Welcome to the UK amateur radio project

The basic limitations of inadequate selectivity and lack of gain of the early receivers led to the development of the 'supersonic heterodyne' principle in around 1920. In the supersonic heterodyne (or 'superhet' as it is colloquially known) receiver, the frequency of all incoming signals is changed to a fixed, fairly low frequency at which most of the gain and the selectivity of the receiver is obtained.

Because this fixed frequency is lower than the signal frequency but higher than the audio frequency, it became known in the early days of the superhet as the 'intermediate frequency' (IF). As the circuits operating at the intermediate frequency, once adjusted, need no further tuning, high amplification and good stability are possible.

In order to convert the signal frequency to the intermediate frequency, a frequency-mixing process is necessary. In the mixer the signal frequency is mixed with the output of an oscillator, the frequency of which is varied by the receiver tuning control. This oscillator is called the 'local oscillator'.

The resulting intermediate frequency is amplified and fed to detector and audio amplifier stages. The output of the IF amplifier is used to provide a voltage, the amplitude of which is proportional to the amplitude of the input signal. This is used to control the gain of the receiver, giving 'automatic gain control' (AGC), to compensate for variation of the received signal.

In order to receive telegraphy (CW) signals, it is necessary to provide a signal to beat with the intermediate frequency to produce a beat note which is audible. This signal is generated by the 'beat frequency oscillator' (BFO) which operates at the same frequency as the IF but is variable about this frequency by about ±3kHz.

Fig 5.1 is a block diagram of the simplest possible superhet receiver. The basic design implications of each stage of a superhet receiver will now be discussed in greater detail.

Mixers

The mixing process of the superhet receiver is shown in Fig 5.2. As is inevitable in the mixing process, two frequencies appear at the mixer output, these being the sum and difference of the signal and local oscillator frequencies. Only one of these is wanted as the intermediate frequency, and in fact only one frequency is accepted by the following IF amplifier. The reason for this will be seen later.

To take a simple numerical example, assume an IF of 500kHz: if the signal frequency is 1000kHz. the frequency of the local oscillator must be 1500kHz; see Fig 5.3(a).

However, a strong signal on a frequency of 2000kHz, that is, a frequency which is twice the IF (2 x 500kHz) away from the first frequency (1000kHz), can also produce an intermediate frequency of 500kHz; see Fig 5.3(b). Hence

Thus two signals, the wanted one on 1000kHz and an unwanted one on 2000kHz, can both result in an intermediate frequency of 500kHz. The unwanted signal (on 2000kHz) is called the 'second channel' or 'image'

This phenomenon is manifest as the reception of two signals apparently on the same frequency and at the same time. The reception of the unwanted signal is known as 'second-channel' (or 'image') interference and it is a 'spurious response' of the receiver in question.

Second-channel interference can only occur when there is a signal on the second channel, which is strong enough to reach the mixer. The most common example is the reception in the 20m amateur band of some of the powerful 19m broadcast stations on a simple all-wave receiver having an IF of 455 to 465kHz (the frequency separation between parts of the 20m amateur and 19m broadcast bands is about 920kHz).

It is most likely to occur when the mixer is inadequately screened and the receiver antenna fed direct to the mixer, so there is no RF stage; in other words there is insufficient selectivity at the signal frequency to reject the second-channel frequency.

As the existence of second channel interference depends on the response of the signal input circuit of the mixer to a frequency which is separated from the resonant frequency of the input circuit by twice the IF, it is clear that increasing the IF will reduce the incidence of image interference.

As a result of the selectivity of the tuned circuits therein, one or two RF stages between the antenna and the mixer will also provide considerable attenuation of the second channel.

