• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Finally Under Water, and Some Shared Wisdom.

H

holli4pirating

[H]ard|Gawd
Joined
Nov 13, 2005
Messages
1,432
I finally setup a water loop in my rig; I've been wanting to for about 2 years. I've gotta update my sig, but it's on my gaming rig. I decided to only put the CPU on water and not the GPU. It's my first try and I wanted to start small.

I've got an Iwaki pump that I found here on the forums, a Apogee GT (which I bowed), a dual heatercore, and the 7/16 Masterclear tubing. I went with a T line to save on space, and the build went pretty well.

My gf works in a lab, so she brought me some distilled and deionized water, and I used Hydrx as well. While I must say it is a rather nice green, I have no windows on the case and I really don't care about showing it off; it just worked out such that it looks nice.

Bowing the block was really rather easy; I took it apart and pulled the old ring. Then I put the base flat on my desk and stretched the bigger ring such that it cleared the pins. Next I took the top part of the block and sandwiched the ring between the top and the bottom. I then applied pressure and poked the ring into place with a flathead until it was seated. This is how I'd recommend doing it, it's real easy.

Getting things set in the case was pretty straight forward, and soon enough I was ready to fill 'er up. I started to fill the T, and then switched the pump on for a second to get the water into the loop. Here's my next bit of advice. Leave several inches of space at the top of the tube, because after the water is drawn in, air will want to get out. When it does, it will force the water up a bit, and mine overflowed a bit. Gotta watch out for that.

My next little insight involves bleeding the air outta the loop. Once it's mostly filled, you've gotta turn the pump off for intervals to allow larger bubbles to form. Then, to get them out all at once, switch the pump on and off repeatedly until the big bubbles get just past the T. When the do, shut the pump off and wait for the bubbles to escape. It took me some time to do this effectively, but once I got it, the process sped up. I also found that tapping on the T itself and the tubing above the T helped to get the little bubbles that stick to the walls out.

Now that I'm up and running again, temps are looking good. I'm running the SMP folding client with 100% usage and 1.42 volts under load an my Q6600 at 411*9. Temps are 43/43/37/38 with little fluctuations. By comparison, I was about 13 degrees higher with my old Geminii. And the AS5 hasn't even set yet. Also, I was able to rig a fan over the PWM, and that's dropped from high 60's down to an even 50. I'm very happy about that, because it had been a concern of mine.

Well I hope my little pointers are of use to others, there are def some things I wish I'd known about getting a loop up and running that I learned by doing. If anyone has questions I'd be happy to answer what I can, and I'll try to get some pics up.

The whole thing
DSC_0001.jpg

The heatercore
DSC_0002.jpg

Fans on memory and the PWM
DSC_0004.jpg

A slot cooler to keep the GTX cool
DSC_0005.jpg
 
Not really sure what you're asking, but the loop is: pump, rad, CPU, T, pump. I also haven't set up shrouds on the rad yet, just my Papst 4114N's at 12 volts blowing at the rad and two high speed 120mm Yates leaning on the rad on the other side sucking air through.
 
Now priming at 1.42 core load voltage and 420*9. This was unstable under air, I think because of either core temps or PWM temps, but its looking good now. Temps are 47/47/45/45 +/- 1.
 
sorry about the random question. haha. i just finished mine today but god damn are there alot of bubbles. the coolant is like a milky color because of that.
 
Ah I see. I found that with that many bubbles, it's best to shut off the pump and let them collect into larger bubbles. Then you can turn the pump on and off to move them without breaking them, and hopefully get them out of the system as a few large ones.
 
oooh I see, i let it sit for a bit and now its all clear! now i have to somehow get the mc8800 heat sinks on haha.
 
I'm glad that worked for you. I think if I was to put my GPU under water I'd get one of those massive D-tek things. Seems way easier than dealing with multiple heatsinks and...well...it's massive.
 
holli4pirating said:
I'm glad that worked for you. I think if I was to put my GPU under water I'd get one of those massive D-tek things. Seems way easier than dealing with multiple heatsinks and...well...it's massive.
Click to expand...

I've seriously considered a UniSink, but they don't seem to have very many fins/surface area..

I am getting frustrated with my Zalman ramsinks falling off every couple weeks, though, so I may end up with one anyway.
 
I think if you were to screw a 120mm fan onto the side of the PCI slots it would provide enough air flow and stay silent. I know fans can be mounted easily like this if you have a long enough screw.
 
but if you get the full blocks and decide you want to get and of the new gpus there no chance that you can switch it over. well unless the cards are extremely similar.
 
The unisink's aren't full cover blocks; they are instead of ramsinks and in addition to a normal GPU block.
 
You must log in or register to reply here.
Back
Top