Thermal pile

[mixedgas]

crude NTC bridge
normally you'd hook a opamp across A-B and feed its output back to Vin,
but thats a pain to engineer without the NTCs being on my bench.

See attached, and yes I cant draw.

Steve

[steve-o]

What kind of NTC? 1 ohm at 25C? 100 ohms? I wish you could do this just with an ole-fashion cad-sulf -photocell... no way I guess -right? What about a solar cell?

[The_Doctor]

NTC? Meaning negative temperature coefficient thermistor? And why negative anyway? Thermopiles use a positive coefficient. If you want a thermistor, you could use a small platinum thermistor, bare chip kind, ceramic with a platinum trace on it, no cladding. (Pt100 sensors have a PTC). Those are fast, but the main problem is that they're very small, unless you want to start spending money. Might be worth it though, they're very stable, the readings will be repeatable. But I'd rather paint a TEC black, you have more surface, and it's a lot easier to do. If you want speed, build a differentiator circuit and mix the output with the main signal. No need to accelerate the fall of output, a TEC changes much faster when it's cooling down.

[steve-o]

NTC? Meaning negative temperature coefficient thermistor?

Yes.
They're flat black and when you shine a laser on it it changes the resistance. Very slow though. Which brings me to a point: What is laser power anyway?--the apparent brightness to the eye? No. So when we want to know the actual power of a laser there's too many variables. Most of us dont work in the 10.6 micron range for light shows obviously so this brings up an interesting question. Should there be a power meter that reads like human ocular senstivity or would that be against all laws of "laserology"?

[mixedgas]

Oh gawd, Dont get us into that lumens/candella $hit.
You can get "corrected" photocells, but why.....

I'd think about a 1Kohm at room temp thermister, NTC or PTC, Mr Wheat s.. tone Bridge dont care@! On a lowl resistance the changes are harder to see.

Stev

[steve-o]

Why not? A power meter for a CO2 laser cant be used to measure 5W of 532...right?

[The_Doctor]

Why not? A power meter for a CO2 laser cant be used to measure 5W of 532...right?

If it's a thermopile, it can. They really are that insensitive to wavelength. Main reason for using watts is it's far easier to quantify. Too many variables in 'brightness', beam width, wavelength, conditions it's seen in. Remember all the troubles people have in working out colour balance equations? Now imagine trying to pin that down well enough to calibrate a meter. Not a chance... People only do it when that's exactly what they want to measure, and it's rarely done in a 'real-world' state. When measuring lumens for publication of data such as for Rebel and Cree LED's, very exact conditions are set up. Far easier for our purposes just to measure the units that all that effort would be founded on if we had to do it. Watts and joules are easy, in comparison.

[Laser Ben]

Actually, a bit of a correction Doc, about the testing of Lumileds Products, I have worked assembling the systems used to test EVERY Luxeon. It consists od a 10" Integrating sphere with a TE cooler for the led. The power/spectra is measured in 100uS pulses at 100mA increments up to the rated power. The power is measured using a calibrated Spectrometer. I know that Cree probably uses an identical setup as the engineer who created the current Lumileds setup left last Christmas for Cree. What I do know is that Cree does NOT test every piece they produce. They test 10 on every 4" wafer. That means about 10 our of 9850 die get tested.

[The_Doctor]

Different issue, but interesting. (I was saying that lumen measures depend on precisely defined conditions of known power, wavelength... And for candelas I think angular distribution and distance to target are also controlled to get specs for those).

I'm not so concerned that Cree don't test every one. If they (or anyone else) establshed that there was significant difference between LED's to indicate that Cree weren't sampling enough to get equality of process, that would be a problem, but it doesn't seem to be. I'll grant that the Lumileds Rebel RB100 is a damn good LED though, better than a Cree XR-E. Tiny, but still easier to mount and connect, probably (Got two, in flashlights, but none to play around with, yet). Never did get round to mounting my XR-E's yet, but I will soon, I want to make a work light to put under the drill mount on a precision drill stand, it's no fun trying to centre a 0.7 mm carbide bit on every hole by eye, with a flashlight in one hand and the machine vice steadied by the other. Can do it ok, but I won't pretend I enjoyed it.

[marconi]

I couldnt help but posting this..
Every time I see the title of this thread, I keep thinking of a steaming pile of dung.. I know, not very scientific of me, but just wanted everyone to know I do have a sense of humor.

I still prefer the power output measurement as this means the most to me.
I guess its left-over from my Ham days of quantitive results.
But at multiwavelengths, this might not be so good.
But, I do like wheatstone bridges..most of all HP and Booton power meters were based on this.

Of all the types laser power meters Ive owned and used..I always go back to using the peltier based ones.