MY WC loop is not performing as well as I would hope

[jph1589]

My bad

Your logic is partially flawed. Water loops should be designed so that it uses the least amount of tubing possible for maximum flowrate. The only time order matters is that the Res/T-line needs to be place before the intake of the pump so that it can be primed.

The difference between the water entering the rad and exiting the rad and entering is only about .1C max. Any more than than you've got flowrate issues. It's more important that the waterblock gets the maximum flowrate possible in order to induce turbulence inside the waterblock so that more water come in contact with the copper and carries away the heat. Better flowrate also has the same effect inside a radiator.

All that heat coming out of my radiator must be my imagination. But if it isn't, I sure don't want all that heat traveling into the cpu first.

 

[Neuromancer]

I figured that the because of the impingement jets in the storm block that straight out of the pump would be best, highest head pressure though same flow as the rest of the loop?

Anyway I can boot windows at 2.75 on stock vcore so I am doing a overnight nonerror checking burnin to see if I can get it stable without needing the high vcore to jump it to almost 50C load temps

Havingthe feeling though that this is going on ebay soon and I am going back to my cheap zalman air cooler. LOL

Tomorrow will play around with fan setups though, to see if a lower CFM but push pull config will help any... or something.

Thanks for all the input guys!

 

[Maviryk]

I figured that the because of the impingement jets in the storm block that straight out of the pump would be best, highest head pressure though same flow as the rest of the loop? ....

Yes that is correct, and it's how I have my loop.

All that heat coming out of my radiator must be my imagination. But if it isn't, I sure don't want all that heat traveling into the cpu first.

It's still disappating heat, hence the .1C difference between the water entering and leaving the radiator. What you're not understanding that even if the water is .1C higher because it's not right after the radiator, the increase in flow rate will make the loop more efficient.

 

[voigts]

There shouldn't be more than about a 10c difference between idle and load if your setup is working right. You also need to realize that your motherboard temps may or may not be accurate. Sometimes air cooling can actually read lower than watercooling depending on where the MB temp probe is located because of lack of airflow over the CPU area with watercooling. I also would check to make sure you don't have any jets clogged in your Storm. If you didn't soak and clean out your rad well when you installed it, you could easily have clogged jets causing the problem.

I wouldn't go back to the Zalman until you figure out what needs to be fixed.

Water within most loops has no more than about .5c difference throughout the loop. That is why most people say component placement doesn't matter. However, if you want to split hairs, then yes, the water is going to be cooler, albeit slightly, going out of the res than coming into the res. I have read of up to 1c difference between the two, the more common being about .3c.

As for the Storm, you probably are better off with it right after the pump.

 

[Rev]

I wouldn't trust motherboard sensors (for the reason that voigts gave), let the overclock be the judge. Are you able to push the system farther than you can with the zalman? Take whichever one gives the best clocks not whichever one the MB says gives better temps.

Since C2Ds have calibrated onchip temp sensors I can personally attest that motherboard readings are BS.

My friend has an e6600 @ stock (1.3Vcore 2.4Ghz) and the Asus probe says he is at 38C full load. TAT (which reads the core temperature) says that he is 52C full load.

My e6600 @ 3.6hz 1.5375Vcore, the Asus probe says I'm at 55C full load whereas TAT says that I'm at 50/48C.

Because he has a big fan blowing near the Asus probe his MB registers lower temps than I do, but in actuality my cpu is much cooler than his, even with my overclock.

 

[myndlessdayz]

How far are the springs supposed to be compressed on the block?

I always went to ABOUT 1/4 - 1/3 compression (a little less) and then a full turn off.

 

[MameXP]

All that heat coming out of my radiator must be my imagination. But if it isn't, I sure don't want all that heat traveling into the cpu first.

Dude, your logic is wrong to begin with.

We're talking about having the Rad b4 the block, not like NOT having the radiator at all. So if you have the radiator after the block, the water going into the block is ONLY slightly warmer due to heat dissipated by the pump.

In a closed up this will reach equilibrium and the temp different is so small to effect the performance.

Are you going to tell me the water coming out of the radiator is the same temp as the water going into the radiator?

NO!!!!! We never said that.

See what you're missing?

 

[Neuromancer]

Yah I was venting when I mentioned going bak to the Zalman. I am beginning to think that is is a borked temp sensor.

