THERMIONIC VACUUM TUBES 63 



29. Tube Noises. Tubes may be responsible for three distinct 

 kinds of noises, (a) Ringing or rattling is due to the vibration of the 

 tube elements and may be eliminated by proper tube construction or 

 by some form of vibration proof suspension for the early stages of the 

 amplifier, (b) Crackling may be produced by high resistance films 

 on the inner surface of the bulb, forming conducting paths between 

 the leads. Faulty electrical contact between the plate, grid and 

 filament and their respective leads is also a frequent source of crackling. 

 Furthermore, in tubes which are well constructed in regard to the 

 points just mentioned, but which contain tungsten filament, crackling 

 may be observed. This trouble is not found in all tungsten filament 

 tubes but, when present, is sufficiently marked to become apparent 

 in a two stage amplifier, (c) In carefully constructed amplifiers of 

 more than three or four stages a noise which can best be described 

 as a hissing or sighing is certain to be present. It appears to be re- 

 lated to an unavoidable statistical variation in the escape of electrons 

 through the grid to the plate. Its magnitude has been found to cor- 

 respond approximately to an output voltage from the first stage of 

 between 5X10 -7 volts and 5X10~ 6 volts. Between these limits 

 the noise is found to increase as the output impedance of the first 

 stage is increased, and also to increase as the resistance across the 

 terminals of the input increases. Its components, above 300 cycles, 

 appear to be of about equal magnitude and uniformly distributed. 

 It is, therefore, impossible at the present time to build amplifiers to 

 handle voltages of less than this order of magnitude, at any rate 

 when the frequencies involved are in the audible range. 



30. Circuit Noises. In general, circuit noises in amplifiers are due 

 to one or more of the following causes : variations in grid and plate 

 batteries, loose contacts and variations in resistances, leakage of con- 

 densers and leakage across the insulating mounting upon w T hich the 

 amplifier parts are fastened, .and external electric or magnetic fields 

 acting inductively on the circuit. The remedy in each case is obvi- 

 ous once the exact cause has been found. To eliminate inductive 

 effects in the wiring it is usually sufficient to run wires in pairs and 

 to shield them electrically, the shielding being grounded. In laying 

 out the various parts of an amplifier it is well to place the bulky pieces 

 at points in the circuit at which they will have as near zero potential 

 as possible. 



31. Singing. Singing, which is one of the most serious troubles in 

 amplifiers, is always due to some form of feed-back. This may be 

 magnetic, electrostatic, or in the form of mechanical vibrations as in 

 an amplifier having a microphone attached to the input and a receiver 



