Waterchilling a 226w peltier for an i7 920

[SS_FireHawk]

I posted this in another forum site and still havent quite gotten enough information on it. Im not sure if i should put this in the peltier section or water chiller section.

I want to use my mcw-452 to cool my mcw-6500 block for 775 processors. Im building a new i7 system w/ a foxconn bloodrage mobo that supports lga 775 socket coolers. Ive read that when properly configured, water temps drop below -10c with the mcw-452. Im gonna use a pa120.3 and a mcp-655 pump to cool the mcw-452; Im just gonna split the line to cool both sides.

Im also going to use a 2nd pa120.3 and mcp-655 for the cpu loop. The cpu loop I want to be split 1/2 cooling the gpu's and a 1/4 line going through the chiller and then into the mcw-6500 to super chill the peltier cooling the cpu.
Please don't tell me it's inefficient and power hungry and all that garbage. I didn't come here to get responses from trolls. Ive used em before. I have a phase change unit already and can readily use it. This is more of a novelty project to see how far i can push this technology.

It will be consuming just under 700w on the peltiers alone lol. That will piss off many a tree hugger. I even have mcw-60 peltier waterblocks for gpus that have a 192w pelt in them. I probably couldnt find a wall socket to power a multigpu quad core system with 1kw worth of cooling wattage being pulled from the wall. I think Id poop myself if it were feasable without having my own 5kw generator outside my window. This is a link to what Ive gotten so far. Feel free to give me some advice or ideas. Im building two identical i7 builds that will have same cpu, mobo and mem.

http://forums.extremeoverclocking.co...d.php?t=329970

[Naja002]

Originally Posted by SS_FireHawk

Please don't tell me it's inefficient and power hungry and all that garbage. I didn't come here to get responses from trolls. Ive used em before. I have a phase change unit already and can readily use it. This is more of a novelty project to see how far i can push this technology.

Until you spelled it out....there was no way for anyone to know what your experience level is, so I don't really think it's fair to call them trolls. Many folks...probably most...are turned away from TECs when they realize what the actual initial costs are going to add up to and what on-going costs are going to be...

Originally Posted by SS_FireHawk

I posted this in another forum site and still havent quite gotten enough information on it. Im not sure if i should put this in the peltier section or water chiller section.

I want to use my mcw-452 to cool my mcw-6500 block for 775 processors. Im building a new i7 system w/ a foxconn bloodrage mobo that supports lga 775 socket coolers. Ive read that when properly configured, water temps drop below -10c with the mcw-452. Im gonna use a pa120.3 and a mcp-655 pump to cool the mcw-452; Im just gonna split the line to cool both sides.




It will be consuming just under 700w on the peltiers alone lol. That will piss off many a tree hugger. I even have mcw-60 peltier waterblocks for gpus that have a 192w pelt in them. I probably couldnt find a wall socket to power a multigpu quad core system with 1kw worth of cooling wattage being pulled from the wall. I think Id poop myself if it were feasable without having my own 5kw generator outside my window. This is a link to what Ive gotten so far. Feel free to give me some advice or ideas. Im building two identical i7 builds that will have same cpu, mobo and mem.

http://forums.extremeoverclocking.co...d.php?t=329970

You don't ask any specific questions, so this is what I see:

Regardless of how well you can cool the 226w on the i7...if you want to OC it to ~4.0...I don't believe that that TEC will handle the load. If you want to run the i7 at stock....or within the capabilities of the 226w...then you should be able to pull something off...exactly what--I'm not sure.

Core i7 920 T.E.C.


Using a rad in the cpu loop is just shooting yourself in the foot. Once water temps reach around ambient the rad will begin acting as a heatsink--drawing energy (heat) into the system from the air passing through the rad--it will begin working against you. You will never get more then a few degrees below ambient--at best. Simple physics.

Post Rad Chiller Concept


The actual temp of the hotside water will increase probably much less then you realize. A serial run from TEC to TEC on the chiller will be much wiser then attempting to parallel (split) the flow. Fluids follow the path of least resistance, so obtaining and maintaining even flow through both parallel runs will be much more difficult then just running a single loop: Rad>pump>TEC1>TEC2>Rad.

