|
An inexpensive six-stage superhet. The small superheterodyne is particularly suited to the needs of the average listener, for it can provide ample sensitivity and selectivity for distant reception, while the inclusion of variable selectivity enables a high standard of reproduction to be obtained for local reception. The performance of the receiver described in this article is exceptionally high, but the total cost of the parts necessary for its construction, including all valves, loud speaker and cabinet, is only some twelve pounds five shillings.
The whole apparatus, including the mains equipment, is assembled on one chassis.
The circuit details
Fig. 1. - The complete circuit diagram of the new receiver shows the novel aerial coupling system. The aerial circuit as a whole is roughly tuned by L1 and L2 and is coupled to the tuned circuit by C1 and C2. The Valves are:- X41, VMP4G, DN41 & MU12.
The principles underlying the design of a small superheterodyne were discussed in last week's issue of The Wireless World, and it was shown that by careful design a very high standard of performance can be obtained with no more than three receiving valves. These valves perform the functions of first detector, oscillator, IF amplifier, second detector, AVC rectifier; and output pentode, and these without any attempt at reflexing. Although even greater amplification could be secured by such means, it has not been attempted in this case, for experience with reflex circuits shows them to be somewhat unreliable and subject to unexpected variation with quite small changes in circuit constants. That the performance obtainable without such means is adequate for general reception, including both local and distant, is well brought out by the curves which accompanied the article already referred to. An average sensitivity of about 150 μV with a degree of selectivity capable of discriminating against the adjacent channel by some 400 times represents a very high performance for a small superheterodyne. It is the more outstanding when it is coupled with a degree of fidelity such that all frequencies between some 30 Hz and 8,000 Hz can be well reproduced.
The complete circuit diagram of the apparatus is shown in Fig. 1, and it will be seen that a triode-hexode frequency-changer is used with an HF pentode for the IF amplifier. A duo-diode-output pentode provides detection and AVC, as well as feeding the loud speaker, and the HT supply is obtained with aid of an indirectly heated full-wave rectifier. Only a single signal-frequency tuned circuit is used, and, with the special system of aerial coupling which has been embodied, this has been found to give freedom from second-channel interference and much greater efficiency than the more conventional pair of tuned circuits, coupled in the form of a band-pass filter.
The aerial circuit consists, in fact, of a pair of coupled circuits, but only one is variably tuned. On the medium wave-band the switches S1 and S2 are closed, and the primary circuit consists of L1 and C2 in series with the inductance and capacity of the aerial itself. The constants are so chosen that resonance is secured towards the middle of the medium wave-band, but an excessive response is prevented by the 25,000 Ω damping resistance R1. A good response over the whole waveband is thus secured; it is greatest in the middle and falls gradually towards the ends of the band. Outside the waveband the response falls, and as this coincides with the range of second-channel frequencies this fixed tuned circuit gives a very appreciable degree of second-channel rejection. The secondary circuit consists of the coil L3 in series with the fixed capacitor C2 and the variable capacitor C3. This circuit is tuned to the frequency of the wanted station and is responsible for the greater part of the second-channel rejection. It is coupled to the primary in two ways; first by the capacitor C2, which is common to both circuits, and secondly by the 'top-end' capacity C1. C2 has a value of 0.005 μF, and as its reactance increases at low frequencies it gives greater coupling at the low frequency end of the waveband than at the high. The effect of the top-end capacity C is just the opposite, and this gives greater coupling at high frequencies than at low. The combination of the two couplings consequently tends to produce constant coupling throughout the waveband. Actually, of course, the coupling is not exactly constant, but it is much more nearly so than it any form of single coupling were used. The capacitor C1 is of very small capacity, and is not a component in the accepted sense of the word, but is obtained in the wiring by the juxtaposition of certain leads.
On the long waveband the switches S1 and S2 are open, and the primary inductance is augmented by L2 to such a value that resonance now occurs in the middle of the long waveband. Similarly, the secondary inductance is increased by L4. No change in the values of the coupling components has been found necessary, however.
