how do hand held IR detectors measure temp of cold objects
Joseph Mack
at 25C inside a hand held gun.
here's one of these devices (for about $30) a Ryobi IR001 SKU 699814
I have no problem with IR detectors sitting in liquid N2 or liquid He
detecting low temps (this is how IR telescopes work), but how does one
of these things at room temp find the temp inside your freezer?
I looked up how IR detectors work
http://en.wikipedia.org/wiki/Mercury(II)_cadmium(II)_telluride
(they have a narrow band gap) and surprisingly to me, considering
Stephan-Botzmann T^4 law, objects at only mildly elevated temps (eg your
roast dinner) have peak output at a wavelength only about 30 times that
of light. They seem to be detecting IR at 5-10microns.
http://www.fas.org/spp/military/program/warning/sbirs-brochure/fig-02.jpg
These IR guns have a resolution of about 2degC.
Do they have a lens? If so what refracts and transmits light at IR?
The specs on these things say that they look at an area of 1:12 through
1:8 (depending on device), so a 1:12 device looks at a disk of 1" when
positioned 12" away. It sounds like the device has a collimator rather
than lens.
If the sensor at 25degC is looking at a cold object at -20degC, is the
sensor getting cool? Why isn't the sensor overwhelmed with 5-10micron
radiation from the inside of the handheld device?
Geostationary satellites provide IR pictures of the earth, measuring
cloud top temperatures down to -40degC. eg
http://www.weatheroffice.gc.ca/satellite/animateweb_e.html?imagetype=satellite&imagename=goes_nam_1070_s_..................jpg&nbimages=1&clf=1
Do these satellites have cooled IR detectors? If so how are they cooled?
Do they have a finite amount of liq N2 or liq He and when that runs out
the satellite has to be deorbited (to make room for the next one)?
Thanks
Joe
--
Joseph Mack NA3T EME(B,D), FM05lw North Carolina
jmack (at) wm7d (dot) net - azimuthal equidistant map
generator at http://www.wm7d.net/azproj.shtml
Homepage http://www.austintek.com/ It's GNU/Linux!
Martin Brown
> More precisely, how does an IR detector detect a temp of -20C when it's
> at 25C inside a hand held gun.
>
> here's one of these devices (for about $30) a Ryobi IR001 SKU 699814
>
> http://www.homedepot.com/webapp/wcs/stores/servlet/ProductDisplay?storeId=10051&productId=100674438&langId=-1&catalogId=10053&PID=1319015&cm_mmc=CJ-_-shopping_com-_-D25X-_-100674438&URL=http://www.homedepot.com/webapp/wcs/stores/servlet/BuildLinkToHomeDepot?linktype=product&id=100674438&cm_mmc=CJ-_-shopping_com-_-D25X-_-100674438&cpncode=23-66403802-2&AID=10368321&cj=true&srccode=cii_13736960
>
>
> I have no problem with IR detectors sitting in liquid N2 or liquid He
> detecting low temps (this is how IR telescopes work), but how does one
> of these things at room temp find the temp inside your freezer?
>
> I looked up how IR detectors work
>
> http://en.wikipedia.org/wiki/Mercury(II)_cadmium(II)_telluride
>
> (they have a narrow band gap) and surprisingly to me, considering
> Stephan-Botzmann T^4 law, objects at only mildly elevated temps (eg your
> roast dinner) have peak output at a wavelength only about 30 times that
> of light. They seem to be detecting IR at 5-10microns.
>
> http://www.fas.org/spp/military/program/warning/sbirs-brochure/fig-02.jpg
>
> These IR guns have a resolution of about 2degC.
>
> Do they have a lens? If so what refracts and transmits light at IR?
>
> The specs on these things say that they look at an area of 1:12 through
> 1:8 (depending on device), so a 1:12 device looks at a disk of 1" when
> positioned 12" away. It sounds like the device has a collimator rather
> than lens.
IR lenses are expensive. It will use some kind of mirror arrangement -
take apart an IR burglar alarm sensor to get a rough idea.
>
> If the sensor at 25degC is looking at a cold object at -20degC, is the
> sensor getting cool? Why isn't the sensor overwhelmed with 5-10micron
> radiation from the inside of the handheld device?
If the thing is designed right then it can measure the amount of in band
radiation coming from the source. Everything near the sensor needs to be
a mirror finish at the active wavelength(s) and the sensor should see
the source reflected in as many directions as possible. There will come
a point where it cannot measure lower temperatures because its own
emissions and signal to noise limitations blind it.
If you remember the first detection of the microwave background
radiation at 4K it had been collected in and bounced off a huge metal
microwave horn at room temperature before reaching the cooled detector.
Indeed one of the initial working hypotheses about the noise was that it
was dielectric material (notably pigeon droppings) in the horn so they
had to give it a very good clean just to be sure.
Part of the noise visible on ordinary analogue TV when mistuned off
channel is due to the microwave background radiation and no component is
cooled there.
> Geostationary satellites provide IR pictures of the earth, measuring
> cloud top temperatures down to -40degC. eg
>
> http://www.weatheroffice.gc.ca/satellite/animateweb_e.html?imagetype=satellite&imagename=goes_nam_1070_s_..................jpg&nbimages=1&clf=1
>
>
> Do these satellites have cooled IR detectors? If so how are they cooled?
> Do they have a finite amount of liq N2 or liq He and when that runs out
> the satellite has to be deorbited (to make room for the next one)?
