Practical Hints on Neutralising

Radio Record, Volume II, Issue 26, 11 January 1929, Page 26

Practical Hints on Neutralising

Considerations when Building

By

Pentode

"| N a previous issue of "Radio Record" the writer gave a short article on the various methods employed for reducing the feed-back produced by interaction between grid and anode coils in a ree@xver. This article will describe, in, it is hoped, understandable language, the effect and remedy of feed-back due to the close proximity of the valve electrodes. If a receiver was designed and built, taking great precautior: to shield and isolate the grid and plate coils, it would still be very unstable ahd would oscillate furiously whenever the tuning systems were in resonance. Oscillation cannot occur without a feed-back, whether inductive or of a’ capacity, and excluding possible causes through the various battery leads common to both coils, we only have the small capacity existing between the grid and plate inside the valve. It is this small capacity forming a coupling between the grid and plate coil through which the feed-back takes place. This has been realised for years and has given food for thought to many radio engineers. Controlling Oscillation. ERHAPS the easiest way to understand the present methods employed to overcome this difficulty is to recapitulate the various courses prac- | tised during the past few years. Assuming that the reader has already read the article in last week’s issue, it will be remembered that a valve will start to oscillate if the field of the grid and plate coils interact. This is known as an inductive coupling. Methods were given for overcoming this interaction by placing the coils at certain angles and also by screening. | Also a method was given for suppressing oscillations by the addition of | damping resistances.

Yet another method which should have been enumerated in. the previous article is the circuits employed in potentiometer control. Just why this controls the valve will interest many — readers, and the reason really lies in the fact that when the grid of a valve | is given a positive potential the resist- |

ance of the grid-filament coil is greatly increased. By resistance the highfrequency resistance is meant. Consider for a moment, the resistance between the filament’ and grid electrodes inside the valve. When the grid is negative the resistance is infinite; when positive, the resistance decreases dependent upon the potential applied to the grid. By referring to the circuit of any high-frequency iunplifying valve it is at once apparent that any resistance, or should we say, partial conductor, between the grid and filament electrodes is directly across the two ends of the grid coil. Thus, with a positive potential on the gsric a damping resistunce is intro-

duced across the grid coil. This resistance is susceptible to variations of erid bias, accomplished by connecting a high resistance, usually 200 or 400 ohms, directly across the filament leads and arranging a sliding arm to run along this resistance. To this arm is connected the grid return.

SUC In practice the grid is made more positive until a balance is reached. The damping introduced exactly balancing the energy fed bac by either interaction between anode and grid coils, or the small interelect ude capacity of the valve itself. It can always be remembered that reaction or feed-back is merely a lowering of the resistance of the grid coil. If, there fore, resistance is added to more than compensate any feed-back present, the circuit will be quite stuble. But to introduce any form of damping lowers the efficiency of the circuit, and the greatest possible amplification per stage is not obtained. The Neutrodyne Principle. LATE in 1925 the neutrody:e principle was in’ uced, and. is still the most popular method c. stalilising a high-frequency amplifyin valve. It must be stressed here that the .eutrodyne principle merely overcomes the feed-back due to the small capacity existing between the srid and plate elect: odes and connections thereto, and does not remedy: any inductive coupling that may exist between grid and anode coils, To obtain a_ perfectly stable receiver by this means, the two

coils will have to be completely isolated by any of the methods enumerated. The neutrodyne principle depends for its action on the application of an equal capacity to that of the valve electrodes and connections only in the opposite direction. The current flowing through the added capacity is out of phase with that flowing through the valve. This is known as neutralizing or balancing the valve. This was first accomplished in th e well-known tuned anode circuit e taking the B positive lead from a t ping in the middle of the anode coll, and connecting a very small conden-. ser across the free end of the coil and grid of the preceding \..lve,. This circuit held the ground for many months, but suffered from .several' drawbacks. It was extremely unselective, and hand capacity was very mar'ced *..h inipulating the anode coil snd aser. These: disadvantages were overcome by using a high-frequency transformer in which the primary impedance. could be matched with that of the valve used and the tuning condenser of the secondary coil was at earthed potential. It is this arrangement, split primary . high-frequenc,; transformer that has-been universally used -throughout the world and is as popular to-day as eyer. The Elstree Six. N November, 1926, the Elstree Six was described and "became a very — notorious receiver in England at that"! time. This receiver, which later took third prize in the Radio World’s Fair at New York, has three stages of neutralised high-frequency amplification, only in this case it is the grid coil which is tapped and connected via a small capacity to the plate of the valve itself. Evolving from. this arrangement, numerous circuits ere published in English journals using the split secondary method of neutralising. This method, found i. perhaps fift,, per cent. of receivers of an English design. is undoubtedly the ‘easiest way to effect neutralisation, but it suffers from one drawback: It is not quite as sensitive as the circuit employing the split primary. ‘The vo.tage applied to the grid is in effect halved, due to the filament connection being in the middle of the grid coil. The Sereen Grid Valve. N the development of the high-fre-quency amplifying valve, the most _ ~~ --

