This absurdly clever "light painting" thermal imaging hack should be very Picaxable => http://www.maxjusticz.com/light-painting-with-temperature/
Thermal imaging flashlight !
- Thread starter manuka
- Start date
IronJungle
Senior Member
Well, if the shoe fits.... I proclaim that as pretty damn awesome.Yuk! I just knew someone would respond with the word "awesome"
Buzby
Senior Member
Clever ? - Yes.
Practical ? - I'm not sure.
The light that the 'real' camera collects is not just dependent on the RGB emitted by the LEDs, it is affected by the colour of the target as well.
So two objects at the same temperature, but different reflectances, will appear as different temperatures.
Or have I missed something ?
Practical ? - I'm not sure.
The light that the 'real' camera collects is not just dependent on the RGB emitted by the LEDs, it is affected by the colour of the target as well.
So two objects at the same temperature, but different reflectances, will appear as different temperatures.
Or have I missed something ?
Pongo
Senior Member
More or less, but the color of the object can only affect the intensity of the reflected light, it can't affect the color (LED's have narrow emission spectra) so the effect may not be terrible compared to the other approximations. I take issue with "an infrared thermometer that reads the average temperature over a small area", hmmm, that sensor has a 90 degree viewing angle so you have to be pretty close to the object to be imaging a small area.
Regardless it's pretty ingenious, and the guy can be justifiably proud of it.
Buzby
Senior Member
A black object, even if it is 'hot', will not reflect any visible colour. It will appear 'dark' no matter what RGB is shining on it.
Yes, but I still can't see the point.
Pongo
Senior Member
True, but black isn't one of the colors so it's somewhat obvious that a black area isn't measured.
Well, you could figure out if the dog was sleeping on the bed when you were out
erco
Senior Member
Different strokes. Perhaps not a "must have" for everyone, but a fabulously creative idea, and a great example of fresh, "outside the box" thinking. I really like it.
IronJungle
Senior Member
That is were I am at with the project. It isn't designed to replace a commercial detector, but I give the guy tons of street cred for coming up with a concept that is, well; pretty damn cool.
mrburnette
Senior Member
Having used the Melexis 90614 in several projects, there is some obvious flaws in this thermal flashlight. First, the part sold by Sparkfun is the Melexis' MLX90614ESF-BAA and the Melexis datasheet for this device specifies 100mS for acquisition settling time (The author indicates that the flashlight must be panned across the room rather quickly to blanket the room for a long exposure with a camera.)
Settling time, sec
MLX90614BAA 0.10
The other rather obvious issue is that the IC comes from the factory set as:
Even the $9 IR thermometers from Harbor Freight are set with the emissivity at 90 - 95.
Oh, for a quick refresher on black body radiation:
PICAXE code for my projects is available somewhere in the Forum and is based on the work of PHAnderson. Pictures are here:
PICAXE-Infrared-logging-Thermometer/ and here:
LASER-HEAT
- Ray
Settling time, sec
MLX90614BAA 0.10
The other rather obvious issue is that the IC comes from the factory set as:
WHY does this matter? emissivity_table
Even the $9 IR thermometers from Harbor Freight are set with the emissivity at 90 - 95.
Oh, for a quick refresher on black body radiation:
It's looks like a fun project to consume an entire Arduino, a $20 sensor, and lots of colored LEDS, but other than that it will not stand up to even cursory engineering scrutiny.
PICAXE code for my projects is available somewhere in the Forum and is based on the work of PHAnderson. Pictures are here:
PICAXE-Infrared-logging-Thermometer/ and here:
LASER-HEAT
- Ray
Last edited:
premelec
Senior Member
I've used a Black & Decker non-contact thermometer 'gun' for a couple of years which can be set for a ref temperature and then scans shine blue [cool] green [within ref] or red [hot] LED at what you are pointing at - there are various features I've liked or not - I subsequently bought a $15. Chinese model with scan recording hottest temperature of scan [and I wanted coolest!] which a also had a pretty much useless RED laser pointer - which can be turned off - to show where it's pointed... response seems pretty fast and if I could decipher the LCD display it could be made to scan walls etc... The Chinese model goes up to about 900 DegF and I pointed it at the sky and got -59 FDeg. Obviously because of emissivity and angle view etc these things are very rough and nevertheless useful. I use mine mostly to track wood stove activity and trend... The cheap whole house temperature camera has yet to appear and playing with non-contact thermometers and scans is fun and educational. I look forward to more such devices...
I recall that Heathkit decades ago made a very useful item for finding cold air leaks etc - an RC oscillator with loudspeaker or headphones where the R was a very small fast response thermistor so you would hear pitch changes as ambient air changes affected it - this could be easily done with a PICAXE READADC to tone output - finding tiny bead thermistors is sometimes hard and mounting them by their tiny wires is delicate... If you want air flow fluctuations as well as temperature variations you can run the thermistor hot by self heating...
I recall that Heathkit decades ago made a very useful item for finding cold air leaks etc - an RC oscillator with loudspeaker or headphones where the R was a very small fast response thermistor so you would hear pitch changes as ambient air changes affected it - this could be easily done with a PICAXE READADC to tone output - finding tiny bead thermistors is sometimes hard and mounting them by their tiny wires is delicate... If you want air flow fluctuations as well as temperature variations you can run the thermistor hot by self heating...
Last edited:
mrburnette
Senior Member
I seem to recall that one of my old Nissan automobiles used a hot-wire airflow (mass) meter as part of the input to the fuel injection computer... apparently, this has better response time than thermistors.
Hot_wire_sensor
- Ray
premelec
Senior Member
They're all subject to dust contamination and such - gotta love wikipedia - they cite cold wire sense units as well! Probably xmas light tungsten filaments could work well too... [now that we've converted to LEDs] - there are some thin film thermistors and the bottom line is - whatever works! I haven't looked at wikipedia for bolometers but I bet that's fun too... {"To measure the heat from a polar bear's seat at a half a kilometer..." as well as moon temperatures}... Happy New Year Ray...
mrburnette
Senior Member
Britianna states that Langley's most sophisticated platinum bolometer could
But... Wikipedia (to have the last word) states
*3: http://www.hao.ucar.edu/education/bios/langley.php
Personally, I think the polar bear's rear would have been a better test subject than the common cow butt, but one must make do with what is at hand.
Happy New Year, too!