Turning now to the oscillator circuit, the triode section of the triode-hexode is used for generating the local oscillations. The tuned circuit is included in the anode circuit of the valve and is shunt-fed by means of the 75,000 Ω resistance R5 and the 0.01 μF capacitor C7. On the medium waveband S4. is closed and S5 open so that the inductance is L6 alone, and it is tuned by the variable capacitor C13, which has in series with it the padding capacity comprising the capacitors C10, C11 and C12. Of these, C12 is the adjustable trimmer and the other two are fixed capacitors, giving a total of 0.0004 μF. Two capacitors are, of course, needed only because 0.0004 μF is not a standard capacity.
On the long waveband S4 is open and S5 closed. The inductance then consists of L6 and L7 in series, the stray circuit capacity is augmented by the trimmer C14, and the padding capacity is reduced by the insertion of the additional series capacity of 0.0015 μF obtained by the two capacitors C8 and C9. It has been found unnecessary to make this long wave padding capacity variable, for it is not very critical. The reaction coil L5 is connected in the triode grid circuit and is fed through the 0.0001 μF capacitor C6, the valve being biased by the grid-current flow through the 50,000 Ω resistance R4.
In this view of the receiver the variable-selectivity IF transformer can be seen on the extreme left, and the fixed coupling transformer between the IF valve and the smoothing capacitor.
The intermediate frequency output of the hexode section of this valve is fed to the IF valve through the IF transformer T1. This transformer has high-Q coils, and they are variably coupled so that variable selectivity can be obtained. The coupling in the second transformer T2 which feeds the diode detector is fixed, however, and as the mutual inductance between the coils of the standard component is slightly too small for the best results in this receiver, the coupling is augmented by the capacity C17 between the high-potential ends of the two circuits. This capacity is obtained in the wiring in a similar manner to C1, and it materially increases the efficiency.
The Detector and Output Stage
The detector is fed from the secondary, and its load resistance R6 is given a value of 0.25 MΩ with a by-pass capacity C18 of 0.0002 μF IF filtering is accomplished by means of the 50,000 Ω resistance R7 in conjunction with the 0.0001 μF capacitor C19. The LF potentials are fed to the pentode section of the valve through the 0.1 μF capacitor C20 and the 0.5 MΩ volume control R8. The pentode is one of the high efficiency type, and anti-parasitic resistances are accordingly included in grid and anode circuits, the grid resistance R9 being given a value of 1,000 Ω and the anode resistance R10 a value of 100 Ω.
The primary of the output transformer is included in the anode circuit of the pentode, and the space-charge grid is fed directly from the main HT line. Grid bias for the pentode is obtained by the voltage drop across the 100 Ω resistance R11. The bias applied to the AVC diode, which is fed with IF potentials from the primary of T2 through the 0.0002 μF capacitor C16, is obtained by the sum of the voltage drops across R11 and R12, the latter resistance having a value of 500 Ω. The diode load resistance consists of the two 0.5 Ω resistances R13 and R14 in series, the latter being by-passed by the 0.1 μF capacitor C23. The full AVC bias voltage developed across the load resistance is applied to the frequency-changer through the filter comprising R15 of 0.5 MΩ and C22 of 0.1 μF and through the 0. 5 MΩ decoupling resistance R2. One half of the AVC bias only is applied to the IF valve in order to avoid distortion on strong signals, and this is obtained by connecting the earthy side of the secondary of T1 to the junction of R13 and R14. The frequency-changer and the IF valve have the same initial bias applied to them. The two cathodes are connected together and taken to the earth line through the 150 Ω resistance R3, which is by-passed by the 0.1 μF capacitor C5. The anodes of all valves are fed from the main HT line without decoupling, since this has been found to be unnecessary, but the screen potentials are taken from a potentiometer connected across the HT supply. This potentiometer comprises the resistances R16, R17, and R18, and potentials of 100 Volts for the IF valve, and 70 Volts for the frequency-changer are available.
