![]() |
|
|||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
![]() |
|
|
Thread Tools | Search this Thread |
Rating:
|
Display Modes |
|
|
#1 |
|
Old Jedi master
Join Date: Sep 2006
Location: Manchester, UK
|
Convection....watercooling....check in
Decided I was tired of using a pump...temps are MUCH better with one but pumps fail and im looking for a new challenge/way of running a rather nice 6000+ AM2 I have here in my workstation.
CPU will run 2GHZ at 1V and 3GHZ at 1.25V so I decided to play a little. First pic the old setup inc DDC and thermochill 120.2. Got bored and stripped that all out, replaced it with a prototype OCZ block and a Black ICE gen2 240 Xflow rad and 2 fans....NO PUMP... Now at first I ran this at 1V 2GHZ fearing it would not handle default speeds with the CPU having +100w TDP...but I was wrong. Presently priming at 61C both cores stable...ambient room is 27C so im expecting 55C ish from a 20C room temp ![]() |
|
|
|
|
|
#2 |
|
Old Jedi master
Join Date: Sep 2006
Location: Manchester, UK
|
Orthos was fine for over an hour, start OCCT and it crashed after 5 mins...pretty brutal piece of software
![]() Vcore now at 1.3V, with pumped water i could do 3GHZ with 1.2V so it shows temps play a huge part in how fast processors will run at what voltage. Testing will continue now at 1.3V with the fans, once in my new workshop I will probably ditch the fans for a totally passive setup as the ambient will be a lot cooler. |
|
|
|
|
|
#3 |
|
Wiseass Extraordinaire
|
Very cool Tony. I wouldn't think that would be possible with some a small amount of water. Have you considered a bong setup or do you still require some portability?
Btw, interesting internals in that OCZ block, is that a recent prototype or something dug up from the archives? ![]() |
|
|
|
|
|
#4 |
|
Old Jedi master
Join Date: Sep 2006
Location: Manchester, UK
|
Not saying how early or late, block will be launched soon for you all to see and it is different to the one in my pic.
No need for a bong, this PC will live in an area that is 15C cooler than this room and may end up being 20C cooler in the winter so i doubt I will ever even need fans on the rad. I will change the leads on the fans to 3pin so i can use the board to switch them completely off dependant on temps, plus im going to have to swap the fan in the old PSU as its noisy as hell now the rest of the rig is in stealth mode ..LOL |
|
|
|
|
|
#5 |
|
Old Jedi master
Join Date: Sep 2006
Location: Manchester, UK
|
Forgot to mention 1.3V fully OCCT stable, ambient 25C, core temp reported 64C max on both cores with smart guardian showing 80C
but we all know thats a little high ![]() Idle is around 46 to 48C which is kinda nice with no pump, I feel a mod to the HTPC coming next ![]() |
|
|
|
|
|
#6 |
|
New Member
Join Date: Jan 2007
Location: Portland, OR
|
that's cool. I'd be interested to see how your temps change if you lay your rig on it's side.
just curious how much natural convection is going on in your lines. |
|
|
|
|
|
#7 |
|
Old Jedi master
Join Date: Sep 2006
Location: Manchester, UK
|
The optimum angle for the rad is 30deg from upright with the top of the rad leaning over the bottom...this is where convection really speeds up.
I may now add a 320w PELT on 5V to the waterblock with a cold plate, this will lower CPU temps around 10 to 15C as seen from when i last tested this 2 yrs ago. not sure why you want me to lay the rig on its side though? |
|
|
|
|
|
#8 |
|
System Fungineer
|
take note, the higher ther delta-T in the loop the more thermal driving head (i.e. flow)
|
|
|
|
|
|
#9 |
|
Old Jedi master
Join Date: Sep 2006
Location: Manchester, UK
|
Ordered some TECS, found a nice 70W that runs on 8.5V so 5V should be kinda interesting with its output higher than a 320W at 5V
. Also it has a 30mmx30mm footprint so direct to heat spreader mounting should be good to go. |
|
|
|
|
|
#10 |
|
New Member
Join Date: Jan 2007
Location: Portland, OR
|
i was more thinking of the liquid convection. I think i might have to pull my heat transfer book out. I think the actual thermal diffusion through the liquid would probably have a bigger impact on the heat dissipated then the natural convection created by the minor density changes in the fluid, but i was curious if there was any noticeable contribution. I can't remember how much water expands with temp right now.
|
|
|
|
![]() |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | Search this Thread |
| Display Modes | Rate This Thread |
|
|