I expect they use thermoelectric cooling if they need to use any active
cooling at all. Cunning designs with heat pipes to shunt heat to the
dark side of the spacecraft might be good enough.
Regards,
Martin Brown
Joseph Mack
> Joseph Mack wrote:
>> More precisely, how does an IR detector detect a temp of -20C when
>> it's at 25C inside a hand held gun.
>>
>> here's one of these devices (for about $30) a Ryobi IR001 SKU 699814
Hi Martin,
> If the thing is designed right then it can measure the amount of in band
> radiation coming from the source. Everything near the sensor needs to be
> a mirror finish at the active wavelength(s) and the sensor should see
> the source reflected in as many directions as possible. There will come
> a point where it cannot measure lower temperatures because its own
> emissions and signal to noise limitations blind it.
I don't know the bit about metal/mirror finish. So that's like the
parabolic dish in a radio telescope then.
> If you remember the first detection of the microwave background
> radiation at 4K it had been collected in and bounced off a huge metal
> microwave horn at room temperature before reaching the cooled detector.
> Indeed one of the initial working hypotheses about the noise was that it
> was dielectric material (notably pigeon droppings) in the horn so they
> had to give it a very good clean just to be sure.
yes know the story well. I've been to the horn.
>> Do these satellites have cooled IR detectors? If so how are they cooled?
>> Do they have a finite amount of liq N2 or liq He and when that runs out
>> the satellite has to be deorbited (to make room for the next one)?
>
> I expect they use thermoelectric cooling if they need to use any active
> cooling at all. Cunning designs with heat pipes to shunt heat to the
> dark side of the spacecraft might be good enough.
so they have a refrigerator dumping out of a heat pipe?
I'm a ham radio operator and have done moonbounce and am familiar with
the concept of the front end FET having a noise temp of 10degK, while
being at 300K, but I have to take it on trust. I can't derive the
equations myself. However I'm less happy about accepting that an IR
detector doesn't detect its own temperature.
My device has arrived and I've been playing with it. It's spec'ed from
-32 to 380C and gets expected temperatures for lights, hot coffee and
-20C for the back wall of the freezer compartment of the refrigerator.
I was quite surprised to point it at cloudless blue sky (the stuff
that's at 9000degK for the visible spectrum personal cameras) to get a
temp of -40C. (What that means when the device is spec'ed to -32C I
don't know, but it must be cooler than -32C). The blue sky is from
Raleigh scattering, not Stephan-Boltzmann black body and presumably has
little 5-10mu IR in it. I expected the coldest part of the sky to be at
90deg from the sun (max polarization), but the coldest part of the sky
was directly overhead (thinest atmosphere?) while the sky above the
horizon was about 5 deg warmer (the murk we look through with a
telescope at low elevation). At night the (clear) sky is only 5 deg
cooler than during the day. I haven't looked at the base of clouds yet.
Pointing at the sun (30deg elevation), only got 50degC. The gun's viewed
area is a 1:12 cone (ie tan(cone angle)~0.1) and with the sun's
diam=0.5deg (tan(0.5deg)=0.009~0.01). Squaring the ratio shows the sun
occupying about 1% of the field of view. With the sun at 5500K, then I
guess I'm in the right ball park.
Walking around a lake just before sunset I got about 20C for the sun,
and 10C for the reflection of the sun in the lake. (the air temp was
just above freezing). Still pointing at the lake, on either side of the
sun, reflecting the blue sky, I got temps of about -10C, rising to about
5C when pointed at reflections of buildings or trees. So water is a good
reflector of 5-10mu IR at angles grazing through to about 30deg.
Surprising to me was the temperature of the unpainted side of metal road
signs (eg stop signs). Regular metal was seen to be its expected temp,
but the road signs (whether facing the sun or away from it in shade),
were -20C or below (possibly I got a reading at -37C but I forget).
After some experimenting, I found I had IR detector pointed up and
seeing the reflection of sky in the metal sign (although the metal was
brushed aluminium and I couldn't see the reflection of the sky in
visible light). The signs were above my head and I had to point the IR
gun upwards to get the temp of the sign. I found a sheet of brushed
stainless steel (phone box) and by angling the IR detector
to see the reflection of the sky, I got -25C or so. When pointing
the IR gun straight at the brushed stainless so it would see the
reflection of my clothing, the temp went back to 5C. So brushed metal is
a good reflector of 5-10mu IR, even if it doesn't reflect well at
visible wavelengths.
Presumably a nice polished mirror finish inside the detector
chamber wouldn't be seen at all.
The collimator has black baffles, presumably for the same reason
we have them in telescopes.
An interesting toy
justbeats
"I was quite surprised to point it at cloudless blue sky... to get a
temp of -40C"
The generally accepted temperature of a clear (night) sky is about
-70C
Hence ground frosts when air temp is above freezing
Joseph Mack
> The generally accepted temperature of a clear (night) sky is about
> -70C
got a reference for that? Here's a graph of the temperature profile
http://www.windows.ucar.edu/tour/link=/earth/images/profile_jpg_image.html
I haven't a clue what an IR detector looking up through this profile
would see. I would assume it would mostly record the hotter atmosphere
at the bottom, but that would depend on the amount of water, which on
a clear night would be low.
> Hence ground frosts when air temp is above freezing
yes understand
Joe
--
Joseph Mack NA3T EME(B,D), FM05lw North Carolina
jmack (at) wm7d (dot) net - azimuthal equidistant map