The mains equipment is of simple type, but is adequate for the needs of the receiver. The mains transformer has windings giving 4 Volts at 4 Amp for the receiving valves and the dial lights, 4 Volts at 2.5 Amp for the rectifier, and 350-0-350 volts at 60 mA. for the HT supply. An indirectly heated rectifier is used, and a 4 μF electrolytic capacitor C25 is employed for the reservoir capacitor. Smoothing is effected by the field winding of the moving-coil loud speaker in conjunction with the 8 μF electrolytic capacitor C24, and after smoothing a supply of some 200 Volts is available. The field winding has a resistance of 2,500 Ω, and is, of course, energised by the current which it smooths. In order to avoid hum being generated in the speaker itself from the ripple on the current through its field a hum-bucking coil is fitted.
A list of the components required for this receiver appears at the end, and it should be remarked that in certain cases, such as the coils and IF transformers, it is important to employ the specified parts. Fixed capacitors and resistances can be of any good quality make, of course, and this applies also to such components as the mains transformer, provided that the physical dimensions are such as to permit their being mounted on the chassis.
Constructing and Operating the New Inexpensive Superhet
Front view of chassis.
A metal chassis with all holes in the correct positions is available for this receiver and its use greatly facilitates the construction. It is necessary to mount the bracket for the variable selectivity control before the trimmers, since they obscure one of the fixing holes. These trimmers, C10 and C14, must not be mounted flat against the chassis; in one case because the trimmer would foul one of the bracket fixing bolts, in the other because the capacity between trimmer and chassis would be too high. Long 6 BA bolts are used for the mounting and three full-size nuts are run on each and securely tightened before placing the trimmers on them, so that the trimmers are spaced from the chassis by a distance equal to the thickness of three nuts.
Care should be taken to see that the electrolytic capacitors, the earth terminal, the coil screens, and all bolts carrying earthing tags make good contact with the chassis, and it is important to make sure that the aluminium-finish to the chassis is removed at the-necessary points. The aerial and PU terminals must, of course, be insulated with the washers provided.
The wiring is straightforward, but care should be taken to follow the original plan not only in the matter of the actual connections but in regard to the positions of the wires. Those wires nearest the base-board will naturally be put in first and for the heater wiring to the three receiving valves No. 16 gauge is used. This gauge is also used for one or two other long leads where its extra rigidity renders it advisable. It should, of course, be first straightened by stretching, otherwise difficulty will be experienced in sliding the insulating sleeving over it. Except in one or two special places, the rest of the wiring can be carried out with No. 20 gauge which is much easier to handle.
Particular care should be taken in the wiring of the detector and AVC circuits for the proximity of certain leads is relied upon to produce the capacitor C17. Similarly C1 is produced by the proximity of two wires, but in this case they are extra wires inserted for this purpose only. A length of No. 16 wire is joined at one end of the aerial terminal and bent into the position shown in the drawings. Another length of the same gauge wire is supported by terminal (1) of the signal-frequency coil assembly and is bent so that it lies parallel with the first. The two wires are then sleeved for the whole of their lengths, using thin walled sleeving, and are bound together with thread to form the capacitor C1.
The adjustments necessary before the receiver will give its correct performance are to the IF circuits and to the ganging. There are four trimmers in the IF amplifier, three for the medium-wave gauging and one for the long waveband - a total of eight. Some form of tuning indicator is a great help, and it can take the form of a proper output meter connected across the output transformer primary or a volt-meter connected between the chassis and the cathode of one of the controlled valves. In the former case, the meter reading will increase with signal strength, but in the latter it will decrease once the signal is strong enough to operate the AVC system. A voltmeter used in this way should have a full-scale deflection for some 6 to 10 Volts.
Ganging with an Oscillator
Underneath the chassis.
