Discussion in 'techPowerUp! Club Forum' started by EastCoasthandle, Feb 8, 2008.
Holy hurricanes, cdawall!!
couple of front views
here are some front views
nidec on the left delta on the right
all of the fans are 4 pin PWM don't know what the hell can control them but still cool
Well, I'm pretty damn sure my tripple rad has a slow leak. I dropped the hell out of it while I was flushing my system, right on its top corner. I've filled my res almost 3 times since I posted my last pic on Saturday. I don't think there was that much air to bleed out. No leaks anywhere. Nothing but undistubed dust in the case and around the fittings. I took the fans off last night, nothing I could notice, except the faintest pinhole of shinier metal that may have been wet. Put it all back together, and have had to fill it a bit today.
I'm guessing it is a leak that is slow enough that the fans evaporate the water, so there is no evidence. Sound reasonable? Anyone ever seen this before?
put some JB weld on it
if the leak is that slow, it will take a few weeks for the water to evaporate in that big amounts... youre sure you just dont missed a lot of air? your setup looks like its not the easiest to bleed!
I hope you're right. It's never been so persnickity, and it's going through quite a bit of water. I just went over it again with a flashlight, nothing visible. I'm wondering if the flushing didn't loosen up some junk that plugged a part of one of my rads, so it's taking longer? Anyhow, I'll keep an eye on it and let you know.
the way your rads are mounted, the backside rad probably traps air bubbles in its top, and needs longer to bleed therefore... the top rad is fine, and shouldnt pose a problem.
Also, if you want to be sure everything is leak free, make the finger test... touch each fitting with index finger and thumb, and feel if everything is dry.... if there is wetness, you probably found your leak
Brillant thanks i've always wanted to watercool now i understand it a lot more
I will go for it
Problem solved!! :shadedshu
Like I thought, the fans were mitigating the situation, so I got out an extra PSU and just powered the pump. Took only a few minutes to find the leak. Lucky that it wasn't fast enough to drip on my video card. The vinigar flush must have been a little to harsh for the old timer.
Got it spliced together, at the moment. Looks like I now have an excuse to uprade from the Swifty wafer thin to an XSPC.
i like my MCR's i will take your old rad if you dont want it tho
Ha! I actually thought you might while I was pulling it. I'll let you know
My idle temps aren't that much different. I'll have to check after some BC2. I don't think it was doing me much good. Looks like some nasty stuff coming out, as well. The guy I bought it from used coolant.
i use dexcool but thats car coolant. considering how many watts your loop puts out i am surprised the single 320 holds the temps...
look into this one as well or a coolgate
that being said i do have a MCR220 (but its the stackable 4 barb model)
Hefty Stuff there MT!
i guess i was terribly wrong
I would have never thought that the performance of the GTs would be big enough to continously dry the leak, and i wouldnt have thought that vinegar could damage a rad that bad... it must have been the vinegar, as there is nothing that would damage the fins otherwise, except maybe physical damage during the cleanup.
How long would you guess did it took, till the first drop formed, without the fans?
Did he take you up on that offer?
Thanks for the leg work. I went with the XSPC from Sidewinder. The EK looked tempting, especially the price, but was 10mm narrower. For some reason, I have always gotten my jazz from PPC, mostly for the selection, but I have heard good things about the owner, Gary, around these parts.
Just long enough to take the four screws out of my rad box, tip my back tripple rad on its back and shake it around to see if there was anymore air, and put it back on. five minutes, tops. Then I grabbed my flashlight and noticed a drop hanging from a blade of the GT, and a drip on the back of my 5870. I was damn surprised. No screw punctures from being too long mounting the fans, just a failure right in the middle.
I like the black and white. Nice looking rig.
It's hard to say for sure, since I had a slow leak, but I think my old Panflos at around 1200rpm performed better than my new GTs at 1850rpm, and were pretty damn quiet.
