US2131669A - Sound reproducing apparatus - Google Patents

Description

Sept. 27, 1938. A. POLIAKOFF' SOUND REPRODUCING APPARATUS Filed May 31, 1935 2 Sheets-Sheet l n u F F 0 w m.

2000 FREOl/E/VC/ INVENTOR ALEXANDER POLIAKOFF B ATTOHN BYS P 1 A. POLIAKOI-=F SOUND REPRODUCING APPARATUS Filed May 31, 1935 2 Sheets-Sheet 2 INVENIOR ALEXANDER PoLIAKoFF ATTORNEYS Patented Sept. 21, 1938 UNITED STATES PATENT OFFICE Application May In Great 1 Claim.

The present invention relates to all sound reproducing devices in which head telephones and not loudspeakers are used to radiate sound, such as ordinary communication telephones, deaf-aid amplifiers, field telephones, aircraft radio receivers and similar installations.

Hitherto, the majority of deaf-aids, both simple microteiephones and valve amplifiers, utilized a single telephone.

Attempts have been made withvalve amplifiers to shape the response curve of the apparatus to compensate partially the patient's tonal deafness" as shown by his audiometric curve. Such a curve is commonly obtained by means of a valve oscillator provided with a calibrated attenuator .or by means of a set of calibrated tuning forks. In both cases the patient hears one note at a time, so that the disturbing or masking efiect by low and middle notes on the high notes in a complex sound is left out of consideration. This effect is known, however, to be large and to increase rapidly with the sound intensity level; curves illustrating the effect may be seen in Harvey Fletcher, Speech and Hearing, 1929, part 3, ch. IV.

A common case of deafness will have as much as 40 decibels greater loss at say 4,000 cycles than at 512 cycles. Witha portable valve amplifier employing a carbon microphone and the usual -magnetic telephone not only is it impossible to compensate for that loss but even to obtain a level response between these frequencies at a reasonable amplification level, since both the microphone and the telephone have large high note losses and the number of valves used must necessarily be small.

In addition, the high notes although present in the telephone of such an instrument will not be perceived by the listener on account of their being masked by the noise in the room and by the low and middle notes in the sound particularly in the region of the principal resonance of the telephone diaphragm. I

Deaf aid apparatus supplying each ear with a high fidelity signal amplified separately to'compensate accurately the audiometric curvesof the.

two ears, which are usually difierent, apart from its prohibitive bulk will not be free from the masking effect and its attendant great diminution in high note perception,

The invention is hereinafter described with ref erence to the accompanying drawings in which:

Figure 1 is a chart showing the curves of a high-pass amplifier relative to a level amplifier.

55 Figure 2 shows a simple way of applying the 31, 1935, Serial-No. 24,445 Britain June 4, 1934 invention to a thermionic amplifier used for deaf aid or ordinary speech frequency amplification.

Figure 3 shows sound output characteristics of the circuit in Figure 2, and

Figure 4 is a circuit diagram ofa portion of the amplifier in which separate valves are used to feed the level and high pass telephones.

The present invention" substantially eliminates the masking of high audio frequency components in a complex sound by the other components of such sound, by the use of two head telephones one on'geach ear, the-low and/or middle frequency components in the sound output of one of which telephones being sufliciently reduced in intensity not to mask appreciably the high frequency components in the sound output of one tact with that telephone. Since there isno masking of high frequency components received by one ear by low and middle frequency components received by the other ear, a given acoustic output of high frequency components in a complex sound thus produces far greater sensation than would be the case if the whole output were applied to one ear only, or were shared between the two cars without frequency discrimination between them. v

An experimental verification of. the last statement has been carried out as follows:-

Two thermionic amplifiers were constructed with common input from a high fidelity noncarbon microphone coupled to the amplifiers by means of a valve common to both. One of the amplifiers had an approximately level frequency response characteristic, while theother had a steeply risingcharacteristic. Two identical mag.- netic telephones were connected in the output in such a manner that, by manipulating a switch, the-outputs of the two amplifiers, could be either combined and shared equally'by the two telephones, or the high-pass and level outputs were delivered to different telephones.-

Figure 1 shows the curve of the high-pass amplifier relative to the level amplifier in terms of in decibels, where V1. and V1; are respectively voltages applied to the first and common grid for v equal outputs from the level and high-pass amplifiers.

The output level of the high-pass amplifier is below that of the level amplifier at all frequencies except 4,000 cycles at which the two outputs are equal.

cies at a lower level than T1.

experiment consisted of making speech in the room containing the microphone unintelligible by a controlled noise radiated by a loudspeaker in the same room, the two telephones sharing equally the outputs of the two amplifiers. On switching over to an asymmetric output, that is with one telephone connected to'the high-pass amplifier only, intelligibility of speech was restored. Noise of different nature such as a single note of SOD-1,000 cycles, imitation petrol engine noise and amplified noise picked up by a-microphone in a workshop were tried with the same result. Further the use of one phone only connected to the high-pass amplifier did not produce intelligible speech, while with both telephones connected to the level amplifier only and the masking noise increased to the point where speech was just unintelligible, intelligibility was not restored by adding the-output-of the highpass amplifier which on account of its low level was entirely lost. Y

To facilitate description, the telephone reproclucing the whole scale will be referred to hereunder as the level telephone and the other as the high pass telephone.

