Cheap! 100mW/700mW 613nm (Orange) Laser Module

[Farbe]

Hey,

I will just chime in and help you understand the concept better:

As for the vacuum vs gas fill:
The gas inside the tec housing is just a normal heat conductor, just like iron would be. (however much worse of a conductor)
So sucking the gas out results in "less conductor" so in less heat conduction overall.
Choosing the gas depending on "heat conduction ability" e.g. thermal resistance will lessen the heat load because the gas will be a worse conductor of heat (thus less heat will "reach" the tecs surface).

The pressure of the gas is therefore directly related to the heat load. As well as chamber design. Imagine the iron example from before:
A 1" long iron bar would conduct heat quite fast. Heat one end with a blowtorch -> burn you fingers fast on the other end.
On the other hand, a 100" long iron bar could be heated with a blowtorch and leave you with cold fingers on the far end.

So this example shows: distance is also a factor in heat conduction.

There is also another problem that arrises, thinning the gas will actually increase the "mean free distance". This counters the "less conductor less heat transferred" part a bit. Have a google for mean free path length, much research to do.

All in all: there are multiple factors to calculating the heat load for a given pressure and a given gas. Its not as easy as: 1W for 1mbar and 0,5W for 0,1mbar or something like that.

I did many many hours of research to pull this off. Its easy to mount a diode to a tec, however getting LOW temperatures is quite a challenge and every little detail counts.

[Laser57]

Thanks Farbe,

I just needed to find the heat conductivity of air at different pressures all the way down to a low enough vacuum. Instead of searching for information regarding thermalelectric coolers, I started searching for the heat conductivity of air in a vacuum and came up with this info:

torr2.png

I ran the above online converter at 10minus5 millibar which is considered a hard vacuum. The graph shows from .00001 up to a full bar, one atmospheric pressure.

From this, aside from radiant heat, I can see there will be very little conductivity of heat through the air if the vacuum is below 10^-4 Bar and lower (the line at near zero there and below). I needed this information to find an affordable vacuum pump which can pump the chamber down enough to have a multistage TEC operate at closer to the full delta T specification. Finding a vacuum pump which can pull down to just .0001 bar (10^-4) or .075 torr isn't cheap and they are big and heavy. Asking Laird, a TEC manufacturer, if they had a multistage unit which could produce -100 C. with a 2.5 watt load in a hard vacuum and they said no, the closest they had would chill to -60 C.

[Farbe]

Yes that seams about right.

Now you just need to optimize the other forms of heat transfer.

[Laser57]

That is where experience is golden. I have none, but hope to. I will have the one you are building for me before I get anywhere close to building my own attempt at it.

[buffo]

I was trying to understand why a Xenon fill would have better performance than Argon... Then I looked up the thermal conductivity and was very surprised to see that Xenon has the lowest thermal conductivity when compared to over 30 common gasses. (Link) Also, unlike some of the other gasses on the list, Xenon is the clear winner across all temperature ranges. Other gasses switch their positions in the ranking depending on the temperature. Huh! Who would have thought?

But yeah, at 200 Kelvin, Xenon's thermal conductivity is 5 times lower than air, and 3 times lower than Argon. So for optimal results, fill / purge the cavity with Xenon first, then seal it up and pull the strongest vacuum you can. That way whatever gas is left in the chamber (however rarified) will have the lowest possible thermal conductivity.

Agree that finding an affordable vacuum pump is going to be a bitch though.

Adam

[Laser57]

That's a great idea to fill with xenon and then evacuate, I suppose filling with argon first would also be better than just evaucuating air, if using a pump which can't get down to a 10minus6 or 5 level, thank you for suggesting. I've been looking at the available vacuum pumps on ebay, but with the unknown condition for most of them, and the high shipping cost to me in Qatar, I will look for something locally first. If new, then I don't need to worry about whether they are just heavy pieces of iron I can't use for anything else. However, I expect anything I will find here will be Chinese junk. I would be happy with a .075 torr vacuum, especially if filled with argon or xenon first, but finding a canister of xenon is another issue for me here.

[buffo]

high shipping cost to me in Qatar

Ouch. Yeah, that's going to complicate matters.

but finding a canister of xenon is another issue for me here.

