[ORE and more industry is realizing the fatiguing effects upon human beings of the various noises to which they are subjected, and with the realization comes the endeavor to reduce, so far as practical, all unnecessary sounds. In New York City for example, a noise commission has been making an exhaustive survey of the sources of city noise and means of its prevention. The most disturbing noises are frequently those generated by some common piece of mechanical equipment during its normal operation. Many machines, such as ventilators,
frequency weighting transformer
power level indicator
variable gain amplifier
TYPE 511 -AM
Noise is picked up in ihc microphone, weighted, and amplified. The power level indicator gives the result in decibels, referred to the background noise level, or any level chosen as reference
TYPE 511 -AM
Noise is picked up in ihc microphone, weighted, and amplified. The power level indicator gives the result in decibels, referred to the background noise level, or any level chosen as reference refrigerating systems, pumps, fans, etc., cause a constant whirring, hissing, or humming sound, which gradually works on the nervous system, causing mental fatigue.
Manufacturers of mechanical equipment are beginning to recognize the importance of quietness in their products.
An intelligent approach to the problem of noise elimination requires a means of quantitative noise measurement. Listening tests are of course of no substantial value in this work. Accordingly, various forms of noise-measuring devices have been developed in order to provide quantitative comparisons of different sounds. An extensive discussion of the problems involved in noise measurement will be found in recent issues of the Journal of the Acoustical Society.
The simplest form of noise-measuring set-up consists of a microphone, an amplifier, and a suitable indicator such as a vacuum-tube voltmeter or an oxide-rectifier meter. Sounds picked up by the microphone are amplified and indicated on the meter, giving a definite method of comparison. The meter should preferably be calibrated in decibels, which are the units generally used to express ratios of sound intensities. Where wide ranges of volume are to be measured, the amplifier gain should be adjustable to avoid overloading on loud noises.
In order to present a true picture of the effect of the noise on the human nervous system, the measuring system must have a frequency characteristic similar to that of the normal auditory system, i.e., the low and high frequencies must be discriminated against. Some such discrimination is unavoidable in a microphone and amplifier system, but the normal loss of high and low frequencies should be accentuated.
Since a fiat frequency characteristic is not desired, an expensive microphone is not generally needed unless very low noise levels are to be encountered. The microphone should, however, show negligible variations with use and age and should have a residual noise level substantially below the lowest noise to be measured. The input transformer or frequency-weighting network should be so designed that its characteristic added to that of the microphone results in a close approximation to the normal ear characteristic.
A noise-measuring set of this type may be composed of several standard General Radio units. The illustration shows a system supplied to a large manufacturer of refrigerating machinery. It consists of a General Radio Type 514-AM Amplifier, a special Type 345-SN Input Transformer, a Type 586-CM Power-Level Indicator, and a microphone. The input transformer matches the microphone impedance to the amplifier and also adjusts the over-all frequency response to approximate that of the human ear. In this particular system, the input transformer is mounted in the amplifier cabinet, and the microphone receives its current from the amplifier filament battery.
The microphone is an inexpensive carbon type, since the noise levels at which it is to be used are comparatively high. The accompanying curves show the over-all frequency response of the noise-measuring system with and without the microphone. It will be noted that the over-all response of the complete system, including microphone, represents a very close approximation to the response of the normal human ear.
Although equipment of this sort is generally used to determine the decibels difference between normal noise level and noise with some machine operating, the equipment can, of course, be calibrated against any arbitrary standard desired. The threshold of hearing at 1000 cycles for an average person is frequently used as a reference level.
A convenient manner of calibrating the noise-measuring set is as follows: a 1000-cycle tone may be obtained from
The frequency characteristic of the system is made to approximate that of normal hearing
The frequency characteristic of the system is made to approximate that of normal hearing a suitable oscillator in conjunction with a loud-speaker. The observer should be so placed that he is approximately the same distance from the loud-speaker as the microphone and reasonably close to the microphone so that it may be assumed that the sound intensity reaching the observer is practically the same as that reaching the microphone. If the volume of sound from the loud-speaker is then reduced until the threshold of hearing is reached, the reading of the noise-measuring set at that point may be taken as the reference level.