efficient is undoubtedly the new screengrid valve. In this valve the ‘plate and grid are separated by the use of an earthed screen interposed between the two electrodes. No neutralising is here necessary as the capacity inside the valve has been reduced to practically nothing. Also the grid and plate leads are brought out to pins on either end of the valve, so that in the wiring of the receiver the capacity from wire. to wire is practically negligible. Practical Hints. NOW that a general outline of the various methods used to obtain stability have been enumerated, the writer will mention a few practical hints that may be of use to builders who experiencg trouble in this direction. reference to potentiometer control, neutralise the valve by any of the ways suggested. With some receivers the capacity needed to neutralise may be in excess of that obtainable on the neutralising condenser. If the receiver is 1..0st stable with all this capacity in, then the condenser is not large enough. Extra capacity can be obtained by connecting short lengths of insulated wire to each terminal of the neutralising eondenser and twisting together. Similarly, the minimum capacity of this midget condenser may be too large. As a remedy increase the capacity of grid-plate leads by twisting insulated wire as before, only connecting in this ease to grid and plate terminals of valve socket. . Neutralising a Receiver. ERHAPS there are some readers who cannot neutralise their receiveks and have given up the job in disgust. To these constructors the following suggestion may.be useful. If the receiver is of the split primary, only use one half of the coil. Instead of connecting the filament to the bottom end of the grid coil, tap this coil in the middle and connect to the filament via a high-frequency choke or resistance, say 10,000 ohms. The tuning condenser still bridges the whole of the coil. Now connect the neutralising condenser between the P terminal of valve socket, and the lower end of secondary or gri«u coil and balance in the usual way, £ nsitivity will not be so good, but the receiver will more easily neutralise. A marked improvement in quality of reproduction should also be noticed. . When adjustments on the neutralising condenser make apparently no difnce to stability, the trouble usually in a feed-back taking place between the coils. No amount of neutralisation will remedy this and the constructor is recoumended to either shield the coils or check the angles they lie with respect to each other. It is as well to remember thit it is the valve capacity which is neutralised and the substitution of different yalues invariably means re-neutralis-ing the set. Run-down B batteries is the cause of many sets becoming unstable after a few months’ use. A remedy.can be effected by connecting a 1 or 2 m.f.d. condenser across the B positive radio frequency leads to earth, or B negative ‘terminal. The high frequency currénts are given a path to earth instead of being choked back by the resistance of the battery. Satisfied. "T HAVE constructeu the full wave :erystal set described in the ‘Radio

Record’ of December 14, and must say that it works very satisfactorily when using one crystal detector. One thing however, that I cannot understand is that when one catwhisker is applied and then the other, the second one tends to decrease the volume. ‘Trusting that you will be able to inform me how this fault may be overcome.‘SATISFIED’." [Correspondents using a nom de plume must sign their names and state their addresses.-Tec. Ed.] REPLY.-Your inability to obtain more volume when using the two crystals in conjunction may be due to one or two small faults. First of all, try

Methods' of Neutralizing, reversing one of the crystal detectors. Just disconnect both terminals and turn the detector round so that the crystal is connected to the wire on which was previously connected the catwhisker. Reverse only one crystal detector, as to revers. both together would only vesult in a similar circuit to the one you have now. ‘fake sure, also, that both detectors are working 100 per cent. One faulty detector would bring the other t ".s on level. See that both coils are tuned correctly and if the condensers are gauged try incorporating a small balancing condenser across either of the secondary coils.-PEN™ ODE.