Mother of All Water Cooling (Pics)

A

annaconda

[H]F Junkie
Joined
Apr 13, 2005
Messages
9,925
Alright Guys as i mentioned in my previous thread that i wanted to add another MCR320-QP-K Triple 120mm fan radiator, to my existing MCR320 radiator + water cooling system.

I am proud to say that my decision was right. After adding second MCR320 radiator my temperature dropped whopping 10C. Yes believe me 10C under full 100% load, following are the pictures with comparison between single radiator and dual radiator.

All the temperatures are been taken with INTEL TAT.

Before the second radiator Temperatures were:

CPU E6300@ 3.2 = Idle 30C, Load = 57C
Gpu 8800GTS 675/2000 = Idle 38C, Load = 42-44C


With the Second MCR320:

CPU E6300@ 3.2 = Idle 25C, Load = 45-47C
GPU 8800GTS 675/2000 = Idle 34C, load = 38C

Well here are the specs of my water cooling Now:

INTEL E6300@ 3.2 with Swiftech Apogee CPU Water block.
EVGA 680i SLI Motherboard.
8800GTS @ 675/ 2000 with Swiftech MCW60 GPU water block.
Swiftech MCP655 Water Pump 12v.
1/2" ID Tubbing from Home Depot :D (20' for $3.99)
Swiftech Micro Reservoir.
6 x 120mm Yat Loon Fans.

and here are the pictures.

1664.jpg


2.jpg


3.jpg


4.jpg


5.jpg


6.jpg


9.jpg
 
yah your temp delta seems really high.. i run a E6300@ 3.3 and my idle is 40 and load is 45
 
Temperatures with Two Radiators. :D 100% load for an hour on both. Room temperature was 75 F on both tests. I think it will be better with lower room temperature.


45-48C.jpg
 
LMAO! Wow....as previously mentioned you should not have been up to 60*C. That being said.... your setup is impressive.


Under load with a 800MHz overclock on my CPU and dual 8800GTX's my load temp is 37*C on my cpu with a single triple 120Rad.

GPU temps are around 50*C load which is understandable.
 
I think the temperature were high under load is because of the room temperature. Like i said room temperature was around 75 F. I like it HOT in winter :D

Guys you should tell me your temperatures using INTEL TAT. if i use NVIDIA nForce then the difference is big as you can see in the pictures, that Nvidia Nforce temperatures are 7 to 10 C less then Intel TAT.
 
Are you guys using TAT for temperature reading. If you wanna match temperatures use INTEL TAT, it is the most accurate temperature reader. It tells you the core temperature, not the temperature from motherboard sensor.
 
With my dual BIP I load at about 51c in TAT with both cores loaded, and RTHDRIBL in the background (the GTS is in loop too).

Room temp of 25, or about 77.

Remount your waterblock?
 
With intel tat's workload tool, and orthos blend running I get cpu 0: 50C cpu1: 50C with my sig overclock at 1.48V.

With RTHDR running also, my 7800GT "loads" at 46C. Speed fan says the NB is running at 30C, but I'm not sure if it's right. Hell the bios says it idles at 18C:rolleyes:

Thats with 1 MCR-220. Willing to bet its your ambient. Mine is a chilly 19C. :p

EDIT: I just checked and I had the workload set to 75% for some reason. Fixed temps.
 
core temp and TAT are much more accuarate that any other temp monitor... they look at the CPU temp sensor not the motherboard sensor under the CPU, i would be most of your reporting temps are not using TAT or Coretemp, so before you report...."your temps were too high to begin with" you should use TAT or coretemp to see your own temp.... Big OCs with decent vcore increase on AIR i would be are around 55-65*C with TAT or CT but 43-50*C with MBM, speedfan etc...
 
yeah i was using coretemp and it was pretty much the same as my mobo (within 1 or 2c)
 
what vcore are you at? I get like a 5-10*C difference between Coretemp and MBM5.... are you delidded?
 
yikes, you're getting some kinking at that loop between the two rads. Maybe get some zip ties and round them out again. Might help even more. that said, there's probably other spots that could help from zipties, but I didn't really look
 
ziddey said:
yikes, you're getting some kinking at that loop between the two rads. Maybe get some zip ties and round them out again. Might help even more. that said, there's probably other spots that could help from zipties, but I didn't really look
Click to expand...

