On Fri, 13 Oct 2000, David VanHorn wrote:

> >
> >Is this a good solution for low power apps, like battery
> >powered where the PIC will sleep until some time, wake up
> >to check timers and input pins, then sleep again?
>
> That's a little tougher. You can get micropower op-amps, and scale the
> resistance values up so little current flows.
>
> It's the old limbo question, "How low do you need to go?"

Warning I know nothing about this..., but that hasn't prevented comments to the
piclist before.

It would seem that some kind of AGC would be required in addition to a threshold
detection. Otherwise, you'd have to have a device sensitive enough to trigger
with low background noise and at the same time in other situations capable of
not triggering with large background noise. That's hard. One way I've done this
for DTMF decoding were the problem is to capture sinewaves in the presence of
noise (people talking), is to build a cheap AGC which varies the threshold on a
comparator. So one side of the comparator has the sound input, while the other
side of the comparator has the AGC controlled input.

Now, a really cheap AGC circuit is a half wave rectifier driving a low pass
filter. As the back ground noise increases, the rectified voltage will increase
and raise the threshold and thus require a larger signal level to cause a
change.

Now for the sound input, you'll most probably want to band pass filter it. For
example, you wouldn't want little bumps to set it off. Perhaps high pass
filtering isn't necessary for your application (but it was for mine).

Scott

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu