• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Bad Water or Good Air?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

LarryRE

Registered
Joined
Apr 16, 2003
Location
Jacksonville, FL
After drooling over H2O results for years, I finally went for it. Trouble is my results are not improved and I am beginning to think I did a good enough job with the air cooled setup that water isn't going to help. Any advice or comments are very welcome!!

The system:
AMD Venice 3200
MSI Neo2
OCZ PC3200 Platinum Rev.2
OCZ Powerstream 470 Power supply
GT6800 with arctic cooler exhausting

Air Cooled setup:
Thermalright XP90 lapped with a 64 CFM Fan
The lapping got me another 4 degrees C cooler.
1 front and 1 side mounted 70 CFM 120mm intake fans
1 rear behind CPU 120mm 94 CFM exhaust fan.

Idle 34 C/ Load 51C

Water setup:
TDX CPU block with the #4 Plate.
Laing D4 pump
Dual 120mm heater core externally mounted
2 80 CFM fans
no GPU or chipset blocks

Idle 33C/ Load 49 C

My overclock was 2680 (10x268) on air and I can't budge the thing with the water setup, nor can I get my temps any lower. I have tried a few careful remounts of the block and purged the air fron the system but no joy on lowering the temps! :bang head

I picked the #4 accelerator plate for the TDX since by all the reviews I could find, it seemed like the best middle of the road choice.

My Fans on the air setup were on a Vantec controller so I had the airflow pretty well balanced.
Now wondering weather I should have expected any real improvement or whether I am missing something on the new water setup. :shrug:

What do you Guru's think????!!
 
are the 2 80mm fans on the heatercore? do u have a decent shroud? does the air coming off the heatercore feel warm? some pics would be nice.
 
lithker said:
are the 2 80mm fans on the heatercore? do u have a decent shroud? does the air coming off the heatercore feel warm? some pics would be nice.

Hey lithker,

I'm digitally challenged on the photography for the moment and already messed up the desciption. I have 2-120mm fans on the heater core.
The air through them does not seem very warm though.
 
The change from idle to load in your water setup seems like way too much. Temps in a water cooling system usually don't change that much from idle to load.

Are the fans on your heater core shrouded? Are the fans 25mm or 38mm thick? Could your radiator be filled with dust? I think you should be getting much lower temps with that setup.
 
Are the fans mounted on a shroud?

A shroud would be a standoff for the fans usually around an inch thick or so that allows the air comming or off teh fans to be directed more evenly.

Also did you bleed the system completely free of air?

And, what are you using to get your temps? If you are going by the temp sensor, the NEO2's on board thermal sensors pretty much suck.

And remember, you still need SOME airflow in yrou case. At least an exaust fan is recomended.
 
I thought the idle to load temps seemed suspect as well. No shroud, the fans are pushing air thru the rad and are mounted right on it. I know about the Neo2 temps sensors being doo doo but this one has seemed very consistent, in that when I have made a change the temps change and then stay fairly constant. I have a danger den resevoir which is mounted at the highest point and it still has some air in the topmost section but I think I got the rest purged.
Could I have mounted the accelerator plate badly? Seemed like it only fit one way.
 
Did you lap the WB? Is the WB mounted to the chip tight enough? Are there any kinks? On board temps or external sensor?

I also hear that the heatspreader may not be mounted to the core very well on the new AMD chips. Your high load/idle temp delta seems to indicate that the heatspreader may be the problem. If you have the balls, remove it.
 
The res should have a bit of air trapped in the top, what you are looking for is to make sure its not sucking down air bubbles into your pump.

Heatercores (for that matter any rad) won't perform as well with the fans just mounted fluch against it. You can get soemthing like a 40% increase from using a shroud. You could purchase one (not sure which core yo have, but Weapon makes shrouds for most of the popular cores). Or you can make a "ghetto" shroud.

Get yourself some aluminum duct tape. Its a tape made from very thin aluminum. Get tome longer posts for your fans (unless they are already long enough to do this). And something to stand them off about an inch to an inch and half form the core. Nuts, washers, whatever works. Mount the fans, the very carefully seal the space around the HC and the fans with the tape.

Its a cheap and effective shroud. I suggest the aluminum tape because its very sturdy.
 
heres a good pic showing how the accelerater plate should be clicky

do u have the pumbing for the TDX right? i.e. inlet over the accelerater plate and outlet on the corner.
 