I bought some Koolance quick disconnects today.. Man they're expensive. 60 bucks for male and female for two hoses. It's going to make taking things apart sooo much easier though.
It will probably make the tacklebox rad even more handy. I've thought about getting some and stashing my rads in my crawl space. Maybe someday.
sounds like its a miniature leak, more like a crack in the copper, than a fullgrown hole.The vinegar probably gave it the rest. Im glad tho, you found your problem, would have sucked to loose hundred of $ for nothing
new water cooler here and i got a question. I got into it by buying a XSPC RASA X20 750 water cooling kit from frozencpu.com. A nice little kit for the price! But now that I have had it for a little while.... I am looking to get even better temps. I have already upgraded the radiator.....To a Black Ice II ( or something like that too lazy to look it up right now )
But my question is.... Which way would I get the most temp drop next? Buy changing out the CPU Block or Changing out the Res and Pump for a higher power pump???? If someone could give some suggestions and even give some examples of possible temp drops with your suggestions.
BTW Will post back tomorrow with pics of my set up. Just cause everyone else is doing it
I don't think much is going to help decrease temps any more. It depends on the gph of the pump I guess. First thing that comes to mind, I suppose would be better fans.
Thanx for the quick response Erocker. Yeah I did that....with the fan upgrade already. Well let me bounce a couple of ideas off you remember still new to water cooling more or less.
1. The other radiator I have sitting around now because I upgraded it......If I added it to the current loop would that lower temps even more? Or would that be over kill or would it cause even more problems. Like worse temps. For example my setup goes.... Pump>CPU Block>Rad>Res ... If I did something like this would it help??? Pump>Rad>Block>Rad>Res
2. So you think a higher flow might help? For example right now I am pushing 750ml if I got like a 900Ml pump would it have a higher flow of water meaning moving the heated liquid away quicker?
3. Right now I am at a low middle end block. Do the higher end blocks produce that much better results that they are priced that high? Or are they just over priced.
4. Last question.... would there be any difference switching tube from 7/16 to 1/2. Or would there be very little like I suspect?
Hell Yeah Go Green Bay Packers!!!
1. Loop order really doesn't matter, because after awhile your water will reach and equilibrium temperature. More radiator surface area is always good, although it may mean you just keep similar temperatures for even longer before the water is saturated with heat. You are always constricted by ambient temperatures. It would, however, allow you to add your gpu to the loop, which usually is more drastic of a reduction in temps than the cpu.
2. There are two schools of thought. One says squirt the water as fast as you can, the other says moderate speed lets the water stay in the rads for longer to leach out more heat. Not sure which, I always adjust my pump to the quietest setting.
3. Not much difference between ultra fancy and mid range blocks, within a handfull of degrees.
4. As far as I have heard, ID has little effect on overall performance. A lot of people think the smaller tubing looks nicer.
EDIT: My temps haven't altered all that much since losing my top rad, even though I just shelled out $70 for a new one. Makes me wonder how much good the old one was doing, I'm hoping for marked improvment from the XSPC. These are my temps after about 35 minutes of BC2. Used to max around 44, although it's pretty cool in the room now at 75 degrees. Usually around 80ish in the winter, due to the wood stove. I'm sure they would get higher with longer play time.
I guess if you want lower temps, there are things that can be done... question is, what a few C are worth to you. for me, they are very important, for example... so i guess i make 2 options you now have, in my opinion:
1. The efficient, cheaper way, probably the most realistic one for you too:
Change the Pump, add both Rads you have to the loop
(more surface will never hurt, i guess my huge ass rad proves that.)
and change the fans to something beefier, 1700-2000rpm+ push fans
(Magmas, Silverstone 1202, Scythe GT and anything that is similar to a San Ace/Delta fan, on a controller or undervolted)
and 1200-1500rpm pull fans should produce a nice result, tempwise, and also shouldnt be too loud. The Temp drop should be also relatively significant, even if not yaw dropping, in comparance from the jump from mediocre air to good water.