Originally Posted by SS_FireHawk

Im also going to use a 2nd pa120.3 and mcp-655 for the cpu loop. The cpu loop I want to be split 1/2 cooling the gpu's and a 1/4 line going through the chiller and then into the mcw-6500 to super chill the peltier cooling the cpu.

The TEC is not going to "pull-down"/chill the water in any kind of "super chill" fashion/manner....it just does not work that way. You will have to allow for pull-down time--because it does take some time. With the Gpus in the loop...I seriously doubt that you are going to get the effect that you seek.

Beer cooler


Best advice, IMO, is to either make 2 loops with the rads cooling the hotsides of the TECs. Or, a single loop...again, with the rads cooling the hotsides of the TECs.

[SS_FireHawk]

Thanks for the reply. I used the word troll as a deterrent for the individuals that just forum snipe without the realization that Ive worked with them and understand to an extent that they are not as efficient as other methods of cooling. Thanks for the criticism though. As for the 226w peltier being unable to cool an i7, I linked the other forum that points to another link I saw of a gentleman using a 226w to cool his i7 at 4ghz. Temps were decent at idle (-10c) but at full load, it reached 28c. Im thinking that; Im guestimating here, the i7 pulls around 180-190w at 4ghz. I would like to know a more exact figure if anyone does know.
Thanks for pointing out the rad in the cpu loop. I wanted to split the line because there is alot of additional power coming from the pump. I wanted a way to slow it down. Should I just buy a much weaker pump for the water chilled fluid and create it's own separate loop? As clarification, the split would have been before it went through the water chiller so the cpu would have had it's own chilled fluid with no other components getting that benefit.

So now Im thinking 3 loops... One for the hot side tecs, the 2nd is for the cpu loop through the chiller, and will be used for the gpu's.

I have to get my ass outta bed and ready for class, thanks for the info. But please do read that link as it kinda goes into what questions I asked b4.

[Hondacity]

the i920 with ht on...@ 4ghz with low vcore..won't pull 150w.... heat output is almost relative to vcore... but if you need like 1.55v for 4ghz stable..the heat output will be alot..

lowering the temps(below 20c) on the i920 helps it alot...

[flak-spammer]

If you want more help this post really should be in the TEC section, more people will have answers to the questions you have.

[Naja002]

Originally Posted by SS_FireHawk

As for the 226w peltier being unable to cool an i7, I linked the other forum that points to another link I saw of a gentleman using a 226w to cool his i7 at 4ghz. Temps were decent at idle (-10c) but at full load, it reached 28c. Im thinking that; Im guestimating here, the i7 pulls around 180-190w at 4ghz. I would like to know a more exact figure if anyone does know.

I assume you are referring to this specific post by Uncle Jimbo (?):

i7 965 and Phenom II 955

If so, then I can tell you straight up that I basically believe anything that Uncle Jimbo has to say about TECs. He's amazing with his knowledge in that area. So, the only question left is the accuracy of his estimated output of the i7 965 @ 4.0.....? I'm not sure if that's accurate or not.

Here's an example of someone that had failure with an i7 920 and 226w:

Swiftech MCW7000-T for Corei7 announced

So, personally I would say that you are borderline at best....

Originally Posted by SS_FireHawk

Thanks for pointing out the rad in the cpu loop. I wanted to split the line because there is alot of additional power coming from the pump. I wanted a way to slow it down. Should I just buy a much weaker pump for the water chilled fluid and create it's own separate loop? As clarification, the split would have been before it went through the water chiller so the cpu would have had it's own chilled fluid with no other components getting that benefit.

The mcp-655 are great pumps. They have a lot of psi for their size, but not a ridiculous amount of flow. I would not swap them. They are very durable, reliable pumps.

Flow really doesn't matter within the 0.8-3.0 gpM range. The 1/4" line will create more frictional loss, but the psi of the pump will handle it...NP.

If you have 2 pumps then what I am seeing is:

1 pump on the chiller loop to the cpu waterblock
1 pump for the hotside of all TECs and both rads.

Shouldn't be a problem.

Originally Posted by SS_FireHawk

So now Im thinking 3 loops... One for the hot side tecs, the 2nd is for the cpu loop through the chiller, and will be used for the gpu's.

