|Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting|
As mentioned in Section 6, tubes (valves) for mains operation have an indirectly heated cathode as the emissive electrode. Apart from this they operate in the same way as battery tubes (valves) and may be substituted for them in all the applications we have looked at in the previous two sections. In fact, the two TRF circuits shown and described could very easily be converted to mains operation with the addition of a power supply unit and a few other components. The great difference is that since power from the mains is cheap and virtually unlimited the tube (valve) can be made to perform much more effectively; in particular very large power outputs may be obtained.
Mains tubes (valves) need negative grid bias just as their battery equivalents, but thanks to their cathodes no battery is necessary. Since the HT current passing through the tube (valve) returns to HT— (normally the chassis) via the cathode, if a resistor is inserted in the circuit a certain amount of voltage drop will occur across it. If the grid is returned to chassis it will effectively be at a negative potential with respect to it, thus providing the necessary bias. The value of the cathode or bias resistor (R1 in FIG. 1) is found in the normal way with Ohm’s law by dividing the required negative voltage by the total anode and screen current drawn by the valve. To provide an AC path for RF or AF currents the bias resistor is decoupled by a suitable value of capacitor (C1), typically 0.1 uF for RF and 25 p to 100 uF for AF.
Types of mains tubes (valves)
Mains tubes (valves) as a whole fall into two broad classes, those with relatively low voltage, high current heaters for parallel operation, and those with high voltage, low current heaters to be operated in series. Generically these are known respectively as AC-only and AC/DC tubes (valves), terms which need explanation.
AC-only tubes (valves) are types which are intended to have their heaters powered by a low voltage secondary winding on a mains transformer in the power supply unit. The actual voltage will normally be, in UK receivers, 4 V or 6.3 V 4 V tubes (valves) were in common use from about 1928 to 1939 whilst 6.3V types were introduced around 1937 and by 1940 (the year that domestic radio production virtually ceased) were set to take over almost completely. In the USA 2.5 V heaters were widely used until around 1935, when 6.3V types were introduced and soon ousted them. The reason for the choice of 6.3 V was quite simple: it permits the tube (valve) also to be used on a 6V storage battery for automobile receivers. The heater current taken by normal domestic type AC-only tubes (valves) ranges from about 0.2 A up to 3 A, whilst a few intended primarily for automobile use were rated at 0.15 A.
We should also mention that a very few AC- only output triodes and pentodes had directly heated filaments rated at 2 V or 4 V. The triodes in particular were mostly employed in expensive receivers and radio-gramophones of the 1930s and immediate post-war period by A.C. Cossor Ltd, Decca, EMI (HMV and Marconiphone) and the Radio-Gramophone Development Company (RGD). They are unlikely to be found in other makes.
The low voltage or LT winding on the mains transformer is sometimes centre tapped with the tap connected to chassis, either directly or via a small potentiometer called a hum-dinger. This is adjusted to minimize any hum from the loud speaker of a set due to slight heater-cathode leakage in a tube (valve) or tubes (valves). This device is normally only found in older receivers and those employing directly heated output tubes (valves). Otherwise it is usual simply to connect one side of the LT winding to chassis.
AC-only tubes (valves) as used in the average domestic radio are intended to work with anode and screen-grid voltages of between about 200V and 250 v whilst high power output tubes (valves) employed in large radio-gramophones and in audio amplifiers may take up to 500 V. These voltages normally are obtained from another winding on the mains transformer, but as this is of course AC it has to be converted into DC — rectified — and then smoothed, that is, have any remaining AC ripple removed. Two main types of rectification are employed, half-wave and full-wave. As the terms imply, the first rectifies only one half of each cycle whilst the second rectifies both halves. The rectifiers themselves consist in the vast majority of vintage receivers of our old friend the diode, exploiting its ability to pass voltage in one direction but not in the other. For half-wave circuits only one diode is required, in conjunction with a single HT winding on the transformer. (A) in FIG. 2 shows how one half of each cycle applied to the anode of the diode results in a DC pulse at the cathode. When a full-wave rectifier is employed with a centre-tapped HT winding and two diodes twice the number of DC pulses appear on their cathodes, which are coupled together. The closer spacing of these pulses (B) make it inherently easier to smooth them effectively, so full-wave rectifiers are used in the majority of AC-only sets with only a few, generally cheaper makes using half-wave types.
