New setup MASSIVEly High Temps!!!
1 Attachment(s)
Well I'm sitting here with my redone flow setup and newly made PCV Res...WTF! My idle temps are even HIGHER than with the old setup with the 2 90¡Æ bends! Being that my mobo's temp sensor is ****ed up and doesn't read below 40C...
XP1600+@1.764(10.5x168)/1.84Vcore = ~87.4watts It's about 27.2C Ambient Idle: 44.0C Water in res: 36.0C N-Bridge: 39.0C Those are the IDLE temps!... Before, with the old "T-Res" setup I would stay around 40-42C@~30-32C Ambient! And max out around 44-45C avg...I'll do a 100% load test in the morning and hope that the sensor is ****ed or something... These temps are ridiculous! 11:38am--- What a NIGHTMARE!!!!!!! It's a blazing 30.8C Ambient in here and I'm IDLING at 48C according to this sensor! My water temps are ~ 39.6C I just don't know HTF my old system where it was: Old-- Pump---(90¡Æbend)-->CPU--->(90¡Æbend)--Rad (Fan blowing in air)---"T-Res"---Pump is killing this system: New-- Pump--->Res(on top of case)--->Rad (Fan Pulling in air)--->CPU--->Pump There's NO 90¡ÆBends in this system at all! Only 2 ~ 60¡Ætube bends from CPU---Pump and Rad---CPU... I'm getting tired of modding and modding just to end up with WORSE results from a seemlying (On paper and in theory) better setup! I just might have to throttle the jets back or go back to the old POS "T-Res" (NOOOooooooo!) Somebody help this fool! Okay thinking rationally now... Here's what I think is the problem 1) With the old setup; I had a 92mm blow hole that didn't have a fan in it, which was a convienent, passive hot air outlet from my 350W PSU. The air coming out of this mofo was pretty darn hot! There are 2 92mm outlets in the rear, just below the PSU's 92mm intake fan(right above the CPU), Of which I had SEALED off with some cardboard and tape. So, with a 120mm--100+ CFM intake blowing in, and a 80mm PSU exhaust and 92mm exhausting the rest of the PSU's heat out the top of the case, there wasn't much air flow. 2) The new setup; I used the 92mm blow hole as a holder for my PVC-Res. The Via 1300 (DIE!) pumps directly up into it, then down (sloping) to the Rad, which now has the 120mm pulling in air, which directly cools the pump in front of it (where before the pump was at the back of the case with no real exposure to incoming air). I took off the tape for the 2-92mm holes at the back. I'm thinking that the hot air from the PSU's internal facing vents is heating up the already hot water from the CPU--->Pump--->Res, BUT the heat core should easily compensate for it, but I gues it's not?! Oh yeah, I forgot to mention that there's another 92mm hole on the side of the case, parallel to the CPU but it has no fan either (too much noise from the fans!) Could it be that the PSU is heating up my res water, which is too hot for the Rad to handle? I wouldn't assume so, but that's NOT what the numbers are agreeing with (DAMN IT!) I'm thinking that I could cover up the vents and let the PSU's 80mm exhaust deal with the pressure build up of air that the interal 92mm it has, but I don't think it would last too long w/ just the 80mm? HELP! |
1 Attachment(s)
Well, I should slap myself! I've found part of the problem: I took off the case sides and low and behold, the temps dropped rapidly!
Sides on: Ambient: 29.6C Water: 39.0C Idle: 48.0C Sides off: Ambient 29.0C Water: 32.7C Idle: 42.0C This still doesn't explain where before, the old system with panels on, Ambient hovering around 30-31C, idle ~40-42C Is the PSU heat causing the rise in temps???! I've got my hand right in front of the vents and the air is warm but not blazing hot... There has to be a way to solve this without adding more fans... since the older system, should (theoretically) have been less efficeint in terms of flow (2-90¡Æbends) and flow direction(Pump--->CPU--->Rad--->T-Res-->Pump) |
1 Attachment(s)
This is what I think might help?
|
errr,
might be pump outlet pumping water into the res... in this case I see pump sucking water through block and rad from the res, despite eventual pressure made by the pump in the res... try to pump the water directly into the block or rad... maybe I'm wrong, but this pump-to-res situation don't look good to me... just a thought... that shroud looks as a good idea. |
Hey I see what you're saying, I was thinking that maybe the pump--->Res is causing the pump to work more, but the pressure is probably LESS than the old set up where the pump had to push water up to a 90¡Æbend THEN through the CPU... where as now it pumps up about 3 more Inches into a 1/2" nip and 5/8" copper tube... I wanted all the air-swirled water coming from the pump to hit the res first in order to lessen the amount of trapped air that would end up in the Rad...
