My Toshiba/Satellite L850 is a great little laptop with only one serious problem:
Most of the time the cooling fan of my keeps on running at high speeds, indicating that its cooling system is poorly designed and its noise changes interfere with my audio recordings. Since I couldn't find a solution to the problem on the Internet (only same complains), I thought of helping myself.
After some testing I discovered that if the TEC is put to work hard, it produces enormous amounts of heat, which requires a big heatsink, strong fan and so, makes them not really useful for PC cooling.
Yet, the moment we find the point at which the TEC produces a useful amount of coldness and moderate amounts of heat, laptop cooling becomes a feasible option.
For instance, I found this point to be, when I connect a (cheap China made) PS of 12VDC/1.25A to my TEC 12705.
Result: On connection the load then drops the voltage to about 7V6 and rises the currant to 2A1, the heatsink gets slightly warm, its (12V nom.) fan is almost inaudible and the PS stays moderately warm. - Ideal!
The cold side becomes finger-hurtingly cold; i.e. more than sufficient to help cooling down my laptop:
With a Dremel Mini I sawed a hole around the laptop's cooling system. Onto its exposed copper cooling rails I spot-soldered a few Chilean 100 pesos copper coins to almost level the surface with the laptop's bottom cover.
Under the wooden shelve board onto which my laptop is placed, I sawed a round hole and (spring) suspended the Pelt unit. Its cold part, consting of a 1/2" round, solid brass piece, is then made to touch the laptop's copper coins.
The suspension of the TEC unit and a bit of thermal paste helps to level-adapt and transfer the coldness to the laptop's copper rails. The laptop's cooling fan then only serves as backup and becomes almost redundant and unemployed... like many of us.
Yes, I agree that that there exist lots of contradictions on this idea...
Well, it's not meant to replace the laptop's original cooling system; just complementing it !!
That way the Toshiba's cooling system /fan works much less and runs quieter.
At this low voltage condensation is not an issue. Besides, any condensation would drip down, back to the TEC; not upwards into the laptop.
I'm just glad having found this effective solution and thought of sharing this idea with people seeking to overcome the same problem, in the hope they might be benefit from it and contribute to improve this idea with some even more innovative ideas...
It's just working fine at the power supply I am using. Any more power will generate too much heat and more coldness I do not need to help the laptop cooling down its copper cooling rails.
That's why I do not have the condensation problems either.
I also learned on thing:
There are lots of sceptical opinions about it on the Internet, but finally one has to gain one's own experience with it than just rely on other people's opinions.
Result: my laptop definitely benefits from it, because I know all too well how its cooling fan was whining at high speeds, before the TEC addition...
So, I really do not believe in doomsday warnings anymore.

The great thing is the Toshiba fan and cooling system is left in original condition and serves as a TEC efficiency indicator and as a back up, should the TEC fail to deliver.
Besides, I insulated the TEC's hot side from its cold side with a thin layer of Styrofoam. Yet, since the hot side has a huge heatsink, its fan is almost unnecessary. So, I reduce the TEC's heatsink's fan with a 40 Ohm resistor. That reduced its noise to almost "0" and at the same time giving the TEC a bit more juice...
In case something goes wrong Toshiba's cooling fan will let me know by running faster at higher speeds... like it used to run without the TEC' help.
I am practical man: once something proves to be working, that's all I need to accept it as logically correct and progress!
Hence, I successfully applied it on my Toshiba Satellite L850.
From my observations I take, that the CPU triggers the laptop's cooling fan. This fan has about 5 velocity stages and according to the data processing demands, the fan adjusts its speed.
Yet, since the TEC keeps the laptop's cooling copper rails cool, the fan only runs for about 1 to 2 Sec at the second velocity stage and then turns off.
Before, without the TEC, the laptop's fan was running at much higher speeds for long periods of time; often for hours.
Also, when I turn the TEC unit off, the laptop's copper rails heat up the normally ice-cold 1/2" round, solid brass piece from the TEC unit gets warm!
If that isn't enough proof for success... I don't know what is!