if the amount of regeneration is adjusted just right for one station, then tuning to another station may result in oscillations, and the regeneration control has to be “dialed back.” Although the regenerative radio is capable of very good selectivity and sensitivity, it is inconvenient to use because of the constant readjustment of the regeneration control.

To increase the performance of a TRF radio in terms of sensitivity and selectivity, multiple stages of the tuned circuits can be implemented. However, tracking of the two or more variable tuned circuits is difficult over the entire span of the broadcast AM band.

There were successful designs such as the 1930’s J.W. Miller Model 570 four-gang variable-capacitor TRF radio. There are “obstacles” in getting this multiple-stage TRF design to work. Careful attention is required to avoid oscillation because there will be multiple stages of high-gain amplification that can leak back to the first stage. Another problem is just aligning the two or more RF stages to track throughout the AM band from 535 kHz to 1,600 kHz. It is difficult but doable. However, if there is mistracking, then there will be a loss of sensitivity.

In about 1918, to provide much better performance than TRF or regenerative radios, a new type of radio architecture was design by Edwin Armstrong. This radio is known as the superheterodyne radio, which includes a mixer, oscillator, and one or more fixed-frequency filters.

Mixing and Beating Signals (aka Heterodyning or Superheterodyne)

Let’s take a look at the superheterodyne architecture first for a typical AM radio (see Figure 10-34). As shown in Figure 10-34, the superheterodyne (superhet) radio has an RF filter circuit for the incoming RF signals, a local oscillator, an RF mixer that frequency translates the incoming RF signal frequencies to a lower frequency known as the IF (intermediate frequency), at least one IF filter and amplifier, and finally, an AM detector/demodulator.

image


FIGURE 10-34 Block diagram for a superheterodyne radio.

The basic principle of a superheterodyne radio is to frequency translate the incoming RF signal to a fixed IF signal and apply fixed-frequency band-pass filtering. Amplifying the IF signal with multiple IF filters provides increased sensitivity and selectivity. Whereas a multiple-stage TRF radio has many variable tuned circuits to provide selectivity, the superhet uses multiple fixed-frequency tuned circuits to do the same job. Fixed tuned circuits are easier to adjust, and once they are aligned properly, excellent selectivity is achieved. The IF signal is then detected via a diode or power detector to provide the audio signal.

A typical standard AM radio has a variable tuned RF filter and local oscillator. There are usually two variable capacitors that track each other such that the resonant frequency of the RF signal minus the resonant frequency of the oscillator signal is a fixed frequency such as 455 kHz. For example, if the RF section is tuned to 600 kHz, the oscillator has a frequency of 1,055 kHz, which is 455 kHz above the 600 kHz RF signal. When the RF and oscillator signals are fed to an RF mixer, there will be signals at (1,055 kHz – 600 kHz) = 455 kHz and (1,055 kHz + 600 kHz) = 1,655 kHz. The output of the mixer will be fed to a 455 kHz IF band-pass filter to pass only the 455 kHz signal and not the 1,655 kHz signal.

Alternatively, if one tunes to a station at 1,000 kHz for the RF signal, the oscillator will track accordingly and provide an oscillator signal at 1,455 kHz. Hence, again, the difference-frequency signal, 1,455 kHz – 1,000 kHz, will be 455 kHz. Once the 455 kHz signal is extracted from the mixer and amplified, demodulation of the amplitude-modulated 455 kHz signal is done typically by rectification or power detection.

Let’s now take a look at the various components that are key to making an AM superhet radio.

Two-Gang Variable Capacitors

For the RF and oscillator circuits, a two-gang variable capacitor is normally used. See Figure 10-35. Generally, the two-section (two-gang) variable capacitors are not identical for standard AM radios. The larger-capacity section (e.g., 10 pF to 140 pF polyvaricon) is used for the RF section, whereas the smaller-capacity section (e.g., 10 pF to 60 pF) is used for the local oscillator circuit.

image


FIGURE 10-35 (a) Air dielectric variable capacitors. (b) Polyvaricon variable capacitors.