With some types of microphones, the residual noise may be enough to cause an appreciable error in reading of the set at the threshold of audibility. Where very weak sounds are to be measured it is recommended that a microphone of the condenser or moving coil type be employed, but if loud noises are to be measured, the following method of calibration will be found quite satisfactory.
The oscillator used for obtaining the test tone should have a reasonably large power output and should be equipped with a calibrated volume control. If this is not the case, a suitable amplifier and attenuator may be used in conjunction with the oscillator. The loud-speaker should be reasonably linear in so far as power output is concerned, that is, doubling the power input to the speaker should double the power output. This of course means that the loud-speaker must not be overloaded. Placing the observer and microphone as mentioned before, the threshold of hearing should be obtained in the same manner and the reading of the calibrated attenuator or volume control on the oscillator or associated amplifier noted. The output of the loud-speaker should then be increased by adjusting the calibrated attenuator until a reasonably loud sound intensity is obtained. An increase of the order of 40 or 50 db is generally sufficient. It is not important that the intensity be increased by any exact amount, but only that the amount of increase be known. The reading of the noise-measuring set at that point, referred to the threshold of hearing, will equal the amount by which the output of the loud-speaker was increased.
For best results, these calibrations should be performed in an acoustically dead room or in the open air at some place where absolute quietness may be obtained. It is also advisable to perform the calibration with several individuals and use the average as a final value. — II. H. Scott
For the convenience of those desiring to construct noise-measuring sets similar to the one described above, the following table gives the list of General Radio equipment which may be used, and the prices.
Type 514-AM Amplifier (without tubes and batteries) . . . $76.00
Type 586-CM Power-Level indicator $75.00
Type 345-SN Transformer (for use in conjunction with single-button carbon microphone to obtain approximate characteristic of human ear) $15.00
If it is desired to have the special input transformer mounted within the amplifier unit and the amplifier input circuit arranged so that microphone current is obtained from the amplifier "A" battery, an additional charge of $15.00 will be made. The complete noise-measuring set as described in the foregoing article, including microphone, microphone case, suitable shielded cable connectors, and all tubes and batteries, will cost $230.00.
A BOOSTER AMPLIFIER FOR 500-OHM LINES
IN modern communications systems, lines and equipment having impedances of 400 to 600 ohms are generally widely used. The high-quality transmission lines associated with radio broadcasting and sound pictures equipment generally have characteristic impedances within this range, and, accordingly, a large percentage of faders, mixers, and speech-input amplifiers have input and output impedances of approximately 500 ohms.
Transmission lines should, of course, he operated at a volume level sufficiently high to minimize the effects of noise and crosstalk which may he picked up from nearby electrical equipment or oilier transmission lines. A level of +2 dl> is generally recommended. Increase in volume level also tends to reduce noises which may he encountered in fading and mixing equipment.
Two of the General Radio Company's new transformers, Type 541-G and Type 541-P, are finding widespread favor among engineers for use in connection with 500-ohm equipment. Type 541-G is a line-to-grid transformer designed for operation into either a single tube or a balanced (push pull) stage. Type 541-P is a plate-to-line output transformer for operation from a single amplifier tube or a balanced stage into a 500-ohm line. Using these two transformers, a line-booster amplifier may be con structed which will provide high insertion gain with an excellent frequency characteristic.
The accompanying diagram shows the circuit and frequency response of a balanced amplifier using two of the new 56-type tubes. The over-all amplification is approximately 20 db. 227-type, 230-type, or 237-type tubes may also be used, if desired, with a slight decrease (approximately 1.5 db) in gain. If a larger amount of output power is desired, 210-type tubes may be used. — H. H. Scott
Prices on equipment used in the booster described above are as follows:
Type 541-G Line-to-Grid Transformer $12.00
Type 541-P Plate-to-Line Transformer $12.00
2 Type 438 5-Prong Sockets . .70 Special Base —• Nickel-plated Brass with Binding Posts $3.00
20 tOO 1000 IIOOO
FREQUENCY IN CYCLES
An excellent frequency characteristic is obtained
20 tOO 1000 IIOOO
FREQUENCY IN CYCLES
An excellent frequency characteristic is obtained
Was this article helpful?