2. Go full bore,eventually WC the board and the GPU, get yourself a HK 3.0 CU or LT,
1 or 2 Pumps with about Laing strenght, a MORA-3 9x120, and full Delta/San Ace/EBM-Papst Push/Pull... should give minimum 10c better load temps on an averagely hot Phenom 2,maybe even more like 15,dependant on the proc, and probably much, much more on an i7.
for me, there was also a third option, i got an old, used, but besides small damages relatively good shaped heatercore, from a 25 year old Renault Espace (Still 100% copper, besides the Steel framing).....4cm thick, 6x3 120mm Fans tall and wide, and about 9-10kilos of copper and about 1 kilo of steel frame. i just had to pay a few € to buy some glasfiber-resin, to fix the crack in the inlet.
I also got me a Laing and a EK Res used in the forums from a very nice UK lad, for a deadly good price...sometimes it happens Because it was a modern, blue impellered Laing DDC
(with a clear square OClabs top tho)
I popped it open, bridged the solder pad, so it became a Laing DDC+
(same electronic and parts, blue Laing DDC are in fact just artificially crippled, by not making the last solder step in the production)
the Heatkiller 4870x2 LT, i got as open box from Aquatuning.com, for 50€, it had fingermarks and discolorations all over it,and a few nuts were missing but after a good scrub with Ketchup and Vinegar-essence, and a few minutes of search in my nut and screwbox, the block was as good as new
The 18 EBM Papst 24v industrial grade Fans, a Friend gifted me, even tho i admit, that was just very big luck and never intended.One Day i woke up, due to the weight of 12 Fans my Buddy threw at me,in a sack, smiling and laughing.
i agree with 2 of your 3 statements:
that better fans and a better pump would help.
but in my opinion, dependant on the budget and enthusiasm of 20MM rain, there are still a bunch of degress left, till (nearly) room temperature is reached.
Also,Ventilation, in my opinion, is 50% of your cooling, from the Experience with my big rad in my badly ventilated room.
If your room is ventilated well, your rigs will be much cooler, logically
You guys both have given me some really great information and I really appreciate it. I guess I have just a couple more questions.
Velvet Wafer 1. You mentioned 2000+ RPM Push fans.... which I have. But my question lies in the way you said it. "Push fans" have you found it better to just push the air through the Rad apposed to a Push/Pull config?
Next question for anyone 2.... Static pressure.... is there a minimum I should be looking for in a fan?
Last 3. With my Rad on the back I have the air flow as follows.....
The reason I did the rad this way was to deflect the hot air coming from the video cards. The rad does not sit exactly above the video card exhausts but close enough that i don't know how it would effect the temps.
So my question is.... which way for airflow.... the way I have it or do you think I might benefit from switching it around???
No, Push/Pull is definetly the preferred option, but not certainly necessary, if your push fans are very strong...would be for price efficieny, not for maximum performance in that case
I wouldnt trust the most ratings for static pressure on fans, often they are way off, very overestimated. I have a good bunch of Fans here, and those that have high pressure at low RPM nearly always have exotic, often upwards tilted blade designs.
Enermax Magma,white Silverstone 1202, EBM Papst... probably also Noctua, San Ace, Delta and the like, even tho i didnt owned one of them till now (Designs are often similar, so is the Performance)
Generally you could say, the fewer the blades,the more RPM and the more agressive the upwards angle of them, the higher the pressure the Fan should produce
In my opinion, good push fans are the Enermax Magmas for sub-2000 RPM, any 2000 RPM+ fan that fits your purse, and if you want to invest in something durable and variable, a few Delta-like fans, preferably under 4000 RPM, 3500 would be fine in my opinion.
I guess you could turn the Fans around to pull air from the direction of the case, if you would install a small selfmade duct, that would ensure, that all the hot air from the VGAs gets directed to the ground, and not upwards
Separate names with a comma.