Plus Zenon isn't exactly cheap, even in the USA.

Adam

[Laser57]

Yes, getting things shipped to Qatar isn't difficult, but Fedex isn't cheap either.

Communicating with Laird about their TEC's, they don't make one with the substrate doping needed to efficiently cool a 2 watt heat load on a four to five layer TEC at temperatures below -50 C., does anyone know which TEC manufacturer does? They didn't offer that info to me.

It's a materials problem when trying to reach from -80 to -100, their TEC's can't handle that much heat load and reach that far down at the same time. I could run the diode at lower power.

Edit: I'm thinking to solve this to go big and use brute force, having a multistage TEC which starts out with a TEC which is much larger at the bottom and thus the top TEC will be proportionally larger too. However, I am sure that induces other problems; higher radiant heat and just more heat in the whole cascade all together, so not sure that will work. Even if I can't get to -80 C. and some other temperature colder, that's fine, I don't give a damn about efficiency except for the extra heat the entire assembly will generate. I'm betting one could spend a career chasing this rabbit. Of course, I could just give up on a thermal electric cooler solution alone and perhaps use a TEC along with a bottle of freeze spray too to bring the temperature down further, but this is another area of research for me, how far can that bring it down? I'm finding that Chemtronics Freeze Spray can bring the temperature down to -51 C. Perhaps anything short of liquid nitrogen won't help enough.

More research to follow, I will report back later if I find something in a can which can produce colder evaporative temperatures. Perhaps hitting the diode with -50 freeze spray (conductively coupled to it, not directly on it) will allow the TEC to reach much colder temperatures at a minimum of solvent use making it economical. I'm hoping most of 2 watts could be taken away using the spray allowing the TEC to reach near loadless temperatures.

Edit again: Adding more.

Dry ice is -78.5 C. I imagine CO2 sprayed on something will be colder than the -50 C. produced by chemical freeze spray, still researching...

Thermal conductivity of gasses: https://www.engineersedge.com/heat_t...vity-gases.htm

[Laser57]

Photo of my own chiller project parts (Farbe is building one for me too, which will be finished first). Of course, as evidenced in this photo I have a long way to go, including a vacuum pump (if I decide to buy one, $$$) which may be over a year from now before I am ready to begin the build. I'm a slow builder, that's why I'm buying one. I was able to buy this heatsink from a Melles Griot DPSS laser from a forum member here, perfect for this!

This TEC6 is probably jumbo oversized for what Farbe uses and likely not doped for the uber low temperatures possible inside a vacuum, but I want to give it a try. If I could find a TEC with the low temp. doping I'd buy it, but so far I haven't found one anywhere, just know it is possible. Base layer is 62 x 62 mm, top 20 x 20 mm.

TEC6a.jpg s-l1600aaaa.jpgs-l1600aa.jpg

Due to the expense of a high vacuum pump, I might fill the chill chamber with Xenon gas instead of a evacuating.

[kecked]

Photo of my own chiller project parts (Farbe is building one for me too, which will be finished first). Of course, as evidenced in this photo I have a long way to go, including a vacuum pump (if I decide to buy one, $$$) which may be over a year from now before I am ready to begin the build. I'm a slow builder, that's why I'm buying one. I was able to buy this heatsink from a Melles Griot DPSS laser from a forum member here, perfect for this!

This TEC6 is probably jumbo oversized for what Farbe uses and likely not doped for the uber low temperatures possible inside a vacuum, but I want to give it a try. If I could find a TEC with the low temp. doping I'd buy it, but so far I haven't found one anywhere, just know it is possible. Base layer is 62 x 62 mm, top 20 x 20 mm.

TEC6a.jpg s-l1600aaaa.jpgs-l1600aa.jpg

Due to the expense of a high vacuum pump, I might fill the chill chamber with Xenon gas instead of a evacuating.

while not perfect, a hand pump used to bleed brakes is cheap and gives a pretty good vacuum. For this use I bet it’s fine.
Xenon is ridiculously expensive. You also don’t need to cool a large area. You could make the enclosure follow the slope of the tec so the vaccuum volume is very small. Harder to make much save a lot on xenon.