The GPUs will only have a hotside. Depending upon which GPUs they are...you shouldn't have any need to run the chiller water through the hotside. I understand what you are thinking, but it should be a total waste. Let me know about your GPU(s).

The hotside of the GPUs can be connected to the hotside of the chiller and run together through the rads....1 loop.

The chiller coldside can chill the cpu...1 loop.

2 loops total.

The biggest problem with the mcw-6500 is that it will either do the job or it won't. You are not going to gain much by chilling the hotside water of the TEC. At best you will get rapid diminishing returns. Basically, the farther you try to push it past it's peak....the faster it's going to go from conductor to insulator....failure. So, the bottom line is: the mcw-6500 is either going to work or not. You're not going to be able to pull off some large gray area of higher temps. At some unknown point it's just going to fail.

I believe that you can use the chiller and the 2x GPU TECs in separate loops to handle the total wattage--depending upon what the GPUs are.

With the mcw-6500--it just depends upon whether it can handle the load or not. You can consider upgrading the TEC with one that has the same dimensions...then it should hold.

Need to know:

GPUs
PC use ("everday", Gaming, 24/7 crunching, etc)
PSUs for the TECs

[SS_FireHawk]

I understand what your saying about the mcw-6500, thats why Im asking questions. I believe it's fairly easy to overshoot the tec with these new quads. I was easily able to do it when I tried to bench my old q6600 running 1.6v @ 4ghz. I was only able to run it on fast benches and not all the time. Ive seen some 300w tecs with the same dimensions as the 226w in the mcw-6500. I just don't know how to take it apart and replace the tec with a more powerful one. Anyone know what type of adhesive is needed to hold the tec onto the waterblock?

[flak-spammer]

It won't be attached with an adhesive (at least I hope not), they are normally squished in between two metal plates (cold plate and hot plate).

[Naja002]

Originally Posted by flak-spammer

It won't be attached with an adhesive (at least I hope not), they are normally squished in between two metal plates (cold plate and hot plate).

I would agree with that. Any "adhesive" should really just be a simple contact cement use to hold the foam in place.....

The unit itself should just be bolted together with TIM in the appropriate locations. You may want to see if you can swap the TEC out by just unbolting the unit, removing the base plate, etc.....without having to remove the foam.


BTW, the Swiftech MCW7000-T was originally designed toward the i7 and best I can tell it still has not made it to market. It was designed around a 417W TEC. The 437w are quite popular, but I'm not sure of their dimensions....

The Swiftech MCW7000-T may very well be an affordable PSU issue and not a TEC issue though...

[flak-spammer]

The 437 is a 62x62mm TEC probably too large. I'm guessing the 226 is 40x40 or 50x50.

[SS_FireHawk]

I was looking at the mcw-7000 but unfortunately it is still just vaporware. The dimensions of 400w+ pelts are usually 62x62mm, a good size larger than my 50x50mm. Like I said, I'm a noob when it comes to "creating" a water block for the tec, even if it is premade... Ive never done it.

Oh and about the gpu's... Im currently running a single gtx 280 clocked at 756/1566/1242. But I am interested in the new Fermi gpu's that are taking forever long to come out. Im thinking I am gonna sell my gtx 280 and run two gpu's. Im still up in the air till I see some performance numbers. I have a 1920x1200 monitor, so maybe two of those nuclear reactors may be a bit much.

Also I need a bit more clarification on this. I can use an mcp-655 to run a 1/4 line and it will have a low enough flow rate move the fluid through the block slow enough to see the chilled water benefits? I don't know how to reword this to make it sound less confusing. Basically Ive heard that you want low flow rate so the chilled fluid has enough time to absorb the heat and carry it away. So if I am reading right, the temps of the fluid exiting the mcw-6500 will still be below ambient temperatures and the radiator will only heat up the fluid. After I get out of class, Ill draw up a "very" basic diagram of some proposed ideas on the loops.

And HTF do I quote previous posts? Lol, it's slightly embarrassing, but I have to figure it out some time :P Oh and btw, yea I did read that thread with Uncle Jimbo almost in it's entirety when I was researching this project. I think I pointed to it in my previous one. Uncle Jimbo's intelligence is beyond me and it seems like I have to reread it like three times before I can retranslate it into something more understandable lol.

