Economy thermal imager?

Fri Aug 07, 2009 6:47 pm   

Phil Hobbs <pcdhSpamMeSenseless_at_electrooptical.net> wrote in
news:7vednQzvUYfK1eHXnZ2dnUVZ_jZi4p2d_at_supernews.com:


I didn't think much about it. That temp would look yellow I think. The red is
right though. I remember being taught to gauge the temperature of a pottery
kiln by the colour. There are even little gauges used for furnaces that are
adjusted till a small element fades to a match with the backgound colour, and
a reading is made of the temperature that way. I think some slack can be cut
re that white. It's wrong, but look at all the colours that pass for white.
You can stick a load of video monitors together in a room and see greater
errors than hers.

Fri Aug 07, 2009 7:46 pm   

Lostgallifreyan wrote:

Nowhere close. You don't even get out of the deep reds until 1500 K or so.

http://hancocktechnologies.com/Support/Black%20Body.jpg

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net

Fri Aug 07, 2009 9:21 pm   

Phil Hobbs <pcdhSpamMeSenseless_at_electrooptical.net> wrote in
news:Y4GdnXaF4Y0CGeHXnZ2dnUVZ_sKdnZ2d_at_supernews.com:


Or you could try this one:
http://wpcontent.answers.com/wikipedia/commons/thumb/b/ba/PlanckianLocus.png/
303px-PlanckianLocus.png

That shows you can follow the curve right down to around 800K and still
expect to see visible red. The only difference is that they resolve a tad
deeper into the red end.

This one shows that it's a matter of how much visible red, not whether it's
there or not:
http://www.techmind.org/colour/colourtemperatureannotated.png
In a furnace there's a lot of energy, even a 'cool' one has enough visible
red to see it.

So it's a question of how much, and also how sensitive we are to it.. Red is
the largest perceived part of the visible sprectrum, I mean it covers more
actual nanometres than the other colours, by far, and when looking at furnace
colours we're not looking at a single line either, which is why I posted that
second link. Also, we're likely to compare the red against a dark background.
(kiln view ports are designed to exclude external light).

Putting it another way, take a small butane flame with inducted air (blue
sharp-tipped flame), it can just about reach 1500K and will melt a copper
wire (ignoring flame colour) with a orangey-yellow ball forming at the end,
but if you try to put that copper on a red hot cooker ring it just sits there
oxidising. The ring IS evidently glowing a strong cherry red, but it's not
anything like the 1356.6K needed to melt copper. And yes, I have thought of
poor thermal transfer there, but I've seen overheated copper soldering iron
bits glowing, but not melting.

I'm sure a lot of this is very subjective but it's still true. That
subjectivity is why the hot-wire gauge is used to make simple meaures of
furnace temperature. A 'red' that glows against one hot glowing body will
look black against another brighter one. And within that very broad range
you'll likely find that every one of them appears red to someone.

Basically even a deep red looks cherry red if you have enough of it and
nothing for it to compete with..

Fri Aug 07, 2009 9:41 pm   

Lostgallifreyan <no-one_at_nowhere.net> wrote in
news:Xns9C60E3699100Azoodlewurdle_at_216.196.109.145:


Meant hotter, not brighter.. (and darker, rather than 'black')

Fri Aug 07, 2009 9:45 pm   

Lostgallifreyan <no-one_at_nowhere.net> wrote in
news:Xns9C60E6D33A5F5zoodlewurdle_at_216.196.109.145:


And just to show how it can't really be said to be either right or wrong,
what colour IS 'cherry' red? Some of them look deeper and darker than the
little red spot seen in a bare IR laser diode... The dark ones taste better
too.

Fri Aug 07, 2009 10:54 pm   

On Fri, 07 Aug 2009 11:29:27 -0400, Phil Hobbs
<pcdhSpamMeSenseless_at_electrooptical.net> wrote:


What color temperature is maroon? I bet her face is red.

John

Sat Aug 08, 2009 6:25 am   

On Fri, 07 Aug 2009 08:28:06 -0500, Lostgallifreyan
<no-one_at_nowhere.net> wrote:


The human eye is very bad in detecting absolute colours, due to the
"automatic white balance adjustment" so you really need a reference
chart at the same illumination level to reliably compare colours.

