The Application of the Screen-grid

Radio Record, Volume II, Issue 36, 22 March 1929, Page 28

The Application of the Screen-grid

Essential Precautions the Constructor must Observe

By

PENTODE

@] 1 has come to the notice of the writer that several constructors have tried out a screengrid valve in their sets without any further alteration of the wiring or coils beyond moving the grid or plate lead and applying the potential to the screen grid. This is all right when done out of pure curiosity to see what will happen, but to expect all the results that the screen-grid valve is capable of giving is bound to end in disappointment. There are many who have heard of all the wonderful results obtainable from the four electrode valve, but who possess commer-cially-made sets and do not care to interfere unless they know what they are doing. The object of this article is to give, if possible, that little help where the writer considers it to be most needed. No doubt it will also help to some extent those who are about to design and build a receiver incorporating this valve, and one or two circuits are given to be used with the screen-grid valve. Without going too deeply into the theory, it is well known that the impedance in the anode circuit must balance that of the valve. Maximum amplification is only obtained in a radio frequency stage when these two impendances are equal. If a valve having a low impedance is inserted in a valye socket of a receiver in which the impedance in the plate or anode lead is high, then the full amplification of the valve is not obtainable. It would need a long article to explain just why this is so, but it can be taken for granted by the reader. _

The Evolution of the Valve. BEFORE dealing with the screened grid valve, let us go over one or two points dealing with the progress of wireless. In America, when broadcasting first began, it was considered advisable to standardise as far as possible, all component parts. ¥irst of all the valve was standardised and the 201A type produced. It was aimed to have one valve suitable for every part of the receiver. Now, to be suitable for audio frequency work the impedance had to be comparatively low. Using this valve in the radio frequency stages, designers had to use a coil of low impedance in its anode, and so the number of turns averaged between five and ten. Fortunately the number of turns for a broadcast coil is usually about 60, and so a voltage step up was possible in the radio frequency transformer, and partly compensated for the low amplification of the valve itself. At the same time in Pngland valve manufacturers were designing valves with amplification factors of 25 and 30. To do this the impedance of the valve was raised. A larger R.F. coil which possessed no step up effect was used in the anode of these valves, and all the amplification had to be produced by the valve. The tuned anode method of coupling was evolved. Now compare these two methods. As far as actual amplification was concerned, both gave about the same, but of the two the American method gave far greater selectivity. All commercial — -as

machines adopted transformer coupling in their R.F. stages and were designed for the 201A valve. The substitution of a higher impedance valve resulted in a serious loss of amplification. Fully 95 per cent. of the receivers in New Zealand are of American design, and it will be now clear why it is impossible to use a screened grid valve without fairly extensive alterations. These valves have an impedance of 150,000 ohms, and it is not likely that it will give anything like full efficiency when substituted for a valve of 6,000 ohms impedance. To get the best out of these valves the anode coil has to be tuned and wound so as to reduce all losses to a minimum. Now look at it from another angle. In a circuit using the three electrode valve the inter-electrode capacities have been balanced and neutralised out. All undesirable capacities in wiring, ete., have been cancelled by the neutralising condenser, but the screened grid valve employs no such devices, nor does it need any, and efficient wiring and screening are needed to exelude all tendencies for the valve to oscillate. The Browning-Drake. N general practice a single stage of screened grid is comparatively easy to build and operate. The use of two stages, however, calls for great care in design, and anyone contemplating substituting two four-electrode valves for two ordinary triodes is steongly advised to either redesign the whole receiver or not to attempt the work. Sooner build a screen-

ed grid "H.F. Booster’ as described in this article, as it adds another stage without the addition of further tuning dials. To owners of Browning-Drakes and other receivers using but a single stage of R.F. the following will make interesting and instructive reading matter, especially if they are thinking of trying out one of these new valves.

Glance at Fig. 2 and it will be seen to be the circuit diagram of the radio frequency and detector stages of the well-known Browning Drake. Neutralisation is effected by taking a tapping from one of the turns of the secondary of the regeneroformer. .The aerial connection being indicated as an arrow indicates that this can either’, be a tapping direct to the grid coil or a small primary adopted. Reaction on the detector valve is obtained by a small adjustable tickler coil. The primary coil inside the regeneroformer is wound with the correct number of turns to suit the valve preceeding. It is this circuit that we desire to alter in the easiest way, so that a screened grid valve can be used in the R.F. stage. Changing Over. HF first essential is to note the performance of the present receiver and before thinking of changing over make quite certain that the following points are quite in order. First and foremost, does any interaction take place between the aerial and regeneroformer coils? "Heaven knows!" will be the answer in most eases, but it can be soon ascertained by the following process. If the set will not nentralise at all it is one of two causes. Hither the neutralising condenser is of the wrong value of capacity or there is an interaction taking place between the two tuning coils. If the receiver is unstable, particularly on the lower wavelengths, it is usually due to an unsuitable neutralising capacity, either too much or too little. Turn the radio frequency valve down and adjust the neutralising condensers until weakest sig-

nal strength is noted. If, on turning on the valve, the receiver is unstable on the higher wavelengths, then the troubl¢ean usually be traced to interaction heétween the two main tuning’ coils. If this is the case, it must be definitely corrected before going any further, Although, by careful arrangement, the coils can be placed so as to prevent interaction, the writer advises screen-

’ ing. It would be easiest to screen the aerial coil, but in this particular circuit it is preferable to screen the one on which reaction is used. If this inter-coil coupling is allowed to exist, it will assuredly cause trouble later on, so it is best to make sure in the first place. Whether it is desired to change the valve or not, this practice of screening the regeneroformer will be found to improve the working of an unstable Browning-Drake; so howlers, please note! .

