View Single Post
  #38   Report Post  
Posted to sci.electronics.repair
Jeff Liebermann Jeff Liebermann is offline
external usenet poster
 
Posts: 4,045
Default Sometimes, you just gotta get brutal ...

On Tue, 27 Mar 2012 21:57:17 -0700, josephkk
wrote:

Mid to far as i understand it. I want to look at thermal maps of
electronics that may be too hot to touch thermally or electrically.


I had the same idea and decided that $2,000 and up for commercial
thermal imagers was too much. My idea was to convert either a
supermarket scanner or laser printer rotating mirror imager, into a IR
camera. Basically, it's a "flying spot scanner" and is identical to
what we were using in early weather satellites to generate thermal
images of the planet.

I've built two failed prototypes so far, but plan to continue. The
major problem is the response time of the IR detector. The guts of a
PIR sensor will work, but is depressingly slow.
http://www.futurlec.com/PIR_Sensors.shtml
http://en.wikipedia.org/wiki/Infrared_detector
Typical thermopile sensors:
http://www.smartec.nl/pdf/DSSMTIR990X.PDF
http://www.smartec.nl/pdf/SMTIR9902SIL.PDF
I was getting about 50msec response time. At that speed, a 500 pixel
line would take 25 seconds to scan. A tiny 25Kbit image would take 22
minutes. This is almost tolerable for a tripod mounted thermal
camera, but not for anything hand held.

There was also a major problem with thermal noise. Cryogenic cooling,
using a Peltier junction device, with some manner of lens fog
protection, will be needed.

The idea of forming an image with inexpensive optics and sensors appeals
to me.


Research old technology on "mechanical television". Then, build
something to work in the optical range. Once it's working as a
visible light camera, switch to IR with appropriate lenses and
filters.

--
Jeff Liebermann
150 Felker St #D
http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558