Digicam IR conversion [NEW: Conversion complete with pix] / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Digicam IR conversion [NEW: Conversion complete with pix]

previous 1 2 3 next

Move Thread

LAN_403

Dr. Shark

Mon May 01 2006, 10:23AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Thermal Imaging:
As far as I know this works with a matrix of pyroelectric crystals, which generate a hugh voltage in response to slight temperature fluctuations, similar to the piezoelectric effect. Motion detectors work by the same principle, I think.
The trouble is that you need optics that are transpartent to heat, which would probably be gallium compounds as in CO2 laser mirrors. You would also need to peltier-cool the sensor.

It might be possible with an ordinary CCD, someone should try it!

HV Enthusiast

Mon May 01 2006, 06:01PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

CCD cameras are only sensitive to near IR light. They will not detect or record heat sources unless the source is extremely hot.

Bjørn

Mon May 01 2006, 07:42PM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

CCD cameras are only sensitive to near IR light. They will not detect or record heat sources unless the source is extremely hot.

The leakage current of the charge wells is directly related to the temperature. It will be possible to detect reasonably small temperature changes at very low resolution if the exposure time is very long. 20 deg C temperature difference makes a measurable change in the leakage current.

Here is the picture to prove it, the purple areas are strongly affected by moderate heat after 30 seconds exposure:

HV Enthusiast

Mon May 01 2006, 11:05PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Bjorn,

That effect is due to proximity heating effects of the nearby electronics (i.e. amplifiers) on the sensor. Thermal imaging of a similar heat source at a distance (optically) from the camera sensor will not be possible with a CCD sensor even under long exposures.

Bored Chemist

Tue May 02 2006, 05:51AM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

How big a distance?
the sun is quite capable of burning your fingers if you focus it on them, even after it has been through an IR filter.
I don't think you wil get good images this way but I guess you should get something.

Penguin7471

Tue May 02 2006, 06:46AM

Registered Member #71 Joined: Thu Feb 09 2006, 08:23AM
Location:
Posts: 63

The talk about thermal 'vision' got me interested.

My Konica Minolta DIMAGE has proved quite sensitive to IR light, even with the filter in place. See this page for photos:

So I turned on the soldering iron, let it heat up for 10 minutes, then focused the cam with IR filter on it in a pitch black room and left the shutter open for 30 seconds (max. exposure time on my camera).

The photo is simply pitch black, with absolutely no detectable trace of any light, even after it is adjusted with software. I have heard myths about 'seeing' a typical clothes iron with an IR filter. I have yet to try this, but I doubt a clothes iron is much hotter than a soldering iron anyway?

Dr. Shark

Tue May 02 2006, 11:30AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

As I mentioed earlier, ordinary glas lenses are completely opaque for FIR. All optics would have to be removed (maybe FIR pinhole camera? just kidding) for the radiation to even reach the sensor. Heck, I might even try this myself, with my butchered coolpix 2500. Now I just need to get a ZnSe window to block out the visible light and pass only the FIR. Guess I have to check eBay...

HV Enthusiast

Tue May 02 2006, 11:43AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Again, the type of IR light captured by CCDs is known as near IR light and is concentrated around 700 to 950nm wavelength. The heat needed to be captured for thermal imaging up around 5000nm. Plus as Joe said, the optics themselves are opaque to this "heat" required in thermal cameras.

Also, the when an object gets hot enough, it does emit light at the wavelengths between 700 and 950nm and this is what your camera picks up, not the thermal heat at 5000nm.

Bored Chemist

Tue May 02 2006, 07:40PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

ZnSe is transparent to at least part of the visible spectrum. Ge or Si filters/ lenses are used for far IR work.
Thin glass is transparent to bits of the far IR spectrum, but you would still need a filter.
With enough power this radiation would warm up the CCD and (whether it's by upsetting the electronics or raising the leakage rate) give rise to an image. Granted, you might need a CO2 laser to get enough "light", but it ought to work.

Ben

Tue May 02 2006, 09:48PM

Vigilatny
Registered Member #17 Joined: Thu Feb 02 2006, 02:47PM
Location: NL
Posts: 158

In order for a CCD to work, as designed, the photon it detects must have a minimum energy(or maximum wavelength) for a given applied voltage. This energy/wavelength is determined by the work function of the material the device is constructed out of and the applied voltage. The applied voltage is limited by the device's construction, too much and it would fail(like overvolting a capacitor). Combined these two factors give the absolute limit, based on the photoelectric effect. The are other limits before this based on S/N.

Within the device heat can also generate electrons(vice the photoelectric effect), sometimes called dark current. But this is different. One way to think of it is that it is caused by the fact that for a given temperature the poisson distribution says there will be occsionally very high energy particles. These electrons are not the same as those generated by the photoelectric effect, as they can solely be generated by the distribution within the device, as opposed to thermal imaging which is representative of the object being imaged. Theoretically you could image something this way, but the first step would involve conducting the heat to the CCD in some manner.

CMOS imagers work similarly, but I believe the limit is imposed by the bandgap of the semiconductor used.

One might be able to use a CCD to see IR/temperature using some clever trickery such as thermoreflectance.

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob