artist needs help- difficult mic circuit

[raphihell]

Before I begin I should say that we are pretty familiar with most
basic electronics but are really stumped here.

What we are looking for is a circuit that would use very little power-
be battery operated and as small as possible. The idea is to a mic
that would be "listening" to the room it would store a charge relative
to the mic signal- or ideally from the small charge generated by the
mic itself- when this value crossed a certain threshold it would
trigger a serperate circuit- lets say a buzzer. the triggering charge
would be reduced to zero and it would all start over again.

In practice what this would mean is that put on a shelf in an a closed
office- the buzzer might go off once or twice a day assuming the
telephone was used - and if there was an office party it would be
going off all the time.

Now the trick here is that this thing should ideally be able to be
"on" for a few months without draining the battery.

We were thinking of a simple biased capacitor charged by a mic- with
the cap discharging at a certain level- but are having problems.

Any ideas?
Thanks
Raphael and Peter

 

[no_one]

why don't you steal the guts from one of those singing bass fish things;
http://www.howstuffworks.com/singing-fish.htm


"raphihell" <eots@riseup.net> wrote in message
news:13be8fca.0408290818.50f53256@posting.google.com...

Before I begin I should say that we are pretty familiar with most
basic electronics but are really stumped here.

What we are looking for is a circuit that would use very little power-
be battery operated and as small as possible. The idea is to a mic
that would be "listening" to the room it would store a charge relative
to the mic signal- or ideally from the small charge generated by the
mic itself- when this value crossed a certain threshold it would
trigger a serperate circuit- lets say a buzzer. the triggering charge
would be reduced to zero and it would all start over again.

In practice what this would mean is that put on a shelf in an a closed
office- the buzzer might go off once or twice a day assuming the
telephone was used - and if there was an office party it would be
going off all the time.

Now the trick here is that this thing should ideally be able to be
"on" for a few months without draining the battery.

We were thinking of a simple biased capacitor charged by a mic- with
the cap discharging at a certain level- but are having problems.

Any ideas?
Thanks
Raphael and Peter

 

[Robert C Monsen]

"raphihell" <eots@riseup.net> wrote in message
news:13be8fca.0408290818.50f53256@posting.google.com...

Before I begin I should say that we are pretty familiar with most
basic electronics but are really stumped here.

What we are looking for is a circuit that would use very little
power-
be battery operated and as small as possible. The idea is to a mic
that would be "listening" to the room it would store a charge
relative
to the mic signal- or ideally from the small charge generated by the
mic itself- when this value crossed a certain threshold it would
trigger a serperate circuit- lets say a buzzer. the triggering
charge
would be reduced to zero and it would all start over again.

In practice what this would mean is that put on a shelf in an a
closed
office- the buzzer might go off once or twice a day assuming the
telephone was used - and if there was an office party it would be
going off all the time.

Now the trick here is that this thing should ideally be able to be
"on" for a few months without draining the battery.

We were thinking of a simple biased capacitor charged by a mic- with
the cap discharging at a certain level- but are having problems.

Any ideas?
Thanks
Raphael and Peter

One approach would be an analog 'integrator' circuit to keep track of
the charge. Unfortunately, integrators are usually built with
capacitors, and its pretty hard to keep a capacitor charged with tiny
amounts of charge over a long period of time like hours or days; the
cap will leak.

Another way would be to count how many times the microphone went over
a particular input excitation, and use that count as the trigger. You
could do this with CMOS logic chips; perhaps a counter chip activated
by some kind of analog level meter circuit on the microphone. Once a
particular bit on the output went high, fire off the buzzer for a bit
using a C555.

A tiny microcontroller, like a PIC 12F675 or ATTiny would also be a
good choice for this kind of circuit; you could use the internal
comparator to wake it up when the mic was active, and keep track of
the count in its eeprom. It could output the 'buzz' as well, so you
could use a cheap piezo disk rather than a buzzer. Thus, it might be a
single chip solution. It would be very low power, but would require
you to write the software for it. Keeping the microphone powered would
probably take the most of your energy budget. However, you might be
able to power the thing using a small solar panel charging a little
rechargable cell. Offices usually have pretty good light.

This is going to be a fairly complex circuit to pull off, I'm
guessing. Good luck!

Regards,
Bob Monsen

 

[Anders F]

"no_one" <no_one@earthlink.net> wrote in message
news:F8pYc.1914$8d1.375@newsread2.news.pas.earthlink.net...

why don't you steal the guts from one of those singing bass fish things;
http://www.howstuffworks.com/singing-fish.htm

I'm sure they're using simple level triggers.

I think Robert is on track in his suggestion of a low power microcontroller.
A mic, a rectifier and a cap before the AD of the uC. The uC should wake up
once in a while to measure the level and integrate. Probably it'll prove OK
to do a discrete mesaurements once or twice per second - allowing the mic to
be powered down.

/Anders