What am I doing wrong here?????
 

Ok, here's my situation. I just upgraded from 3/8" Maze3 and DD gpu block to a 1/2" WW block, Maze4 gpu block ,and I converted the whole system over to 1/2" tubing and fittings. After all this, my temps actually INCREASED by about 2-3C. The pump is an eheim 1048, and I'm using a 10"x 5 1/2" heater core with 2 120mm Enermax fans. My idle temps before were ~40C and now they are ~42C, according to my temp probe and my motherboard monitor. This seems really high for a barton 2500+ at 1.7v, with ambient room temps at ~27C. I know the block is mounted correctly, I used PCM+ for the TIM and I've even lapped it a little bit (DTEK did a horrible job lapping the base on the WW).

So how could my temps go up, I just don't get it. Should I try a different pump? Something like a Mag3, that has a higher head and flow rating? I'm at my wits end with this setup I don't what else to try. Any suggestions would be GREATLY appreciated.

I allways find that temps vary at lot after a remouting, (something I've had to do a lot recently for various reasons), air in system/thermal paste settling whatever. I would check back in a couple of days of use and see what the temps are then.

~ Boli

BigAl what are load temps before and after? Idle temps are useless.

List everything that changed, including thermal paste, pumps, placement of fans/radiators, etc, etc.

There's just so many variable when you do a complete system change, and just describing the system isn't that helpful. Do you have any pictures of your new setup?

What were you using previously for TIM material? I've seen discussion on this board (sorry no thread pointers) that suggested that Phase Change TIM's were not as effective as ASIII, Ceramique, or other good paste TIMs, and that MIGHT be enough to account for the difference if you were using a paste before.

Gooserider

The only things that changed were the blocks, the tubing and the fittings. I kept the same rad and pump, I just changed the fittings on the rad. I also tried remounting the block several times and tried increasing the tension on the mounting screws. I too thought the PCM+ might have been a factor, so I swapped it out for AS3 and the temps went up ~1C so I switched back. The load temps before I changed the blocks and fittings were ~46C and now they are 47 - 48C.

I'm not even that concerned about the slight increase in temps from the changes I made. What bothers me is my overall temps seem way too high before and especially after the changes, given the setup I have. Heck, my buddies exos puts my stuff to shame, and that's just sad. I know the WW is the best block on the market (except the cascade, which I'm already on the waiting list for), and I should have plenty of cooling with a rad as big as the one I've got. Any suggestions guys? :shrug:

Isn't the WW a high flow block that requires a higher flow pump? The lack of head on the pump may explain some of that...

Oh, and the TIM shouldn't really find a 2-3C diff from one paste to the other should it? Are you mounting block with enough EVEN force to push it flat against the die?

One other thing you may have overlooked though.. (serious). When taking out a block, you may have loosened some crud. The crud floats around in a system. The WW is vulnerable to crud in the fins over the die. If you had some crud, (growth particles or loose teflon tape) it may have become lodged in the WW. Change your water and clean the system out. Check for goop in the fins.

I was under the impression that the WW didn't require a lot of flow, but a higher head due to the impingement. Either way I would've figured the eheim 1048 would've been sufficient. I did clean everything out just 2 days ago when I set this cooling loop up in a new case, but still no difference. The block should have even pressure, I turned each of the four mounting screws the same number of turns, in a criss cross pattern to ensure it was an even pressure. Also, The difference between the AS3 and the PCM+ was only ~1C, not 2-3C. Thanks for the suggestions though, I'm no sure what else to try.

silly question, but when you upgrade to 1/2, you are giving the 1048 a real run for its money. i'm not saying that it can't handle it, but do yourself a favour, and chk if you have some flow.

i once did some upgrading in tubing diameter and although the overal flow resistance dropped, the 1048 had real trouble priming the airbubbles out of the loop. look at it this way, the overal volume of displaced water might increase a little but the velocity of the water will drop after increasing tubing diameter, with the velocity dropping at +/- the same percentage of your tubing increase (+/- 1/3), it might be the proverbial drop that spills the bucket. for all you know you might have some airpockets in crucial component areas (rad, wb?).

That's along the lines of what I was thinking. It seemed like that pump was having a really hard time getting the air pockets out of the lines. I had to rock the case back and forth pretty rough to get the tubing to even completely fill with water. I know this isn't the best way to determine flow rate ( I don't have an inline flowmeter for my 1/2" tubing yet) but just looking in my res (tank-o-matic plug-on), there is so much suction from the pump it looks like a whirlpool in there. Whereas before moving to half inch, the suction from the pump barely caused a ripple in the water. So I figure there has to be more flow than before, or do you think that there is just less resistance and my velocity went to crap. I'm leaning towards getting a Danner mag3. Does this sound like a good idea?

I'm also having less than stellar (reported) temps on my WW/1048/amd rig. http://forums.procooling.com/vbb/sho...&threadid=7420

What are your water temps at load? What program are you using for load (i.e. prime, cpuburn, toast) ?

How many turns over recommended did you crank the thumbscrews?

Where is your temp probe?

I didn't bother to lap mine, but I may try that next.

well, in my case i had just switched 2 HC's from serial to parallel, the parallel setup decreased velocity and improved flow. temps however didn't budge. i solved the priming air problem by installing a 1250, but temps didn't get better. so, had i stayed serial my temps would have been the same as with a 2x strong pump and parrallel hc setup.

what this could mean for you is that if you would put the CPU block and GPU block in serial ... (dunno if they aren't already). don't take me as a guru on this, i have no proof whatsoever that airpockets could be the cause of the temp problems, just seems like the only explanation.

edit: i know it might seem harsh, but why not sell your gpu block? i'm ocing my 9500pro -> 9700pro to very high numbers with a p3 HS and 5V cpu fan. this would solve all your flow problems in 1 go.

Ruiner, I'm using prime to get my load temps. As for the thumbscrews, I've tried everything from the recommended number of turns to cranking down the thumb screws so hard the springs were fully compressed. When I got them really tightened down hard it made a few degrees difference, but I don't like putting that much pressure on the core. My motherboard is an NF7-S, and it's cpu temp monitoring is absolutely worthless. So I use a hardcano 9 , and I mounted the thermal probe on the bottom of the cpu. If I were you, I'd check that block to see if it's flat and has been lapped properly. The baseplate on the one I got was way too rough, not to mention it was pretty convex. I've read where several other people got bad WW's from dtek too.

:(

Have the bad WW's been all polytops? How many others are we talking about?

Well mine's a polytop, but the others I don't know about. I don't know how many really, but I've read on a few other forums where people were mainly complaining about thier baseplates having really rough grooves in them.

If the base-plate is "grooved", even in a small way, then that is what will be causing the issues. It will need to be lapped smooth.

I wonder if DTEK is having some quality control issues?

Maybe I missed it but what are the water temp & air temp