AMPLIFIER NOISES

Hawke's Bay Tribune, Volume XXI, Issue 243, 26 September 1931, Page 14

AMPLIFIER NOISES

Sometimes, in order to obtain a desired degree of selectivity, the input signal energy is reduced by the use of a very small aerial, the cutting down process being compensated for by providing high r.f. amplification (states an exchange). This is satisfactory up to a point, but there is limit to the extent of amplification which may subsequently be carried out It is possible to construct a receiver which in three stages of r.f. amplification would increase the signal strength at least 30,000 or 40,000 times before the detector is reached, but if it was worked to its maximum sensitivity on a weak station there would bo subsidiary or “background” noises, probably as loud as the signal itself. These disturbances could be analysed into two parts, one, a loud rushing noise as of tho flow of water, and the other a series of sudden noises. The continuous sound is generallyrefrred to as “mush,” and although sometimes it may be partly due to disturbances external to the set. it mostly originates in the set itself. This may be demonstrated by arranging that the aerial coupling and sensitivity may be varied. If a strong station is tuned in with the set operating well below its maximum sensitivity both kinds of noises will be present, although both will bo effectively drowned out by tho signal. If the aerial coupling is loosener! so as to decrease the signal input, and at the same time the sensitivity is increased so that the output is constant, the intermittent noises will remain of the same strength relative to the signal, but the rushing sounds will increase compared to the signal. This would indicate that the sudden sounds originate outside the set, while the others arise within it. The cause of this internal disturbance is perfectly well known The process of electron emission is not unlike that of the emission of steam from boiling water. Although the rate at which steam issues from rhe spout of a kettle may ho substantially constant, there are variations in the nuinoer of übbles arriving at the surface of the water. Tho plate current thus cannot be assumed to be a perfectly steady flow, but must he considered as the total of a perpetual stream of millions of independent electrons all fired out independently as shots from a gun. This was pointed out some years ago by W. Schottky, who gave the name “shot-effect” to this variation. Atmospherics will cause momentary oscillations in an aerial iireuit. and those tiny hut sudden changes of plate current will have a somewhat similar effect on a succeeding tuned circuit. Naturally the disturb-

ances caused in the first valve will be magnified by the action of the second valve, and still more increased by the third and succeeding r.f. amplifiers, if any. The later valves will also add their contributions, but the first will be the main cause of the trouble, because of the subsequent amplification Ultimately the magnified disturbances will reach the detector, and they will be rectified to produce that toneless sound, which will, of course, be again increased b the a.f. stages. The effect will be much greater in receivers having tuned circuits in the plate circuits, such as the tuned anode or band pass type of circuits. The effect is most noticeable in sets working from frame aerials where the signal voltage is naturally of a low figure and practically the only assistance towards eliminating it can be obtained by making the aerial as efficient as possible for by this means the signal voltage is made larger compared to the voltage created by the shot-effect.