Just pulled the trigger on first WCing, would like some input.

D

devil_trigger

Gawd
Joined
Jan 24, 2007
Messages
726
Alright guys, I'm about to build my new computer finally, 2500k, 16gb RAM, and 7970 and I just bought all of the WCing parts for it, I would just like you guys to go over it and see if I missed anything.

One thing I forgot is a way to drain this setup when its together, should I add in a t-line somewhere towards the bottom? Also, the res will have the fillport and inlet at the top, so should I have added another barb and some length of hose to help with filling?

10ft primochill 1/2in ID 3/4in OD
EK 7970 acetal block
10x worm clamps
250ml FrozenQ liquid fusion V res
Swiftech Apogee HD CPU block(comes with barbs)
Silver killcoil
Swiftech MCP655-B
Swiftech MCR420
3x bitspower 90 degree barb fittings
3x bitspower straight barbs

Now I'm going to warn you that I have some MAD MS paint skills, so beware!

I was going to use the 3 90 degree fittings on the outlet from the radiator to the res, and at the top of the res inlet, and the other one for the inlet to the vid card. The pump will be on the opposite side of the case as the res, near the window, with the inlet facing the res so I'm hoping it will make that bend.

photo2-3.jpg
 
Last edited:
Better to have the water flowing from the radiator to CPU to GPU then to reservoir then to pump back to radiator.

Flow into reservoir looks good but if you change everything else be sure you enter from at high or equal position then let it flow out of reservoir at bottom or opposite end.
 
From all of my researching on WCing for as long as I can remember I've gathered that loop order doesn't really matter, just to be sure to have the res right before the pump.
 
devil_trigger said:
From all of my researching on WCing for as long as I can remember I've gathered that loop order doesn't really matter, just to be sure to have the res right before the pump.
Click to expand...

Completely true. The order does not matter. Ofc the reservoir need to be before the pump, but other than that it's irrelevant.
 
Idef1x said:
Completely true. The order does not matter. Ofc the reservoir need to be before the pump, but other than that it's irrelevant.
Click to expand...

to add to that, and it could just be common sense, but that the res should also be above the pump so that the pump isnt working against gravity.
 
Idef1x said:
Completely true. The order does not matter. Ofc the reservoir need to be before the pump, but other than that it's irrelevant.
Click to expand...
Check

staradderdragoon said:
to add to that, and it could just be common sense, but that the res should also be above the pump so that the pump isnt working against gravity.
Click to expand...

And check.

Any other opinions guys? Like should I add in a t-line for a drain port somewhere or is there some other way I could drain it? And should I also add a fill tube at the top of the res to make it easier for filling?
 
Last edited:
Anybody? I will be building this next week so I want to try to get as much straightened out as I can. Need to know if I should add a t-line somewhere for draining and about filling the res.
 
No need for t-lines, it's just convenient to have.

When I had the original Swiftech Micro-res, the way I drained the loop was by placing a bowl under the reservoir, and taking the tubing off that connects it to the pump, and draining it that way. In order to fill the reservoir, I opened the top port, put in a funnel, and filled it.

T-lines just make things simpler. They're definitely not needed.
 
What about the water in the rest of the loop though? Will it backflow out of the pump and go everywhere?

I'm only concerned about filling it the way it is because there is going to be a pretty small amount of open space between the top of the res and the top of the case and the rad.
 
Not initially. The reservoir would need to empty first to allow air into the loop, which will then allow the backflow. The way to counter this is to either keep the tube above inlet port on the reservoir, or empty out the tube first, allowing air to enter the system. It's critical here to not open the fillcap, so that the only way for air to enter the system is through the pump, and gravity pulling water out of the outlet on the reservoir will pull the water out through the system.
 
Ahh ok, gotcha now, sounds good. I guess I'll also just use a small full and run a little tube into the funnel from outside the case for filling.

Thanks for the help!
 
devil_trigger said:
What about the water in the rest of the loop though? Will it backflow out of the pump and go everywhere?