The standard method of smoothing the HT is carried out by a combination of large value capacitors and one or more LF chokes, called in this application smoothing chokes. The voltage from the rectifier filament or cathode is first applied to a reservoir capacitor which evens out to a considerable degree the DC pulses. The result is DC which is almost, but not quite, steady since it still contains a small AC ripple. The purpose of the smoothing choke is to filter this out, which it does by presenting a low resistance to DC but a high impedance to AC. On the other side of the choke another smoothing capacitor finally removes whatever tiny ripple remains. Most ordinary domestic receivers needed only this basic choke plus two capacitors arrangement but some large radio-gramophones and amplifiers employed more than two chokes and three capacitors.
A very popular arrangement during the 1930s and 1940s was to have the smoothing choke in the form of a field coil in a loudspeaker employing an electromagnet instead of a permanent type. This was doubly useful as the electromagnet generally was superior to the permanent magnets of the day and its efficiency did not wane with age, whilst it also saved the cost of a choke. Some sets employing double smoothing used the field plus an ordinary smoothing choke.
In FIG. 5 L3 is the field coil of the loudspeaker and L1 its speech coil, fed via L2 from the secondary of the output transformer Ti. There is a danger that any remaining ripple voltage in the field would be picked up by the speech coil and reproduced as a hum; it is the job of L2 to prevent this. Known as the humbucking coil, it consists of a few turns of wire wound on the same former as the field coil but in anti-phase to it. It picks up the same ripple voltages and by feeding them to the voice coil cancels out the others.
The introduction of high capacity; small size electrolytic capacitors prompted some firms, notably Philips/Mullard, to adopt resistance smoothing in which the choke or loudspeaker field is replaced by a high wattage resistor. The smoothing properties of the latter are very limited and there is very great dependence on very large values for the capacitors.
Whatever smoothing arrangement is employed in a receiver a certain amount of voltage is bound to be dropped across it. Because of this the HT winding on the mains transformer is designed to supply a higher voltage to the rectifier than will be employed as HT in the receiver, etc. Typically, a set with an HT line rated at 250V will employ a 350—0—350V HT winding, allowing for 100V or so to be dropped along the way.
The vast majority of sets deploy the smoothing choke or speaker field in the positive HT line with the centre tap of the transformer and the negative terminals of the smoothing capacitors taken to chassis, which acts as the HT— line. This is called positive smoothing, but a few sets placed the smoothing components in the negative line. This feature is mainly to be found in sets from the middle 1930s, especially those made by EMI. The negative terminal of the reservoir capacitor is isolated from chassis and is negative with respect to the latter by the amount of voltage dropped by the choke or field. This negative voltage may be used to bias the tube (valve) as an alternative to cathode bias, especially when directly heated output types are employed, or to act in certain types of automatic volume control, to be discussed later.
‘Swinging choke’ smoothing
Any HT supply employing the kind of choke- capacity or resistance-capacity smoothing just described is subject to a certain amount of voltage variation due to fluctuations in the current drawn by the tube (valve). To combat this a modified type of smoothing circuit is sometimes employed in which the reservoir capacitor is omitted and a special type of smoothing choke is fitted, the impedance of which is directly affected by the load current. The absence of the reservoir reduces the initial no- load output at the rectifier by perhaps as much as 100 V, but the subsequent variation due to current drain is greatly reduced. As far as the writer knows this device was used in only one UK domestic receiver in the late 1940s, but even then it was speedily replaced by conventional smoothing. It is as well to be aware of the swinging choke, since it may well be encountered in high power amplifiers.
Directly or indirectly heated?
Although FIG. 2 shows two separate diodes in the full-wave circuit, in practice they are almost always combined in one envelope. From the point of view of their output there is no need for them to be indirectly heated but this may be an advantage in actual operation. Directly heated rectifiers start to work almost immediately a set is switched on, some time before the indirectly heated tubes (valves) begin to draw HT current. Note that the theoretical DC output voltage is equal to the peak value of the AC input, i.e. 1.41 times the RMS value. Although this is not usually realised fully in practice the DC voltage may well be significantly higher than the AC input, particularly during the initial warm-up period of the other tubes (valves), and may cause harm. For this reason indirectly heated rectifiers were developed having similar warm-up times to the others in the set.
