DIRECT-COUPLED AMPLIFIER

Radio Record, Volume III, Issue 48, 13 June 1930, Page 28

DIRECT-COUPLED AMPLIFIER

ee Loftin-White System FULL CONST RUCTIONAL DETAILS

By

Megohm

EFADERS have already been shy igiven an outline of the excelPAilent qualities of the LoftinWhite system of direct-coupled af. jamplification. By its use a piwéy'icain of 300 is obtained by the 26 2) |employment of two valves only -a UY224 and UX245, so that ample volume is obtained from any satisfactory pick-up. The amount of apparatus with which this result is attained is comparatively small, and thus expense is kept down. It is well to remind constructors that as in other amplifiers, the last: valve governs the volume of undistorted output. For the average constructor and listener this amplifier is a revolutionary change.- Because, combined with simplicity and compactness ,it gives even amplification over.a greater range than the ordinary working frequencies. With the 245 valve it gives volume and quality at least equal to that of a 210 valve, which is always considered as giving high grade results in connection with an ordinary receiver.. With a 250 valve in the last stage, still greater volume may be handled. The disadvantages of audio transformer amplification are numerous, in-

cluding self-capacity, wave-form distortion, hysteresis, saturation effects, resonance, ete., and though research has found means of reducing. many of these, the presence of iron in the circuit is always recognised as a drawback. , Resistance-coupling has its own disadvantages, but they are not so numerous. The new system eliminates most of these troubles, and its achievements depend chiefly upon the performance of the last valve and the loudspeaker. These two latter are now sufficiently improved at least to do some justice to the, new system-when a screen-grid yalve is used for the output, a flatter characteristic results, but a large power valve leaves little to be desired, so far as reproduction outside the laboratory is concerned. The Apparatus Reviewed. SMALL power-pack comprises the greater part of the total apparatus, and enables the amplifier to be completely operated from the a.c. mains, whilst if so desired may also supply: filament and high-tension current for the r.f. stages of a receiver.. An illustration gives a general idea

of the amplifier as now being retailed in commercial form by Messrs. Fear & Co., of Wellington, who also supply kits or parts for home assembly. The compact size as indicated by the measurements should be noted. At the left is the screen-grid valve, the 245 in the centre, and at the right the 280 fullwave rectifier. Behind: the screen-grid yalve is: a knob which is rotated *o balance out the hum, and the metal box behind the valves contains the power-pack. The wiring, resistances, and by-pass condensers are under the metal base. It is not likely that the amateur eonstructor will build his power-pack in so small a space as that occupied by the commercial product, especially if current is also to be supplied to r.f. stages,. but it need be no less efficient on that account. It is absolutely essential for the power-padk to be completely shielded with metal, and if an aluminium chassis is used the cover for the power-pack may be of sheet-iron, which is the best for the purpose. The wiring lay-out shown ‘should not be departed from to any great extent because long leads are to be avoided as much as possible. Operation of the Circuit. AN idea of the working of the circuit is obtained by examining the fundamental circuit, in which batteries are used for illustration, and the chain of resistances is represented by the one resistance, R. It will be seen that the circuit of the "B" battery is completed

through the valve from plate to filament, and through the resistance, as shown by arrows.. | The upper end of the resistance connecting to the negative pole of the pattery is the most negative portion of the whole circuit. ‘The battery voltage is distributed across the internal resistance of the first valve, and across the fixed resistance. There is a voltage drop in the resistance when plate current flows, whereby the lower end is more positive than the upper, and this places’a negative bias on the grid of the next valve. , Variation in signal strength in the first valve causes its plate impedance to vary, as there is a continual change in voltage distribution ‘between valve and the resistance, the value #f the latter being constant. When the plate impedance is large, the drop across the resistance is small, or the opposite condition may prevail, so that this continual variation is being impressed upon the grid of the second ~ valve. Only a very low voltage is required upon the plate of the first valve, a condition which reduces possible, trouble from microphonic noises, ete. Referring to the full theoretical diagram, the path of the plate current for the power-valve may be followed from the plate to the filament to point A, where it divides, part going through the .5 meg. resistance to the plate of the 224, whilst the greater part returns to B minus through the resistors to the left. By either route the voltage drop will be equal.

