Screened Grid Set

Sun (Auckland), Volume I, Issue 309, 21 March 1928, Page 14

Screened Grid Set

DETAILS OF CONSTRUCTION AS promised last week, the design for a successful 4-valve receiver is given hereunder. The results were sufficiently good in the final set to warrant its recommendation to all who wish to experiment with this latest development. One can conservatively claim it to be as far above the average Browning-Drake as the B.D. was above the average 5-valve set.

key to success is in the coils. Taking the theoretical aspect first, the amplification gained depends on the impedance of the output circuit and increases as the impedance increases. A few figures obtained by W. James, a well-known radio engineer in England, will explain this. For example, with an impedance of 600,000 ohms the amplification on a calculation basis is 90, with 300.000 ohms it is 67, and so on, down to an impedance of 10,000 ohms, which gives an amplification of only 10. Having got this basis to work from it becomes necessary to design the coupling coil in the anode circuit. The coil has an increasing tuned anode resistance as its ordinary resistance is decreased, and with a coil made of Litz wire to cover the broadcast wave band, an impedance of 250,000 ohms may be obtained. Various coils were tried, about 12 in all, before the theoretical aspect was considered, a foolish and time-wasting procedure, but desire to try results must be pleaded as the excuse. Starting off with a space wound coil on a 3in. former with 20 gauge enamel wire, the results were decidedly promising, so a move was made in the direction of a dealer and a pair of screened formers and cans were purchased and carefully wound with silk covered wire. The results were most disappointing. Next some salenoids were made on the same formers, but space winding was used, an improvement which was lost on putting the cover on the can. The effective tuned resistance of these

coils, which give good results with ordinary valves of the G.P. or small power class, is too low for the new valve. Plain solenoids spaced and unspaced were tried with varying results and the writer eventually ran to earth a reel of L.E.W.C.O. Litz wire and from that moment the set has gone ahead. Mr. James in his laboratory results gives 100,000-150,000 ohms for a wellmade coil using solid wire, i.e., an amplification of from 30-40, about. The plug in a commercial coil or a coil wound with fine wire, has an effective resistance of 40,000 to 50,000 ohms, and a consequent amplification of 19, which results would have saved much time and money if the job had been started at the right end. The aerial coil may be an ordinary spaced wire solenoid, but if Litz is available it should be used. The valve was mounted by using valve pockets placed in strips of bakelite and fitted with slotted brackets to enable the valve to be removed. An aluminium screen was made with a hole through which the valve fitted quite snugly, the three-pin end being on the aerial side, and the two pin end on the detector side of the screen. By this means only filament leads go from one side of the screen to the other, and if a metal panel is used (as it was) this was made the negative lead, and for the X lead a piece of flex was lead under the partition which is, of course, at earth potential, and' the valve is placed so that it is in the same place as the screen in the valve and this method gives sufficient screening witnout loss. The coils were: LI, 10 turns 26 D.S.C. on small 2ln. former, which slipped inside L2. L2 50 turns 27/42 Litz D.S.C. L 3 50 turns 27/42 Litz D.S.C.

L 4, 12 turns 36 D.S.C. wound on 2in. former, and mounted on a spindle to rotate |in. away from the end of L 3. The tickler method was chosen as being the simplest and if carefully made gives a nice smooth reaction. Cl and C2 .0005 low loss condensers fitted with vernier dials. A point to be ‘noted is that the grid leak must come, not across the condenser, as is usual with the R.F.

transformer, but to the X filament at one end and y the grid at the other. The detector and R.F. valves were controlled by the same rheostat as the detector used is not in the least critical, and this gave an excellent volume control. The components used in the final set were: Detector valve, P.M. 6. First audio valve, P.M. 6. Second audio valve, P.M. 256. Radio audio valve, S 625. Cl and C2, .0005 general radio. Audio transformers, R.I. multiratio. Valve sockets, general radio. Coil former, ribbed ebonite 3in. Rheostats, R.I. dual. Dubilier leak and condenser. The aerial coil was mounted on a clip base so that it could be pulled out and a loop plugged in. The valve was tried as a detector, but was not remarkable. Recent journals show remarkable results using the UX 222 as a first audio, and the writer hopes to give data for using this stage in conjunction with the present R.F. stage. A point with regard to the use of Litz wire is that very great care must be taken to pick up every single turn of the fine wire. If a single strand is not picked up the effective resistance in the tuned circuit will be so lowered that almost any inefficient coil could replace it. One last point of interest is that the use of a small aerial primary inside the secondary of the R.F. grid circuit is the increase in selectivity obtained. Mr. C. R. Russell has several times mentioned this method in the sets he has published in the Christchurch “Sun,” and the result obtained is certainly worth while. Of course, there is no necessity to use the components mentioned as any really good article will give satisfaction, j Messrs. Radio, Ltd., informed the