Apparently the Draper point at 798 K seems to be some standard for
visibility of hot objects, but I have not found any references how
this is actually determined.

To get a good general overview how the radiation behaves on short
wavelengths, it is a good idea to plot the black body radiation on a
log/log scale, e.g. figure 10 in
http://ceos.cnes.fr:8100/cdrom-98/ceos1/science/baphygb/chap3/chap3.htm

Look at the 800 K (Draper point) curve, which has a peak at about 3,5
um and look at intensity at 0.7 um (the nominal limit of human
vision), the magnitude is about 5 orders of magnitude (50 dB) below
the peak.

At 1500 K, the peak is at 2 um and the intensity drops only 1.5 orders
of magnitude (15 dB) at 0.7 um but the absolute level at 0,7 um is 5
levels of magnitude (50 dB) above the value for 800 K at that
wavelength.

Compare this with the sensitivity of the eye
http://hyperphysics.phy-astr.gsu.edu/hbase/vision/efficacy.html

For normal illumination levels (photopic vision) the nominal 700 nm
limit is somewhat arbitrary, since the sensitivity is doubled (+ 3 dB)
every 10 nm when going from 770 nm down to 670 nm, a total increase
about 30 dB. In that wavelength band the 800 K black body radiation
intensity drops about 12 dB and the 1500 K radiation drops about 5 dB
in that region, so the stimulus is strongest near the shortest end
(670 nm) of that band.

However, the absolute level of 800 K and colder objects is very low
and a dark adapted eye (scotoptic vision) is required, however the eye
is insensitive to deepest red at these levels. Practical values
starting at 700 nm and the response doubling every 10 nm down to about
580 nm.

The scotopic vision is of course black and white and if in a dark room
an object is heated, a glow will be observed, when there is sufficient
power below about 700 nm, but you can just tell that something is
glowing, but you can not determinate the colour due to the scotopic
vision.

When the temperature is further increased, the absolute power levels
below 770 nm are increased significantly and sooner or later the
photopic vision will be smoothly activated and it becomes possible to
determine colours.

The level at which the transition occurs, depends of the absolute
level reaching the eye and hence also of the angle of view.

Paul

Sat Aug 08, 2009 3:30 pm   

Paul Keinanen <keinanen_at_sci.fi> wrote in
news:pp2q75tunrt3tr1fncp0t6ee6p4ou9isod_at_4ax.com:


That's what I was getting at (rather verbosely). That and the fact that
cherries actually have so many colour variants you can plot most of the
visible black body spectrum with them.

Sat Aug 08, 2009 3:49 pm   

Paul Keinanen <keinanen_at_sci.fi> wrote in
news:pp2q75tunrt3tr1fncp0t6ee6p4ou9isod_at_4ax.com:


That's interesting because when I looked at that page I saw I was apparently
wrong when I said that to spot a dimly glowing overheated soldering iron bit
it helps to use the scotopic vision. It IS less sensitive to light at
longwave red. Even so it did seem to help, and your comment that scotopic
vison is needed seems to back this up. Maybe it has more to do with the
effort of trying to see it, balancing the two types of vision by arranging
the object at certain angles to our centre of vision. Maybe our brains can
make more of it by this comparison than by either type of vision alone.

Anyway, looking at the rest of what you wrote, it seems that a claim to see a
glow from something as 'cool' as 500°C is viable (though the standard you
mentioned states 10° higher), if seen in darkness. I thought I'd managed it
at 450°C once but I'm not prepared to back that up. Next time I have a
controlled way to set that temperature I'll try it. I did once try this with
a kiln observation hole but I can't remember what the temperature was when it
became visible. It was thirty years ago, and I think I got distracted and
revisited the kiln too late to see the first light anyway.

Sat Aug 08, 2009 5:10 pm   

wikipedia for John William Draper
http://en.wikipedia.org/wiki/John_William_Draper
says:

In 1847 he published the observation that all solids glow red at
about the same temperature, about 977 F (798 K), which has come
to be known as the Draper point.

It has a couple of references if you want more info.

--
These are my opinions, not necessarily my employer's. I hate spam.