Hahaha, those kinks arn't too bad, you should see my loop... One of my tubes is 60% closed.
 
My VCORE is 1.45v. No there is no kinking between radiator, where do you see it.
 
The tubing connecting the two rad's. It's not that severe as a "true" kink, but it looks like it would affect flow a little, and for the effort it takes to put two zip ties, seems like it's a "might as well" thing at worst
 
Wow your temps still seem real high - on AIR (Arctic Cooling Freezer 7 Pro) I was getting high 50s Orthos load (w/TAT) with my E6300 @ 3.22GHz.
 
As I stated, I had a major kink in my loop... 2 of them actually.

Adding zip ties did nothing for my temps,

But you might as well...
 
Those are not the kinks, it was just the angle of the shot. It looks perfect i just made sure again.

But thanx for the heads up.

Guys there is a small problem, when you shuts down the computer all the coolant come down to the reservior, because it is down in the bottom right next to pump. It caused some overflows too.

I was thinking changing the radiators setting. Instead of putting them over the top of each other put them right next to each other , and make a big sandwitch.

What do you guys think? I know some might say the hot air from one radiator will effect the performance of other. But there is no hot air coming from the back of both radiators, so it might work?

I think i have to check the contact on the CPU block one more time. Any other suggestions please?
 
i think it would work. theres not enough "hot" air coming out of the first rad to affect performance.... IMHO
 
 
Stacking radiators works like stacking waterblocks...

Each block heats the water up more and more, reducing the effectiveness of the next block down the line. Unless the water gets WARMER than the copper waterblock, it will always take away some heat (under normal operating conditions).

Same thing works for air. The warmer air will just take LESS heat away from the 2nd radiator, reducing it's effectiveness... Maybe about 20%?

And HOLY HELL Ghost. That's a big rad. get some pics up :D
 
ehhhh? whats up with using wire to hold the two rads together?

surely you couldve come up with something better looking than that. :p
 
ozziegn said:
ehhhh? whats up with using wire to hold the two rads together?

surely you couldve come up with something better looking than that. :p
Click to expand...

Other wise the radiator will fell down. This way they are stacked perfectly and balanced.

I cannot believe that air cooling users are getting the same temperature as i am. I think i have to do some changes.

My current loop is:

Reservoir -> Pump -> GPU -> CPU -> Rad 1 -> Rad 2 -> back to Reservoir.

I think this might work better ?

Reservoir -> Pump -> CPU -> Rad 1 -> GPU -> Rad 2 - Reservoir.

But this will increase the tubbing length, unless i re-locate the second radiator neer to the GPU.

What do you guys think?

So you guys are saying my temperatures are still High?
 
annaconda said:
I cannot believe that air cooling users are getting the same temperature as i am. I think i have to do some changes.
Click to expand...

I top out in the 54C to 56C range with Orthos running my C2D chip @ full load with air cooling.

surely I would think that watercooling would give much better results than what I'm getting.
 
the only thing I could think of is maybe your CPU waterblock isnt making optimal contact with the CPU's base.

(I'm just taking a guess because I am obviously not a watercooling expert.)
 
ozziegn said:
the only thing I could think of is maybe your CPU waterblock isnt making optimal contact with the CPU's base.

(I'm just taking a guess because I am obviously not a watercooling expert.)
Click to expand...

I will definately check it today again, i checked it before at least 2 times in the last 2 months. Now a guy in this forum mentioned we should use difference screws for the apogee block, because stock screws don;t put enough pressure on the cpu.

I did that also, but just to make sure i will double check tonight to see the block is making good contact.
 
That's a crazy setup anaconda :cool:
You've been in my thread about the Apogee, and I'm sure it's the problem. Mine looks fine (okay, maybe the base is incredibly slightly convex but nothing a tiny layer of AS5 wouldn't fill) but the temps were just bad with it. I know someone who got an Apogee ages ago and swears it works really well (although he does talk a lot of crap) so I dunno what's up. Did you try the uncapped nuts or extra washers? Lap the block maybe, mine is almost imperceptibly
 
Yes i just put some uncapped nuts, and i re applied the AS5, as i checked the last contact was perfect also, but we will see how it goes.
 
You must log in or register to reply here.
Back
Top