Your thoughts on this CPU water loop ?

[Grooveriding]

I've fixed the turbulence issue, but I'm still having temperature issues imo. I brought them down some though. I'm seeing about 59 to 65 at load depending on the core with 1.39V. Still not where it should be, imo.

Going to have to play with it some more when I have more time. No idea what the issue is of course I redid the loop and changed it around some, will post a picture sometime soon. Burnt out on this whole thing for now. :awe:

 

[OVerLoRDI]

Hmm.. Those temps seem a bit high for a 3x120mm rad. Post pics when you can and ill see if I can notice anything.

In the meantime you might have a big bubble in the radiator. If you haven't already try rotating the case.

 

[Kenmitch]

I've fixed the turbulence issue, but I'm still having temperature issues imo. I brought them down some though. I'm seeing about 59 to 65 at load depending on the core with 1.39V. Still not where it should be, imo.

Going to have to play with it some more when I have more time. No idea what the issue is of course I redid the loop and changed it around some, will post a picture sometime soon. Burnt out on this whole thing for now. :awe:

Seems to me that 6cores 12threads @5ghz(1.4v) = some monster heat!

Your temps look perfectly fine if you take into account the massive amount of heat your chip is pumping out at those clocks!...Unless your stated temps are from something whimpy that doesn't stress the cores HARD!

What are you using to load your cores to check your temps?

 

[Grooveriding]

Seems to me that 6cores 12threads @5ghz(1.4v) = some monster heat!

Your temps look perfectly fine if you take into account the massive amount of heat your chip is pumping out at those clocks!...Unless your stated temps are from something whimpy that doesn't stress the cores HARD!

What are you using to load your cores to check your temps?

I'm starting to think that may be the right of it after seeing some other WC results on highly overclocked SB-Es. I've seen other reports of people seeing high 60s and even 70s for SB-Es on 1.4V or so. Maybe they are just that hot

I stress for core heat with small FFTs in prime and/or LinX.

Going to put the CPU to bone stock and see what my temps are like. Thanks for the input everyone, will toss up a pic of how I have it setup now this evening.

 

[Kenmitch]

Maybe they are just that hot

They do run very hot just like the other SB's once you get the clock up there!

Your current clock + air cooler + Intel Burn Test = 3,2,1 your chips would throttle down!

I think it's one of the resons you see alot of 4.4-4.5ghz overclocks as it seems to be the point temps start jumping alot. But on water it's like a couple of degrees or so.

The amazing results from water come in the GPU department which I'm pretty sure you found out already! That and the turn on your rig and what you hear is what your gonna hear during idle, stress testing, gaming, etc.

Playing around at 4.8ghz with HT on during an overnight run of the beta prime w/AVX support my hottest core did peak at 67*C. I'd tend to believe that with my chip and the stock cooler at stock speeds my temps would have been alot higher!

I don't game anymore so my video card just idle's while gurgling under water. I did download the AVP and Stalker CS benchmark demos last night just to see what temps are under that kind of senario and I guess to see somewhat the stability of the rig with loads other than the normal stress testing.

These were at 4.6ghz on my 2700k. The results were pretty much the same when ran at 4.8ghz today.

AVP hottest core was 41*C with GPU of 36.5*C
Stalker CS hottest core 43*C with GPU of 37*C

My GPU in sig is clocked to 1000/1300 (1.163v)

 

[rgallant]

nice set up ,
-on your next order you could get one of these.
http://dazmode.com/store/index.php?main_page=product_info&cPath=158_159&products_id=611
-it's hard trouble shooting water cooling without knowing the water temps.
-high water temps could mean ,you need more fan , rad or air temps into the rad is high .
-lowish water temps and a hot cpu [ in your case] could be a bad mount.
-having a fan controller with temp. monitoring comes in handy.

btw daz is a good guy to order from ,I asked about spare o rings for QDC's last year and he did not carry them , but they showed up on his site not too long after that.

 

[nyker96]

oooo those white tubes, theyu are soooo beautiful, they are just drawing me in ... must resist

 

[Grooveriding]

Here is a crooked picture of the new setup, almost the same, but a little different.

So basically a few changes are that the block was rotated 45 degrees to put the inlet on the bottom and outlet on top, the outlet from the block now comes into the top of the reservoir, rather than the bottom and I am now using the outlet on the top of my pump - rather than the side outlet.