Normally both types of rectifier have their filaments or heaters supplied by another LT winding on the mains transformer, well insulated since it too is up to 500 V DC above chassis. The vast majority of indirectly heated rectifiers have the cathode internally connected to one side of the heater, but a very few were made with separate, highly insulated cathodes enabling them to work from the same LT winding as the other tubes (valves). Their use is principally confined to automobile receivers but they did occasionally appear in domestic sets. Their use is to be avoided if possible since a breakdown of insulation between cathode and heater, with one side of the latter earthed, results in a dead short from HT to chassis and the probable burning-out of the mains transformer. One particular well-known radio manufacturer discovered this to its cost in the early 1950s.
The voltage doubler
A third type of rectifier circuit, found almost exclusively in very early AC-only sets, is the voltage doubler. A single HT winding is used to supply simultaneously the anode of one diode and the cathode of another. One diode passes only the positive halves of each cycle, the other diode the negative halves, the resulting voltages being developed across two reservoir capacitors wired in series. Each of these receives a charge notionally equivalent to 1.41 times the RMS input to the anode or cathode, and as they are added together to give the HT output to the receiver the combined voltage should be 2.82 times that of the AC input. In practice this is seldom obtained except at light loads, but there is a useful step-up.
Voltage doublers are associated mainly with the alternative to the diode, the ‘metal’ rectifier which takes advantage of the fact that certain combinations of metals, such as copper and copper oxide, will pass voltage in one direction only. The principle was exploited mainly by the Westinghouse Brake & Saxby Signal Co. Ltd, which was the major manufacturer of metal rectifiers during the 1920, 1930s, 1940s and 1950s. It is said that voltage-doubling circuits were favored with metal rectifiers since each of the two used had only to work at half the total HT voltage, thus reducing stresses on them.
The conventional metal rectifier employed small copper washers oxidized on one surface sandwiched between lead washers of similar size and much larger thin metal plates used only to dissipate the heat set up in the rectifying process. The number of washers and cooling plates employed depended on the working voltage rating; rectifiers for use in radio sets tended to be large and rather cumbersome. More compact types with smaller, thicker discs and plates appeared in the late 1940s, whilst in the 1950s a radical new version was developed. This was the so-called contact-cooled rectifier in which the discs were contained in a thin aluminum box, the latter being bolted down firmly to the chassis of the receiver through which the heat was dissipated.
FIG. 8 Bridge circuit employing metal rectifiers, represented by symbols 1, 2, 3 and 4. The flat bar is the cathode and the arrowhead the anode.
Contact-cooled rectifiers were responsible for a brief revival of a fourth type of rectifying circuit called the bridge. This uses four diodes to give the same kind of output as a full-wave rectifier but with only a single HT winding. As will be seen from FIG. 8 the diodes are arranged in two pairs, each of which is wired with the anode of one connected to the cathode of the other. The four are then arranged so that the remaining anodes are joined together and the remaining cathodes are treated similarly. The AC input is applied to the two junctions of anodes and cathodes so that during each cycle one of each pair of diodes must be conducting in one direction or the other. The positive side of the output is taken from the combined cathodes and the negative from the combined anodes. The bridge circuit had long been employed in battery chargers, using conventional disc type rectifiers, and continues to be popular in this field. Its reign in radio receivers was limited to a few makes for only two or three years, and it is most likely to be found in imported continental sets.
The mains transformer in a true AC-only set isolates the ‘works’ from the mains itself and consequently the chassis may be earthed for safety as with other metal electrical appliances.
Different mains supplies
In Section 1 it was shown how many early power stations supplied their customers with DC. In fact in 1926, when the National Grid was authorized by Parliament, there were 223 different DC-only power stations as against 156 supplying AC only. A further 103 stations supplied both DC and AC. The range of voltages was astonishing: 24 different DC voltages and no fewer than 35 for AC of between about 100V and 250 V. Most AC supplies were at 50 Hz but 26 stations supplied others ranging from 25 Hz to 100 Hz.
