[Techtoolslist] Using Flir android thermal camera to track down short

[Dave McGuire]

On 9/10/19 10:58 AM, John Robertson wrote:
> What is interesting is that a number of digital cameras will work in the
> IR range, so I've thought of getting a visible light filter whenever I
> again order some lens from Edmonds to play with some of my cameras to
> see how far they actually see. I have used camera phones for checking IR
> diodes (including laser diodes on CD players) and the older iPhones
> worked for that - which was handy for troubleshooting trough optos.

  While CMOS and CCD imagers can be very handy for visualizing the
near-IR output from IR LEDs, LASERs, and such (750-900nm or so), this
approach won't get you very far for thermal imaging.  The IR response of
CMOS and CCD imagers doesn't extend up into the thermal IR wavelength range.

  The "infrared" spectrum is a broad segment of the EM spectrum, 750nm
to 1mm wavelength.  This is particularly broad, especially when compared
to the width of the human visual spectrum, which ranges from ~375nm to
~740nm for most people.

  Most consumer-grade Si CMOS and CCD imagers have useful unfiltered
response up to about 1050-1100nm (1.05um-1.1um) or so.  The type of IR
radiation that's useful for passive thermal imaging consists of
significantly longer wavelengths, from ~6um to ~15um...starting at ~6x
the wavelength at which typical silicon junction photosensitivity drops off.

  Due to CMOS/CCD sensors' lack of sensitivity in this region, thermal
imagers are usually implemented using a 2D array of thermopiles,
essentially tiny stacks of thermocouples, one per pixel, with solid
germanium optics.

              -Dave

-- 
Dave McGuire
McGuire Scientific Services, LLC
New Kensington, PA