Also I have mounted the rad on top with three fans on top of the case, they're actually on the outside, blowing into the rad and those two fans below. There was no room for a third due to the reservoir. I'm going to get a custom shroud to put on the top of the case so the fans are not exposed, probably this one : http://www.koolance.com/water-cooling/product_info.php?product_id=691

I should of gotten a smaller reservoir I am learning the more I've had to work around this big one I actually should of gone with a bay reservoir for aesthetics and ease of use. I avoided that because I have a front fan and my video cards need air or they'll get too hot. Once I get some new video cards and put them under water I am going to get a bay res that also holds my D5 pump and sell this one.

So with these adjustments and what was probably a poor mount on my block I am getting better temps. Idles are now 28-32 and prime load is 65-72. I've read up on a lot of forums and it does seem that a SB-E with the voltage I am giving mine (1.48) gets hot even under water. For context using this chip with 1.38V under a NH-D14 got me to the high 70s under load. Running the chip at stock I get 46-52 depending on the core under load.

WC is definitely something I think you get better at with practice. For my first try I am pretty happy with it. It is really a pain draining and re-assembling though. Not interested in doing too much of that Looking forward to getting some video cards under water as well, the system should be quite quiet.

I'm also planning to sleeve all my power cords with white sleeving and black heat shrink. I was considering getting some MDPC sleeving as I've heard it's good quality and has the look I want, but it's just too expensive imo for what it does along with shipping. Going to get a spool of this and some heatshrink and do it for under $100.

 

[Kenmitch]

Here is a crooked picture of the new setup, almost the same, but a little different.

Looks much cleaner that way. :thumbsup:

Did you try the cpu block the original way you had it and rotated when you redid the loop? Looks more natural to me with the outlet up and seems like it would work best anyways.

White hoses for the win! I like white the best for some reason.

What cured your turbulance issue?

Did you try the upper rad fans as all exhaust? Seems to me with good airflow it wouldn't make a diff at all and would keep the MB cooler at the same time.

 

[OVerLoRDI]

Looks good.

I'm still surprised at how much heat that cpu makes. For comparison 2x 6970s at 1.3v barely break 40c on my 3x120mm rad. So the heat production of that cpu at those clocks is greater than two overclocked 6970s by a significant margin. Wow...... The waves of heat coming off that rad must be incredible. What is the power draw like?

 

[Grooveriding]

Looks much cleaner that way. :thumbsup:

Did you try the cpu block the original way you had it and rotated when you redid the loop? Looks more natural to me with the outlet up and seems like it would work best anyways.

White hoses for the win! I like white the best for some reason.

What cured your turbulance issue?

Did you try the upper rad fans as all exhaust? Seems to me with good airflow it wouldn't make a diff at all and would keep the MB cooler at the same time.

I think it was putting the res inlet up top rather than down below next to the outlet, but on the other hand it could of been air caught in the system that was cleared from draining/refilling.

Once I flipped the block I never went back I would of liked to try each individual change on its own, but the work involved with that was just too much imo.

I was going for ideal temps, so I assumed pulling in fresh air would be better than pushing air from inside the case through the rad.

Looks good.

I'm still surprised at how much heat that cpu makes. For comparison 2x 6970s at 1.3v barely break 40c on my 3x120mm rad. So the heat production of that cpu at those clocks is greater than two overclocked 6970s by a significant margin. Wow...... The waves of heat coming off that rad must be incredible. What is the power draw like?

I have no idea on draw, I have no means of measuring it. I can't find numbers for power consumption at the voltage I am running. For 1.37V here they found it to be 183W I'm using .1 more than they are.

 

[Kenmitch]

I think it was putting the res inlet up top rather than down below next to the outlet, but on the other hand it could of been air caught in the system that was cleared from draining/refilling. Once I flipped the block I never went back I would of liked to try each individual change on its own, but the work involved with that was just too much imo. I was going for ideal temps, so I assumed pulling in fresh air would be better than pushing air from inside the case through the rad.

That does make sense on the intlet positioning as it most likely got rid of the whirlpool effect.