I'm only concerned about filling it the way it is because there is going to be a pretty small amount of open space between the top of the res and the top of the case and the rad.
Click to expand...

Take the tube off the reservoir, not off the pump, keeping the end of the tube above the pump.
 
Tsumi said:
. . .T-lines just make things simpler. They're definitely not needed.
Click to expand...

Resevoirs just make things simpler. They're definitely not needed.:D

I use a T line with a fill port from the top of my case, its pretty hassle free and what people say about bleeding the system is greatly exaggerated.
 
Your diagram is perfect. It wouldn't get any better. A single 2 fan rad has so much cooling power with the right fans you can cool 2 GPUs and a CPU with one 2 fan rad all day long.

Don't over bake this, Think simple and take any recommendations from the people here with junior experience with a truck sized grain of salt. Not saying anyone in particular on this thread is junior at water cooling but I have seen very very many amateur comments that can get a lot of people in trouble with water cooling on this forum especially concerning radiators and setups. I even see one guy recommend to another member that he get a very slow low flow rate pump so the water has more time to contact the parts. This is a HUGE NO NO. let the learning begin!

and ....

semiconductorslave said:
Resevoirs just make things simpler. They're definitely not needed.:D.
Click to expand...

Yes you are right. Resovoires just make things easier for 2 reasons:
1. Lets air bleed off the loop and bubble to the top of res.
2. Gives an extra insurance against gas permeable water loss through tubing such as Tygon and other synthetics. Yes you lose coolant over time (miniscule amounts) using Tygon but the tubing is about as good as it gets.
 
Looks pretty standard to me. Although I'd get some 45 degree fittings on those tight turns, but since you're using worm clamps the situation of the tubing popping off during high temps shouldn't happen.
 
I use a T-line and added a T-Line Res so I don't have to worry about topping off. With the tubing going to my pump there is plenty capacity.

Here is my loop order: (I apologize for my cable managment, its a work in progress;) )

Dual pumps to 120mm Rad, to CPU, GPU, 120, and back. Tline is fit right before pumps and tube goes to resevoir fill port on top of case.



Heres a view of my T-Line Res from the back of my case. Oval used to be two 80mm fan holes in PSU back plate (my PSU is right below):


Port on top behind 240mm radiator grill:

 
Alright guys sounds good. I just took a trip to Microcenter and picked up a "cheap" case to transplant all of my soon-to-be HTPC stuff into and just finished the operation, successful! Now I have my main tower all gutted and ready to start planning and building. I got everything so far except the GPU block and the res, will be here monday.

My main concern now with the rad on top is that there is already a 120mm blow hole but its definitely not going to line up with where the rad is going to go, so its going to look a little goofy, hopefully I can pick up a grill to cover it up.

Thanks for all the suggestions guys, I will try to update this thread with progess although I'm not that good at documenting it. I like to get into a groove and not stop to take pics lol.
 
Well I'm going to have to cut too, what I was saying is that there is already 1 120mm hole, but I'm putting a 4x120 rad up there. Its not going to line up with the size and fan placement of the rad.

Like this:
Say this is the top view of my case and the O is the single fan hole
|......O....|

And then with a 420 rad
|.OOOO.|

You can see that the hole placements will overlapse, so I'm not sure how its going to look.
 
You could always get a rectangular piece of aluminum for the top that was wider than the hole and then cut your radiator cut out into that.



I used to have dual 80mm front fans, but used a blank square plate of aluminum and cut a 120mm into that. I used nice black finish screws to mount the plate, came out looking good.

I've also seen replacement lian li top pieces that are solid. You might have to buy a rivet gun at home depot, I did and they're a lot of fun.

http://www.lian-li.com/v2/en/product/product04.php?cl_index=2&sc_index=45&ss_index=74

http://www.frozencpu.com/cat/l3/g30...g_Case_Upgrades-Lian-Li_Top_Panels-Page1.html



 
res-pump-cpu-gpu-rad-res...