Sat Aug 08, 2009 11:50 pm   

hal-usenet_at_ip-64-139-1-69.sjc.megapath.net (Hal Murray) wrote in
news:OrednYczvpQRLODXnZ2dnUVZ_j1i4p2d_at_megapath.net:


I'm not sure that citing it as any guide to precision matters much. I just
tried that test I mentioned, I used a Weller WSD-81 'station' and an LR-21
iron, set at 450°C. I could see it before my eyes were fully dark-adapted,
but could not resolve colour. I could see it pretty much equally with
photopic or scotopic vision, done by arranging wide angle changes off-axis
from central vision, but I could only resolve any detail when it was at the
centre. Not much either, I couldn't see the tip shape very clearly but I
could see the silhouette of the heavy guard coil wire as it passed across it.

What did surprise me was that a small near-IR sniffer I built could not see
it. It's not very sensitive but I thought it ought to if I could, given the
stuff I've used it for before, so eyes are obviously rather good at this.
Couldn't see a thing at 400°C though.

Sun Aug 09, 2009 7:50 am   

On Wed, 22 Jul 2009 16:55:30 -0700 (PDT), "miso_at_sushi.com"
<miso_at_sushi.com> wrote:


I know what thermal imaging is. I do not need a primer.

Sun Aug 09, 2009 8:15 am   

In <Xns9C6288007E4Azoodlewurdle_at_216.196.109.145>, Lostgallifreyan wrote:

This appears to me that you were seeing the very-dimly-glowing tip and
maybe the likely-slightly-hotter heating-element-area mostly with scotopic
vision.

450 C is 723 K. The surface over the heating element was likely
somewhat hotter, likely mid to upper 700's K.

The 798 K "Draper point" appears to me to be the threshold of achieving
significant stimulation of photopic vision to the point of seeing color
more than gray.


What wavelength range is your "near-IR sniffer" good at, and what
radiant power density in that wavelength range is lowest it will sense?
If you report this, I can figure temperature necessary to achieve this.


I am not surprised - that is 673 K.

I consider myself optimistic at thinking that I may be able with best
dark adaptation to dimly see with blurred outlines and no color,
incandescence at 700 K.

I calculate the following values of candela per square centimeter to
photopic vision corresponding to the following temperatures low enough
to achieve glow so dim as to be seen primarily by scotopic vision despite
spectral content at wavelengths long enough to favor photopic:

775 K: 3.25 E-7 (s/p ratio is .131 on a scale where 555 nm
monochromatic yellow-green = 1)

750 K: 1.22 E-7 (s/p ratio is .119 on the above scale)

725 K: 4.27 E-8 (s/p ratio is .107 on the above scale)

700 K: 1.40 E-8 (s/p ratio is .096 on the above scale)

===============

Temperature at which s/p ratio achieved by a blackbody is "unity" as in
same as that of 555 nm yellowish-green monochromatic narrowband light:

2093 K,

which I have determined that a USA-usual 120V 100W lightbulb of
"Big 3 brand" and rated to last-on-average 750 hours and to
produce-on-average 1670-1750 lumens at 120 volts, to achieve at around
53-54 volts, maybe give or take another volt.

I am aware that a few lower-current longer-life vaccum-containing
incandescents have a fair chance of having color temperature this low or
slightly lower. However, it appears to me that more-usual for tungsten
incandescent lamps with design current low enough to be served better by
vacuum than by argon-nitrogen gas fill, along with design life expectancy
around 2,000-3,000 hours, is for color temperature to be not far from
2360 K (which achieves s/p ratio of 1.16 on scale where 555 nm achieves
unity).

The USA-usual 120V 100W incandescent with rated average life expectancy
of 750 hours, rated to produce 1670-1750 lumens, and having CC-6 or CC-8
filament, and of "Big 3 brand", appears to me to typically have color
temp. of 2870 K (2865 K by a slightly older definition revised by a
redetermination of one of 2 constants in the "Blackbody Formula"), appears
to me to achieve s/p ratio of about 1.42.

I hope the above supports my impression that human photopic and scotopic
vision have nonlinearities that differ from each other. I sense that
there is dynamic range compression achieved by both, less for human
photopic vision than for human scotopic vision.

