DIY Thermal LPM for under $50

Xplorer877 said:

^^^
Looks great man! Good job!

I've been lagging it a bit with this project but as of today I've ordered all the needed parts. So I should have mine done soon.

-Tony

Click to expand...

Xplorer877 said:

I got TEC today but I'm still waiting on a few things.

BTY, the TEC has a HOT side and a COLD side when it's powered. Normally the HOT side will be heatsinked when it's used for cooling. But in this case we're essentially using it backwards. So the COLD side will get adhered to the heatsink correct? Then the laser heats the HOT side?

-Tony

Click to expand...

MarioMaster said:

It won't matter which side you use, depending on which side you heat, the polarity of the output signal might be reversed but you can just swap the wires.

If you want to be sure, just put the tec on a heatsink and apply heat with your finger and check the polarity of the output with a DMM.

Click to expand...

HIMNL9 said:

Ehm ..... not totally right, sorry .....

First consideration to do: the reading surface of a thermal sensor is NOT resistant to anything, to infinite abuses ..... one thing is using a 200W thermal head with ceramic plate, another is using a 2W reading head with special ultra-thin plate ..... using an 1W FOCUSED to a burning point on one of the sensitive (and delicate) low-power high-resolution plates, can easily end in damaging the surface of it, that, considering the usual cost of thermopiles, is not exactly a funny thing .....

The same thing is more valid for TEC sensors, that have a coating made with paint, usually ..... regardless the quality of the paint, and the presence of the ceramic substrate for draw the heat, it can be damaged also more easily from a focused beam with burning ability (i made the same with one of my tests, too, thinking the paint was good ..... it was "high temperature paint for exhaust tubes" for motorcycles, and they reach a high temperature, so i thoughd it was resistant to burning lasers too, but i burned it on the TEC surface with only 200mW of 405nm, focused for burn, so, just imagine) .....

The first "good" rule, then, is: always try to defocus the beam a bit, when you read a laser ..... if you have a small reading surface thermopile, like the ophir ones sold with some LPM also here, or a small TEC sensor, like the lasersbee one, is always a good thing to defocus the beam to AT LEAST half the diameter of the sensor, also more, if the laser is hold with a stand ..... this prevent unwanted plate / coating damages

Second consideration to do: there is a phisical difference from a thermopile and a TEC cell, that gives you different results in heat distribution and heat draining from the measuring area, so the point where you target the reading area IS important for a correct reading, other than the reading time .....

In more clear words: the TEC cell have the plates made in ceramic (ok, ok, a special ceramic, but i suppose you don't need the formula ), that is more thick and highly thermally conductive, much more than the thermopile reading plate ..... also, the cells are distributed on all the back surface of the plate ..... from the counterpart, the structure of a TEC cell, also if not used for cooling, is much, much more heat transfering than a thermopile reading plate ..... this made it relatively insensitive about the exact point where you shine your laser, cause all the heat transfered to the plate from the laser, is spreaded really quickly on all the surface, and influence in almost equal mode all the cells (except than shining it on a corner, but also this is someway compensated) ..... so, almost regardless than the fact that you shine the laser exactly on the center or not, the reading is practically the same (some minimal differences are always presents, but are so small that does not influence the reading in a really effective way) ..... at the same point, anyway, the more quick transfer of heat from the reading plate to the "cold" plate, due to the big mass of the junctions, tends to "equalize" slowly the temperatures, falsing the reading in long-cycles readings ..... anyway, this does not have a big influence reading low-powered lasers, only reading high-powered ones, indicatively from over 500mW and up ..... this can be, anyway, minimized using a big or very efficent heatsink on the "cold" plate (a thermal compensation can be helpful, too, for long readings, but is not indispensable at hobby level)

On the other side, a thermopile have a totally different structure, especially the sensitive ones that are able to read very few milliwatts ..... builders had to become to a compromise from sensitivity and dimensions, and this means that the reading plate is, sometimes, extremely thin and delicate, made from special materials for maximize the heat transfer speed and at the same time kept small enough for become influenced from very low powers ..... but, you know, infinites does not exists ..... regardlass from what material you use, there's no materials that have an infinite heat transfer speed ..... and also, for the constructive principle of those sensors, the reading elements are a lot of bimetal thermopiles made with very thin metal plates, connected in serie for increase the reading, and placed all around the circumference of the plate .....