It will be seen from these figures that AC-only receivers made for general sale throughout the country had to be readily adaptable to suit individual mains supplies. The voltage problem was catered for by having the primary windings of mains transformers tapped for different inputs, sometimes every 1OV between 200V and 250V. Some early sets also had tappings for 100/130 V but usually listeners on these low voltage supplies had to order special versions of a receiver. Most transformers accepted between 50 Hz and 100 Hz quite happily but it was often necessary to have special versions for 25 Hz. In any case, the greater time spacing between positive pulses in the latter called for improved smoothing for the HT, and again special versions of receivers were offered.
By 1930 there was near parity between the number of AC and DC power stations and by the end of the decade the ratio was about 5:1 in favor of AC. However, it has to be remembered that it had mostly been the more affluent communities which were first supplied with electricity, and that the vast majority of these were DC. They remained as a very significant potential source of sales income for radio firms. Initially DC-only receivers were made but were swiftly outmoded by sets capable of working on either AC or DC without adjustment, save perhaps that of a voltage tapping. The tube (valve) used in AC/DC or Universal receivers have
their heaters wired in series — a chain — and although each may have a different voltage rating they all must pass the same current. There was no general consensus amongst tubes (valves) makers on the latter, which may range from 0.1 A to 0.3 A. The heaters are supplied from the mains via a voltage dropping resistor usually known simply as a dropper. The value of the latter depends both on the total voltage of the tube (valve) heaters in the chain, their current and the mains voltage.
Droppers produce heat which has to be dissipated into the air space within a receiver cabinet. Whilst this is of no great significance in the average table set it can be highly inconvenient if not downright dangerous in a small cabinet. This problem was solved by the use of a special kind of mains lead (line cord in the USA) incorporating a resistive element. The latter consisted of thin resistance wire wound on an asbestos core and surrounded by a woven asbestos sheath. When it plus one, two, or occasionally more conductors were enclosed in cotton braiding it was no thicker than a conventional mains lead. The resistance value was generally 60 ohm per foot for 0.3 A rated lead and 100f per foot for 0.15 A. The length of the lead was calculated to suit a particular tubes (valves) line-up at an arbitrary voltage supposed to represent the half-way mark between the minimum and maximum mains voltages in use. In the UK this meant 225V and in the USA 117.5V, and a certain latitude on the part of the tube (valve) with respect to heater voltage was supposed to take care of discrepancies. The inevitable heat generated by any type of resistor is dissipated harmlessly over the entire length of the lead, which should run just pleasantly warm.
It has been the writer’s ill-fortune to have read in recent years a tremendous amount of nonsense written by ill-informed persons regarding the alleged dangers of resistive mains leads, from its heating effect of the fact that it contains asbestos. Yes, a resistive lead will run too hot if too much current flows through it, but this is not the fault of the lead per Se. It can usually be traced to a bad repair or to an owner shortening the lead for the sake of tidiness. Again, it is conceivable that if one were regularly to chew the resistive element over a period of many years the asbestos might cause harm, but few rational people are likely to indulge in this habit. The truth is that a resistive mains lead, properly installed, poses no more danger than any other type, although naturally, as with any other cable, it needs to be examined for worn insulation and exposed conductors.
The barretter is a special kind of electric lamp having an iron filament in a hydrogen-filled bulb. It has the curious but useful characteristic of passing the same current at a wide range of applied voltages, and if placed in series with an AC/DC tubes (valves) chain it automatically adjusts for different mains voltages. In view of this convenience it is rather surprising that the popularity of the barretter was limited, probably because it cost more than a conventional dropper and also required a special holder.
Rectifiers in AC/DC receivers
Only a half-wave rectifier can be used in AC/DC sets, wired as shown in FIG. 9. On AC supplies it delivers pulsating DC to the reservoir capacitor to be smoothed as in an AC-only set, whilst on DC it acts simply as a low value resistance. Good heater to cathode insulation is, of course, of prime importance (the cathodes in some early AC/DC rectifiers were so thick as to take as much as three minutes to warm up from cold).