Fig 5.4 shows a typical mixer/oscillator arrangement using transistors. Bipolar transistors, FETs and MOSFETs are all suitable for this application, and the typical arrangement shown here is a Colpitts oscillator, the collector supply being stabilised by a 6.8V zener diode. The oscillator output is fed to a buffer stage to provide isolation between the oscillator and the mixer. The mixer uses a dual-gate MOSFET which is particularly suitable for mixer applications, having two gates. The input (RF) tuned circuit is between gate I and earth, and the output (IF) tuned circuit is between the drain and the 12V supply. The oscillator voltage is applied to gate 2.

The receiver local oscillator has the same requirements of frequency stability etc as the VFO in a transmitter. The discussion of VFO stability in Chapter 4 is therefore equally applicable to receiver local oscillators. The situation is complicated by the necessity for the receiver local oscillator to be switched to cover a number of frequency bands.

Generally the local oscillator frequency is on the high side of the signal frequency. The reason for this is as follows.

Assume a receiver tunes to signals in the range 1500kHz to 4500kHz and has an IF of 1000kHz. The signal frequency range has a ratio of 3 to 1 (4500 to 1500) and because

, the change in capacitance must therefore have a ratio of 9 to 1, say, 20pF to l80pF.

If on the low side, the oscillator would have to tune from 500kHz to 3500kHz, a frequency range of 7 to 1, requiring a capacitance range of 49 to 1, say 20pF to 980pF.

Alternatively, if the oscillator is on the high side it would need to tune from 2500kHz to 5500kHz, a frequency range of 2.2 to 1, requiring a capacitance range of only 4.8 to I, say 20pF to 96pF. A capacitor having a range of 20pF to 96pF is obviously very much more practical than one which has a range of 20pF to 980pF.

Tracking

The tuned circuit of the local oscillator must maintain throughout its tuning range a constant frequency separation equal to the IF from the mixer tuned circuit. This requirement is known as 'tracking'.

The need for tracking arises because the oscillator and mixer tuned circuits cannot be identical in inductance and capacitance. For example, for a signal frequency range of 5 - 10MHz with an IF of 500kHz, the mixer tuned circuit must cover 10MHz to 5MHz (ratio 2:1), while if the oscillator is on the high side, its tuned circuit must cover 10.5MHz to 5.5MHz (ratio 1.9:1). Thus the oscillator tuning capacitor often has a smaller capacitance than the mixer capacitor.

The wider the frequency range, the more difficult tracking becomes; in practice the optimum solution generally considered is that tracking should be correct at both ends of the tuning range and also at a point near the middle.

Tracking is generally achieved in the better class of receiver by the careful adjustment of a small trimming capacitor in parallel with the oscillator tuning capacitor at the high-frequency end of each range, and the inductance of the tuning coil (by means of a dust core) at the low frequency end.

RF amplifiers

RF amplifiers, ie tuned amplifiers operating at the signal frequency, are employed in the majority of high-quality receivers and also in transceivers. Their tuning is ganged with the mixer/local oscillator tuning control.

Basically, an RF stage improves the sensitivity of the receiver, ie it increases the signal/noise ratio. The additional selectivity resulting from the extra tuned circuits may be advantageous in a number of ways, ie the chance of second-channel interference is reduced, as is radiation from the local oscillator via the antenna. This additional RF selectivity is always useful.

The older receivers with an IF of around 465kHz always employed two RF stages and the second-channel interference then only became unacceptable above about 30MHz. If the IF was 1.6MHz, one RF stage could be considered to be adequate.

The Intermediate-frequency amplifier

The function of the intermediate-frequency (IF) amplifier is to amplify the output of the mixer before demodulation; it is a tuned amplifier, it operates at a single fairly low frequency (the IF). Hence high gain and stability are easily achieved, in fact it is the IF amplifier which provides virtually all the selectivity and most of the gain of the superhet receiver. Its importance is therefore obvious.

The selectivity of the IF amplifier can be achieved by means of tuned circuits or bandpass filters. It is desirable to be able to change the bandwidth of the IF amplifier to suit the signal being received, ie from about 2.7kHz ('wide') for SSB to about 300Hz ('narrow') for CW.