I have been bruning in for over 8 hours at stock vcore, and am down to 41C CPU load. Which is only 10C up from idle temp. But more importantly I did the diagnostic touch test last night, and nothing in the loop is really warm at all. Including the base of the storm.

Think I am going to stop the CPU burn and try some orthos now

EDIT: Just stopped the burn and in about 60 seconds temp dropped 11C so loop must be doing its job...

 

[QuietIce]

According to their site the AS5 takes several hot/cold cycles to set, and the cold cycle is to room temp. But even w/out that those temps are good ...

 

[Graystar]

(Takes off hat...prepares to toss...)

The order of the components won't make a measurable difference.

The amount of tubing, +/- 2 feet, won't make a measurable difference.

All that really matters is proper installation, fan performance, and temperature of the cooling air. If you have a good fan, and you're sure that ambient air is being used to cool the rad, then I would agree with others that say installation might be the key factor.

I also wonder about that electrical tape setup. I'd say get rid of it for now. When you're satisfied that any issues have been worked out, try it again.

 

[Neuromancer]

I was just doing another touch test, because I bumped up the vcore to burnin at 2.85 GHz, back up to 48C load already, and I noticed something. Actually I have the pump going intothe Rad and then into the block. Not sure How I switched that around without noticing.

I would drain and try another configuration out but I really dont feel like doing that again. No more water additives nor money for anything till the 1st

 

[Maviryk]

So... are you sure you're putting the water from into the center of the waterblock and not the side? Using the Apogee yes?

 

[Graystar]

So... are you sure you're putting the water from into the center of the waterblock and not the side? Using the Apogee yes?

I believe he is...

Routing is Pump out to center block barb.. offset barb to RAD.. rad to res, res to top of pump.

He's got an MCP350. By "top of pump" I presume he means the barb that appears to route to the center of the pump, as it is raised and nearer the "top" of the pump. This is the suction end.

Also, the output has an arrow showing the direction of flow. Kinda hard to mess up.

 

[Maviryk]

Yes... kind of hard to mess up...

But then again OP just said he has the pump running to the rad and THEN to the block... contrary to his previous post...

So... must double check... human error...

 

[[O-CuK]Marci]

I'd suggest getting a temp probe infront of the rad's air intake and measure the temperature of the air that is being used to cool the radiator. This will have a large impact on the final idle and load temps. If rad is rear mounted, with fans pulling thru, you could be using warm case-vented air or PSU-exhaust air to cool the rad... the main reason I ain't a fan of dangling rads off the backs of cases...

 

[Neuromancer]

So... are you sure you're putting the water from into the center of the waterblock and not the side? Using the Apogee yes?

*Ding* Give the man a prize!

And kick me for being a freaking moron.


I do not know how I did it, but that post made me go double check. The tubes are slightly entwined right now... (I pulled itout of the case as a whole loop and put it on top of the TV, so when I set it on the motherboard, had to position things so there was no pulling and that meant twisting a little to prevent kinkage) So how, some way, mounted the hoses on the block backwards Wasnt even any to blame

Well I suppose its for the best. This way I have an excuse to drain and take apart the loop before mounting it in the case again

Thanks again everybody for all your help. Hopefully I will post back later with improved results.

 

[Maviryk]

Someone owes me a beer.

That was the first question I asked too!

Anyhow, glad we could help.

 

[Graystar]

*Ding* Give the man a prize!

And kick me for being a freaking moron.

Oh man! That sucks. And to think...I backed you up! I feel sooo let down...

Here ya go Mav

Hey now we get to test a pet theory of mine (that's one that based on even fewer psuedo-facts than my other theories!) I bet performance will be the same!

 

[Neuromancer]

Holy Hannah!!!

Well l decided to do RE-Do of the entire case since I had it apart.

Added some UV effects to it, and gotta say, I never thought I was a UV fan before... but I am now

Replaced the Denki with the stock 120MM case fan, and load temps are hitting 42C but its completely silent Might have to wire up the Nidec BetaV again as a push fan temporarily just to see how cool this thing will run (wont do it permanent though dont want to burn out the CM fan.

 

[Neuromancer]

Okay.. different issue.

I tried replacing the coolermaster fan with a faster (and louder) Nidec Beta V and the temps did not change! I thought the BIX rad needed high powered fans to cool effectively?