For AM band and shortwave (SW) radios from 530 kHz to about 30 MHz, twin-gang variable capacitors are generally used. When a twin variable capacitor is used for the AM band, the oscillator circuit has a series capacitor between the oscillator coil and one section of the variable capacitor so that the maxiumum capacitance is cut approximately in half. For example, twin 270 pF polyvaricon types are common among the amateur radio community. To use this for an AM radio, connect a series capacitor whose value is between 300 pF and 330 pF to VC1B of the variable capacitor as shown in Figure 10-36 (right side).

image


FIGURE 10-36 Schematic of a dissimilar variable capacitor and a twin-variable capacitor for AM radios.

Oscillator Coils

For most transistor radios, the traditional oscillator coil is a tapped transformer. The tapping in the winding is not a center tap. Thus there is a low-side tap that is identified by using an ohmmeter. From the center pin to one of the outside terminals will show higher or lower resistance than the other outside terminal. The two terminals that include the center pin with the lower resistance are the low-side tap.

Typical oscillator coil part numbers for AM band transistor radios are 42IF100, 42IF110, and 42IF300. All these coils have five pins, three on one side (including a tapped winding terminal in the middle) and another two on the other side. The transistor radio oscillator coils and IF transformers have a slug tune adjustment that requires a special screwdriver so that the brittle ferrite slug is not damaged while it is turned (see Figures 10-37 and 10-38). Typically, the color code for AM radio oscillator coils is a painted red tuning slug.

image


FIGURE 10-37 Transistor radio oscillator coils.

image


FIGURE 10-38 Alignment/adjustment tools.

IF Transformers

The transistor radio IF transformers are similar to the oscillator coils. However, typically, they have internal capacitors across the outer pins of the tapped winding.

NOTE The schematic drawings in this book will not include the internal capacitor in the IF transformers. When a particular IF transformer part number is specified, unless otherwise noted, assume that there is already an internal capacitor across the outer pins on the side of the coil that has three pins. Also note that oscillator coils do not have internal capacitors because they are generally connected to variable capacitors.

The slugs are color coded for the Xicon IF transformer as:

     • Yellow for the first IF filter, 42IF101 or it may be substituted with 42IF301

         • White for the second IF filter, 42IF102 or it may be substituted with 42IF302

     • Black for the third or last IF filter, 42IF103 or it may be substituted with 42IF303

The first and second IF transformers are interchangeable and can be substituted for each other. However, the third IF filter, 42IF103, should not be replaced with a 42IF101 or 42IF102 part. The primary to secondary turns ratio of the 42IF103 is the correct one to deliver the proper voltage to the AVC (automatic volume control) circuit. If a 42IF101 or 42IF102 IF transformer is used as the last IF filter for the detector and AVC circuit, there will be insufficient AVC voltage, and thus the AVC circuit will not be effective in turning down the gain when a strong signal is received. Like the oscillator coils, the IF transformers also have low-side taps.

For superhet radios similar to the ones in my book, Build Your Own Transistor Radios (McGraw-Hill, 2013), see Figures 10-39 to 10-42 for examples of radios using the oscillator and IF coils.

image


FIGURE 10-39 Four-transistor superhet radio.

image


FIGURE 10-40 Four-transistor radio prototype circuit.

image


FIGURE 10-41 Tuner section of an eight-transistor radio.

image


FIGURE 10-42 Audio amplifier section of the eight-transistor radio.