BIGBEN HELP ME! I know you're around! :mad: :cry: |
I think you have all the info in front of you, you're just not seeing it.
The only way that your water would drop 7 degrees, is if your rad suddenly started running better, and that means that the fan is pushing more air through it, since you took the cover off. Sounds to me like you have an airflow problem. If you have 3 fans in your PSU (which one is it, btw), where one fan blows in (from the bottom) and the others blow out to the front and back, then I'd give some thought into reversing the front fan. For the lovers of the negative case pressure, you have negative air pressure in your PSU, but you have positive pressure in your case, and it's preventing your rad fan from being efficient. Reverse than fan and you have the exact reverse situation (may not be good for the PSU though). Alternatively, you could remove (or disable) the front PSU fan. |
I think that you are loosing a lot of power (pressure) in that res in setup like this... Air can be compressed and water cannot, so your res acts like an expansion chamber (something used in central heating for the houses). I think that youre loosing lots of pressure this way, so the flow is impacted...
|
Look V12|V12, whatever reasons you had for the current direction of water flow in your system, it DOESN'T WORK!
Just reverse the flow: pump --> block (or rad) -->rad (or block) --> res --> pump. The res should ALWAYS be the last thing in the circuit before return to the pump. The way you've got it, all the pressure is lost in the res, water trickles down through the rad (which has high flow resistance) due to gravity, has to go upstream again to the block (only manages that because it is coming from a higher point (res) to start with), before being sucked down by the pump. So like Puzzdre said, first of all JUST REVERSE THE FLOW, ok? |
1 Attachment(s)
How am I loosing that much pressure when the pump ALSO acts as a vacuum and sucks water through the system... That's how pumps get there power, they PULL in a lot and it gets ejected as fast as the outlet will allow until it gets bogged down. While some mixed air-bubbles are getting sucked down the outlet of the Res I can SEE the velocity of them and they are flying down that mofo! Okay I'm not going to argue the physics of pump blah, I appreciate ALL the constructive criticism! ;)
Okay I hate when this happens, the "why didn't I think of this before" deal: Here's a quick solution that might work? |
Quote:
About the Air flow prob... With the OLD-SETUP I had the case panels on and the fan Blowing in air through a 120mm hole in the front of the case into the rad and case... then I switched it so that the Fan was pulling air through the Rad which help alot, but the pump was still pumping preheated water to the CPU CPU-->Rad--->Res--->Pump---> = Lower temps than now: Pump--->Res--->Rad--->CPU---> How? Even with the case panels off it still idles higher than the old system with the sides on! :( |
Quote:
2 thumbs up !!!!!! good luck! |
Why don't we start by separating the problems on their own:
Problem #1- airflow problem through the rad Problem #2- order of cooling components #1: It's not going to be easy to fix, because what you want to do is make sure that that 120mm fan can run unrestricted. Your first hint here will be with that side hole: is it sucking air in, or blowing it out? (It's blowing out, I know). Option 1: add exhaust fan (above the PCI slots). Now you probably don't want to do this, because you don't want more fans/noise. Option 2: duct the rad fan exhaust to the bottom/side. It'll be a bit of work, and it'll involve cutting up the case some more, but it will take care of that problem. Option 3: add an opening to the back of the case. Just cut a hole. Option 4: stack another 80mm fan, outside the PSU. Not the best idea, but it might help some. It probably won't resolve the rad fan problem though. #2: just try it, and stick to what works best for you. Everybody's got a different opinion, and although the rad-before-CPU should work best, I've seen it disproved. |
Slap me if I'm wrong, but I seem to remember that running a centrifugal pump, it is a lot more efficient to push than to pull. With the res before the pump there is no resistance to water being sucked into the pump ... that is the nature of the res. With the res after the pump, turbulence is created in the res causing a dramatic loss of pressure. If you thought your 2x90deg bends were bad, that's nothing compared to what you introduced by putting your res where it is. You need to turn your pump around and have it pump the other way. You can blow all the air you want over that radiator, but it won't make up for the dramatic drop in flow rate through the entire system. Swap flow direction, and be amazed when your temps drop by 6-9C (estimated). That isn't a theory, that is based in solid mechanics. If you flip it, good temps will come.
|
Okay I see what you mean about the pressure lossage: The flow rate (though I don't have a meter) is still at least above 200GPH, I ASSume!lol, The pic that I posted about the changed flow route is what I"ll try, seems to be the best solution, though the Radiator will gobble up some flow rate too, probably not as much as my Res...?