writer that the Cossor screened grid valve is expected shortly, as are also samples from the other makers. THE NEW VALVE Another aspect of the screened grid valve is its use as a first stage amplifier after the detector. Here the screen is used as a space charge grid and biased to the extent of about 1£ volts negative, while the inner grid is placed at a potential of about 22£ volts positive. The amplification factor when used in this connection with resistance or impedence coupling is in the neighbourhood of 150. A plate voltage of about 180 is necessary on the plate, and the external screen in the case of the UX 222 is essential. Through .the courtesy of Messrs. Abel Smeeton, Ltd., The Sun was able to get a sample of the UX 222, and a rough amplifier was connected up, using a circuit given in the March number of the “Radio News.” The difficulty is getting the correct coupling impedance. This journal uses an autofcrmer made by connecting the primary and secondary of an audio frequency transformer in series and they recommend the use of a special transformer as being essential for getting good results. Though in most cases any similar unit of good make will do, there seems some foundation in the claim in this instance, as a multi-radio R.T. transformer was used and the results, using various ratios, showed a very marked difference. The volume was excellent, but considerable experiment is required in the matter of C voltages and resistances in the coupling

circuits. Its use in America in short wave sets is quite the vogue, and the writer heard that Ivan O’Meara, the well-known Gisborne amateur, is now being heard by many American short wave enthusiasts to whom he had previously been inaudible.

A NEW RECEIVING CIRCUIT At a recent meeting of the American Institute of Radio Engineers, Dr. Frederick K. Vreeland read a paper outlining the basic principles of a new development which permits reception of the whole of the side bands transmitted from a broadcasting station without loss of selectivity.

It is well known that the ordinary circuits designed to give selectivity do so at the expense of tone. A broadcast signal of speech and music includes each side of the main or carrier wave frequency, modulation bands having a width of about ten kilocycles, assuming the highest speech frequency transmitted to be 10,000 cycles, Thus the total width of the transmitted band is 20 kilocycles. Now in an ordinary circuit the response curve rises to a sharp point, the maximum being at the carrier wave frequency, when the circuit is accurately tuned. Thus ultra-selectivity will neglect these side bands to a greater or lesser degree, and quality and tone will be sacrificed. In the case of ordinary wire telephony, filters are generally used to separate adjacent frequency bands. Filters could be used for wireless work if the bands were definitely fixed, but it must be remembered that in the field o t broadcast reception it must be possible to adjust the band of reception in the frequency scale, or to “tune” the receiver. The circuits must therefore be simple, and preferably adjustable by means of a single control. Dr. Vreeland’s invention makes use of a “band selector” for which fidelity, selectivity and simplicity are claimed. In essence it consists merely of setting up two small and extremely simple balanced circuits in front of each radio stage or in front of the detector if there are no radio frequency valves. The principle comprises a system of reactances so related to each other that they are mutually balanced, not merely at a single frequency, as in the case of the ordinary tuned circuit, but also at any frequency within a given band. At any frequency outside this band the reactances are not balanced; the unbalanced reactance is high and the band selector unit cuts off very sharply at the limits of the band. Roughly, the unit consists of three coils, A, B and C, all connected together at one end, and the centre one, B, connects to the free ends of A and C, via two capacities operated on the gang principle by a single control. The two condensers have small compensators in parallel to bring them into resonance. A, B and C are inductively coupled to each other, A is inductively coupled to the aeriaJ coil, and C to the periodic grid coil of the radio frequency or detector valve. These Vreeland circuits are said to be free from existing H.F. circuit patents and several prominent manufacturers are negotiating for the operating rights.