Mitsubishi ML520G71...Red Holy Grail or Flashlight Fail ??

[steve-o]

Where are you getting your TECs, MrPlanters ? You got a good source for good-uns?
(ps- nice poem CDB lol)

[planters]

I like to order the TECS from this source(www.customthermoelectric.com/), but more from habit than as the result of some monumental research.

The real trick is to use a TEC (s) that possess the highest total Q max because no matter what the final cooling requirement, either delta T or heat load ( and here it is certainly delta T), the efficiency is a lot higher and you are on a steeper part of the performance curve than if you push a TEC toward its Q max.

[steve-o]

That makes sense. So, if you are dissapating 4W from a laser diode, a 9 watt TEC should cover it.

[planters]

No

I will try to post the video this weekend. But, the TEC that I am using is rated with a Q max of 400W! If you go to the site I listed and pull up the performance curves for a few different coolers then assume for the sake of argument that you want to remove 10 W (4 diodes @ 2.5Vx1A) and you want a delta T of 35C. You will see that the larger the maximum total capacity of the cooler(s) the less power (waist heat) that will be used. The only limit is the available diode-mounting-plate surface area and to some extent, cost.

I have followed my own advise and compared curves and there appears to be a sweet spot with the coolers and selecting a cooler with a Q max 4-6 times the expected heat load is optimal

[steve-o]

According to their site, Q=W, which I don't agree with. What you are saying is that wattage dissapation , (yes differential delta) , should be 4x-6x times which equals (your example of 10 watts) 10x4=40 or 6x10=60 watts , where do you get 400?? (oh and btw waist, as in waist-heat, is spelled waste! )

[planters]

I just HAiTE it when spell check lets me down. I've been using it long enough...it should know what I mean!

Regarding the 400W Q max, this is the largest standard TEC they sell and I happen to have a couple on hand. So, not optimized, just free. The performance is less than optimal, but it is better than pushing a much smaller device close to its max. After looking through their list, I like this one 12711-5L31-05CL. It works well at 12 V. It has a little extra capacity to handle heat leakage from the environment ( I am arbitrarily doubling the 10W) and it is cheap, so I am going to order 4 and if a pair doesn't cut it I can always add two more and/or tweak the PS voltage to fine tune.

[djpsych]

Will these work: http://www.ebay.com/itm/140461703212#ht_1265wt_773 ?

[planters]

Yes. I have used Marlow TECs on an observatory project many years ago and the quality was very good. I don't know what your heat load will be ( hell, I'm not sure what my heat load will be), but these are in the ball park in terms of Q max. I would order 4 or 5 (in case you drop one). Plan to allow for mounting between 1 and 4 under the diode plate once you see how they run.

The choice of TEC is not critical. A very large range will work at the expense of a LITTLE loss of efficiency. You might want to invest in a cheap panel thermometer for permanent installation or at least a lab style meter to dial in the set up.

[planters]

I finished the video and the link is below. Just a note; even though I say this once in the presentation be aware that all the power output readings are 10% lower than the actual output due to the losses through the microscope slide.

[LaNeK779]

a full watt of precious red out of a single chilled g71 diode!!
great news sir planters! thanks for doing this