The radio in Figure 10-39 uses a mixer-oscillator circuit via Q1. The oscillator is a common-base amplifier with the signal being fed back from the collector to the emitter. Using the low-side tap coupled to the emitter of Q1 via C2 allows the oscillator tank circuit from VC1 Osc and T1 to be essentially unloaded. If, for some reason, the circuit does not oscillate, try switching leads 6 and 4 of T1. The base of Q1 is fed to a stepped-down voltage of RF signal via L1’s secondary winding. The RF signal is thus added to the base-emitter voltage of oscillator Q1 to form a combination mixer-oscillator circuit. The oscillator’s frequency varies between about 990 kHz and about 2,050 kHz and typically has at least 200 mV peak to peak at the emitter of Q1. The RF signal at the base is much smaller than the oscillator signal, and thus the oscillator signal drives Q1 into distortion to provide an amplitude-modulation effect on the RF signal, which is also a form of multiplying the RF and oscillator signals. Thus, at the output of Q1, there are signals whose frequencies are related to the RF signal, the oscillator signal, and sum and difference frequencies of the RF and oscillator signals. The IF signal is extracted via T2 and is amplified by IF amplifier Q2. A second IF transformer is connected to the output of Q2 via the collector of Q2. Demodulation of the AM signal is accomplished by half-wave rectifiers D3 and D4. The two diodes are used not only for rectification but also to establish an initial positive AVC voltage at the anode of D4 that supplies a bias voltage (R4 and C4) to IF amplifier Q2. When the signal strength is high, a negative voltage is added to the initial positive AVC voltage, which results in a smaller positive voltage that in turn biases a lower collector current to Q2. This then lowers the gain of Q2 to establish AVC (automatic volume control). The audio signal from D4 is coupled to volume control and audio amplifier Q3. The output of Q3 is fed to interstage audio transformer T4 on the center-tap primary winding (e.g., 10 kΩ center-tapped primary, 600 Ω secondary). From the secondary winding of T4, it is coupled to audio power amplifier Q4. The output collector signal from Q4 is fed to T5 (120 Ω primary and 8 Ω secondary windings). A speaker or low-impedance headphone can be connected to the secondary winding of T5.

We can achieve more sensitivity and selectivity by adding an extra IF transformer and amplifier. Also, we can provide more AVC range by applying the AVC voltage over two stages, a mixer and the first IF amplifier (see Figure 10-41).

In this radio, there are separate oscillator and mixer circuits. The oscillator circuit via Q1 is again a common-base design, as previously described in the four-transistor radio. A separate RF mixer is provided via Q2, which is a common-emitter amplifier that is overdriven with an oscillator signal at >200 mV peak to peak into its emitter. The base of Q2 receives a tapped-down RF signal. The tapping down or stepping down of the RF voltage is essential to keep the RF tank circuit from being loaded down. The output of the mixer is fed to the first IF transformer T2 and first IF amplifier Q3. The output of Q3 is connected to a second IF transformer and IF amplifer T3 and Q4. Demodulation of the AM IF signal (converted RF signals to IF signals) is done via diodes D3 and D4. The audio signal is filtered via R5 and C5 to provide AVC signals to the bases of mixer Q2 and first IF amplifier Q3.

The audio signal level is adjustable via volume control VR1 in Figure 10-42 and is fed to the first audio preamplifier transistor Q5. A second preamplifier transistor Q6 is connected to interstage audio transformer T5 that provides a push-pull or balanced signal to the bases of Q7 and Q8. The outputs of Q7 and Q8 are connected to output transformer T6 to provide an audio signal to a loudspeaker or headphone.

To reiterate, in the four-transistor radio, the oscillator circuit also serves as a mixer. Sometimes this circuit is called a converter circuit because it converts the incoming RF signals to IF signals all in one step. When the RF signal is much smaller than the oscillator signal, it can be added or piggy-backed on top of the oscillator signal. The transistor “sees” the two signals, RF and oscillator, across the base-emitter junction, and because of the nonlinearity of the transistor, RF mixing action is achieved. The IF signal is extracted from the collector of the converter transistor, even though the collector signal has a combination of a strong oscillator signal, the small RF signal, and sum and difference frequency signals from the RF and oscillator signals.

In the eight-transistor radio, Q2 serves as the RF mixer. In both cases, where the oscillator’s signal is >200 mV peak to peak, the conversion transconductance at the intermediate frequency is just the same as the small-signal transconductance of the converter (Q1 in Figure 10-39) or mixer transistor (Q2 in Figure 10-41) based on the collector quiescent current ICQ. The small-signal transconductance of a common-emitter or common-base amplifier is gm = ICQ/0.026 volt.


Second Project: A Superhet Using the MK484/TA7642 Chip

By using a one-transistor converter circuit with an oscillator coil and IF transformers, we can build a superhet radio with the MK484/TA7642 chip. Previously, the chip amplified RF signals in the TRF radios, but this time the chip will only be amplifying the IF signal (see Figures 10-43 and 10-44).

image


FIGURE 10-43 A 1.5 volt superhet radio using an MK484/TA7642 chip.

image


FIGURE 10-44 Picture of the MK484/TA7642 superhet radio.

A mixer-oscillator circuit via Q1 provides about 200 mV peak to peak of oscillation signal into the emitter of Q1. The oscillator voltage amplitude can be increased by lowering the value of R1 (e.g., 1.8 kΩ → 1.5 kΩ or 1.2 kΩ). However, this also raises the conversion transconductance and may overload the MK484/TA7642 IC. The collector output of the mixer-oscillator or converter circuit is coupled to two IF transformers T2 and T3. The two IF transformers are coupled via C4, a 470 pF capacitor, to allow for some “independence” between the two coils so that improved selectivity is achieved and filtering out of the oscillator signal prior to inputting to the MK484/TA8642 chip. If there is too much oscillator signal along with the IF signal, the AM envelope detector in the MK484/TA7642 chip will output distorted demodulated audio signals. The output of T3 is coupled to the input of U1, which then amplifies the IF signal and provides an AVC signal via R3 and C5. Operation of this radio is at 1.5 volts. Diodes D1 and D2 provide a voltage reference so that Q1 provides a stable oscillation signal even if the battery drops in voltage (e.g., from 1.5 volts to 1.2 volts).

Parts List

         • C1, C3, C5, and C7, 0.1 μF film or ceramic capacitors

         • C2, 0.01 μF film or ceramic capacitor

         • C4, 470 pF silver mica or ceramic capacitor

         • C6 and C8, 1 μF, 50 volt electrolytic capacitor

         • D1 and D2, 1N914 or 1N4148 silicon diodes

         • R1, 1.8 kΩ, ¼ watt 5% resistor

         • R2, 10 kΩ, ¼ watt 5% resistor

         • R3, 100 kΩ, ¼ watt 5% resistor

         • R4 and R5, 1 kΩ, 5% ¼ watt resistor

         • Q1, 2N3904 or 2N4124 transistor

         • T1, 42IF100 oscillator coil

         • T2 and T3, 42IF101 455 kHz IF (intermediate frequency) transformer

         • U1, MK484 or TA7642 IC (integrated circuit)

         • L1, 680 μH antenna coil with a 470 μH high-side tap

         • VC1, two-gang variable capacitor, 140 pF and 60 pF


An Observation on Common-Emitter Amplifiers

In this chapter, we see that when a grounded or AC-grounded emitter amplifier is overdriven, it can be used as a power detector for a standard AM signal and also used as an RF mixer in superheterodyne radios. So not only does it work as a “fuzz” guitar audio circuit, but it also has other purposes in radio circuits.

As a mixer with sufficient oscillator signal drive (>200 mV peak to peak), the conversion transconductance = IF signal current/RF input voltage is very high compared with other mixers and is equal to the small-signal transconductance ICQ/0.026 volt.

When we look back at Figure 8-23, we see the effect of a small signal riding on top of a larger signal. The smaller signal is amplitude modulated, which amounts to a multiplying effect. A simple common-emitter amplifier serves well as an RF mixer, but it is not a balanced mixer, and the oscillator and the RF signal also appear at the output along with the IF signal.


References

  1. Ronald Quan, Build Your Own Transistor Radios. New York: McGraw-Hill, 2013.

  2. K. Blair Benson and Jerry Whitaker, Television Engineering Handbook. New York: McGraw-Hill, 1992.

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.84.7.255