If you have more then 1 rad id do it differently... in my current set i use this

res-pump1-gpu1-gpu2-rad1-cpu-pump2-rad2-res
 
Like I said earlier, I've read all over the place that loop order doesn't matter, as long as the res is before the pump. If I do CPU before GPU, it will make my tubing run longer, which isn't necessary.

EDIT: I do have a quick question though. Since my res is going to have the outlet on the bottom and inlet at the top, do I have to fill the res completely so that the water isn't "falling" so-to-speak, into the res? Or will it be ok even if its filled a little more than 3/4 or so?
 
Last edited:
you want to avoid air bubbles from getting in your system, and if having the water fall through your res causes bubbles then you will need to fill the res up all the way.

Now if you fill the res up all the way you still need somewhere with air so when the water heats up/expands it has something to compress
 
So do you guys think it will be easier if I flip the res so that both the inlet and outlet are on the bottom and just run the extra tubing from the rad to the bottom of the res?

Also, update! Just finished cutting the holes for the rad on top, the middle holes look like the newegg symbol.:p

2727e1b7.jpg


18f86c33.jpg
 
Sure it does...rad before res ... otherwise your shooting hot fulid into the res... res is a sitting area ...fuild should enter already cooled...
 
Um, well if you look at my loop order my rad is before the res.

Another update! Got mostly everything mounted and the tubing run.
6b9a1566.jpg


Does anyone have any suggestions for covering my cuts to the brace near the top? Its just covered with white electrical tape right now lol.
 
id change that loop to cpu then video card...gpu is easy to cool ...

find rubber oval gromlets..if you cant putty would work... not great looking tho

also make sure you have that inlet right on the cpu.. i cant see it well in the picture looks backwards..usually the left side is inlet
 
auto stores have black trim for door edges to protect the paint that you could use, or you could get black rubber vacuum hose from auto store and cut a slit in it and push it around oval.
 
If flow rate coming out of your pump is not an issue, then the order (CPU then GPU or vice versa) does not really matter. However, the reservoir should immediately feed into the pump. I have tried different iterations of order and at best, there is a 1C difference. What I have noticed is that having multiple serial radiators did helped some (2-3C without fans or 4-5C with fans) as long as your pump can handle the flow restrictions.
 
The CPU inlet is right, I double and triple checked, its just that the block is rotated to allow an easier bend from the GPU to the CPU.

I'm liking the black vacuum hose idea, or maybe somethings like the plastic tubing for cables.

Like I said, I've been reading about WCing for a couple years now and the differences in block order is almost negligible, maybe 1-2C. I'm not an extreme overclocker or someone that needs the lowest temps, I'm not going to care about 1-2C, I'm mainly doing this for the experience.:)
 
SDNYC09 said:
Sure it does...rad before res ... otherwise your shooting hot fulid into the res... res is a sitting area ...fuild should enter already cooled...
Click to expand...

completely untrue.

The water at any single point the loop is going to be within a couple degrees of the water at any other single point.

Example: The water leaving the hottest block in your loop will be within a couple degrees of the water exiting the radiator.
 
Quote:
Originally Posted by SDNYC09 View Post
Sure it does...rad before res ... otherwise your shooting hot fulid into the res... res is a sitting area ...fuild should enter already cooled...
completely untrue.

The water at any single point the loop is going to be within a couple degrees of the water at any other single point.

Example: The water leaving the hottest block in your loop will be within a couple degrees of the water exiting the radiator.
Click to expand...




From hands on experience ... it does matter .. i had cpu-res and my res was hot to the touch...res should provide a small time for the fluid to sit...
 
Alright guys, I am now leak testing!

ee253ec4.jpg


Man I should have gotten a flow meter, its really hard to tell that the liquid is actually moving, but I can hear the pump humming away so its got to be pumping right? Lol.

Also, all of these little tiny bubbles attached to the tubing and the res will eventually work themselves out right?

6ace3026.jpg


53895b5a.jpg
 
yep, those tiny bubbles will eventually disappear. You can see any movement on the surface of the water in the res? That's how I can tell mine is moving.
 