As for computer use, I like to bench and game with it. I generally don't fold, do any encoding or any industrial work with it. I see it more as a hobby. I take it apart 1-2 times a year and try something new. And I apologize that my thoughts are so random and all over the place in my posts

[Naja002]

Originally Posted by SS_FireHawk

I was looking at the mcw-7000 but unfortunately it is still just vaporware. The dimensions of 400w+ pelts are usually 62x62mm, a good size larger than my 50x50mm. Like I said, I'm a noob when it comes to "creating" a water block for the tec, even if it is premade... Ive never done it.

There are some 50x50mm 400W on ebay right now. The 226 Watt Peltier should be 50x50mm also. I cannot speak for the quality, efficiency, etc of each unit though.

Originally Posted by SS_FireHawk

Oh and about the gpu's... Im currently running a single gtx 280 clocked at 756/1566/1242. But I am interested in the new Fermi gpu's that are taking forever long to come out. Im thinking I am gonna sell my gtx 280 and run two gpu's. Im still up in the air till I see some performance numbers. I have a 1920x1200 monitor, so maybe two of those nuclear reactors may be a bit much.

I seriously doubt that those 192w's will handle a Fermi. 280 I'm not sure though. There are usually so many wattage numbers thrown around that it's hard to tell what is actually correct.

Originally Posted by SS_FireHawk

Also I need a bit more clarification on this. I can use an mcp-655 to run a 1/4 line and it will have a low enough flow rate move the fluid through the block slow enough to see the chilled water benefits? I don't know how to reword this to make it sound less confusing. Basically Ive heard that you want low flow rate so the chilled fluid has enough time to absorb the heat and carry it away. So if I am reading right, the temps of the fluid exiting the mcw-6500 will still be below ambient temperatures and the radiator will only heat up the fluid. After I get out of class, Ill draw up a "very" basic diagram of some proposed ideas on the loops.

Flow doesn't matter. Between 0.8-3.0gpM...it really doesn't make much difference. That was shown a long time ago. The faster the water travels through the chiller--the faster it travels through the waterblock. The less heat/energy it gains at the waterblock and loses at the chiller. The reverse is true also: The slower it moves through the chiller--the slower it moves through the waterblock. The more heat it picks up...the more heat it loses. In the end, the system will reach equilibrium regardless of flow rate.

What a lot of people don't understand is that there is only going to be like 1.5C or less difference in temperature from the water going into the block and the water coming out once the system reaches equilibrium. Prior to equilibrium is pull-down time. The delta (difference) between the water leaving the chiller and the water leaving the block is, again, not going to be very much at all....that's why it takes time to pull-down to = equilibrium.

Some of these concepts are difficult for some folks to get their head wrapped around and seem to go against common sense. The truth is though: Once you "get it"--it all starts making perfect sense.

Originally Posted by SS_FireHawk

And HTF do I quote previous posts? Lol, it's slightly embarrassing, but I have to figure it out some time :P Oh and btw, yea I did read that thread with Uncle Jimbo almost in it's entirety when I was researching this project. I think I pointed to it in my previous one. Uncle Jimbo's intelligence is beyond me and it seems like I have to reread it like three times before I can retranslate it into something more understandable lol.

At the bottom right of each post you should see a quote button and a multi-quote button. Use the quote button if you want to quote 1 specific post, and the multi-quote button if you want to quote more then one post--"mutli-quote" each post then click reply.

If you want to split up a post like I've done here...then that will require manual labor that you can pick up later.....

Originally Posted by SS_FireHawk

As for computer use, I like to bench and game with it. I generally don't fold, do any encoding or any industrial work with it. I see it more as a hobby. I take it apart 1-2 times a year and try something new. And I apologize that my thoughts are so random and all over the place in my posts

Ok, the usage just gives us an idea of what your goals actually are, and from that we may see/understand things that the OP doesn't. It just helps us help you.


I really cannot see a benefit to trying to use a TEC waterblock on the Cpu and the chiller. I may be wrong, but I just don't see you gaining any/enough benefit to make it worth the extra hassle. I think that the chiller will handle the i7 and that the 226w won't. Personally, I think I would just focus on the chiller with a standard waterblock. Maybe someone else will come along and clarify that for us...

