3M Novec 7000 Group - Page 4

There is actually a company that is selling Novec-cooled blade servers today: www.iceotope.com
Video here: http://www.youtube.com/watch?v=N3gCavl2Y6U

The solution seems well designed. Drip free, completely sealed and low pressure (0.2 bar). Only 80 watts required to cool a 20kWatt rack full of blades. What do you guys think? Is this a viable implementation of the technology?

Regards,
Kevin

Would love to see if Novec could be used in a closed loop environment similar to a H100 but without the pump. Put the rad in a position higher than the cpu block and let the liquid boil and re-condense in the rad.

Excellent Idea, I start reading about this possibilities like 2 weeks ago. Now I'm waiting for parts to come up and build first pumpless system. My goal is build highly effective cooling system, with low to 0db noise level. I will post some pictures after building it up, with some temps reading. May be I can build them for sale in future.

Hello everyone! I apologize for my lack of knowledge on this subject. I've read through everyone's posts on this thread and it really sounds like this chemical has a very cost effective and practical use, although its usefulness just hasn't been fully "unlocked" so to speak. Again my comments are vague and ambiguous at best, and I apologize for that.

I've been working on a theory of my own yet on a consumer level. Extreme overclocking set aside would using this chemical be practical if used on a water cooling line?
Lets say we take out the triple and quadruple radiator/fan stacks we see a lot of in a loop and replace it with a concept unit. This unit consisting of waterblock:peltier:heatsink:Novec 7000:heatsink:peltier:heatsink:fan. (heat transfer from left to right or bottom to top) All of this in a semi compact tube. This tube looking somewhat like a resevoir size "Bacta Tank". ;-)

My theory proposes that the Novec controls the "spiking" in heat. With using a vertical set-up you could actually see the liquid level in the mini-tank fluctuate. We go from major fan configurations to using just a small fan at the top of the mini-tank. Nothing is submerged. Were just capitolizing on Novec's low boiling point and cooling the usual array of components i.e. CPU, GPU, RAM,& MOSFET, and keeping everything at a steady average of ~37C.

Any questions or comments are wanted. Please be critical. I'm considering on building a mockup and testing everything out myself once my funds come in. I just hope my idea doesnt prove futile and a waste of time.

Anytime you are dealing with phase change in a closed loop you're dealing with pressure, which could work but it would be a bit more complicated than what you have described. I don't think standard watercooling components would be much use.

Phase change issues aside, this wouldn't get you very good temps. The best blocks give you a CPU temp 30-40 degrees above your coolant temp while at load. The water in a good loop never gets more than about 10C above room temperature, so by operating at 37c (or more, depending on pressure) your temps will be worse.

Neat idea though!

so running the water in the loop past a chillied water block wouldnt essentially chill the water itself?

Oh wow, I completely misunderstood your idea the first time I read it. I had it in my head that you wanted to use novec as a coolant in your loop, sorry!

I'll be honest, I don't know enough about peltiers to say how plausible it is... But what makes you think the heat sink + novec + heatsink combo will perform better than a simple heat sink in ambient air? The second peltier?

I was thinking of using the Novec inbetween the two peltier to help control heat "spiking". Say there is a sudden jump in heat from the gpu or cpu going into full load, I would think that with something like this in the loop somewhere it would help level out the overall temp. So instead of seeing the temp jump to an immediate 70C or 80C the peltier in the line would actually cool the heat spike faster than a regular fan/rad or a small heatsink on the peltier. This idea certainly isnt efficient on the pocketbook by any means but i think it would make for an interesting show inside the case as well as serve some minor function.

Oh and sorry for the confusion. I'm horrible at explaining things in adequate detail.
Edited by flamingfury472 - 12/15/12 at 7:44pm

Well, in my experience watercooling heat spiking isn't a problem-- water temperature changes over the course of minutes, not seconds. I think you're better off trying to improve your overall ability to move heat.

That said, you could be on to something-- maybe there's a way to use TECs and novec 7000 to increase the heat differential between your heatsink/rad and the ambient air, allowing you to dissipate more heat with less airflow. Maybe a series of TEC chilled waterblocks, hot side immersed in a small tank of novec, with a condenser at the top?

Yea thats exactly how pictured the setup to begin with. Maybe not for the same reasons but hey thats how most great things are found; on our way to finding something else.
I just wish i had the cash to fund such a setup. If you see anything like that get put into play, please let me know. Theres always satisfaction knowing your idea works even if you cant build it yourself.