The tuned circuits are designed as 'coupled pairs' (see Chapter 2). An IF transformer is such a pair in a screening can (see Fig 5.5). A typical IF amplifier may consist of two or three such stages in cascade.

The value of the IF will depend upon the selectivity required and the need to minimise image interference. These requirements are incompatible, ie low image interference requires a high IF whereas high selectivity requires a low IF. One solution to this problem is the 'double superhet', having two different intermediate frequencies. The first is fairly high, typically 1.6 to 3MHz for good image performance; this is then converted by means of a second mixer and local oscillator to a second IF which is low to provide high selectivity. The second IF may be as low as 50 to 100kHz.

The achievement of wide and narrow bandwidths in such an amplifier presents difficult electrical and mechanical design problems.

A better IF selectivity characteristic can be obtained by the use of a 'bandpass filter'. One such version is based on the use of quartz crystals; two, three or four pairs of crystals carefully matched in frequency may be used. The older general-purpose receiver used a crystal filter which employed a single crystal in conjunction with a phasing capacitor. This simple arrangement gave a nose bandwidth of less than 0.5kHz and so was very useful for receiving telegraphy. However, the skirt bandwidth, being determined mainly by the tuned circuits, was often very wide.

Another form of bandpass filter is the mechanical type. This is a mechanically resonant device which receives electrical energy, converts it into a mechanical vibration which is then converted back into electrical energy at the output. The mechanical vibration is set up in a series of six to nine metal discs by the magnetostrictive effect.

Filters of the bandpass type have a much flatter top to the selectivity curve and shape factors of 1.5-2.5. They are made in various bandwidths from 0.3-10kHz at frequencies of commonly 455kHz and 3-9MHz. Such filters of the desired bandwidth can be switched into an IF amplifier, and the design of the IF transformers used then becomes relatively unimportant as the overall selectivity is determined by the filters. They are compact but tend to be expensive. This type of filter has a sufficiently steep characteristic (ie low shape factor) to filter out the unwanted sideband of a double-sideband signal and is therefore the basis of the filter method of single-sideband generation. It is also used in the receive function of the SSB transceiver.

The detector

The purpose of the detector is to rectify or demodulate the output of the IF amplifier, in order that the modulation originally superimposed upon the carrier wave at the transmitter can be recovered as a varying direct voltage, which can be amplified and converted into sound by the loudspeaker.

For optimum performance, each mode of modulation requires that a particular demodulating circuit is used in the receiver.

If the carrier is unmodulated, as in telegraphy, it is necessary to mix with the IF amplifier output another signal of a slightly different frequency which is generated by a 'beat frequency oscillator' (BFO) in order to produce a difference frequency in the audible range, ie an audible beat note which is then recovered by the detector.

Diode or envelope detector

The simplest and most commonly used detector is a single diode operating as a half-wave rectifier as shown in Fig 5.7. The output is developed across the resistor (the diode load) and then fed to the following audio amplifier. This arrangement is also known as an 'envelope detector' as its object is to recover the modulation envelope. It is the normally used circuit for the detection of an amplitude-modulated signal and, in conjunction with a BFO, for CW telegraphy.

Detection of an FM signal

The ratio detector is a circuit which has been widely used in both entertainment and amateur receivers. The basic circuit is shown in Fig 5.8.

Because the primary voltage is injected into the centre tap of L2 by L3, the voltages at the ends of L2 depend on the phase difference between L1 and L3. At resonance, at the centre frequency, the voltages applied to the diodes are equal. As the frequency increases, the voltage at one end increases and decreases at the other end. The reverse occurs when the frequency decreases from the centre value. The voltages at the ends of L2 are rectified by the diodes and so the rectified voltages appearing across C1 and C2 vary, giving rise to an output at point A.

The output is therefore proportional to the ratio of the voltages which appear across C1 and C2. The total voltage across C1 and C2 is held constant (ie its amplitude does not vary) by the capacitor C3, which is at least 8uF.