No lapping on the WB, it was really smooth and yeah I'm quoting the on board temps, it's all I've got. They are likely not accurate but have been pretty consistant over time. And yes I do not have the balls to remove the IHS. The WB is tight enough that the springs are completely depressed and a half a turn passed that. I will triple check my air purging but don't think I have any left. Damn.. what you said about the IHS makes sense. That is if I really have all the other basics covered. How much difference would a shroud make? I will also try putting on an additional 120mm fan that's a screamer to see if my air flow can make much difference.
 
A shroud can make a huge difference.

Think about it this way.

Heat exchange is all about passing the mediums across eachother as quickly as possible in the largest space possible.

If you limit the speed or the space you loose cooling potential. By mounting the fans directly on the core you create dead spots anywere the fan blades do not cover. So you have limited your sureface area. Moving the fans off the radiator and pushing the air through a sealed chamber allows it to use the full breadth of the radiator.
 
lithker said:
heres a good pic showing how the accelerater plate should be clicky

do u have the pumbing for the TDX right? i.e. inlet over the accelerater plate and outlet on the corner.

Thanks for that link. Think I did the plate right. As for the plumbing, I have the pump connector which is pointed up (the pump output I hope) straight to the center barb on the TDX. It seemed as though that was the pump output!?
 
Jas said:
That would be the outlet yes.
Thanks Jas!
I guess I need to see about the shroud. I would not have expected a big difference but certainly see how that could help. The thing that bugs me the most is the big idle to load temp swing. I would think the lack of a shroud would more limit my total cooling ability but not impact the temp swing that much.
Well the boss is kicking me out now. Nice he let's me screw around at work so I must oblige him. I'll pick this up at home and try to get some pic's for you all. Thanks
 
Have the fans pulling air through the radiator instead of pushing. Pulling air through provides less resistance than pushing. You should see a temp difference from simply changing the direction of the fans. A shroud will also help your temps as previously stated. Not only will a shroud eliminate the deadspots at the hubs of the fans, but it will allow you to use the full surface area of the radiator that is going to waste around the fans.

Good luck, and I hope you conquer those temps. :p
 
LarryRE said:
Thanks Jas!
I guess I need to see about the shroud. I would not have expected a big difference but certainly see how that could help. The thing that bugs me the most is the big idle to load temp swing. I would think the lack of a shroud would more limit my total cooling ability but not impact the temp swing that much.
Well the boss is kicking me out now. Nice he let's me screw around at work so I must oblige him. I'll pick this up at home and try to get some pic's for you all. Thanks
Yeah, make yourself a shroud, it will help temps a good bit. I think what you said is right though, the overall load/idle temp delta isn't affected by the lack of a shroud much. I could be wrong though.
 
It's the MSI Neo2 board that's holding you back.

Very average power supply to the CPU. The board doesn't like you pushing the CPU too hard and having the CPU place a large drain on the board's ability to supply power.

In fact, for many overclocking scenarios, it is the motherboard's power supply circuitry that is at fault probably, oh, about 80% of the time, when things don't overclock much better on water than air.

Edit - having said that - there is still something definitely wrong if your temps are only improving by 2-3C between water/air.
 
Cathar said:
It's the MSI Neo2 board that's holding you back.

Very average power supply to the CPU. The board doesn't like you pushing the CPU too hard and having the CPU place a large drain on the board's ability to supply power.

In fact, for many overclocking scenarios, it is the motherboard's power supply circuitry that is at fault probably, oh, about 80% of the time, when things don't overclock much better on water than air.

Edit - having said that - there is still something definitely wrong if your temps are only improving by 2-3C between water/air.


Well I have to say I have gotten really great intelligent feed back on this. I actully can confirm Cathars' comment on motherboard power circuitry. The Epox 8RDA+ mobo I had got major improvements in overclocking from putting Bbig heatsinks and a fan on the mosfets. I do have heatsinks on the mosfets on this one but it hasn't done squat. So 2 probs...a mediocre overclock which likely won't improve much even with the temps solved and my temp problem. That makes a lot of sense. This temp problem really has me bugged!!! I was always reluctant to get involved with water so it's now bit my *** by not working well. :cry:
 
Back