Sometimes, you just gotta get brutal ...

Nope. Two different makes, two different owners and both had been in service and previously working just fine (apparently) for a long time ...

Arfa

Yeah, it was a bit of a last resort. The print on the power rail was actually quite good quality and size, so I wasn't expecting to get a track burn up. With only 4 volts on there, I was actually expecting something to just get hot to the touch or possibly smoke gently. But yes, you are right. There was certainly a good pyrotechnic display and a lot of smoke considering how small the cap was, in very short order. I'll give your soda straw idea a try next time.

Arfa

What worked the best was a sheet of black construction paper.

The idea is the same as in a telescope. You don't want anything that reflects light or heat. My first attempt was a brass tube and brass washer to fit over the lens. Even painted flat black, the IR thermometer would read the temperature of the brass tube, and not the heat source. I substituted ordinary paper, which proved to be IR transparent. I also stupidly left the brass washer, which again caused local heating. Eventually, the gears operating the brain engaged, and I realized that I needed something that was opaque to IR and had minimal thermal mass. Black construction paper was the best I could find. I wanted it corrugated to minimize reflections, but that proved to be difficult to build.

Several variations were built. The one that worked best was a conical like construction, that covered the entire lens on the IR thermometer end, and narrowed to almost a point on the other. However, attachment of this arrangement to the IR thermometer proved fragile, so I tried a soda straw shape, and washer cut from a shipping box. For light proofing, I buried the mess under a layer of black electrical tape.

I suggest you try the cone first as it's easiest to build.

I just ordered a B&D TLD100 heat leak detector, which seems to have more resolution over its narrow -30C to 150C range, than the wide temperature range common IR thermometer. $50.

Drivel: In the past, I've ranted on building several IR flying spot scanners using bar code readers and various pyrometers. I never really finished. I demonstrated that it worked, but was very slow to respond. I'm tempted to resurrect that project and build a real far IR imager.

Some ideas:

Moving the camera for scanning sucks.

I used some matt black rubber tubing .about 5mm diameter as a shroud but as I say the resolution , down to SMD dimensions , seemed to be no different with or without the tube

Mine has a short metal tube, or more like a cup, that is reflective. It seems to me that a reflective tube would work better at conducting the IR from the endpoint.

A reflective tube is better for conducting IR to the sensor. The problem is that it also picks up plenty of IR from the sides. The point of the black construction paper tube was to measure only the IR that goes directly from the hot component to the sensor, not the junk that comes in from other components. Ideally, that could be done with a lens that focuses to a point. Lacking one of those, the tube is a tolerable 2nd best.

It=20

IR=20

Now what do they make lenses out of for that wavelength range?

?-)

Germanium.

Near, mid, or far IR? Heat is mid to far IR. Unmodified CCD digital cameras do near IR. The ideal material for mid IR is germanium. However, that's rather expensive.

Much cheaper are various plastic formulations, that pass IR.

For tinkering, the IR motion detectors (PIR) lenses are cheap and easy:

There are also a tangle of band pass, low pass, and specialty filters.

Or, you can just be weird:

Jeff Liebermann wrote in news: snipped-for-privacy@4ax.com:

often,they use curved mirrors instead of lenses,for IR.

and sapphire for a window. ;-)

Hows about the lens from a an old CD laser ? Just thinking out loud ...

Arfa

Maybe. The common DVD laser diode is 632nm, which is in the visible light range. Laser printers and CDROM drives are 780nm, which is near IR. Since outside light sources do not enter into a CDROM drive, there's no need for an IR bandpass filter. I couldn't find anything on the optical characteristics of the CD or DVD lens. The dark ones might be a possible bandpass filter. The clear lenses are probably not.

DVD lasers operate at visible red as you say, whilst CD lasers are near IR. CD laser lenses often appear to be optically coated, but for what reason exactly, I'm not sure.

Arfa

Reflective coatings are usually to minimize reflections and/or to improve signal to noise ratio by reducing the optical bandwidth.

That may work well, but it is rather spendy. Any other materials?

?-)

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.

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

?-)

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.

Typical thermopile sensors:

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.

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, you are soo off the beaten path.

?-)

appeals

I have been thinking along somewhat different lines. Not close to real working ideas yet.

?-)