Subject:
|
Motion detector
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 9 Aug 2000 16:39:43 GMT
|
Original-From:
|
John Barnes <barnes@sensors.comNOMORESPAM>
|
Viewed:
|
716 times
|
| |
|
|
I recall seeing some stuff about the Eltec 442-3 pyroelectric sensor on
lugnet a few
months ago. I was somewhat put off by the device cost - $40 however.
Well, I noticed that my local grocery store now has a motion/intruder
detector made
by GE on the shelves for $20, so I bought one last night and took it to bits.
There is a pyro-sensor right in the middle of a circuit-board jam-packed
with parts.
It is a three pin device with a common/ground pin, a supply pin and an
output pin.
There are three ICs all with the markings deliberately obliterated.
However, I traced
the circuit out and it is obvious that the 14 pin chip is an LM324 (quad
op-amp) clone.
The circuit is quite simple. Two of the LM324 amplifiers bring the sensor
output signal
up to a usable level, the third is used as a comparator to trip the
tone/alarm generator
and the fourth forms a stabilizer for the power to the sensor.
The second chip is an anonymous 16 pin device, which together with the
little 4 pin
(tone generator?) I think are only used for the alarm or chime functions.
Pin 8 on the LM324 provides a very suitable point to pick off an output.
The case
comes apart very easily and I ran two wires out through the hole which can
be used
for an external 4.5 volt supply, (The unit runs on 3 x AA batteries.), a
common from
the battery negative side and signal from the diode next to pin 8 on the LM324.
The signal normally sits at about 1.5 volts, drops as a warm object appears and
rises when it disappears from the field of view. If you stand very still in
front of the
unit, it will slowly return to its quiescent output, but it is very
sensitive to even the
smallest movement.
The field of view is very wide, at least +/- 45 degrees, maybe as much as
+/- 60.
I ended up fitting a small paper tube over the sensor to limit it to a
narrower acceptance
cone - glass lenses do not work in the 8 - 15 micron wave band! I also
tried using
a technic gear motor to add a "chopper" in front of the tube to try to get
an relative
temperature difference signal, rather than just a change, but the bandwidth
of the sensor
is really high enough to make it usable.
I can see this device being transplanted into some glued together bricks to
make a
more aesthetically "compatible" sensor brick. The sensor and a suitable
dual op-amp
can be built to work within the RCX's available supply current limit for
active sensors.
I will work on that over the next few days.
A pyroelectric sensor is a really handy way to either avoid bumping into or
to find people.
Also, although I haven't had a chance to try it yet, but the "ac" nature of
the sensor I suspect
makes it a good candidate for being mechanically scanned to find "target"
heat sources. I
have yet to determine if the signal will go the opposite way if cold
objects are scanned. Perhaps
a "cold beverage" could be identified too!
JB
|
|
1 Message in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