[SS_FireHawk]

Well I think that Im in denial lol. I want to use that 226w just because it's there lol. So in other words you thin that just going straight chilled water will do a better job than trying to hodge podge the two together. I can definitely see the 226w becoming an insulator if that is the case. Now in using just the straight chiller, would it be recommended to have a reservoir of fluid after going through the cpu so whatever is already below ambient stays nice and cold b4 being cooled again and sent through the loop? Throw an insulator around the reservoir of course. I haven't done anything with chilled water b4 so it is a new process for me. What is the recommended loop to ensure the fluid temperatures are as cold as possible? I'm sure insulating the lines would help. Crap i gtg back to class, Ill post l8r

[zipdogso]

I have read a lot of the threads referenced here including uncle jimbo's, who I also hold in regard, but if you look around some people are rating an overclocked i7 at up to 250w (can't remember the speed.) the problem really is that your 226w TEC is only 226w at full power with with a Dt of 0 i.e. no cooling, that is the Dt =0 bit. the 12726 is a 15.6v unit (standard for 127 thermocouple unit.) so yo you run at the generally accepted 12v you are only going to get a max cooling of about 180w this is assuming absolute 100% thermal transfer which of course is pie in the sky. I am not sure 170-180w is sufficient overhead on the max wattage of a low rated i7 to get a good level of cooling and of course it is no good what so ever for one that goes up to 250w. 180w cooling of a 180w load wont get you sub zero.
I don't know..... I don't have an i7 but I don't think you get good results in fact I even think someone in the threads you mentioned said it was not enough.

In my opinion since the max operating wattages are getting do high it is getting increasingly difficult to get a single large TEC for direct die cooling. In my opinion as someone who has played with TEC's you are better of building a TEC chiller then you can multiple TEC's undervolted to give you virtually any power....within reason and don't forget the hotside wattages ! (the hotside is roughly 2 - 2.25 times the cooling power.) and of course you can get an efficient unit with a COP of more than 2...even 3. That said though if you had a phase unit and a TEC chiller running at the same COP the phase unit would have a bigger Dt and therefore be much colder. In my opinion as someone who has played with TEC's if you can afford phase or live where it's cheap... use it and forget TEC.

As for the waterblock for a TEC watch out for waterblocks with a raised contact area because the raised area needs to completely cover the TEC surface, waterblocks with flat copper bottoms are the best. and you don't want any flimsy hold-downs TEC's should ideally be pressured to 150-300 psi a lot of them will bend with this pressure. Clamp the TEC between a waterblock and a copper plate then apply to the CPU. Normal CPU pressure is about 75 psi. If you can't get get a suitable water block just clamp the TEC between 2 copper plates and fix your waterblock to one side of that.

[Naja002]

Originally Posted by SS_FireHawk

Well I think that Im in denial lol. I want to use that 226w just because it's there lol.

Trust me, I know where you're coming from....

Originally Posted by SS_FireHawk

So in other words you thin that just going straight chilled water will do a better job than trying to hodge podge the two together. I can definitely see the 226w becoming an insulator if that is the case.

If the 226w cannot handle the load alone....then the chiller is not going to help the situation. I think we can agree that the 226w is not going to cut it, so all it will do is be a roadblock in between the chiller and the cpu. Game over.

Originally Posted by SS_FireHawk

Now in using just the straight chiller, would it be recommended to have a reservoir of fluid after going through the cpu so whatever is already below ambient stays nice and cold b4 being cooled again and sent through the loop? Throw an insulator around the reservoir of course. I haven't done anything with chilled water b4 so it is a new process for me.

The system will reach equilibrium with or without a res. The only difference will be that a res will add to the pull-down time (more water mass requires more heat removal requires more time).

Again, once at equilibrium the temp difference in the loop really is not going to vary more than 1.5C or less. The temp coming out of the chiller and going into the water block is going to be the same. The temp coming out of the water block and into the chiller are going to be the same. Now the difference between the 2 halves of the loop should not be more than 1.5C or less. So, there is no already chilled water, because for practical purposes--it's all the same temp (within 1.5C).

Try looking at it like this:

the water going into the chiller is 21.5C
The water coming out of the chiller is 20C
The water going into the water block is 20C
The water coming out of the water block is 21.5C

The average loop temp is 20.75C

And around and around it goes....with or without a res--makes no difference. Equilibrium isn't about temps or mass....it's about energy in = energy out.

Originally Posted by SS_FireHawk

What is the recommended loop to ensure the fluid temperatures are as cold as possible? I'm sure insulating the lines would help. Crap i gtg back to class, Ill post l8r

Once you start to get below ambient you will need to be concerned with the dewpoint and condensation...and, yes, that will require insulation in order to protect your components.

With 1 exception, loop order does not matter. That exception is: the pump goes after the rad or res. Res if there is one, rad if no res. The Chiller/Cpu loop without a res will not matter where you place the pump.

[SS_FireHawk]

okay... so now the real problem.... Which pc am I gonna stick it in? I have mine which is a mountainmods 18x18x18 case with 6 fans up front and 4 in the rear (3 lined vertically, one on the mobo side). My phase change is inside that one.

My gf wants the Danger Den acrylic case with the 240mm wc loop. Idk if that will suffice in cooling that beastly mcw-452 properly. I have a spare pa120.3 but obviously.... it's not gonna fit in there. What should I do?

[Plan.B]

Originally Posted by SS_FireHawk

What should I do?

post a pic of your gf.

[SS_FireHawk]

This is her page: http://www.myspace.com/angexlynne

She started playing the Sims games on my computer and found out how crappy playing on a laptop really is. Then she decided to play F.E.A.R and she liked that. And then she likes the Batman game and Devil May Cry 4. It was kinda like the snowball effect. She also finds it interesting when i tear apart a computer and put it back together. She likes fixing things. So for Christmas, she is getting a bunch of parts and she's gonna build her own computer, but of course it can't suck, it's in my house.... and I don't want her on my computer when I could be on it lol. I'm selfish. She wants a water cooled computer so it can have pink liquid and she wants pink fans and lights. A girly computer with a hairy chest.

So what ur telling me Naja is that there is no need for a reservoir besides having a fill port on the chilled side. Would it not give it more cooling capacity or am I thinking too much like traditional water cooling? The dew point in Las Vegas is pretty low which is nice. not to say that condensation can't occur, it's just alot lower than most places when avg humidity is between 3-5%. The heat wrecks havoc in my computer room during the summer... The electric bill goes way up while I try to keep the room at a moderate temperature.

[Plan.B]

Originally Posted by SS_FireHawk

she decided to play F.E.A.R and she liked that. And then she likes the Batman game and Devil May Cry 4. She also finds it interesting when i tear apart a computer and put it back together. She likes fixing things. So for Christmas, she is getting a bunch of parts and she's gonna build her own computer, She wants a water cooled computer so it can have pink liquid and she wants pink fans and lights.

and she's a Pisces, ya lucky bastage

[Naja002]

Originally Posted by SS_FireHawk

okay... so now the real problem.... Which pc am I gonna stick it in? I have mine which is a mountainmods 18x18x18 case with 6 fans up front and 4 in the rear (3 lined vertically, one on the mobo side). My phase change is inside that one.

My gf wants the Danger Den acrylic case with the 240mm wc loop. Idk if that will suffice in cooling that beastly mcw-452 properly. I have a spare pa120.3 but obviously.... it's not gonna fit in there. What should I do?

If you don't have the ability to move the phase or get it moved...then it's stuck in that case. Phase on the cpu and TECs on the GPU(s) and your good to go.

Talk her into a different case, or just give her plain WCing. She'll have 1 or more GPUs to cool also I would think....

Originally Posted by SS_FireHawk

This is her page:
So what ur telling me Naja is that there is no need for a reservoir besides having a fill port on the chilled side. Would it not give it more cooling capacity or am I thinking too much like traditional water cooling?

It doesn't give more cooling capacity in regular WCing. The capacity comes from the blocks ability to transfer the heat to the water and the rads ability to release the heat to the environment. Nowhere in there is anything to do with the amount (mass) of fluid. You will need a way to fill/drain the system, so that generally leaves a res or fillport. The choice is genuinely up to you....one is not better than the other. The same principles apply here.

Originally Posted by SS_FireHawk

The dew point in Las Vegas is pretty low which is nice. not to say that condensation can't occur, it's just alot lower than most places when avg humidity is between 3-5%. The heat wrecks havoc in my computer room during the summer... The electric bill goes way up while I try to keep the room at a moderate temperature.

