Cooling high water cooled temps

Look, fei, again: speed up your fans to 9V. It will make little extra noise but make a big difference in temperatures. Honest. Go on. Try it.

EDIT: D'oh! Should have read your first post better:

OK, my bad...

 

Before trying another pump you could remove a component at a time from the loop and see where the resistance is coming from. I have an AC setup using 6mm ID tubing, multiple blocks and an aqua-stream pump and the water entering my res (last thing in the loop) moves a lot faster than what you have described.

 

I also noted that. I'm using an AC setup with both 6mm ID and 8mm ID, and the water comming into my res sprays. Your pump is a lot more powerful than mine and DRB's, so there has to be something choking the loop.

Try hooking the pump directly to the radiator and see if the flow is even going in and comming out. There could be something wrong inside it.

 

ok i've found out something really wierd.
i got hold of a ddc to test.

well i put it all together and tightened the cpu block down as far as i can,
temps was reading the same so i tightened the block some more and it made temps raise slowly to about 90c!

so i thought ok i'll loosen the block abit while the com was in bios and the temps droped and stayed at 38c!

for some reason tightening the block down too much made temps read higher.

i'm gonna try the same thing again tonight using my 1250 pump.

btw i've only got the cpu block in now as i've got rid of the nv68, and have tried b4 with fans at more volts

 

Cool!

 

Interesting... perhaps overtightening it temporarily and then loosening it corrected some block seating problem or something. Although it is hard to visualise how. On the other hand, perhaps the added pressure caused a warping of the mobo or socket and hence some weird artefact of the temp sensing signal. Or perhaps it caused the flux capacitor to overload on Ziemen particles creating a Tyken's Rift phenomenon. Er...


OK, I can't help my curiousity: how can your room temperature be 25-27C?

 

there has to be something cloggin the system for the water to run that slow

The G4 is clear and the NV68 is out ot the loop leaving the rad, tubing, and pump. I suspect the rad. It could have a giant airbubble in it or debri but i dont know how that could have got there

well good luck

 

well that was when it was real cold outside and i had central heating on pretty much full blast.
room's dropped abit to about 24c now with no heating on.

 

i've run the every component by it self and its only the G4 that stop's flow quite alot. i've opend it a few times and not found anythign blocking it and even blasted it with compressed air, even tried a second G4 with same flow results.

the DDC does pump the water thru abit faster but not by a great deal

i can try get some vids of what the flow is like if i have time

 

mod it!

Ok, fun testing at stock eh?!! Try to get permission (if borrowed) to 1/2" flowmod the DDC. That might get the pump to spit a bit more water!

 

I don't think flow is the problem, really. I have 1.6 litres per minute and my temps are just fine...

I am a bit bothered about your internal case temps. 38C in a Stacker? Seems a bit high to me. In your current config that would be close to the temperature of the air drawn through your rad, so you hardly get any cooling at all. But you said you tried the rad outside of the case as well. It all just doesn't make sense. All I can think is that the temperature reporting is off.

In any case 38C idle is not bad, but you will still have to check it under load (BIOS temps are invariably idle temps).

What ID barbs are you using, by the way?

 

The front bays of the stacker are all mesh and iirc it has two large mesh panels on the sides so 38C is very high. With that amount of 'breathing holes' surely convection alone would dissipate the case temp a bit as its a large temperature difference between the case and ambient.

 

Actually, with all the mesh, air goes static, and temps are worse, unless you situate the fans properly. Problem is, in a regular case, air pressurizes, and so it is forced to go where we want it. You can't really develop any kind of pressre, but instead have to push hard enough to make currents-a much more difficult job.

 

OK, but you're not using the same pump or blocks are you? The G4 is really quiet restrictive, I think only the NexxxosXP may be more restrictive.
The more pressure that G4 gets, the better I would think (barring any other problems)

BTW- Are you using that Johnson pump currently? Isn't it a brushed motor?
What's the MTBF under continuous duty?

Thanx.

 

Are you trying to say that the mesh dissipates the airflow? Exit path shouldn't matter, as long as hot air leaves the case and cool air gets in.

Yeas, but that's not the point. Fei reports a flow that sounds roughly similar to mine. An Eheim 1250 puts out 10% of its rated capacity (317 gph) at high resistance (say tests at overclockers.com) , which is about 1.89 litres per minute. The G4 is a really good block and should be able to perform at least as well as my Phaze II blocks. I use PF-5080 but that shouldn't make a lot of difference. And he has a way bigger rad to play with.

My pump is now a Davies-Craig EBP --very similar in performance to the Swiftech MCP650 (and 50.000 hrs MTBF) . Ditched the Johnson: very powerful, and robust, but very noisy, and way too short MTBF on the motor.

What keeps bothering me is that high internal case temp. Perhaps he just needs to take some external thermometers to it and see just how accurate that BIOS reporting is...

 

Well, to me that is the point! can't compare systems even if same components are used... too many variables. I highly doubt the 1250 is pushing 1.89L through his system, isn't the 1250's dead head at 7'?
[edit] brainfart! OK let's say 1250 is pushing 1.89lpm .... wouldn't say 5-7lpm be more optimal? I think 1.89lpm is pissy no? I need more coffee!!! heh.

The G4 properly pressurised would blow our blocks outa the water!!

I'm not saying the 1250 is the problem.... but with his level of restrictiveness it may be the weakest link.

[edit] Yeh I remembered the thread at OCAU about the Johnsons. heh I said Johnson!

 

case temps may be to blame, or there may be a faulty temp sensor.

As for the case flow, it's hard to pressurize open air, isn't it...

 

This Procooling review suggests that the G4 should perform on an equal level with its predecessor, the SS Cascade at 0.5 GPM (1.89 litres per minute) although admittedly it fares a bit worse at lower flow than that. Still, his setup should work...

1.6 lpm seems sufficient in my dual Opteron setup with a teeny-tiny BIX.

I'm just getting confused now... Perhaps he should check the rad again... Coffee sounds good.

 

I agree, coffee would be great.

Waiting till tonight, i found a coffeeshop that makes God's own coffee not five minutes from my place... /end hijack

The G4 is a helluva block, i had to go doing some reading about it, but you should NOT have those probs.

However, do me a favor and check the temps of any MOSFETs nearby...

I had an instance of heat wash driving my koolance crazy...

 

well on this nf4 dfi ultra-d board is the PWM IC suppose to be case temps?
on stock cooler atm
that's reading about 40 currently
chipset reading 41c
cpu idles 38c. but them very second i start the In-place large FFTs test in prime85, cpu makes a huge jump to about 52c like instantly! and slowly climb to 64c, stopping it makes the temps fall straight away to about 48c and slowly cool back to 38c

smart guardian and mother monitor is reading the cpu volts stable.(not fluctuating or anything)

i've ran some rad flushing stuff thru the 120.3 for a hr or so just in case.
i'm going to chuck the G4, 1250, 120.3 back in later........been bleeding over night and pretty much all the air is out