If a test oscillator be available, connect a fixed capacitor in series with its high potential output lead, and set it to 465 kHz. Connect the earthy output lead to the chassis and the other to the grid of the IF valve. Roughly adjust the trimmers on T2 for maximum response. Then adjust the output of the oscillator so that the indicating meter reads a convenient value; if no meter is used, and the ear is relied upon, keep the input very small so that the signal is only just audible. Then carefully adjust the two trimmers to their optimum settings. In this connection it should be noted that the secondary trimmer will have little or no effect upon the indication of a voltmeter, and this circuit will usually have to be adjusted aurally.
When satisfied with the adjustments to T2, connect the output of the oscillator to the grid of the triode-hexode and adjust the trimmers in T1 roughly. Then set the selectivity control for nearly maximum selectivity, that is, nearly fully rotated in a clockwise direction, and carefully adjust each of the two trimmers in T1. A much smaller output from the oscillator than before should be required, and it will usually pay to check over the adjustment of all four IF trimmers before proceeding to the ganging. When satisfied with the adjustments, set the selectivity control at optimum. This is readily done by ear by choosing the setting which with a weak input gives the loudest signal in the speaker.
Transfer the oscillator output to the aerial and earth terminals, using a standard artificial aerial if available, if not a 0.0002 μF capacitor between the aerial terminal and the high potential oscillator lead. Set the test oscillator to 1,500 kHz is and the tuning control at zero. Stop the receiver oscillator from functioning by joining terminal (1) of the oscillator coil to earth by a lead fitted with spring clips. Set the oscillator to give a large output of 0.1 to 1 Volt and adjust the trimmer on C3 for maximum response as indicated on a voltmeter connected across R3. Then set the test oscillator to 1,400 kHz and tune the set to it by the main tuning control. Reduce the output of the test oscillator and remove the short-circuit from the set oscillator. Now tune in the signal exactly by the adjustment of the trimmer on C13 only.
The next steps are to replace the short-circuit on the set oscillator and to set the test oscillator for large output at 600 kHz. The receiver should then be tuned to this signal using the voltmeter as an indicator. When this has been done, reduce the oscillator output, remove the short-circuit, and tune in the signal by the adjustment of C12 only. To ensure the most accurate ganging, a return should be made to 1,400 kHz and the trimmer on C13 readjusted in the manner already described.
The Long Waveband
Drilling, assembling and wiring details for the chassis.
When the medium wave ganging has been completed it is the turn of the long waves. Here only one trimmer, C14, needs adjustment. This is best done by setting the test oscillator at some 250/300 kHz with a large output and the set oscillator short-circuited as before. Tune the set to the signal, using the voltmeter as an indicator, reduce the oscillator output and remove the short-circuit. Then tune in the signal by adjusting C14 only. This completes the ganging and no further adjustments should prove necessary, save, perhaps, that it is advisable to check the adjustment of the trimmer on C3 with the aerial on which the set is to be used. This is easily done by tuning in a low wave-length station on the medium Waveband and adjusting this trimmer for the strongest signals.
Where a .test oscillator is not available, the ganging must be carried out on signals. This is not quite so easy, but is perfectly practicable. At first screw up the trimmers on T1 and T2 fully. Then unscrew the secondary trimmer on T2 three complete turns and the primary trimmer two and a half turns, the bottom trimmer on T1 one and a half turns, and the top trimmer two turns. Unscrew C14 fully and C12 about three complete turns. Screw the two trimmers on the gang capacitor fully home and then unscrew that on C3 half a turn and that on C13 three-quarters of a turn. Set the selectivity control for low selectivity (anti-clockwise) and the wave-range switch for the medium waveband (also anti-clockwise).
It should then be possible to tune in a signal and it a voltmeter be connected across R3 as an indicator, a strong signal is an advantage; It the ear is used to judge signal strength, only weak stations should be used. Having obtained a signal, adjust the four IF trimmers as accurately as possible; as the signal strength increases, increase the selectivity and readjust the trimmers in T1, repeating the process until the final adjustments are made with the control set for nearly maximum selectivity. The IF circuits are now in line with one another, but it is not yet known whether they are tuned to 465 kHz or not. If they are not, the ganging may not hold accurately.
