Another Simple Laser Power Meter

[MarioMaster]

I tried this with my 2 stage cascade refrigeration system and two stock aixiz modules, it does make a pretty noticeable difference, but then it's getting a lot colder than a TEC would

 

[HIMNL9]

Just some note about using TECs for LPM.

Remember that, opposite to thermopiles, TECs are thermal transfer devices ..... i mean, a thermopile is made intentionally for AVOID the more possible to transfer heat from the reading disk to the base plate (the serie of thermocouples wires at the border of the disk are thinner than hairs, so they can read also low differences in temperature without change the "base" temperature, that is the reference one) ..... instead, TECs are built with massive junctions, with the purpose (ofcourse) to transfer the temperature from one face to the other with the better possible efficency.

This means that, also when used as sensors (unpowered), they transfer heat from one side to the other from 1500 to 2000 times more than a common thermopile ..... this don't mean that they cannot be used as LPM sensors (see as example the good Laserbee one), this mean only that you cannot build a LPM that can be used stable for long times measures, cause part of the temperature that is produced from the beam on the "hot" face, is transferred to the "cold" one, slowly equalizing the temperatures (and reducing the reading in the meantime).

Another thing, is important to never touch the heatsink too, cause also your fingers change the temp and false the reading ..... i made some tests with my smallest TEC and a little heatsink from a chipset (a small old one), and for long time reading the output value was changing ..... with a 50mW green, (pushing out 66mW), keeping it pointed on the TEC for long time, the reading was slowly but constantly decreasing (it gone from 65 to 45 mW in approximatively 4 minutes) ..... and keeping the heatsink with 2 fingers when reading, in a second test, it passed from 65 to 30 mW in approximatively 1 minute (cause my fingers was heating up the heatsink)

There are basically 2 ways for avoid this, or, at least, for try to reduce the effect of this thermal derive ..... the more complex one, that involve the use of a second TEC, and a thermostatic circuit, is valid only if you're planning to build a very stable unit for lab use ..... the second solution is to use a big heatsink with a decent thermal mass (and i mean BIG, like, an old CPU one, more is massive, better it is) ..... placing the TEC on it, placing some insulant foot (gum or felter) on the side that you use as bottom, and overall left it stabilize with ambient temperature and not touching it during the measure, can contribute to reduce significantly the thermal derive, and maybe permit you to make some mid time reading (2 or 3 minutes, maybe), without too much errors.

Another improvement can be trying to repair the more possible the TEC from air flows ..... a sealed chamber is not applicable, cause the front glass reduce the reading, but a piece of tube (plastic or cardboard), with a plate glued in front and a hole in the plate can help a lot ..... like this:

..... just some crazy ideas, as usual

 

[Bluefan]

Mariomaster, are those two equal red diodes? the cooled one on the right seems to me more orange. At what temperature is it?

I did have a try with a 20x20mm TEC, these things are indeed sloooowwww. First try was with a 1mm thick copper plate in front to spread the heat more even. After 1 minute it was within 10%, after 1.5 minutes within 2% and stable after 2 min. Drift was about 0.5 mW, but this includes changing ambient light, so the actual drift may be lower. So even this large TEC was pretty acurate, it responded pretty fast to a led key chain 15cm away. I didn't shield light away in any way.

Second try was without the copper plate, in 40 seconds acurate to 5%, 60 seconds 1%, stable in 90 seconds. Drift even a bit smaller, close to 0.25mW.
It's slow, but not that bad actually. I'd love to have a smaller tec!

Edit: HIMNL9, TECs are not made to conduct thermally, this would defeat it's purpose of keeping one side hotter (or colder) than the other. They're made to conduct electrically, for several amps or more. Too bad you can't do that with a high thermal resistance, which is why TECs conduct heat so good.

 

[Asherz]

That's some good info there HIMNL, cheers.

I've ordered a two piece 15mm TEC 7W cooler, they were only cheap and after talking to bill it looks like these are going to be way to big.

BUT... I think I have managed to source these for a very reasonable price, and you can buy only 1 if you want, and they have one that goes down too 2mm in width and height

 

[HIMNL9]

HIMNL9, TECs are not made to conduct thermally, this would defeat it's purpose of keeping one side hotter (or colder) than the other. They're made to conduct electrically, for several amps or more. Too bad you can't do that with a high thermal resistance, which is why TECs conduct heat so good.

Sorry, maybe i've expressed bad what i mean.

I don't mean that TEC is made for conduct heat on purpose when unpowered (like heatpipes or metal) ..... i meant that, due to their constructive technology and phisic characteristics, they conduct heat much much more than thermocouple junctions in the thermopiles, ALSO when unpowered, so they still transfer part of the heat generated from the beam to the other surface, causing errors in long time readings.

If you want to have a practical demonstration about what i mean, do an easy test with a small TEC, also without a laser ..... connect the TEC as an LPM sensor to your DMM, then just place a finger on one of the surfaces, and wait ..... with the temperature of your finger, the effect is much more quick than with a laser, so you can see really easily that, keeping the finger on the TEC, it start with a reading, then slowly change, when the heat from your finger is transferred through the material of the junctions to the other side and tend to equalize it (you can never get zero reading, cause there's always ambient dissipation, but you must be able to notice the thermal derive the same).