The invention is not dependent on the use of a special receiver such as a condenser or a moving coil telephone in the high pass position or even on the use of thermionic amplifiers, and may be applied to the simplest microphone systems employing ordinary magnetic telephone receivers.

The output of the high-pass telephone alone must not be confused with a frequency corrected output, since in general the level of low and middle frequency components will be below the threshold of audibility so that the high-pass telephone alone will not usually provide intelligible audition.

As shown in Fig.2, two telephones T1 and T2 which may be identical are connected in the output, T1 direct, T2 through a condenser C1 the impedance of which is high at low and middle frequencies compared with Z1 and Z2. If Z1and Z2 are the impedances of T1 and T2 respectively and Z1=Z2, then the sound output of T2 will have a rising characteristic and will be at all frequen- -If Z1 is made higher than Z2 the sound output of T2 will be higher than of T1 at high frequencies and lower at low and middle frequencies.

Curve I, Fig. 3, is the sound output of T1 with T2 disconnected. Curve 2 is the sound output of T1 with T2 in circuit, Z2 being slightly smaller than Z1. Curve 3 is the sound output of T2. There is no. increase in the total acoustic output on connecting T2 in addition to T1, but merely a transfer of half or more of the high frequency output to the other ear.

The curves in Figure 3 should be contrasted with those published for dual speaker combinations, the object of which is to reproduce a greater range of frequencies than can be covered by a single speaker and in which the introduction of the high note speaker greatly increases the total high note output.

In my invention the low and middle notes particularlythe latter are reduced in the high-pass telephone to a low intensity level. Thus my invention does not consist merely in the use of two telephones, one of which reproduces high notes better than the other. It is common in dual speaker'dombinations for the output of the two units to be approximately equal at 1,000 cycles. The headphone equivalent of such a pair of speakers may have slightly higher intelligibility in one telephone should replace correction, that is, shaping of the sound output frequency characteristic to suit approximately definite types of deafness or special requirements of listeners of normal hearing acuity situated in noisy surroundings. It is claimed however that with a ivenavailable high note sound output in a complex sound the application of the invention results in'a far greater sensation being produced by the high notes.

It has been found useful to control the frequency response of the output of the amplifier feeding telephones T1 and T2 by means such as an intervalve transformer described in U. S. Patent No. 1,996,685.

In practice it has been found that sufficient low and middle frequency attenuation in the output of T2 may be produced by means of a simple series condenser to meet the requirements of most deaf cases. Where space permits, however, a proper fixed or variable high-pass filter may be used, the construction'of which is common knowledge. The use of separate thermionic valves to feed T1 and T2 need only be adopted with cases requiring a very large output of high notes. In such cases the two output valves V2 and V2, Figure 4, have their input from the valve V1, and the high-pass output from the valve V2 to telephone T2 may be obtained for example, by using two intervalve transformers M1 and M2 with their primaries in parallel through a small condenser C2. The transformer M1 being of normal design, feeds the grid of the level ouput valve V2 while the transformer M2 is of the type described in' U. S. Patent No. 1,996,685, the response curve of which is made more or less sharply rising with frequency according to the setting of the associated potentiometer. The function of the small condenser C2 is to prevent the primary of M2 unduly lowering the impedance in the plate circuit of the valve V1 for low and middle frequencies.

In some cases it is possible to do away with a filtering arrangement by using receivers inthe high-pass position which by virtue of their electrical or mechanical construction have a naturally rising response characteristic. A condenser receiver for instance connected straight across a magnetic receiver will in general reproduce high notes only.

It has moreover been found that with impaired hearing the high-pass telephone should generally be placed on the ear with better high note hearing, which in the majority of cases is also the more sensitive ear at all frequencies. The invention has then the additional advantage of putting more energy into the bad ear than the good ear so that loud noises produce less distress than would have been the case if a single telephone were used on the better ear.

So far mention has been made principally of deaf aid amplifiers, the invention is however of considerable value in sound reproducing systems used by persons of normal hearing acuity.

The existence of large high note losses in the ordinary communication telephone are of course well known. The user of a telephone in noisy surroundings may be usefully regarded as sufier-- ceivers one for each ear and arranged in difierent circuits, one receiver having a uniform frequency response of low, middle and high frequency components and means for altering the frequency response in the second receiver to cause the low and middle frequencycomponents to be reduced to an extent that they do not mask the high fre- 5 quency component of said second receiver.

ALEXANDER POLIAKOFF.

Images