The next step is to tune in a low wavelength station and to adjust the trimmer on C3 for maximum response. If a station below some 220 metres can be found and identified, set the tuning control so that its Wavelength is indicated by the dial and tune it in by the adjustment of the trimmers on C3 and C13. Then find a station at the other end of the waveband and adjust C12 while rocking the tuning control backwards and forwards over a few degrees until the optimum combination of settings is found. A return should then be made to a low wavelength and the trimmer on C3 readjusted. This completes the medium wave gauging provided that the intermediate frequency is correct.
Checking the Intermediate Frequency
Voltages and currents.
The question of superheterodyne whistles will be dealt with later, but it may be remarked that a whistle is quite probable when receiving a station having a frequency twice that of the intermediate frequency. That is to say, when the intermediate frequency is 465 kHz a whistle can be expected when receiving a station on 930 kHz. This forms a useful check upon the intermediate frequency, and the station upon which the whistle occurs should be identified so that from a knowledge of its frequency the intermediate frequency can be found. If the adjustments be widely out so that accurate ganging cannot be obtained, a number of whistles may be found, and in this case the wanted one is likely to be the strongest.
If it be found, for instance, that the whistle occurs at 1,000 kHz, the intermediate frequency is one-half of this, 500 kHz. This is much too high and all four IF trimmers should be screwed up some-what until the whistle occurs at 930 kHZ. On the other hand, if the whistle occurs at, say, 860 kHz, the intermediate frequency is 430 kHz, which is too low, so that the IF trimmers must be further unscrewed. Once the correct frequency has been found, the adjustments can be proceeded with in the certainty that they are final.
The long wave ganging is easy, for C14 is the only trimmer to be adjusted and its optimum setting is very near to minimum. It is best done on a station towards the middle or lower end of the waveband, such as Luxembourg, and C14 must be adjusted while rocking the tuning control backwards and forwards over a few degrees until the optimum combination of settings is found.
On test, this receiver was found to give a very satisfactory performance indeed. The sensitivity proved adequate for the reception of all worth-while stations with quite a modest outdoor aerial, During the daylight hours and in the heart of London stations such as Fécamp, North Regional, Cologne, and Brussels on the medium waveband would not only give full volume but produced a detector input well beyond the threshold of AVC. On the long waveband, Huizon, Radio-Paris and Luxembourg gave signals which anyone could be excused for mistaking for locals.
Even with optimum coupling in the IF transformer, giving quite a high degree of selectivity, the quality of reproduction reached a high standard and with the lower selectivity possible for strong stations it proved to be exceptionally high. Mains hum was found to be negligible, and the selectivity entirely adequate, while background hiss was inaudible.
Very few whistles were found. On the long waveband a trace of one whistle could be detected when the receiver was mis-tuned by some 5 kHz to 10 kHz from Droitwich due to the oscillator harmonic beating with the London National. The whistle could only just be heard, however, and was inaudible with the receiver correctly tuned to the station.
On the medium waveband also the whistles were few. As already mentioned, one is to be expected at 930 kHz and is due to feed-back of the second harmonic of the intermediate frequency which is generated in the detector. With such a compact layout of components, it is hardly practicable to avoid it. Another whistle may be found at 1,395 kHz, due to feedback of the third harmonic, but this is likely to be less prominent and was not noticeable in the original receiver. Second channel whistles and the multitude of whistles due to kindred effects which are found in some superheterodynes are completely absent.
Conclusion
The complete receiver and loud speaker housed in a CAC cabinet.
In conclusion, it may be said that the receiver is one which will give a very satisfactory performance indeed, judged from all points of view. Even if no account were taken of its cost, its performance would be a remarkable one. When the low price of its components is taken into consideration its performance becomes really outstanding.
List of Parts
|