Since the HT voltage in AC/DC sets is likely to be somewhat less than that of the mains input, in general UK tubes (valves) were designed to work with anode and screen voltages of around 200V maximum. In the USA, where mains voltages were in the 110/120 V range, tubes (valves) were designed to work efficiently on under 100 V HT, but some could also accept up to 200V when used in UK receivers. Standard choke-capacity smoothing may be found in better quality AC/DC sets, especially those designed for 200/250 V mains where some voltage drop may be tolerated. Energized loudspeakers too occasionally were employed. Cheap 200/250 V sets and most 110/120V midgets had the HT for the output tubes (valves) anode taken directly from the rectifier cathode, with a resistor of a few thousand ohms feeding the rest of the set, in conjunction with large value reservoir and smoothing capacitors. The loudspeakers in such sets seldom had a wide enough frequency response to reproduce 50 Hz hum!
An alleged improvement on the above arrangement came into vogue in the 1950s. It featured a special output transformer with an overwind on the primary through which the HT current from the rectifier cathode passed, supposedly to smooth it. It is not unusual for the overwind to fail, and if an ordinary transformer is substituted it is almost impossible to detect any difference, so perhaps this represents a defeat of hope by experience.
Apart from the differences in the power supply section the circuitry of AC-only and AC/DC receivers does not differ basically, although there are, of course, wide variations between makes and models.
Most UK manufacturers continued to use the chassis in AC/DC sets for the negative HT line, which meant that it had to be connected to one side of the mains input. In UK and in most other countries, one side of the mains — the neutral — is permanently connected to earth, with the other side — the live — at 230 V above it. This means that should the live side of the mains be taken to the receiver chassis it will be at 230V above earth and therefore dangerous to touch. Radio manufacturers were well aware of this and with a few dishonorable exceptions took elaborate precautions to prevent owners from coming into contact with any metal work whilst operating their sets. These included close fitting cabinet backs, covers over chassis fixing bolts and wax fillings over the grub screw used to fasten control knobs onto spindles; these must always be replaced carefully by anyone who has cause to dismantle an AC/DC set. In addition care should be taken to wire the mains lead so that the neutral is taken to chassis.
The better class of American midget sets had the HT— line isolated from chassis, although this was still employed as a return for RF currents. The result was that the chassis could not become live whichever way round the mains was connected.
The writer has also seen a great deal of rubbish expounded by know-all alarmists about the alleged inherent dangers of AC/DC sets. The fact, borne out by statistics, is that AC/DC receivers are no more dangerous than any other electrical appliance and are very probably a good deal safer than some when installed and handled properly.
Part AC-only sets
This oxymoron applies to certain sets in which the tube (valve) heaters are supplied by a transformer, making them suitable for AC mains only, but with HT drawn directly from the mains as in an AC/DC set. As a general rule this arrangement was confined to small cheap sets and to firms intent on cutting costs to the bone, possibly as a prelude to going out of business. The same remarks regarding live chassis apply.
Shown above is the circuit of an actual ‘straight three’ AC receiver, the Cossor model 378 which appeared in 1936. It has some interesting features which illustrate many of the aspects of design mentioned in the text.
The aerial input is either to the primary of the RF transformer formed by L1, L2 and L3, or when Si is closed to the top of the secondary, which is the MW tuning coil. The switch would be labeled ‘local’ and ‘distant’ for different reception areas. S2 shorts out the LW coil on MW. The dotted line enclosing the coils indicates that they are mounted in a metal screening can. C16 is the first tuning capacitor and C17 its trimmer. V1 is the variable-mu pentode RF amplifier. Its screen grid is supplied by the voltage divider comprised of R1 and R2, and is decoupled by C3. R2 returns to chassis via R4, a variable resistor in the cathode circuit of V1 which acts as a gain control by varying its grid bias. C4 is the cathode decoupling resistor.
Tuned—anode coupling is used to pass the amplified RF signals on from the anode of V1. This is relatively rare since it involves the fixed plates of the tuning capacitor C19 and of the trimmer C18 to be at HT with respect to their moving plates. R5 and CS decouple the HT feed to V1 anode.
L4/L5, the anode coupling coils, form the primary of an RF transformer with L8/L9 as the secondary. Switches S3 and S5 short out the appropriate LW windings on MW. These coils plus the reaction windings L6/L7 also are contained in a screening can. R6 and R7 are ‘stopper’ resistors incorporated to improve the stability of the circuit. C? is the grid capacitor and R8 the grid leak for the detector (V2), a straight RF pentode. Note that the bottom of L9 does not return directly to chassis but via S6, which on radio is closed, together with S7. The latter shorts out the cathode bias resistor R11 and by-pass capacitor C8 for V2, bias not being required for a grid leak detector. It is needed when the tube (valve) is operating as an ordinary amplifier, however, so for gramophone record reproduction (with a pick-up plugged into the sockets marked P U.) S6 and S7 are opened, bringing R11 and C8 into play and returning the grid of V2 to chassis via the pick-up, which would have been a magnetic type with a DC resistance of about 2000 ohm.
R9 is the screen grid feed resistor for V2, decoupled by C6. R10 is the anode load and C9 the RF by-pass capacitor. C10 is the AF coupling capacitor, R12 the grid return for V3 and R13 another ‘stopper’ C12 is second RF by-pass capacitor, an unusual addition. Note that V3 is a directly heated pentode so a humdinger (R14) is wired across the filament. The centre tap of this is returned to chassis via R15, causing the filament to act as a cathode and to provide negative bias for the grid. Ci3 is the by-pass capacitor for R15.
The output transformer in the anode circuit of V3 has a fixed tone control capacitor wired across it. The sockets marked EXT LS are for an extension loudspeaker, not a good idea since any wiring to the latter will be at around 250 V DC above earth.
L10 is the speech coil of the moving coil loudspeaker. The way in which L10 and L11 are shown wound on a common iron core indicates that the former is a humbucking coil and the latter afield winding. It does duty as a smoothing choke for the HT supplied by the full-wave rectifier V4, in conjunction with reservoir capacitor C14 and smoothing capacitor C15.
Note that the wiring to the heaters of V1, V2 and V3 is not shown but indicated by the letters a and b adjacent to them and to the LT winding on the mains transformer. The letters also show that the two scale lamps are fed from the same source.
The circuit of the Cossor Model 379 shown above provides an excellent example of how an A C-only set may be adapted for AC/DC working with a minimum of circuit changes. In fact, other than the tube (valve) having 13 V O.2A heaters there is no difference in the design as far as V2 grid and screen grid apart from the omission of the pick-up sockets. An addition in the anode circuit is a decoupling resistor R10 and capacitor C9, the purpose of which is to remove any very slight AC ripple on the HT line remaining after the smoothing process. C10 is the RF by-pass capacitor and C11 the coupling capacitor feeding the auto-transformer Ti which boosts the AF signals travelling to the grid of V3 via the ‘stopper’ R12, due to the low HT
Note that V3 is again an unusual type in being an indirectly heated power triode. R13 is the cathode bias resistor and C13 its by-pass capacitor. The circuitry in the anode of V3 is almost the same as in Model 378 except that the extension loudspeaker sockets are omitted.
V4 is the indirectly heated half-wave rectifier with the HT taken from its cathode and the anode supplied from the live side of the mains. S6 is the on/off switch and noise suppressor choke L14. C16 is fitted to suppress ‘modulation hum’, an obtrusive noise which appears to be tuned in with strong stations and is particular prevalent in AC/DC receivers. L13 is another noise suppressor choke; it was common practice to have them wound on a common former ‘back to back’ so that electrical noise impulses in the one cancelled out those in the other, and vice versa.
Note that the heaters of V1, V2, V3 and V4 are all wired in series and fed from the mains via the dropper R16, which has three tapping for different voltages. Note also that the heater of V2 is last in the chain before the chassis. It is a convention in all AC/DC sets to place the detector tubes (valves) in this position in the chain as a precaution against hum caused by heater/cathode leakage.
The scale lamps are wired into the lead from neutral main to chassis, thus carrying both the heater and HT current. They are shunted by R15, which helps to protect the lamps from surge voltages when the set is switched on from cold and the resistance of the tube (valve) heaters is lower than normal. It also permits the set to remain working even if one or both of the lamps should go open circuit, which otherwise would break the heater and HT circuits. Positioning the lamps here has other advantages: the lampholders are at low potential with respect to chassis and the passage of both HT and heater current through them assists in keeping them bright when the tube (valve) have fully warmed up.