DIY Thermal LPM for under $50

[diachi]

Hmmm, I have most of the components for this - might need to give it a go.

Do we have a source for small TECs?

 

[MarioMaster]

Question on the components. You list 3 10k pots, but there are only 2 shown in the circuit. R6 in the circuit shows as a regular 10k resistor.

R8 is the third potentiometer, it's used for the battery indicator.

Hmmm, I have most of the components for this - might need to give it a go.

Do we have a source for small TECs?

Tetech has a good selection for reasonable prices.
I would recommend TE-31-1.0-1.5 - TE Tech Products
It's a 14.8mm square module



Standard Modules - TE Technology

 

[gregma]

thank you on the mod. You also need to add r6 to the pcb component section. It doesn't list the 10k ohm resistor

 

[diachi]

R8 is the third potentiometer, it's used for the battery indicator.



Tetech has a good selection for reasonable prices.
I would recommend TE-31-1.0-1.5 - TE Tech Products
It's a 14.8mm square module



Standard Modules - TE Technology

Awesome, thanks!

 

[Benm]

How do you calibrate these things though?

I reckon you could shove a resistor dissipating 0.1 to 1 watt onto the laser facing side of the TEC and measuring that result to get a baseline, but how does that compare to results using real lasers?

Also, what absorptive paintjob do you do on the TEC?

Designing a scaling amp to turn the tec output into mv on a multimeter is fairly trivial when you know the correlation, especially when it's linear. The practical construction and calibration are harder nuts to crack.

 

[paul1598419]

The paint I used is Rust-Oleum high heat matte finish. Black, of course. It advertises heat resistance to 2000 degrees F. It also seemed to work well for a wide range of powers. That is really the key.

 

[MarioMaster]

How do you calibrate these things though?

I reckon you could shove a resistor dissipating 0.1 to 1 watt onto the laser facing side of the TEC and measuring that result to get a baseline, but how does that compare to results using real lasers?

Also, what absorptive paintjob do you do on the TEC?

Designing a scaling amp to turn the tec output into mv on a multimeter is fairly trivial when you know the correlation, especially when it's linear. The practical construction and calibration are harder nuts to crack.

Yeah this project requires a stable laser of a known power to calibrate it, but even without a known laser you can at least measure the relative power differences between your lasers.

One could also send the completed LPM out to someone who has a calibrated LPM. It's all up to what you want to do.

 

[gregma]

Ok. All parts have been ordered. Should be delivered in the next 2-3 weeks. I'm ready to go! Got duplicates of just about everything except the TEC. I ordered this TEC. I looked at the TEC you recommended MM, and though it was only $14.00, when I went to check out, they included a $71.10 "surcharge", and $15.37 shipping cost. It came out to just over $100.00 for one TEC!! The more product I ordered, the lower the "surcharge" became.

 

[paul1598419]

gregma, that is a very tiny TEC at 8.4mm X 8.4mm. I have always ordered 15mm X 15mm. I hope you will be able to focus all the light from your lasers on such a small surface area.

 

[ARG]

8.4mm ones will work, just have to be more careful with the coating. They'll have a better response time anyway.

 

[gregma]

gregma, that is a very tiny TEC at 8.4mm X 8.4mm. I have always ordered 15mm X 15mm. I hope you will be able to focus all the light from your lasers on such a small surface area.

My understanding after reading many lpm threads here is that any portion of a laser beam has the same power. So if you only get half the beam onto the TEC you will have the same power reading as if you got the other half. The power of the beam does not increase or decrease based on how much of it is tested. Of course I'm very new to this so I could be wrong.

 

[paul1598419]

The other thing that occurred to me is the overall TEC p-n junction density. A 15mm TEC had 225 square mm, while the 8.4mm TEC has only 70 square mm. The number of p-n junctions in the package is very important to the power it will be able to endure and also the sensitivity.

