infrasonic detector

thank for your replies.

I think the frequencies are below 5 Hz, and in anycase from 0 to 15 Hz.

I don't understand the micros problem and about the range

no german !! i'm italian so i think english is better. My microphone (the laptop one) has 2 pin, so i think is god for this purpose.

What do you mean with "with easy selected microphones for 50 ct" ?

Freq. below 1Hz ? Is it possible ? About the Op Amp in the schematic do you think Microchip's MCP6004 could be a good choice ?

I have not checked through all of your circuit but I think you will have difficulty finding a microphone which works below 5Hz. Almost all microphones have some kind of 'leak' round the diaphragm so that they are not affected by long term changes in air pressure which are much larger than the pressure changes caused by sound at normal levels.

There is another issue with the common low cost electret microphones - the built in amplifier has a finite input impedance which together with the source capacity of the electret cell sets the low frequency cut off point of the electronic circuit. Since this is built into the microphone you can't do much about it.

The microphone built into a PC will not work at 5Hz.

Do you have any idea of the magnitude of the pressure changes you are trying to detect - if you do I might be able to suggest a suitable transducer.

MK

Hi,

Don't worry, I built a few of such devices,they all works fine. But I had to select the micros. From 100 only

The frequencies I had to detect where below 1 Hz. Remember, that worked fine with easy selected microphones for 50 ct.

The problems with the microphnes where jut the leakage of the closed room inside of a mic without a direction characteristic (how to say this in correct english?) This is then the first point to figure out, whether your mic may be OK for your system. And additionally there have to be no C inside the mic, you need a 2 wire mic. A 3-wire mic may also be ok, but may be not ;-)

no problem, we may continue in german ;-)

Marte

For response down to 0 Hz, you might try an absolute pressure transducer, like:

These are 8USD from Digikey. They will respond from 0Hz to about 160 Hz. You can monitor the weather at the same time.

HVAC systems can create a lot of sub-audible noise (oxymoron?)...

Best regards, Spehro Pefhany

They are good for the 5 volt supply stages after the signal has been amplified a bit. Notice the input noise voltage for .1 to 10 Hz, on page 3:

A quieter front stage at sub audio frequencies might be AD820:

Are you asking for advice on construction practice?

Probably won't have much response below 20 Hz.

Not if it is clean and the opamps are designed to operate at that voltage.

I think you might download FilterPro to design a better filter with gain than you have shown.

But if you can characterize that noise, you may be able to detect changes in it caused by people moving around (partially blocking door ways, etc.).

Once in the late seventies, when I was working on subwoofers, I asked the the chief developer at Perl Microphone Laboratories in Sweden about such microphones. He said it was no problem at all to tailor-make an electret microphone for any (reasonable) frequency range you want. He offered to make one "measurement grade" microphone for 5 - 2000 Hz, for about US$250.

The Panasonic capsules I'm familiar with have very good low response. I don't know the typical cutoff, and each is likely to be different. A cardiod type will not work. The omni capsules are somewhat sealed, but must leak in order to equalize the pressure over a short time. If the response must be flat, which i doubt, then they would have to be each calibrated.

greg

In working with these, you will notice ventillation pressures changing depending on the situation. THis may cause a problem with the detector system. These sounds which you can hear or feel at times, are around 2 - 5 Hz. or more. A person walking through a doorway will create this also, but will just be a single shot.

greg ]

Hi,

may be, may be not, You will know after testing ;-)

We made a testbox there was a low frequency generator with a few mW power and a subwoofer in a closed box. Then we drilled a little hole in it, where we stick the mic inside the box, sealed with a little rubber. We used a trapezoidal waveform to drive the loudspeeker @ 0.2 Hz and compared the measured result with a go nongo schema. It tooks about 5 seconds for a good selection. You see it very fast, whether the signal stays on the top or it falls down, because of leakage inside the Mic

Well we had very low loss.

Every OPamp will be good for your frequencies, Have a look @ supply voltage. LM324 will be not bad I guess, but I didn't look at your scematic today...

Marte

Large.

thanks, so i'm going to change OP AMP due to the voltage noise

oh, but how can I use that ? Like a microphone ?

John, Would you know where a non-AES member might obtain a copy?

...Jim Thompson

Thanks, John, You just saved me $20 ;-)

"associated amplifier has zero input impedance and a V out/I in response rising with frequency at 20dB/decade throughout the audio spectrum" was the information I needed.

I think I'm just going to "listen" for the school bus, with FLAT Vout/Iin ;-)

...Jim Thompson

Maybe a dumb question, but what would a woofer loudspeaker be like - used as a microphone for infrasonics?

I am not sure if it is sensitive enough for your use (a full atmosphere of pressure change produces only 2% of the excitation voltage as an output voltage), but you would use it somewhat like a microphone. Its advantage for low frequencies is that it compares the air pressure to an internal sealed vacuum, so its response really goes down to 0 Hz.

It contains a 4 resistor strain gauge bridge that needs an excitation voltage source and produces a pair of outputs, each at approximately half of the excitation voltage. When pressure increases, one output changes voltage in the positive direction and one changes in the negative direction.

You need a first gain stage that amplifies the difference between these two signals. You could just use one of the outputs, but amplifying their difference removes most of the errors caused by voltage ripple on the excitation input, and doubles the signal strength.

Go to