I don't blame you for not wanting to mess around with it once your back up and running. I'd think that outlet up would be best as it looks like it's the only way to get air out. Tilting works also but as time goes on a little bit of air can still accumulate in pockets due to those tiny little air bubbles that like to cling on.

The intake vs exhaust thing plays more of a part with a case that has poorer airflow. I just go the exhaust way on my rads as my case has plenty of airflow. No change without side panel on so figure it doesn't hurt me and will keep my MB cooler.

I'm currently debating on picking up a Corsair 500r and swapping my rig to it. I like the basic look and the black interior.

Now ya just gotta get some gpus under water and you'll be set

 

[Grooveriding]

Hey thought I would update and throw out a caution to people considering an EK Nickel plated HF block. The particular model I have is the EK HF Acetal-Nickel Plated Rev4 Block.

So I was tearing my whole loop and system out of my 800D tonight because I have some time tomorrow free to cut it up and mod the case to accomodate a rad down below and open up the grills up top for more airflow for the top rad.

Anyways when I pulled off the CPU block I got a glimpse in one of the ports and saw corrosion on the block! Only 3 weeks old! I am not mixing any metals either and was painstakingly careful about cleaning out my new rad, fittings, block, pump - heck even the tubing got clean water ran through it.

It is covered with brown spots and a little green discoloration as well. There are also these very small white spots developing on the block. Here are a few pics, forgive the trash quality, I am terrible at taking good photos.

Sorry if the pics are massive, not sure on the res. I don't have photo editing software on the PC I am using while my main PC is in pieces.

I contacted Dazmode where I bought the block and asked for a refund to put towards the purchase of an XSPC Ray Storm special edition copper block. I think this qualifies as poor manufacturing ? I can't see how I could of caused this.

I tried to clean it off with a soft toothbrush and it doesn't come off. Used a little 99% pure alcohol as well and nothing.

Crappy manufacturing ? I'm no metallurgist, but know my chemistry quite well, and this looks like a corrosive reaction to my eyes from what I am assuming is either a use of impure metals or very poorly applied nickel plating that flaked off from the flow.

 

[OVerLoRDI]

Ek had issues with their nickel plating last year and are now using a different process. It's possible that the block you had was of the older manufacturing technique.

I'd be surprised if it was galvanic corrosion, your loop doesn't look like it has mixed metals. The rest of your loop looks good?

 

[chimaxi83]

Yea their plating wasn't high quality, leaving a lot of exposed copper (microscopically speaking) and the exposed copper/nickel was causing some corrosion. That might be it.

Try using some ketchup and that soft toothbrush to clean that off. The little bit of acid in ketchup is pretty great at cleaning blocks, may give you a closer look at the surface.

 

[Minerva]

You might want to check out this regarding EK's junk plating!

http://www.realredraider.com/vbulletin/showthread.php?t=47917

You would be fine served with their copper blocks though.

I would not trust their so called EN process either. There's a lot of old stock out there that could get a sticker slapped on it and you know the rest.

 

[Grooveriding]

You might want to check out this regarding EK's junk plating!

http://www.realredraider.com/vbulletin/showthread.php?t=47917

You would be fine served with their copper blocks though.

I would not trust their so called EN process either. There's a lot of old stock out there that could get a sticker slapped on it and you know the rest.

Wish I had read that before I bought the block. EK gave me an RMA after a bit of persistance. Their first suggestion was to send the block, at my expense, to them in Slovenia for 'assessment and polishing' I had to shoot that down as ridiculous considering the cost of shipping there with an insured and trackable method.

The store I buy from didn't have any straight copper EK blocks in stock so I went with the XSPC Raystorm. Temps are a few degrees lower with it, so I'm happy

 

[OVerLoRDI]

Wish I had read that before I bought the block. EK gave me an RMA after a bit of persistance. Their first suggestion was to send the block, at my expense, to them in Slovenia for 'assessment and polishing' I had to shoot that down as ridiculous considering the cost of shipping there with an insured and trackable method.

The store I buy from didn't have any straight copper EK blocks in stock so I went with the XSPC Raystorm. Temps are a few degrees lower with it, so I'm happy

Wow Ek, wow. Yeah shipping to solvenia makes sense....

Already was planning on avoiding their nickel plated stuff, might just avoid all their stuff from now on.