SDNYC09 said:
From hands on experience ... it does matter .. i had cpu-res and my res was hot to the touch...res should provide a small time for the fluid to sit...
Click to expand...

False. Our blocks and radiators are not efficient enough to make loop order in any way/shape/form important. A plastic reservoir is not going to radiate heat at all, as plastic is an insulator, so letting "hot" fluid sit in a reservoir won't cool it off.

I don't care where you measure from, loop temperature between blocks or radiators or in the reservoir will not differ more than 1°C, assuming you have enough pumping power. Water does not spend enough time in a block or radiator to gain or shed enough heat to make a measurable difference in temperature.
 
tip the case back and forth , give it a shake to loosen the bubbles... they will all clear out in time... after a few hours open the top screw to the res let some air out...i would have put a flow meter...just to make sure the fluid is moving...
 
SDNYC09 said:
From hands on experience ... it does matter .. i had cpu-res and my res was hot to the touch...res should provide a small time for the fluid to sit...
Click to expand...

I hope you are joking. At a flow rate of a gallon+ per minute.. How long do you think water sits in the res?

Your radiator hold more water than the res, is designed to remove heat from the water, yet the water entering the radiator is barely warmer than the water exiting the radiator.
 
depends on how big your res is 1...2... from my experience like i said MY experience of 6 years of water cooling ... from cpu block directly to the res i was not comfortable with the res being noticeably hot to the touch...changing the loop order to include it passing through (My setup) 3xGTX rad then to res..corrected that for me...

In his setup your right it may not change it noticeably... in my prior setup res-pump-rad-5870-5970-cpu- chipset -res..the heat was noticeable...
 
Last edited:
Here is one of the best simple write-ups I've seen on loop order:
thebski @overlcockers said:
Through extensive research and questioning here and on other forums, let me debunk some myths that you seem to have in your mind.

I need the "coolest" water going into my blocks:

As some have indicated, loop order literally does not matter. The water inside your loop is like a body of water that is going to be within a 1-2C range anywhere you measure it inside the loop, as long as water is moving.

My intuition may not be exactly correct, but I think of it like heating up water in a pan that has a hot plate on the bottom but somehow has the sides of the pan actively cooled. In your loop, the cpu block acts as the bottom of the pan (heated by a hot plate), and your radiator acts as the sides that are cooled somehow.

Now imagine this set up with someone stirring the water. Is the water going to be hotter at the bottom and cooler at the top? No. It will be one unified temperature because someone is stirring. There is a certain amount of heat going into the pan via hotplate, and a certain amount being extracted via cooled sides, so eventually the water in the pan is going to reach an equilibrium temperature. The point being though, that the temperature is going to be the same in the bottom of the pan as the top of the pan because the water is being stirred.

It is exactly the same in your loop. Your loop is a body of water. You may not think of it as a body of water because it isn't a container of water like you'd generally think of a body of water, but it is still a body of water that is formed in an odd shape contained by your tubing, res, etc. As long as the pump is going and water is moving, the water is going to be mixed around in your res and other parts of your loop to where it will essentially be the same temperature at any point in the loop.

What does all this mean? You could put 4 GTX580's and an i7 3960X OC'ed to 5Ghz in order and then follow that with 4 huge radiators. You could also have a radiator between each component. You might think having a radiator in each component would cool better because each block is getting "cooled" water. Not so. The whole body of water is cooled. The whole body of water is also heated, but it's moving so that's where the equilibrium temperature comes in. It does not divide up into sections.

Also, water doesn't really spend enough time in a block to "heat up" enough to really see a difference. It's in and out and might heat up 1C, but it's the continual passing of water that is at an equilibrium temperature that cools it.

All that said, if your pump fails and water stops moving the water in that block will heat up in a heartbeat and you will have major heat issues. A water block doesn't act as a very good air cooler :).
Click to expand...


so.. if you can notice one part of your loop being hotter than other parts.. you need more flow.
 
You must log in or register to reply here.
Back
Top