 

[Benm]

Yeah this project requires a stable laser of a known power to calibrate it, but even without a known laser you can at least measure the relative power differences between your lasers.

One could also send the completed LPM out to someone who has a calibrated LPM. It's all up to what you want to do.

What would be against calibrating it from a known heat power source?

You could, for example, mount a SMD resistor to the front of the TEC and run that off a known voltage, so you'd know the power as well. As this is not much different from shining a laser onto that TEC, you could perform a course calibration that way.

Let's say you'd get a 1 ohm, 1% resistor and affixed that to the tec, appling 1 volt at obvously 1 amps. This would be a decent reference to how hot the laser facing side of the tec would get when hit with a 1 watt laser.

This is not a very accurate way of going about things, but i think with some practical procedures could be good enough to calibrate it within 10% or so. 10% (in)accuracy isn't much of a performance compared to professional calibrated meters, but it's good enough to tell bunk lasers from proper ones.

I'm not sure i can get that same coating material here, but probably something similar. At these price ranges i'd be happy to build a copy of the design to see how it fares.

 

[paul1598419]

I don't know that it would work the same as the resistor is dissipating heat in all directions even though you have only one surface against the TEC.

 

[gregma]

The other thing that occurred to me is the overall TEC p-n junction density. A 15mm TEC had 225 square mm, while the 8.4mm TEC has only 70 square mm. The number of p-n junctions in the package is very important to the power it will be able to endure and also the sensitivity.

You are correct, density of the junctions is very important. However having a larger surface area does not equate to having a more dense number of junctions. An 8mm TEC with 49 junctions is more densely packed than a 14.8mm with 49 junctions. Unfortunately I haven't seen any TEC datasheets that quantify the junction density nor even the count.

 

[MarioMaster]

Ok. All parts have been ordered. Should be delivered in the next 2-3 weeks. I'm ready to go! Got duplicates of just about everything except the TEC. I ordered this TEC. I looked at the TEC you recommended MM, and though it was only $14.00, when I went to check out, they included a $71.10 "surcharge", and $15.37 shipping cost. It came out to just over $100.00 for one TEC!! The more product I ordered, the lower the "surcharge" became.

Wtf, I never had any surcharge from them when I ordered before, but that was a few years ago. I guess they go out the window then for a TEC source.

My understanding after reading many lpm threads here is that any portion of a laser beam has the same power. So if you only get half the beam onto the TEC you will have the same power reading as if you got the other half. The power of the beam does not increase or decrease based on how much of it is tested. Of course I'm very new to this so I could be wrong.

Not the case here, if you have a laser beam with a power of 1 watt, if you pick off half of the beam you only have 500mW. Were that not the case you would break the laws of thermodynamics =P

So for anyone skimming, the ENTIRE beam must be on the TEC for an accurate reading.

What would be against calibrating it from a known heat power source?

You could, for example, mount a SMD resistor to the front of the TEC and run that off a known voltage, so you'd know the power as well. As this is not much different from shining a laser onto that TEC, you could perform a course calibration that way.

Let's say you'd get a 1 ohm, 1% resistor and affixed that to the tec, appling 1 volt at obvously 1 amps. This would be a decent reference to how hot the laser facing side of the tec would get when hit with a 1 watt laser.

This is not a very accurate way of going about things, but i think with some practical procedures could be good enough to calibrate it within 10% or so. 10% (in)accuracy isn't much of a performance compared to professional calibrated meters, but it's good enough to tell bunk lasers from proper ones.

I'm not sure i can get that same coating material here, but probably something similar. At these price ranges i'd be happy to build a copy of the design to see how it fares.

It's a good idea, but you'd need to be careful to not muck up the surface of the TEC before painting it, that and you don't know the absorption percentage of the coating.

You would also probably want to use wire as thin as possible to reduce heat conduction through the power leads. (that and the 1 volt would need to be measured at the resistor to avoid the voltage losses from the wires)