|
How the Barretter functions.
Osram Barretter Type 301.
Electrical energy from the mains is cheap compared with that from either dry batteries or accumulators; and with the widespread standardisation of 230 Volt 50 Hz mains, the occasions when one is unable to obtain a connection for mains-driven apparatus are few. But for the more refined types of apparatus the public mains have the serious disadvantage of being liable to sudden fluctuations in voltage. An accumulator will maintain practically constant voltage over 80% of its discharge, and even dry batteries (provided they are not overloaded) suffer only a small and steady decline of voltage while in use. But even very good supply mains, where the voltage may never vary more than 4% on either side of its normal value, may quite probably jump suddenly from one extreme to the other; in such a case there is an instantaneous shift of 8%, a change which would take some hours to occur with a battery supply.
This has long been recognised as a troublesome effect to be combated in laboratory instruments, such as valve volt-meters, apparatus for amplifying direct currents and voltages, etc. But it is only recently that there has been a likelihood of the trouble becoming apparent to the ordinary user of radio apparatus in two different cases. The first example is in the all-wave superheterodyne receiver, for, unless the designer has exercised great skill in working out the oscillator used for frequency changing, its frequency is likely to shift with variations in anode or filament voltages, and by an amount which may necessitate retuning at the higher frequencies. The second example is in the television receiver; here there is a slight risk that when receiving very weak signals the adjustment of the synchronising may be dependent upon mains voltage.
So we may expect to see a gradual increase in the use of devices for stabilising power supplies in commercial broadcast and television receivers in proportion as the apparatus is called upon to satisfy more exacting performance specifications. The simplest device is the barretter, whose function is to maintain a constant current in spite of changes in applied voltage; it works equally well on AC and DC. In construction a barretter consists of a filament of iron wire enclosed in an atmosphere of hydrogen, and its mode of action can be understood from a study of the effect of temperature on the resistance of metallic conductors.
Suppose we take four wires, one each of copper, tungsten, iron and eureka, all of them having a resistance of 100 Ω at freezing point (0 °C); then at boiling point (100 °C) we shall find that the resistances of the first three have changed a great deal, the new values of all four being: Copper, 142.8; tungsten, 151; iron, 162; and eureka about 100.1.
Fig. 1. Voltage/current characteristics of three different conductors; these graphs show the effect of temperature, (a) being cold and (b) and (c) being hot.
It does not matter whether the wires are heated internally or externally; so that if we steadily increase the voltage applied to a conductor, as soon as the current is great enough to make the conductor appreciably hot its resistance rises, and the current through it then increases less rapidly than the applied voltage. This is illustrated by Fig. 1, which shows current plotted against voltage for three different conductors: (a) a cool-running resistance, (b) a tungsten filament electric lamp of the vacuum type, (c) a tungsten filament gas-filled electric lamp. (The tungsten filament is, of course, not pure tungsten, but an alloy with other metals which has been found by the manufacturers to be stronger and generally more satisfactory than the pure metal. This is mentioned because an alloy usually differs from the pure metal in resistance, and even more so in the temperature coefficient of resistance; the figures given in tables for pure tungsten cannot, therefore, be applied to numerical calculations on tungsten filament lamps.) It will be noticed that the flattening out of the current curve is greater for the gas filled lamp than for the vacuum lamp.
Barretter Characteristics
Fig. 2. Showing how current through a barretter is maintained sensibly constant over a wide range of working voltages.
Returning to the barretter, it can be seen from the figures that were quoted above that iron is a suitable material for a current-limiting resistance working on this principle, for it has one of the highest rates of increase of resistance with temperature in addition to a fairly high specific resistance. Hydrogen is chosen for the gas filling because it is a very good conductor of heat, besides being free from chemical action on the iron filament. The completed barretter has a characteristic which shows a nearly constant current over a voltage range of about 3:1, as shown in Fig. 2. The resemblance of the first two-thirds of this characteristic to that of the gas-filled lamp (Fig. 1, curve (b)) is easily seen.
In universal and DC mains receivers the use of a barretter in series with the valve heaters, in place of a simple voltage-dropping resistance, may be regarded as normal practice to-day. One of its advantages is that it avoids the necessity for change of tapping in order to adjust the receiver for mains of different voltages between, say, 190 and 260. It also tends to be more compact than a resistance of similar wattage, since its filament runs at a high temperature; but it must not be forgotten that it is actually dissipating the same power, and therefore requires good ventilation to avoid damage either to itself or to neighbouring components. A typical example of the use of a barretter in this way is to be found in the 'DC Quality Amplifier' (The Wireless World, February 24th, 1938).
The action of a barretter seems at first rather mysterious, for its purpose is to maintain constant current, and this it is supposed to do owing to the fact that an increase of voltage causes its temperature, and therefore resistance, to rise rapidly; but its temperature can only increase as a result of a greater current flowing through it, and this increase of current we want to avoid. The fundamental point to consider, however, is that the heat lost from the iron wire by cooling must always be equal to the heat generated in the wire by the current. The factor which helps is, then, that the heat generated is proportioned to the square of the current and to the resistance of the wire, and as the resistance is increasing with the current, the heat generated is proportional to the nth power of the current, where n is quite a substantial number. Consequently, the change of perhaps 5% in current, which is permissible over the working range of the barretter, can produce a very much larger change, of the order of 100%, in the temperature and resistance of the barretter filament, and this suffices to keep the current variation down to its low value over a good range of applied voltage. It is only over a certain temperature range, however, that there is a suitable relation between heat input and output. Both above and below this range (which is just below red heat) the current increases more or less in proportion to the applied voltage.
|