 

[brtaman]

Nice write-up HIMNL9...agreed. :beer:

 

[MarioMaster]

Mariomaster, are those two equal red diodes? the cooled one on the right seems to me more orange. At what temperature is it?

They are both the same diode, stock 5mW red axiz module diodes. As seen on the multimeter, the cooled module is at -100F

 

[brtaman]

They are both the same diode, stock 5mW red axiz module diodes. As seen on the multimeter, the cooled module is at -100F

Damn nice cascade then! -75 C ?

What gasses are you running? DIY? Do you also/used to OC? I had some pretty nice results with my single stage (WR for some time :santa

 

[MarioMaster]

Damn nice cascade then! -75 C ?

What gasses are you running? DIY? Do you also/used to OC? I had some pretty nice results with my single stage (WR for some time :santa

Yeah, I used to - back when you actually had a shot at world records without gobs of cash at your disposal. PC phase systems are what got me interested in HVAC as a profession which I am currently going to school for

The cascade has r1150 and r290 (for oil return) on the second stage, then r502 on the first stage. I didn't build that one, it's a commercial lab unit that I got in a trade from a PL member. I have built single stages though, one of which I had down to -56F :san:

 

[brtaman]

Yeah, I used to - back when you actually had a shot at world records without gobs of cash at your disposal. PC phase systems are what got me interested in HVAC as a profession which I am currently going to school for

The cascade has r1150 and r290 (for oil return) on the second stage, then r502 on the first stage. I didn't build that one, it's a commercial lab unit that I got in a trade from a PL member. I have built single stages though, one of which I had down to -56F :san:

Nice man.

I quit for the exact same reason...-.- My final OC generation was the age of the 6800GT and single core A64's (754/939)...ah the good ole days.

 

[Benm]

remember, the higher the temperature-difference, the more effective the heat transfer! so a hot host will get rid of the heat better. if then the diode is even cooled down by a tec, it might be quite effective! sure, again, the tec doesnt like it hot..

Then again... you will quickly arrive at the point where the host gets too hot to handle for portable operation. This is one of these things where size does matter, no matter what way you try to take around the problem

The -75C comparison picture looks cool - color shifted way to the orange. I suppose you could replicate the result using dry ice instead of expensive cooling equipment - for those that want to give it a try

 

[MarioMaster]

dry ice wavelength shift (it's not quite that noticeable to the eye)

 

[Krutz]

wow! thats an impressive shift to the orange! on the photos it looks like pretty close to 635nm already!
i will use a 20mW 635nm to begin with, though.
chris (heruursciences) had those badass 620nm (ish) diodes, which have to run at low temperatures, and i read that the regular 7mW 635nm diodes from ebay go up to red when they are overdriven to 20mW (which they do constantly, in my tests) because of heat..

i will report back, eventually!

..and like the many info here!

manuel

 

[ZRTMWA]

Bumpitty Bump!

This was a good thread! I think I've found a link to the TEC's you had billg519:

http://www.kelk.co.jp/english/thermo/pdf/thermo.pdf

Note: To view that PDF you make have to download the Japanese font' symbols package from adobe, it took like 2min to download.

Anyways, I think you had the in Table 1: Micro Modules, Single-stage standard, MH Series, Model Number KSMH023. You can see the sizes of this module match up in the second chart.

Supposing we could order any of the TEC's from this chart would a smaller one be better or maybe a Two stage, multi purpose, or multi stage TEC? I have no idea what these are, just trying to throw out ideas!

I'm not an electronics guy so I'm having trouble understanding how you would make one of these with the 50uA and multiple resistors to get a multirange meter. What is a multi range meter first of all?

Edit: This company may have gone out of business . The website noted at the bottom of the PDF file cannot be found currently.Guess we'll have to keep looking on ebay.

 

[Asherz]

Nice link ZRTMWA,

I got my 20mm TEC from ebay through the post a few weeks ago, I slapped it on a CPU heatsink and sharpied it black.

I don't have a laserpower meter so there's no accurate way of saying if this worked or not, but I have a green O-like module that supposedly was 171mW's, I used that as calibration and with my PHR@120mA I got a reading of 91mW's, this seems a little on the low side but plausible.

I then decided to measure my argon, and that's where the readings go a little wierd.. I couldn't get the internal meter to match my meter at all, I should of been getting a reading of 32? But got something ridiculously lower.

Without a power meter it's hard to tell if this works.

BTW, here is a link I found to a 2mm! TEC that looks perfect, and doesn't have to be ordered in large batches! (you can order just single's)

00411-9G30-20CN

looks good, I think when I get some more money and time I'll order one.

 

[lasersbee]

If you want to make a LPM... save yourself some money...
the TEC is a bit too small... 0.088 " X 0.167"
That is barely large enough for a laser beam..
The Specs link are down... but at that size I don't think
there are enough PN junctions to put out any usable current..
I may be wrong...

Jerry