Right, but you do need to understand dewpoint in relation to condensation. How it all works out is unseen. I intentionally run my rigs chilled to sub-ambient, but maintain everything above the dewpoint. So, I don't use insulation. I simply found the lowest water temp range that worked 24/7/365 for this environment without causing condensation. IIRC, it's ~60-64F.

[SS_FireHawk]

Originally Posted by Naja002

If you don't have the ability to move the phase or get it moved...then it's stuck in that case. Phase on the cpu and TECs on the GPU(s) and your good to go.

Talk her into a different case, or just give her plain WCing. She'll have 1 or more GPUs to cool also I would think....

Im looking at some more mountainmods cases for her. It sounds homosexual but i wish they had more "pink" stuff to use lol. Maybe not the full size ones, but atleast something larger. I can remove that phase out of my case, but it has holes drilled into the the bottom of it that I used for mounting. Or I can get one heffer sized case and throw both rigs in the same one for some crazy water loops. It's a love/hate relationship for this part of the build phase.

Originally Posted by Naja002

It doesn't give more cooling capacity in regular WCing. The capacity comes from the blocks ability to transfer the heat to the water and the rads ability to release the heat to the environment. Nowhere in there is anything to do with the amount (mass) of fluid. You will need a way to fill/drain the system, so that generally leaves a res or fillport. The choice is genuinely up to you....one is not better than the other. The same principles apply here.

Ive always learned that having 4 liters of fluid is always better than having 2 liters because there is more mass available that can absorb heat. I do understand that that has it's limit. I compare it to using a 5 gallon jug with no radiator and a pump circulating fluid. The computer can't cool the entire ocean. But you did say "regular wc'ing" lol.

Originally Posted by Naja002

Right, but you do need to understand dewpoint in relation to condensation. How it all works out is unseen. I intentionally run my rigs chilled to sub-ambient, but maintain everything above the dewpoint. So, I don't use insulation. I simply found the lowest water temp range that worked 24/7/365 for this environment without causing condensation. IIRC, it's ~60-64F.

I was just making reference that it makes it much easier for me to run these setups because of the low dew points. I understand the relation very well. Ive only had condensation occur once after my computer was shut down and I saw water leaking out of my phase change retention nozzle. I immediately unplugged it from the wall and modified the retention nozzle with a rubber lining that shrouded the socket. I haven't seen that issue since

I talked to my gf yesterday, I asked her about what she wanted in her computer... She said she wanted it to work and look nice. I told her I would build one for her that was as strong as my 1.9ghz amd x2 laptop. She was like "F that. Whats the point in building one then? I want one that works, plays games, looks nice, and" I don't remember what else she said, I became uninterested because I was drinking beer. Ill look around some more for a proper home for the setup

Ha, I got the quotes to work lol

[Naja002]

Originally Posted by SS_FireHawk

Ive always learned that having 4 liters of fluid is always better than having 2 liters because there is more mass available that can absorb heat. I do understand that that has it's limit. I compare it to using a 5 gallon jug with no radiator and a pump circulating fluid. The computer can't cool the entire ocean. But you did say "regular wc'ing" lol.

Well, if we were looking to store heat...then the mass would matter. But we're not. We're looking to remove and relocate the heat. How would "storing" it in the loop help that? It wouldn't. Think about it.....

[SS_FireHawk]

Alright, another question... Is it possible to cool two i7 920's oc'd in the same loop with the mcw-452? Similar to how multi gpu's are done. I am thinking of buying a larger mountain mods case and throwing both pc's in there. Like I said before, I'm just brainstorming ideas, trying to figure what works and what doesn't. Feel free to throw ideas my way. I do have to build two computers :P

[Naja002]

Well, again, it depends upon the variables. If a mcw-6500 will not handle an i7 at 4.0...then 2x 226w won't handle 2x i7 @ 4.0.

Will the mcw-6500 handle an i7...yes, I believe it will, but at what clock? So, the same plays out with the mcw-452 x i7x2. Will it handle them to your satisfaction?

[SS_FireHawk]

Alright so maybe I'm confusing a couple things. I thought that the temperature change was "minimal" after exiting a water block. Ill see if i can find more about it.