I sense that human scotopic vision achieving greater dynamic range
compression than human photopic vision does, is the explanation for seeing
barely/hardly/minimally/dimly - incandescently-hot objects with color
"degrading to gray" (my words) as temperature decreases from about 800 K
or so to mid-700's K or so.

- Don Klipstein (Jr) (don_at_misty.com)

Sun Aug 09, 2009 11:01 am   

don_at_manx.misty.com (Don Klipstein) wrote in
news:slrnh7sqbd.qk8.don_at_manx.misty.com:


That figures, but does scotopic vision also have resolving power at the
centre of vision? I was definitely seeing better imaging there..


Agreed. I'd have tested that, but the iron only went so far.. :)


Not really sure.. I built it from various RF plug parts and a couple of
button cells driving directly in series, an phototransistor and a high-
brightness green LED that happened to put out well at very low currents.

I think the phototransistor is an IR type from Maplin, 3mm package with
spectral peak response at 880 nm. Current varying betweem 0.1 mA and 20 mA
(which is why it seems so suited to direct LED drive). Those currents are
specified for 1000 lux and Vce of 3V but it saw less by one LED Vf, and its
sensitivity was further reduced by an IR pass filter made from the head of a
dark IR LED. I'd cut it off, polished (crudely, on a paper laid on glass),
then placed nose to nose to focus from a 5mm diameter region for best
coupling into the phototransistor while excluding any visible light.

This thing has been generally useful for seeing IR in various conditions
where it it would be within low visible intensities if it were visible at
all.


Yeah, just confirming that I couldn't. I could have tested at what point
between the two I lost sight of it, but it would taken a long time and not
proved a lot.


Likely enough, I think. When I tried the 400°C test I thought I had seen
something but I tried the obvious test of sheilding it with my hand, and the
consciousness of the heat being shielded and the slight phosphenes (or
whatever they're called when slight neural generated effects occur) dominated
it. Even if I did see it it was lost in noise and loss of focus. So it was
doubt, rather than absolute nonseeing, so I am sure that at 425°C it ought to
be visible.


I read all that but I don't think I could have done it. Lets just say I make
an unusually good guinea pig. :)

I agree with the claim that scotopic vision has greater dynamic range though.
That fits my experience directly.

Possibly related to your comment on colour degrading, or perhaps not, is the
IR lased diodes. When I have seen them (5mw single mode types, looked at
directly at range greater than 1 foot, almost certainly eye-safe conditions)
they go beyond the wine or rubt reds of visible types. The impression is of a
maroon colour. (Yep, maroon, I ought to have mentioned it earlier in this
thread, no? > And what makes THAT interesting is that it implies a bluish
interpretation! Interesting given that deep violet at the other end hints at
a reddish interpretation of what is just a very deep blue. Life is full of
cylic/linear transforms, but this is the single weirdest one I know.

Sun Aug 09, 2009 8:14 pm   

In article <Xns9C627A4B01391zoodlewurdle_at_216.196.109.145>,
Lostgallifreyan wrote in part:

<SNIP because I only have time at this moment to respond to one point:>


This reminds me of sensing some barely visible IR LEDs as emitting a
slightly orangish shade of red. I have seen their spectra with a
diffraction grating, and I know that I am seeing only very long
wavelengths past 750 nm, generally past 800 nm. I have seen the slightly
orangish shade of red in the long wavelength spectral region shown by the
diffraction grating, with the shorter wavelength end of the spectral band
appearing to me closer to pure red in color.

I have heard of this being called "color reversal" or "infrared color
reversal".

As for the soldering iron tip becoming more focused but still appearing
gray when viewed with central vision: I suspect that there are some rods
in central vision, in smaller and more closely spaced clusters to achieve
better resolution. Possibly photopic vision plays enough of a role to
increase sensation of resolution, however. However, I do see dimly
gray-appearing incandescent objects becoming invisible when viewed
entirely with the most central degree or two of vision. And when I look
at such an object that partially falls into the supposedly rod-free
most-central degree or two of vision, I sometimes tend to see the whole
thing anyway - maybe I get prone to "seeing things" under such conditions.

- Don Klipstein (don_at_misty.com)