If, after screening, reaction is difficult to regulate, either increase the B+ det. voltage slightly or add a few more turns to the tickler coil. Now let us turn our attention to Figs. 1 and 8. These are two circuits of a single stage of R.F. using the screened «grid valve. It will be seen that Fig. 1 merely uses the tuned anode method of coupling instead of the R.F. transfor-

mer. The screening grid is given a positive potential and also needs a bypass condenser of at least .5 mfd. The grid leak on the detector valve, instead of bridging the grid condenser, runs directly to A-++, otherwise the high voltage potential from the B_ battery would be impressed on the grid. Reaction has been omitted to make the diagram more simple. This circuit is the most efficient. Fig. 3 is a diagram of a modification of the preceeding circuit, and is the

x one most readily adaptable for the average set. A radio frequency choke in the anode of the S.G. valve allows the direct current to flow, but offers a more or less complete barrier to the radio frequency oscillating. These pass through the small coupling condenser to the grid of the detector valve. No alterations to the grid leak are necessary, as the coupling condenser pre-

vents the direct voltage from passing. Compare this circuit with that of Fig. 2 and points of similarity will be apparent. Instead of the primary windind, a R.F. choke is used and the coupling is obtained via the small condenser. If a 8.G. valve is used with the plate, connection being made on the cap on top, the change over is quite simple, the only additions to the set, besides the valve, being a R.F. choke, a bypass condenser and a mansbridge condenser of at least .5 m.f.d. capacity. These should be screwed down in convenient places so that all wiring will be kept short. A word about the R.F. choke. If the best results are to be obtained this must be of good quality. Generally speaking, an English R.F. choke will be found to be suitable, as the only purpose, generally speaking, for which a choke is used in America is to obtain reaction by the throttle control method, in which case this component need not have such a high degree of efficiency. Constructional details for making one’s own chokes have been given in the "Record" from time to time (eg., caecyainnansaaes

December 21). The control grid lead is connected to the same pin, as in ordinary. triods, and no alteration is here necessary. This is, of course, speaking of the valve on which the plate connection is on the top. Certain makes differ in this respect, and allowance will have to be made when witing up. The plate connection of the valve socket corresponds to the screening grid, and this will have to be taken from the regenroformer primary and connected to B-+-45 and the .5 m.f.d. condenser. With a length of flexible wire connect the top cap of the valve to one end of the R.F. choke and also to one side of the coupling condenser. The free end of this condenser joins to the grid end of the detector grid coil, The other end of the R.F. choke goes to B+135 and to one side of a 1 or 2 m.f.d. Mansbridge condenser, the other side of which is earthed. It is really quite simple to do the change over, but it is hopeless to try without sticking to an approved circuit. If, when completed, the receiver is unstable, one of the coils will have to be screened, preferably the one ems bodying reaction. en |

Tuned Anode Coupling. Y=2t another method of coupling is shown in Fig. 4. This is really better than the one using tuned anode, Fig. 1 and the diagram Fig. 3. The tuned anode suffers from one drawback. It often suffers from objectionable body capacity effects. The best cir-

cuit of all would be one using a 1 — 1 ratio R.F. transformer with’ both the primary and secondary tuned. This would add too many tuning controls, but it has been found that if two coils are 4oupled very tightly together and one tuned, the other also has the characteristics of the tuned coil. So

if the primary and secondary were close wound, each having the same number of turns, the secondary being tuned, a very efficient coupling would be produced. In practice this can be accomplished by space winding one eoil and winding the other coil in the spaces. Fig. 4 would be an excellent

circuit to try for anyone who would go to the trouble of preparing the coil. For broadcast band, 60 turns on each coil, one being tuned with a .0005 m.f.d, condenser would be found suitable, Reference was made for a screened grid booster earlier in this article, and. Fig. 5 gives the theoretical diagram for this. The whole unit can be made

in a very compact form and stand alongside the mother receiver. Two R.F. chokes are employed. Both must be of good quality and very efficient over the broadcast band, as the whole unit depends upon this. The terminal marked "to grid of first valve" must be connected by a length of wire, as

short as possible, to either the grid tuning condenser. The wire, to lessen direct or to the fixed vanes on the first capacity effects, need be no larger than 24 or 26 insulated, and, if necessary, to shorten the length can be led through a hole drilled in the end of the cabinet.