Direction of a sound source by phase-interferometry

[nephriticus]

Hi. For the final project of my laboratory course a direction finder circuit has been given. The circuit will work on the principles of phase-interferometry. (2 microphones and a speaker driven with a certain frequency are to be placed so the voltages of the microphones have a phase difference according to the angle of the speaker.) I thought that summing the voltages could be a way to get the phase difference information as the amplitude of the resulting signal will vary according to the phase difference. I suppose there should also be a noise filter and amplifier. To get a clean sinusoidal waveform. Then this output should be measured by a multimeter to get the information to a computer program which should directly give the angle of the speaker as the output. What do you think of this method? Should there be any other elements? In which ordering of the circuit parts i get the best result? (There should be an audio amplifier and a pre-amplifier according to the specifications given.)

 

[audioguru]

Echoes will mess up the results.
Plan the spacing of the microphones and the frequency so that you don't end up with a comb filter that has multiple lobes.

 

[nephriticus]

actually the spacing of the microphones is also given in the specifications (i have forgotten to write it there.) it is half of the wavelength of the incoming sound waves. The frequency will be also be specified. (yet i don't know it)

 

[nephriticus]

I am rather curious about the type of amplifiers i should use. In the specifications it says audio amplifier is the main part of the circuit. But after searching in the internet I understand that audio amplifiers are used to amplify the power of the signal such that it can be used in big speakers. However in this circuit we don't need an audio output. So I can not understand why we have to use two amplifiers. Since a digital multimeter is reading the amplitude does it make any difference even if no amplifier is used in the circuit?

 

[audioguru]

The frequency response of most multimeters is flat for 50Hz and 60Hz mains frequencies. At higher frequencies the scales are different from what they should be and the measurements read low.

The microphones have very low levels and each needs to have a preamp. The outputs of the preamps can be summed then filtered. The multimeter can read the filter's output if the frequency is not too high for the multimeter.

The speaker needs to have a power amplifier driving it.

 

[nephriticus]

I suppose the frequency will be above 20kHz. Does it need a second amplification to be accurately read by the multimeter?

these are the exact specifications given:

Specifications and Requirements
1. The loudspeaker should be at least 2 meters far from microphones.
2. The amplitude of the sinusoidal wave which is applied to the loudspeaker should be equal
or less than 5 Vpp.
3. The distance (d) between the microphones should be Lambda/ 2. Consult your assistant for
operation frequency.
4. The direction angle calculation should be done using “Agilent-VEE”.

Allowed Circuit Elements
You may use a loudspeaker, microphones (without embedded pre-amplifier), transistors, op-amps,
resistors, capacitors, inductors, diodes in your circuit. Other ICs (Integrated Circuits) are not allowed.

 

[audioguru]

I suppose the frequency will be above 20kHz.

The half-wave distance between microphones is very small. Less than the diameter of most microphones. The frequency is too high.

Does it need a second amplification to be accurately read by the multimeter?

Most multimeters cannot measure the level of frequencies as high as 20kHz. The frequency is too high.

Most speakers cannot produce frequencies as high as 20kHz. Some tweeters can, but continuous power will burn them out because they are very fragile. Maybe a piezo tweeter might survive. The frequency is too high.

 

[nephriticus]

After i posted it i calculated the wavelength... It was too small. Anyways as it says there i am going to consult the assistant for frequency. One more thing; Can i model the microphone as a sinusoidal voltage source in series with a resistor (or a capacitor) (while modeling the pre-amplifier). And if so what could be the value of the resistance/capacitance.

 

[audioguru]

Don't you know which type of microphone you will use?
Some are low impedance, others must be powered and are a higher impedance.
1) Dynamic with coil and magnet.
2) Electret with a high voltage and a FET buffer built-in.
3) Condenser requires 48VDC.
4) Crystal might not be available anymore.
5) Carbon is really old and was used in telephones.

 

[nephriticus]

It'll be an electret microphone without embedded pre-amplifier. I googled electret microphones but only saw phrases saying it is high-impedance. Since the output of the microphone is going to be sinusoidal, i wanted to model it as a sinusoidal source in series with an impedance...

 

[audioguru]

An electret microphone has a FET impedance converter inside. It needs to be powered. It uses only about 0.5mA so a 10k resistor from 9V is fine.

 

[nephriticus]

I designed a circuit for 2kHz operation frequency. (Half of the wavelength is 8.5 cm this time. ) The output voltage peak depends on the phase difference of the two microphones. Do you think with the real components it will work? Is my modeling of the microphones right?

 

[audioguru]

Your filtering opamp is an oscillator with positive feedback. It is a filter if it has negative feedback like this:

 

[nephriticus]

Actually we have studied that circuit as an active RC filter. The frequency response is in the attachment.

 

[audioguru]

Tell your teacher that your "filter" oscillates.
It has positive feedback that makes it oscillate.
A filter is supposed to have negative feedback.