Introducing... The "Differential Three"

Radio Record, Volume IV, Issue 33, 27 February 1931, Page 16

Introducing... The "Differential Three"

Adding a Screen-grid R.F. Stage to the " Differential Two" _

By

CATHODE

SiN proceeding to the, threevalye version of the "differ"ential" sexies, the writer has again sought to avoid changes to the existing portion as far as possible. ~ _- Reference to. the theoretical diagram -of fig.’ 1 will show that. the general design of : the existing -receiver remains unchanged, except for one thing-namely, that. instead of, the: aerial lead . connecting with a-small coil inductively ‘coupled to the tuning coil, the output of ‘the high-frequency stage . is conveyed via & small: fixed condenser direct to the grid‘end of the tuning coil. The reason for this change is not difficult. to see. . In ordér to secure maximum ‘amplification,- a screen-grid valve is used for the radio-frequency stage, and it is a condition for the successful operation of this type of valve that the impedance in its: plate circuit shall be very high; comparable, in fact, to the high impedance of the valve itself. Now, with the small and loosely-coupled aerial coil in the plate ¢ireuit ‘of the screen-grid valve, the impedance would be quite low, so that this change is very necessary. With the new connections, the plate impedance is practically that of the tuned circuit as, for reasons which have previously been discussed in these columns, the parallel path presented by the high-frequency choke is of so high. an impedance as scarcely to affect matters at all.

. There is very little else that need be ‘discussed from the theoretical standpoint. : The rather unusual tuning arrangement Whereby a small fixed con: ~ denser may be plugged in in series with the tuning condenser so as to reduce its capacity was fully discussed in ¢on‘nection with the "Differential One." The reader would be well advised: to look .out the issue of December 26 again and refresh his memory -on this little -point.. The! resistance R interposed in the negative filament lead to the .screen-grid' valve is intended not so much to reduce the filament voltage applied to'this valve (although it has this, effect) as to provide .automatically a small negative bias to the control grid. Owing to the very high amplification factor of the usual type of- -sereen:. grid valve, it is found that, except for very high plate voltages, the negative bias of 14 volts, which ig the minimum that can be provided by the CO battery, is too high for optimum results. No harm is likely to result from reducing the filament voltage a trifle, however, and this drop in voltagé: can appropriately be employed for providing negative bias. It might perhaps be

more accurate to say that the filament is ‘positive with respect to the grid, but the effect is identical. The operation of this means of obtaining a negative grid bias will be better understood from an actual ex-¢ ample. Suppose that we are. using a 6-volt valve which normally takes a filament current of 0.1 amps. Such a valve will operate just as well with an actual filament voltage of 5.4. Now, if we interpose a resistance of 6 ohms in the negative filament lead the actual , voltage on the filament is reduced to 5.4 and the drop in voltage across the resistance automatically provides a grid bias of 0.6 volts. A 4-volt 0.1 amp. valve might be provided with a resistance of 4 ohms, which would reduce the actual voltage weross the filament to 3.6 and provide a bias of 0.4 volts. Valves of the 2-volt type usually take a filament current of: 0.2 amps., so that-a resistance of 1 oh suffices to give a bias of 0.2 volts i this case. , It is not necessary to use a proper filament resistor for-this purpose, as no great accuracy is necessary. If a scrap of resistance wire of approximately the required resistance can be secured from an old rheostat or fixed

resistor this can be used as the actual connecting wire between the negative filament. terminal.on the yalve-holder and the A lead. It will be seen that this arrangement is shown in the praetical wiring diagram given in Fig. 2. Screening has been reduced to the absolute minimum. It has been found that the inexperienced constructor is not at all happy about assembling and wiring components in the confined space of a screening box. Moreover, with the extremely short leads employed, there is no difficulty with regard to oscillation when the ordinary type of screen-grid: valve is used. It is just possible that, with an unusually efficient valve, oscillation might be experienced. .In such a case, however, all that is necessary is to interpose a resistance of a hundred or two hundred ohms in the lead to the control grid of the screen-grid valve. ‘This will at once qnell the oscillation without having any adverse effect on "amplification. Do not make the resistance any higher than is necessary, however. The screening actually consists of a' simple flat vertical shield six inches high and of the same width as the baseboard. Hither copper or aluminium may be used, the latter having the advantage of not corroding. It is, of course, permissible for the -advanced constructor to use more complete screening if he so desires. The constructional details which would otherwise be necessary have to a large’ extent been covered by the previous articles on the "Differential" receivers. Thus it is unnecessary to describe the little plug-and-sockcy system for bringing a fixed condenser in series with the tuning condenser, as this has been fully explained in the issue of. December 26, Similarly, the necessary coils are exactly similar to those previously described except

