microphone bias, low noise pream?

[ahgu]

low noise mic bias

I want to detect far away audio/voice sound from couple meters away?
How can I design a circuit to do that? what mic should I use?
what bias voltage?
And what kind of amplifier?

my ADC input range is 2.8v p-p.
And the programmable gain (PGA) is from 12dB to -34dB.

thanks
Ahgu

 

[flatulent]

preamp mic ssm2135

Using a parabolic reflector will do more than anything to improve sensitivity.

Select a microphone element with the highest sensitivity. They are rated at an output voltage caused by a standard sound pressure level input.

Use a step up transformer to transform the microphone impedance to the optimum amplifier source impedance. The optimum source impedance is the noise voltage density divided by the noise current density. The best performance in this area is JFET amplifiers.

 

[klug]

Hi,

You can use directional resonance microphone, look to the picture:

**broken link removed**

Microphone directional system is made of aluminic pipes with 10 mm diameter. Every pipe has own resonance frequency.

L(centimeters)=330/(2*F)

It is for internal diameter of 10 mm.

For voice band it will be enough 7 tubes.

Description of this construction is available on Russian here:

http://kiev-security.org.ua/box/8/27.shtml

If you will need some details, I can explain them for you.

Best Wishes! klug.

 

[ahgu]

what material is the reflector?

What material is the parabolic reflector? Where can I find those?

are you saying a big transformer? I hate to use transformer in my design.

Do you know where I can find a circuit diagram?

thanks
Ahgu

 

[flatulent]

parts

The transformers are rather small. You may not need one.

Parabolic reflectors are more frequently used in microwave systems. They reflect sound waves just as well as radio waves.

I do not know how to get them. The design with the resonant tubes may be easier to make if you cannot get the microwave reflector.

I have seen some ads for the entire system advertised in Edmund Scientific catalogs. They are claimed to be used for listening to birds. This is because such audio systems are not legal in many places for listening to humans down the street.

 

[ME]

Try to look at page 9 in this datasheet for Analog Devices SSM2135 OpAmp, there's an application note for designing a low noise microphone preamp. You can use an electret microphone for this design.
https://www.analog.com/en/prod/0,2877,SSM2135,00.html

 

[cozturk]

hello everyone,

Very good mic is a piezo xtall buzzer (5cm diameter).

It is ideal for voice only recording and bug. (not for music)

connect it high gain preamp. tehere is no "hiss" and "humm" because it is capacitive material

 

[nostrada]

does it works with ultrasonic range?

 

[nostrada]

if it works only with human voice can we adapt it with more tubes to increase bandwidh

 

[klug]

To increase bandwidth you need to add pipes, if you would look on the link, then you will see a table at the bottom of that page. There are calculated length for these pipes. Longest pipe #1 has 920 mm (180 Hz). Shortest one, pipe #37 has 20 mm length (8.2 kHz).

It seems that 8.2 kHz is upper frequency for 10 mm diameter of pipe, to obtain higher frequency you will need to narrow pipe's diameter, but as I think it will be wrong to combine narrow and wide pipes together, as you will have difference in Q - in this case you will need to make another step between working frequencies.

At the bottom of these tubes is located small parabolic reflector with mike in the focus. I think, you will need to make some acoustic isolation at other side of this reflector and in front between the tubes, it will allow to escape some acoustic noice from other directions. It will sutable to use some foam matherials there.

It is interesting that such system is bidirectional, and you can put some speaker in the focus of such construction. In this case you will radiate sound only in one direction. There is a prison near our local market and I frequently see a lot of people communicating with prisoners through fence, they are talking by loud voice. I think such system with directional acoustic receiver and transmitter will be very sutable for them. It is some funny and sad example of application for such system.

It is interesting, is there any software emulator for such acoustic resonance constructions?

 

[nostrada]

I don't understand russian and I would like to know is there a relation between the pipe's diameter and his lengh?!?
--->sensitivity and selectivity

 

[cozturk]

Very good mic is a piezo xtall buzzer (5cm diameter).

It is ideal for voice only recording and bug. (not for music)

connect it high gain preamp. tehere is no "hiss" and "humm" because it is capacitive material

It have 1Khz voice band width. Perhaps it can be used for ultrsasonic

 

[jourval]

I work 2 years in small company that developed
microphones for some special application, mostly - security.
I check a lot of them, include number of sound concentrators.
1. Resonance tube concentrator will not work for high frequencies
(more than 15 kHz), because holes between tubes will be re-
sonators too and this increase background noise and decrease
directivity. Small diameter tubes is not solution because tubes
wall cross-section area will be bigger (relatively) and you lost acoustic
transparency at all. Also this kind of concentrators are very sensitive
to noise of wind.
2. Parabolic concentrator is OK for high frequencies. It is difficult to
built real parabolic, as a simplest approximation you can use
any metal (not plastic ! - not work) spherical basin or something
like from kitchen utilities shop. Make some calculations - were is
center of this sphere and place in this point your microphone
(of cause it will be back to signal source). The bigger diameter
give you more audio gain.
3. Electret microphone + preamplifier from Analog device (see above)will be fine.
Believe me - you can't build preamp. better, at least with better
signal to noise ratio. Variable gain amplifier must be with
positive gain (0-40 dB), negative gain is unuseful at all -
you'll decrease signal, but noise will be the same. Instead of VGA amplifier
may be better use 2 -3 fixed gain (x100) amplifiers cascaded with AGC
cells between them. AGC cell - it is resistor+small signal mosfet connected
as feedback controllable divider ( more signal - fet opened - signal divided etc)

 

[abba]

where is it?

 

[ahgu]

Jourval,

Can you give me decent design for electret+preamp?

I have a codec that does -20dB to 12dB, what kind of preamp should I have?
since you mentioned -dB is not good. 12dB is rather small.

Also, can I get away with some cheap general purpose OPAMP instead of expensive analog audio opamp? I am only doing voice grade stuff, but need to detect sound from couple meters.

thanks
ahgu

 

[MCMC]

Hi ahgu

Find in this link **broken link removed**

looking for mic preamp and if yours interest for RS soundmeter mod.

mcmc

 

[jourval]

No problem, but I need more details
for calculations.
What is real bandwith , voice or more?
What is working distance 5m or 50 m ?
You need only understand speech or
recognise voice - it is not the same etc.

 

[ahgu]

voice band

4khz, I am sampling 8kz.
5m working distance.

I need to understand speech. like a phone.

thanks

 

[gil]

Hi Klug,

Very interesting posting.
You have successfully triggered my curiosity: How can a bunch of empty tubes could be so good in capturing dim sounds??.
I downloaded a translated version of the page in Russian (thanks Babel Fish). I also have found an article written in English that, perhaps; is the original one. It was originally published in 1965.

**broken link removed**
**broken link removed**

With the Russian article, plus these two articles ones, and some more about sound waves, I‘ve been working hard trying to understand the physics behind the project. Still I have a part of the puzzle to be solved: You said that for voice, only 8 tubes are necessary. However according my calculations (might be wrong) with tubes of 10mm of diameter, temperature of 15ºC, a correction of 0.8D for diameter and with an acoustic Q=15, the only way that I have found is accepting that the system will work over the fundamental frequency and the first 6 overtones (n= 1, 2, 3, 4, 5, 6, 7). Is that right?
Do you know the right way to calculate the bandwidth of the acoustic system?
Best regards
Gil

 

[wang.mickey]

Very good mic is a piezo xtall buzzer (5cm diameter).

It is ideal for voice only recording and bug. (not for music)

connect it high gain preamp. tehere is no "hiss" and "humm" because it is capacitive material