Exploring Below Ambient Water Cooling

Page 19 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
No more room on my desk mate so I'm thinking the chiller will be on the shelf the other side of my monitor with four feet of tubing between it and the Pc,
I may not even need another pump but I'm prepared to get one if I do,
three pumps, five resses, three Radiators, three blocks, nice little loop hehe
**Edit, Three Radiators insert^
Moto
 
Moto, Please keep us informed in this thread, with pictures, I'm sure there are many keeping an eye on this thread that are curiously considering similar possibilities, I know I considered going that route myself. Ryan
 


Yes, I've had the plan for a while, I intend to hook my Gf's rig up to the chiller as well at some point so it'll be a shared res once I W/c her rig (other upgrades priority though)
but I like to get the idea sorted in my head and plan everything out well in advance of buying anything
so a basic loop order (As it is now) is:
Res/pump in main Pc>Gfx1>Gfx2>External Radbox Ex360one>Ex360two>Respump in radbox>Cpu>Rx240 on roof of main Pc>T-VirusRes one> T-Virus Res two>Return to Res/pump,
The chiller will be plumbed in after the graphics cards and return to the first Ex360, using quick disconnectors, this allows me to remove the radbox or chiller from the loop at will depending on cooling/silence needs/wants,
I already have two temperature sensors recording ambient room temp and the temp in one of the T-virus resses, I'll add another to the chiller to monitor its water temperature
The main Pc and radbox can already be run independently of each other which makes flushing/bleeding a walk in the park despite the complexity of the loop,
you can't see much of my tubing due to Ocd cable management as a lot of it goes through panels, hidden in a roofbox I made etc
Keep watching and I'll link the build log in here once I begin :)
@Ryan, you know theres a log coming mate, thats half the fun of a project in my eyes :)
and if you want some pics in here as well thats not a problem
Moto
 
Seeing as how I am now cooling my CPU and both GPUs in the same loop, these temperatures are totally acceptable.


With a 23c ambient my CPU idles at 30C, keep in mind that's with 2 GPUs in the same loop, with the CPU overclocked to 4.5ghz Hyper Threading running all 8 threads with the 2700K.
Click to expand...

"One Sniper means a dozen!" Quote from the game, Bulletstorm

Any hardware changes affect the overall and I switched CPU water blocks to the XSPC Raystorm.

This new Rad/TEC cooling solution is very consistent setup the way it is, CPU idle temp @23c ambient room temperature is 29c, GPU1 is at 31c and GPU2 is at 29c.

Under stress load the CPUs max reached temperature is 52c and neither GPU goes above 39C, IMO considering all of this is on the same cooling loop is impressive.

I decided to run a test to see what would happen if the TEC failed so I turned the power off to the Peltier and reran all the stress testing with only the 360 radiator keeping the system cool.

CPU idle went to 35c, GPU1 33c, GPU2 30c
Stress Load CPU went to 57c, GPUs 43c

So the TEC energized delivers a 6c lower CPU idle temperature and a 5c lower CPU load temperature, an approximate 2c lower idle between the 2 GPUs, and a 4c lower GPU load temperature.

I'm sure at this point there are some of you questioning if a 5c CPU and 4c GPU load temperature is worth running the TEC in the first place, well the jury is still deliberating on that one, but I will be giving a closing statement on this peltier cooling adventure.

You may possibly be able to reproduce these same TEC cooling results using 2 high performance quad radiators without the TEC?

Initial purchase price comparatively about the same, more cooling fan noise, but completely eliminating the TEC electrical load from your electric bill?

The option advantage I do presently have is, if I'm just surfing the net and posting here at THGF I can run without the Peltier energized, and just power it up when gaming. 😉

Ryan
 
Same brain in a different skull man, :)
instead of settling back on the cooling throne, you say 'what if?'
I recall you removing rads as the peltier was 'too' effective and wanted the same possibility for my own loop, I just wanted it instantly interchangeable and replaceable hehe,
Can't wait to get some numbers of my own up once its here and modded into things
Moto
 
ArthurH shared with me that he had removed his tubing insulation and the results were a definite increase in temperature, so I decided from his input to insulate the return line from the CPU/GPU loop back to the reservoir.

I originally had only insulated the line from the TEC assembly directly to the CPU, but now all of the CPU/GPU tubing is insulated plus I also insulated the Reservoir my CPU idle temp dropped from 29c to 27c. :)

Woop De Do!

Does that mean earth shattering results?

No it doesn't.

At least from my experimentation you all can learn what not to do!

Statement Coming!

 
But again, what works well in one system, may not be the optimum for the next,
I hadn't planned on insulating any tubes further than putting them through my 2"" ducting and see what the first set of numbers say, that will be my baseline measurement,
all components hooked up, radfans off chiller off and see how it performs as passive,
then change things and record the results, eventually find the best setup for summer and winter (must write down!)
😛
Moto
 
Regarding this thread from the very beginning to where it is now.

The Ice Chest cooling was very successful, allowing below ambient cooling down to literally 0c water temperature, with condensation becoming an issue at various temperature levels which was fully covered.

The RadBox cooling the GPUs was very successful allowing a combined GPU load temp half of what the stock air cooling yielded.

However this Peltier/TEC cooling solution is not something I am going to recommend, the negatives far outweigh the benefits it brings to the table vs standard water cooling, or anything this thread previously covered as the below ambient gains are just too costly.

Someone had to try it and answer some of the missing questions, so I Guinna pigged myself as the scape goat, and what's been learned is at my expense and now I'm going to share that with you.

TEC cooling is not Computer mainstream cooling today for good reasons, first is the electrical added expense to reach the below ambient levels, for the TEC to properly perform it has to be pulling a serious wattage load constantly, requiring it's own power supply.

Cooling the hot side of the TEC is a challenge in itself, seeing as how it's best performance comes from allowing the hot side to get hot, the best efficiency for the TEC I purchased is around 100c which is actually 212f, the maximum heat range for my TEC was 125c which is 257f, way past the temperature to boil water.

The only way to truly cool the hot side has to be some type of heatpipe cooler because to cool it with water is not a good idea, your water cooling hardware is not actually meant to deal with that much heat, my XSPC water pump got so hot I burnt my finger on the metal output spout. (That is no Joke!)

The radiator got so hot I couldn't put my hand on it, at that point the peltier hot side temperature reading on the outside was 160f, I estimate about 180f in the center of the peltier, cooling that much heat is serious and in itself takes a serious solution, negating the benefits of this type of cooling.

So for the record, I do not suggest anyone take this route to cool your computer and I'm asking ArthurH to close this thread and leave it stickied because the first of this thread is viable solid information.

Feel free to PM me for any unanswered questions you may have, but since this thread title is exploring below ambient cooling this TEC cooling reaching below ambient for any dependable solution is just not worth it.

As I have zero intentions of giving up, experimenting with this TEC cooling solution, when I have more viable below ambient information to share, I'll have the thread re-opened and continue posting in it.

Thank You All that have kept up with it's progress!

Ryan



 
OK I'm back in below ambient business and using the TEC/Peltier cooling, actually I now have a combination setup, an insulated reservoir, (Small maybe 2 gallon size Rubbermaid Ice Chest Cooler), feeds two valve controlled lines.

The output line from the pump is Y'd to two separate lines.

One branch line straight to the CPU allows TEC cooling only, which is below ambient cooling.

The other branch line runs through the 360 radiator to the CPU allowing standard water cooling, which is above ambient cooling, so basically now the best of both worlds.

Regarding why I asked Arthur to close the thread as I have explained above, was because running the TEC on the feed side directly to the CPU is a bad idea, I didn't know until I tried it.

Water Cooling the hot side of the TEC is also a bad idea, it gets too hot for water cooling hardware, I guess that's a missing reason as to why it's not so popular!

Cooling the TEC with a really good heat pipe air cooler is the way to go!

The TEC on the return line just before the insulated reservoir is the way to go!

The TEC needs a buffer, and the reservoir becomes the buffer to store cold, that's very important and it will make a tremendous difference performance wise.

Getting what you want from the TEC is a balancing act, you have to let the hot side get hot, for the cold side to get cold.

I know that sounds crazy but that's how it works, you have to have a heat pipe cooler that is on the level of the TRUE (Thermalright Ultra Extreme), or the Noctua NH-D14, you won't get away with some cheap cooler the TEC will burn up if it is allowed to get too HOT.

I've rebuilt my RadBox twice since closing the thread, believe it or not, Ripleys!

Presently the TEC is chilling the water in the reservoir which is presently 13.4c, my ambient room temperature is 22.2c, my CPU idle temperature is 17c, so this thread is officially back on below ambient track!

The small cooler I'm presently using is a lot better insulated than the large cooler was, and there's only about a gallon of water in the reservoir, there's at least a quart of water in the lines and radiator and all the lines in the CPU loop are insulated.

There is a second entire independent loop cooling the GPUs and they are being cooled by a 240 radiator.

I'll have to take some more pictures to show my present setup, as I was too busy straightening it all out to worry with pictures!

Ry

 
Hi Moto, I was going to shoot you a PM the thread was re-opened but I see nothing gets by you here and you found it. :)

I'm really interested to see how that chiller works for you?

I ran some tests last night gaming and playing a FPS "Bulletstorm" with settings fully maxed the temperature of the water increased by 0.2c in 30 minutes, so if that stays consistent it would take 2 1/2 hours of game play for the water temp to increase a full 1.0c.

Not bad at all seeing as how all those temperatures were not only below ambient but also below any CPU thermal throttling, (Another topical benefit to below ambient that the majority of THGF Sandy Bridge users are clueless about! :) ).

That gaming temperature increase with the water in the cooler acting like a buffer is direct proof as to why running the TEC straight to the CPU was not successful, it had no recovery time, and no water mass working in it's cooling favor.

The cooler reservoir also allows running the TEC at a cooler hot side operating temperature, and that is controlled by the fan speed on the heat pipe cooler, I know that sounds weird, none the less it's 100% true.

That is a testament as to how good the Thermalright TRUE actually is, that it can make such a cooling difference and control the hot side temperature with no voltage adjustment to the TEC just cooling fan speed air flow.

That's why I said you need a good quality heap pipe cooler, basically any heat pipe cooler that can actually be used in passive mode, and that narrows the field down considerably.

To get the results I'm getting now, I keep the hot side around 130F (Measurement taken with thermal probe at the heat pipe entrance to the base, actual TEC temperature is probably about 10F ~ 20F higher, the probe only measures in Fahrenheit), that gives me a cooler running TEC if I allow the TEC temperature to increase say to 180F by slowing the cooling fan speed I could stop the gaming increase at a room heating trade off much worse than the 130F.

Of course in the winter time it would easily heat my office! :lol:

I'm laughing but that's not a joke! It outputs that much heat!

Ryan

 
I think we covered that earlier somewhere in the thread Ry, I told someone that its not about cooling the heat from the chip, its about providing cooler water for the loop and your results bear that theory out obviously :)
Thermal throttling becomes almost irrelevant at this level I reckon, efficiency of the loops means that the Cpu is never going to get anywhere near limits which doesn't so much 'improve' performance as preventing performance from being choked (same result but the definition is pertinent I feel)
and you've seen that the larger mass means it holds cold better/longer meaning the TEC has to work less, less heat, less powerbill hehe
I like the control you have over the hotside as well, very nifty work there,
I think the only ways I can alter anything on the chiller is by limiting voltage or varying the mass (Adding Ice if really necessary😛), but thats three options I can tweak at leisure once I get some numbers
the other control I have is obviously the fans, they can be off, low,maxed, there really is quite a lot of playroom in my loop so I'm looking forward to this as much as anyone, all parts ordered so hopefully be getting my mod on soon,
Moto
 


Regarding the thermal throttling if the water temperature is low enough to keep the throttling below it's activation threshold, then your statement is 100% accurate.

Regarding the power bill however it is a constant draw no matter how the hot side of the TEC is manipulated by cooling fan speed, the TEC is still drawing the same power.

It is a power Black Hole!

The one advantage I do have is the TEC would be drawing even more power if it was getting it's spec'd 15v DC voltage, but it's only getting 12v.



Better read the PDF manual for that chiller you ordered. 😱

Ryan
 


Once you get started it's like eating crepe pancakes at IHOP!

You're hooked! :lol:

The possibilities and discoveries are like feeding coal to an old locomotive, it never gets enough!

Curiosity they say killed the cat, but for us!, curiosity and discovery are driving forces, at least for me it is.

The more I learn of what this TEC cooling can do, the more I want to get further into it.

Did you know if you can externally apply the hot and cold to a peltier plate it's opposite reaction is to produce electricity?

Ryan :)
 


Haha, nope, I certainly didn't know that!

I was wondering if it was possible to harvest the waste heat to somewhere else, like say, a storage tank. That way, we can get hot water for bathing without the need of a heater.
Say the storage tank is 20 liters, it would be able to absorb quite a lot of heat from the TEC.

Letting that kind of heat dissipate into our rooms and making our air conditioners work even harder seems like a waste.
Is my idea feasible? =D
 


Man that's a great idea, I'll have to give that one some serious thought! 😉
 
I have a 4 or 5 gal cooler from Canadian tire and really impressed how cold it keeps things, so I’m thinking maybe I could cool flowing water. While searching out info on the Pelteir cooler I ended up here.
I want to cool the water in my boats live well down to about 40 degrees or less from about 70 degree lake water. It holds about 4 gal of water and I have 30 amps at 12 volts to play with from my outboards charging system. I can cool the hot side with lake water, insulate the tank and run lake water through the cold side into the fish tank. Cool fresh lake water running into the live well would be just awesome.
There are some discouraging opinions here. Is this idea worth trying out?

Thanx much
Mikey
 
Mike they seriously frown on you putting your email address in the forums so you might want to delete that, I've already written it down myself anyway.

I'll be glad to share all I've learned experimenting with the Peltier cooling and I believe besides a few problems to overcome you can do what you want, you'll definitely have the available power you'll need, that's one obstacle out of the way right there.

Sorry, I caught this when my wife and I are going to one of the local paint supply stores, so I'll get back to sharing regarding what you intend to do when we get back.

Ryan
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom