Filter Circuit for Human Speech

Figure 1 

The circuit has been designed to develop a speech filter that will improve the signal processing circuit for optimizing speech recognition.

• TL072 – a low noise JFET input operational amplifier with features such as common-mode input voltage range, high slew rate, operation without latch up, compensated internal frequency, high input impedance at the JFET input stage, low noise, low total harmonic distortion, protected from output short circuit, low input bias and offset currents, wide common-mode and differential voltage ranges, and low power consumption
• High Pass Filter – an electronic circuit that allows the passage of high frequencies while opposing any unwanted low frequency components
• Low Pass Filter – an electronic circuit that allows the passage of low frequencies while reducing the amplitude of frequencies higher than the frequency response limit of the system

The operation of the circuit lies on the automatic determining of an audio signal by obtaining a relative measure of signal in a selective range of frequency and controlling the path or passage of the audio signal. The human speech is consisted of components such as a buzz and a hiss. The passage of air from the lungs over the vocal cords results to the formation of a buzz which has a fundamental frequency range of 80 Hz to 240 Hz. The effect of various resonant cavities and the articulation of the tongue provide a hiss outcome that covers a wide range of frequencies that extends well above 5 KHz.

The human speech can cover a frequency range of 300 Hz to 3 KHz. This range of frequency is internationally known for the transmission of speech through telecommunications networks. The circuit is used to capture this frequency range and reject those which are not within the range. To do this, two active second class filters were used for critical damping of signals. The first filter is built around the region of IC1A as it performs the operation of a high pass filter that rejects low frequencies below 300 Hz. The high frequencies above 3 KHz are being rejected by IC1B as it functions as low pass filter. The components that make up the circuit like the resistor should be made of metal film with 1% tolerance since they are used for greater precision, while the capacitor should be made of polystyrene because of their low dissipation factor, excellent temperature stability, low distortion, and much higher degree of stability and reliability than other types of capacitors.

Figure 2 

The theory behind this speech filter circuit can be applied to those filters that can be attached to a radio receiver to pass selectively the speech or music depending on the demands of the user. In telephone communications, the speech filter may be found in the processing of signals from a microphone used for hands free phones where it is intended to produce a clearly intelligible and strong signal. They are also being utilized primarily for use in hearing aid devices, voice input microphones, and similar devices to increase the range of interpersonal voice communications and improve the signal-to-noise ratio by enhancing the sound quality and electronically reducing the noise.

Source:users.otenet.gr/~athsam/speetch_filter.htm