I'll tell ya what though, having the res on top and external with a clear top makes the tranquil, water running down a steam sound, it's kinda neat... Now that I think about it, I'm going to add a little Candy to it and drill in a BLue or Green LED (though I'm not for candy, it would look nice at night). I might also try the old flow route of having the Pump-->CPU---Rad-->Res...since that would yield the most flow-rate through the block... but It just takes SO much time to drain, take apart all the tubes, re-grease and seal stuff etc... If anyone's handy with MS-Paint and could throw out a couple flow designs AND I'm going to need to find a spot to put a drain valve or something, cause right now it's going to be a BEAR to drain! I didn't think about that when I started (DOH!) THanks all for the helpful replies! :D |
it could be just the hieght the pump has to push against. before everything was almost level, now the res is way up at the top. has to pump all that up there, and then push through everything else.
If you dont want to go back to an everything level set up, then do what others have told you and just reverse the pump (or rather put the outlet in the in, and vise versa) |
Need more passive blowholes :) ...
|
Quote:
I was about to flame you for writing that, then I read it again, and I agree. The centrifugal pump is pretty lousy at pulling (sucking) water, that's why it needs to be primed. All they really do, is throw water around. I think that keeping a pump at the bottom of the loop is best, and that's what most of us do, but on the other hand, in a closed loop (which this is not), I have a hard time believing that it matters at all. I agree that this splashing around in the res does not really serve any purpose, except for aesthetic (it looks cool). It is however restrictive, to a certain extent. Reversing the flow should fix that. Good call. Alternatively, you could take that copper tube, and cut it in half along the length (leave a small ring to attach it), and leave the opening pointed towards the res outlet. |
V12|V12: the new drawing you posted with the new circuit is perfect: all in the right place and as simple as possible. Arrange it like that and your rig should be a lot better.
Edit: by the way, you can't have the two PSU fans blowing in opposite directions (= struggling, noisy fans = no airflow = baking PSU). The best setup is the original one: air enters the case at front bottom, through the rad, warms up & rises, leaves the case through the PSU at the top back. Don't obsess about cooling the res. |
im spurised that no one even opted for a "T" with a bleeder/filler tube going to the res. That way your system is still a small level loop, water can be intorduced as needed, air bubbles will still get out, and all is happy.
Although you would have no splashing in the res. |
Quote:
As for the fans, theres one pulling from the bottom and one pulling that air out on the left side of the pic. The other portion of hot air, is pressurized air that the 80mm can't handle from the 92mm on the bottom, which leads to the extra air flowing out of the PSU inner vents and up against the plexi-bottom of the Res.... I'm going to start on the rerouting tonight... can someone PLZ get me some MS-Paint pics of a potential bleeding route...somewhere near the bottom plz! :) Thanks all! Happy coolin+g!:D |
1 Attachment(s)
Don't know if this would help, but i think so: your ambient is pretty high, but if it was lower, i think this could do the job, and combining it with reversing the flow i think you would get some pretty nice temps: the trick is (of course) that your rad isn't anyway near the inside of your comp, which usually is a lot hotter than ambient.
|
Okay, I re did the flow to the last pic that I posted and the temps were about the same as the previous one, not as high, but I did see some 41-43C idles at around 30C ambient with one side of the case off... Damn I thought that the old way was bad, with the pump---cpu---rad---T-Res... this shit sucks, well not sucks, but ain't what I thought it was going to be! Even though the flow rate is slightly less, but the cooler water coming from the Rad should have compensated for it...but I guess not. I just can't think right now, it's like I've got a cooling-mental-block. It's a simple fix, which isn't dawning on me, yet... I'm going to take BBen's advice and chop off some of the Res-lnlet tube, that might be restricting the flowrate... argh!
|
Quote:
|
Quote:
In an open loop, which is what this is, a centrifugal pump "throws" the water out, if it happens to be sitting in the housing (i.e. it's primed). |
All times are GMT -5. The time now is 11:47 AM. |
Powered by vBulletin® Version 3.7.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
(C) 2005 ProCooling.com If we in some way offend you, insult you or your people, screw your mom, beat up your dad, or poop on your porch... we're sorry... we were probably really drunk... Oh and dont steal our content bitches! Don't give us a reason to pee in your open car window this summer...