Detection of an SSB signal

The detection of a single-sideband (SSB) signal necessitates the insertion of a signal into the detector to simulate the carrier wave which was suppressed in the transmitter. This signal is generated in the receiver by the 'carrier insertion oscillator' (CIO).

This function can be fulfilled by the BFO of an AM/CW receiver, and by the use of the usual diode envelope detector, reasonably satisfactory results may be obtained. However, the diode detector system has directly opposite requirements for optimum CW detection and optimum SSB detection. A very small input signal from the BFO is preferable for CW whereas SSB detection requires a much larger BFO signal. As the BFO injection voltage is never adjustable, it should be set to suit whichever mode is of most interest. A large BFO voltage is likely to affect the operation of the AGC system as discussed later in this chapter.

The 'product detector' is the preferred circuit arrangement for the resolution of an SSB signal. This is a mixer circuit, and one of several varieties is shown in Fig 5.9. The frequencies involved in this mixing process are the receiver IF and the frequency to which the BFO (or CIO) is set. The BFO frequency will have been adjusted to produce an acceptable audible beat frequency. The mixer output frequencies are the audio frequency required (difference) and the sum of the intermediate and BFO frequencies -the following stages will not operate at the sum frequency. The circuit shown uses a FET. The IF amplifier output is connected to the gate and the BFO/CIO injection voltage is taken to the source.

The product detector is also a very effective demodulator of CW telegraphy signals. A further advantage is that the BFO injection voltage necessary is small and is the same for SSB and CW.

Thus the modern all-mode receiver will include two detectors, ie a diode envelope detector for AM and a product detector for SSB/CW. The VHF receiver is most likely to also include a ratio detector for FM.

Beat frequency oscillator

The BFO is a conventional oscillator which operates at the IF of the receiver. Its frequency is generally variable by ±3- 4kHz by means of a front-panel control in order to provide the audible beat note discussed earlier, and to enable this note to be set at a frequency which is acceptable to the operator. The BFO is switched on and off by a front-panel control. Coupling between the B FO and detector is very loose: 5pF or so, or even by stray capacitance.

Carrier Insertion oscillator

The carrier insertion oscillator generates a signal to simulate the carrier wave which has been suppressed in the transmitter. It also performs the same function as the BFO when receiving telegraphy.

In order to achieve the frequency stability necessary in an SSB system, the CIO would be crystal controlled, a separate crystal being used for each sideband. The crystal frequencies are typically ±1.5kHz from the intermediate frequency.

Automatic gain control

Automatic gain control (AGC) refers to the control of the gain of the receiver in sympathy with the strength of the received signal. The object is to ensure that the output of the receiver remains constant or nearly so, irrespective of the incoming signal strength which may undergo considerable variation due to propagation conditions (fading) or simply due to the relative signal strengths of the several stations operating in a net.

The basis of the operation of an AGC system is as follows. As the received signal strength increases, so does the receiver output, and a sample of this is taken from some point in the output stages and fed back in such a way as to reduce the overall gain of the receiver. As the signal fades or a weaker signal is being received, the output falls and a lower control voltage results, hence increasing the receiver gain.

The gain of the IF amplifier is controlled by feeding the AGC voltage to the base of each transistor in order to vary the emitter current and hence the gain.

The point in the receiver from which the AGC control voltage is taken depends mainly on its complexity, and inevitably it is after the IF amplifier. In a simple receiver the AGC voltage would be taken from the detector diode circuit. A separate diode to develop the AGC voltage, fed via a small capacitor (say, 33pF) from the same point as the detector diode, provides a more flexible arrangement from the design aspect and is generally to be preferred.

The simplest arrangement is shown in Fig 5.10. R1 - R4 form a divider which provides the diode with a small forward bias to make it more sensitive to weak signals. Any increase or decrease in voltage at point A due to changing signal strength will be applied to the base of the first IF amplifier transistor. Any audio component is filtered out by R2 and C1.