— iat the reaction winding L8 should omitted, both on the broadcast coil don the short-wave coils, Constructors will have noticed that the aerial coil Li on the. old coils is not used now that the high-frequency stage is in service. This winding may be removed from the coils if desired, although it will do no particular harm if left on. Another point which will doubtless be observed is that it would have been possible to use four-prong coil formers and sockets "instead of the five-prong type. It has been thought wise to retain the fiveprong coils, however, so as to avoid any possibility of the coils being inserted in the sockets intended for the valyes or vice versa. Doing the Job. ‘THE addition of the high-frequency stage will take but a short time. The first step is to remove the existing leads from aerial and earth terminals to the coil socket. ©Do not interfere, however, with the lead from the frame of the tuning condenser to the coil socket. "The metal screen ‘may: then be serewed down to thé baseboard, a flange being bent at the bottom for this purpose. It is as well to drill a couple of holes through the screen before it is mounted, one low down for the filament lead to the hf. yalye-holder, the other fairly high up .80 "as. to come as near as possible to the cap (ie, the plate connection) on’ the" screen-grid valve. If these .holes can be fitted with insulated bushings so much the better; otherwise the leads may be protected by insulating sleeving. The high-frequency choke may now B mounted in the position shown. One end is connected to the B- terminal to which other connections are already made; the other end is fitted with two leads, one a piece of: flex terminating in a tag or clip for connection to the plate of the screengrid valve, the other a stiff connection to one terminal of the 0.001 fixed condenser C3, the other terminal of which is connected to the G terminal of the existing coil socket as shown in Fig.- 2. This fixed: condenser is _ supported solely by its wiring, which must thus be fairly stiff. The remaining components must now be. mounted and wired up in accordance with fig. 2, The lead from the P terminal of the valyeholder (really the connection to the screening-grid now) is passed under the baseboard to the second B-++ terminal (ie, the one which has remained unused up to the present), the voltage to be applied to this terminal should be that recommended by the makers of the §&.G,

a valve for the screening-grid; usually about 60 volts. | The lead from the A- terminal should make good contact with the shield. From there it continues to the "earth" terminal, from which one lead goes to the new coil socket and tuning condenser, and another lead,

consisting of a short piece of resistance wire, makes connection with one of the filament terminals on the valveholder. The other filament terminal on the valveholder is supplied from a convenient terminal on the old coilsocket, a short: insulated lead passing through the screen. The two or three remgining connections are so clearly shown in fig. 2 as to render further description unwarranted. The screen-grid, valve must, tnless very substantial modifications. are made, be of the Hnglish type, having the plate terminal at the top. This is. important. With this limitation, the valve may be of any good make, bearing in mind the remarks made earlier in this paper anent the filament’ resistance.

Trying Out. ITH a suitable screen-grid valve and both coils in position, the receiver may now be given a try-out. The improvement: on broadcast ‘wavelengths should be.very marked indeed, and the distance-getting qualities of the receiver will be appreciated. On the short waves the increased amplification will not be so apparent, but there should nevertheless be a substantial improvement. Those listeners in close proximity to the local station may. find some difficulty in clearing it; to such we recommend the use of a very successful wavetrap to be described in the 1931 "Guide." The method. of coping with unwanted oscillation in the screen-grid stage in the unlikely event of this occurring has already been dealt with. It will occasionally be found that the removal of the aerial damping from the detector coil renders the reaction control itself less smooth. It is highly unlikely that this will occur, too, but if it should it is well to remember that the original smoothness ean be restored by slightly reducing ‘the number of turns on the reaction eoil. As a final word of eaution, see that the high-frequency choke is above reproach. This is perhaps the most important component in the whole reeeiver and an inferior article will spoil the performance of the set’ completely.