This cause a difference in the reading for different zones of the reading surface ..... cause the heat transfer from one place to the whole plate is not infinite, the thermal resistance of the plate itself, causes a higher reading when you shine the beam on the borders of the plate, where the bimetal junctions are directly posed, and a minor (but more correct) reading, when the beam is shined exactly on the center of the plate, cause in this way the reading is the effective power reading of all the heat, equally transferred to all the bimetal junctions (yes, is not that your laser is more powerful if you shine it on the borders, you are simply doing the reading in the wrong way, doing this )

So, for a correct use of a thermopile, is required, usually, that the beam is defocused at least on half of the reading surface, and that it will be directed and kept more centrally possible on the reading plate.

And as final consideration, regardless if you are using a thermopile or a TEC or a piezothermic element, all these elements, more or less, are sensitives also to thermal radiating power (purely heat) generated from the measured devices ..... so, for a real power reading, is not enough to correctly "zero set" the meter and shine the laser on it ..... the better and more precise reading is obtained following some simples, but indispensables, rules:

1) no air currents around the sensor (it can draw heat from the hot or the cold face, falsing the reading).

2) never keep the sensor with the hands (the heat added from the hands, slowly change the difference of the temperatures from the two faces, making the reading imprecise and incostant).

3) no big lamps or heating elements around the sensor (not too good making a reading, having a 100W halogen lamp half meter from the sensor, or similar )

4) NEVER keep the lasers close or touching the reading assembly (the heat from the laser body can false your readings, also 40 / 50% sometimes ..... i always laughed seeing sellers that shows on their sites 400 or 500 mW green lasers, and "proof" the output power showing a pic where the power is read keeping the laser in contact with the reading head ..... i can bet whatever you want that, in these cases, at least 40%, if not more, than the readed power, is infrareds AND heat from the laser body ) ..... a decently precise reading can be obtained keeping the laser AT LEAST 20cm away from the reading head ..... also the double, if possible, is also better

5) ofcourse, frequently check the zero set of the meter ..... thermal derive, thermal transfers from the sides, changes in temperature of the ambient, also just the radiant heat added to the items from the body of the reader, in extreme cases, always changes the thermal model scheme of the ambient and reading setup ..... if you have the patience, and really want that your measure is the more correct possible, check the zero set after each measure, and if the time is not a problem, do at least 3 measures for each laser you're measuring, and made a medium value from the 3 measures (i know, this is almost a lab procedure, but, it's the more correct possible, for hobby level ..... we don't have NIST labs in the pocket or in a desk tire, after all, so we must do that what we can, with that what we have ..... )

(sorry for the long post )

Click to expand...

lasersbee said:

Excellent explanation of the heat transfer and active surface area
differences between Radial type Thermopile like the OPHIR 20C-A

Thermopile Heads and the Wafer type Thermopile like we use on the
LaserBee products..:gj:


@ lazeerer and Xplorer887.... good job on your Heads...
Any test results yet...:thinking:


Jerry

Click to expand...

lazeerer said:

ThanksMeans alot coming from some one that makes meters at the level you do., Ive played around with the LPM but not really tested yet because i need to fully calibrate it.

Also I will come pair the 2 different TEC's when the new one comes in. I put together a much cleaner board then my first attempt.

The Tec Right now that iam using it works awesome when reading my 445nm 1amp. laser and reads quickly but with my over spec'd 5mW green laser it takes a little time for it to read the output of the laser. So i figured with the smaller TEC and i believe more junction i will get a faster reading.

So when i get the new TEC i will test both of the TEC's. I have sent 1 laser to 2 different member here that have accurate meters. it is on the second member right now So i figured that should be enough accuracy to calibrate my LPM.

Here is a short video showing it working very good:"Just click the image"
It is not calibrated so it is way off it will be calibrated as soon as i receive my laser back and i will post some more videos also with the new TEC.

Hope you guys like.

Click to expand...

jib77 said:

Im getting some prototype PCBs made. The board is only 1" x 0.5" so it will fit in the slot of a TO-220 extruded heatsink. They should arrive in a couple of weeks.

Top Layer:

Bottom Layer:

Heatsink:
Mouser Link

TEC I'll be using:
TEC Link

Click to expand...