Effective AGC for CW reception presents a number of difficulties. The rectified BFO voltage may well reduce the gain even in the absence of a signal. For this reason AGC is often switched out of operation (by S1 in Fig 5.10, ganged to the BFO on/off switch) when receiving CW signals.

A receiver intended for CW/SSB reception will invariably employ a product detector. This provides much better isolation between the locally generated BFO/CIO voltage and the AGC circuit, and hence AGC on CW reception is much more effective. The AGC voltage in an SSB receiver is sometimes obtained by sampling and rectifying the audio at some point in the audio amplifier. There is not much to choose between the two methods.

By suitable design, an AGC system can provide a characteristic which exhibits little change in output level (less than 4dB) for a very large change in input signal (90-100dB). However, a more significant characteristic, particularly for SSB with the intermittent nature of the signal and its syllabic variations, is the speed of operation of the AGC system. The AGC must take effect quickly: the attack time must be of the order of 2ms but the release should be much slower, about 200-300ms. These times are governed by the time constants (ie products of resistance and capacitance) in the AGC circuit.

Audio stages

The audio side of the communication receiver is conventional in every way, bearing in mind the restricted audio bandwidth necessary for communication purposes. The audio power output is normally 1-2W peak to a small loudspeaker located within the receiver cabinet. Generally provision is made to plug in a pair of headphones at the input side of the output stage.

The inclusion of some form of additional selectivity or filter in the audio chain is not uncommon in the more complex receiver, particularly in the older general-purpose receiver.

This may take two forms, one being a sharp notch filter, in which the notch can be tuned across the audio band. The gain in the notch is very much reduced and so it can attenuate a particular interfering frequency. Alternatively a sharp peak of amplification at a particular frequency, say 1000Hz, may be provided and, by careful adjustment of the BF0 to give a 1000Hz beat note, the overall selectivity for CW may be improved.

Calibration oscillator

This is a crystal oscillator arranged to produce a high level of harmonics and operating usually at 100kHz. It provides a calibration 'pip' every 100kHz throughout the receiver tuning range. Generally provision is made to adjust the tuning scale or the pointer slightly to enable the calibration to be corrected at the 100kHz points. In the better class of equipment there are facilities for checking the accuracy of the crystal frequency against a standard frequency transmission.

Noise limiters and noise blankers

Much electrical interference to reception arises from the short pulse of energy radiated whenever a spark occurs, be it from a faulty switch or a car ignition system. The noise limiter is an arrangement of diodes which clip off those interfering pulses which exceed the modulation level in amplitude. The level at which clipping occurs is normally adjustable. The noise limiter is a simple and quite effective circuit.

The noise blanker is a much more complex circuit in which the interfering noise pulses are selected, amplified and detected. The resulting waveform is then fed back into the receiver via a gate circuit. Thus the interfering pulse waveform is 'blanked' out before it reaches the output stage of the receiver.

Squelch circuits

'Squelch' is the name given to a facility which is normally part of an FM communication receiver (or transceiver). The object is to switch off the audio output of the receiver in the absence of a signal or when the incoming signal strength is inadequate for satisfactory communication, ie when the receiver is just at the maximum range of a particular transmitter. Thus the annoying hiss produced by the receiver in the absence of a signal is eliminated. The level at which the squelch circuit comes into operation is normally adjustable.

Signal-strength meters

Most commercial receivers now incorporate a signal-strength meter (S-meter). Normally this consists of a sensitive milli-ammeter, often in a bridge circuit, which is used to monitor the AGC control voltage. This of course varies in sympathy with the incoming signal strength.

The meter is calibrated in S-units up to S9 and decibels up to 40 or 60 above S9. There is no generally agreed definition of an S-unit (it may be 4 or 6dB) or of the zero point of the meter. Unless an S-meter has been specially calibrated against a signal generator on each band, no great reliance should be placed on its readings