Hi,
I'd like to add another question to your discussion of pir's.
I have difficulty understanding the workings of these devices.
My Q is: Is is it possible to use such a cheap PIR detector for remote temperature measurements? Or should I use the $200 ones?
Is it possible to get a linear output for say: freezing to boiling water?
>I have difficulty understanding the workings of these devices.
>My Q is: Is is it possible to use such a cheap PIR detector for remote
temperatu
>re measurements? Or should I use the $200 ones?
>s it possible to get a linear output for say: freezing to boiling water?
I dont know that these can be used for temperature measurement. My understanding
of the innards of these is the sensor has a number of separate areas on it, and
the lens is designed to focus the IR onto these areas. The sensor works by
detecting motion between the sensor areas, i.e. each area generates a non-DC
signal, which is amplified/detected. If more than one sensor area generates a
large enough signal for long enough then the output is activated to signify
movement.
I suspect that to detect temperature values, you would need to start with just
the sensor, and build your own DC amplifier for which you would need to generate
your own calibration procedure.
I have not actually used these sensors, this info is just what I have gleaned
from articles and discussions, so it would be a case of digging inside one
unless someone else on the list has actual experience with them.
The Non-Contact-Thermometers use expensive sensors for wide temp range
and good linearity a/or repeatability (years ago I bought few of those
sensors around $20 each). The PIR sensor is cut to have a best response
curve around the ambient temperatures, that goes from -10 to +50¡C (I
believe). The sensor itself should generate an analog output related to
the ambient IR. If you use a right lens you could narrow the IR target
area. Just the sensor itself should give you an analog type signal based
on the target IR. The PIR circuit itself, try to compensate and feedback
different levels in such way to discriminate small changes in the
received IR, independently of the ambient average IR received.
So, probably you could use just the sensor for your intentions, but I am
not so sure if it will be so linear in a wide range, including 100¡C
boiling water or above. In any way it would require a front-end
amplifier.
| Alan Pearce wrote:
| I have difficulty understanding the workings of these devices.
| My Q is: Is is it possible to use such a cheap PIR detector for
| remote temperature measurements? Or should I use the $200 ones?
| Is it possible to get a linear output for say: freezing to boiling
| water?
>The Non-Contact-Thermometers use expensive sensors for wide temp range
>and good linearity a/or repeatability (years ago I bought few of those
>sensors around $20 each).
I have a Fluke IR thermometer probe - it looks as if it uses a similar
sensor to the cheapo motion detectors. The major difference is that it has
a mechanical shutter that 'chops' the incoming IR signal at about a 5 Hz
rate. There is a little 68hc11 in there as the controller.
I think that something similar might work just fine using a sensor from a
motion detector. The shutter will be the problem.
I was under the impression that those PIR sensors usually had an IR
bandpass filter on the front which is centered on the IR from an object
in the temperature range of 90 to 110 deg F, for living animals. Will
this still pass 0 deg F to 212 deg F well enough?
> >The Non-Contact-Thermometers use expensive sensors for wide temp range
> >and good linearity a/or repeatability (years ago I bought few of those
> >sensors around $20 each).
>
> I have a Fluke IR thermometer probe - it looks as if it uses a similar
> sensor to the cheapo motion detectors. The major difference is that it has
> a mechanical shutter that 'chops' the incoming IR signal at about a 5 Hz
> rate. There is a little 68hc11 in there as the controller.
>
> I think that something similar might work just fine using a sensor from a
> motion detector. The shutter will be the problem.
>
> dwayne
>
>
> Dwayne Reid <.....dwaynerKILLspam@spam@planet.eon.net>
> Trinity Electronics Systems Ltd Edmonton, AB, CANADA
> (780) 489-3199 voice (780) 487-6397 fax
>
> Celebrating 16 years of Engineering Innovation (1984 - 2000)
>
> * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
> Do NOT send unsolicited commercial email to this email address.
> This message neither grants consent to receive unsolicited
> commercial email nor is intended to solicit commercial email.
>
I think there is one thing that is going to keep you from EVER being able
to measure temperature accurately using IR, and it is called emissivity.
The emissivity of an object determines how much energy it radiates at a
specific temperature. For example, two objects at 100degF, but one with
emissivity of 0.2, the other with emissivity of 0.8, will look radically
different to the IR detector. IIRC, numbers like 0.2 to 0.8 are quite
reasonable in the real world. The US Army battle uniforms have a
low-emissivity treatment that makes the soldier less visible to IR
viewers.
Reflection of ambient light will also be a problem. I worked on
pyrometry systems for measuring temperature of turbine engine blades at
Pratt & Whitney, makers of the world's finest jet engines(;-). We had to
use special filters and two sensors to cancel out the effect of flame
light reflecting off the blades. You could use narrowband filters, but
flame, sunlight and incandescent lights all have significant content at
the wavelength you are trying to measure.
The accuracy of our system was only +- 5degF at 1500 degrees. Higher is
better - it got better than +-2 at 2500 degrees, but below 1200 degrees
we couldn't even get a measurement. Of course, we were doing high speed
measurements (400 Ksamples/second), and we did achieve some pretty
remarkable things, like mapping the temperature of EACH blade in two
dimensions while the engine was running!
Today's sensors are sure to be much better, but I doubt you'll ever get
much accuracy at liquid water temperatures without using pretty good
equipment and knowing a lot about your target.
With that said, I know that CCD cameras (particularly the 'low-lux'
variety) are pretty sensitive at near-infrared. Perhaps by knowing the
background temperature and using high and low peak detectors, you could
extract some reasonable measurements. Of course, you'd either have to do
it in the dark or have a pretty good IR bandpass filter.
planet.eon.net>