A bias is developed across the 50,000 ohm resistor with respect to the grid of the 224; this bias is too low, about 14 volts, but is balanced up by the 425 ohm resistance, giving an opposing voltage of about 12 volts, leaving a working bias of about 2 volts. This resistor is not included in the wiring diagram shown, but is shown in a separate diagram of this portion, showing the flat type of resistances, and should be included if the amplifier is to be used for radio reception, which in most cases it will be. Another justification of its inclusion is because the home-built powerack will give a slightly higher voltge than the amplifier only, so that other plate supply may be taken without unduly dropping the voltage. In«clusion of this resistor will slightly , raise the voltage applied to the screen Of the 224. An important feature of the system is the self-adjusting characteristic which automatically adjusts bias on the 224, In radio reception a strong carrier would increase the plate current, increasing the drop across the

.5 meg. resistance, making the grid of the 245 more negative, and reducing its plate current. With the 425 ohm resistor in circuit, the 245 decreased plate current is flowing, which results in a decreased positive bias to balance that developed in the 50,000 resistor, so the actual bias on the 224 is increased, and conditions revert to normal. For pick-up: reproduction, the amplifier operates with one side of the pickup connected direct to the grid of the 224. For local reception a tuned coil with loose-coupled aérial may. be used for selectivity, or if this is not desired. a single-circuit tuner may be used. When adding r.f. stages, certain precautionary measures have to be taken, and these will be dealt with later. [t appears strange at first that it is possible to conductively couple the plate of the first. valve to the grid of the second, when the first is positive and the second is negative. They are really at the same potential, but the plate of the 224 is positive, and the grid of the 245 is negative, each to its respective cathode, each latter being maintained at a different potential. As this amplifier is amplifying at 50 cycles, there must be no hum. The 50,000 ohm cathode resistance has to be by-passed, so the condenser is made to act two parts at once, by selecting a point on the series of resistances where the phase of the mains hum opposes the phase of the hum on the grid; we can connect the high side of the cathode resistor through a 1 mfd. condenser to this point, and reduce the hum to practically inaudibility. The required position is a little above the grid return, so a 200-ohm potentiometer is introduced and easy variation obtains. The exact point is sometimes critical, and should not be passed over quickly. Once set, the position should remain correct, only to be readjusted to suit another valve. By properly proportioning the voltage on the first valve we have a very sensitive detector that will. handle heavy signals without overloading. It will be as sensitive as a grid-leak detector, and will carry a greater load than an ordinary power detector. As long as the screen-grid does not draw grid-current, its input resistance’ is very high, and it is only when input signal voltage approaches in magnitude the negative bias value that grid-cur-rent may be drawn. With the automatie bias the grid is kept negative even on strong signals. In actual prac

tice the 245 will overload before the grid of the 224 becomes positive. Thus we have conditions that are almost lacking in the usual detector circuit. The voltage gain of the 224 and 245 combination has been stated by a good authority as being 450. ° "THE standard Loftin-White circuit is given in its entirety, wherein will be noticed slight differences from other published circuits, but. whilst small changes are made, chiefly to suit a difference of voltage, the general: principle remains the same. The object of the condenser connected to the potentiometer arm is to balance out the hum, and this it does very successfully, making the output "dead quiet" so far as hum is concerned.

The 224 acts as an amplifier, but also is a detector when required, and when preceded by rf, stages, functions as the detector, . With two screen-grid r.f, stages, ex. cellent results are obtained. When a Screen-grid r.f. stage precedes the 224 detector-amplifier, a by-pass condenser of at least .02 mfds. should be Placed across the low potential end of the secondary coil and "B-." ‘The bias connection from the high end of the 425 ohm resistor connects to the so udary coil to bias the grid of the 224. For the operation of the amplifier for radio reception a little care ig required to prevent feed-back coupling between the output and the aerial. (Continued on page 380.)

Material Required ney 1 UY 224, L UX 245 1 UX 280, Rectifier. 2 Condensers, 1 mfd., 500 test; two 800 test. _ & Condensers, 2 mfds., 800 ‘ test. \. 1 Potentiometer, 200 ohms. 1 Wire-wound resistance, 5500 ohms. 1 _ Wire-wound resistance, 425 ohms. Resistances, 100,000, 50,000, 25,000, .5 meg.; and mounts for same. 2 UX sockets, 1 UY, Connecting wire, 6yds. Output transformer (?) 4 Terminals. ; 1 Balancer resistance, 30-60 ohms. Stalloy, 2 doz. lin. by 3ft. or) stock size, eut (transformer). Stalloy, 2 doz. lin. by 3ft. or _ Stampings (choke). 1 Ib. 36 swg. enamelled wire 2lb. 30 swg. enamelled wire. 10 doz. 18 dee. 12lb. 32 swg. enamelled.

The L.W. Amplifier (Continued from page 29.) Shielding of the last valve, short or shielded output leads, and other similar precautions should be attended to. The inclusion of a r.f. choke in the output circuit as shown in the theoretical diagram is a simple precaution. The .0005 fixed condenser shown dotted may help matters. Alternatively, it may run direct to the shielding instead of to the 245 centre-tap. An output transformer is shown in the diagram, but its use is governed by the same rules pertaining to all amplifiers. Its use is recommended, even with a magnetic speaker, because 30 m.a. is far too great a current to put through speaker windings. (To be concluded next week.)