I'm running Prime95 and Memtest89 v4,2 on my 1st build (also my 1st NEW computer since a 1994 starmax apple clone... i'm always running two gen behind... still waiting to play Crysis for 1st time)
I purchased a h50 and Antec 902 case 3 years ago, then put off the build because of reported heat issues on core i-7 and gtx 400-500 series cards.
Continuing to limp alng with a c2d e6600 on an intel mobo in a generic box, that I got from CompUSA's training room, when they went under (2007?). I download usenet vids and game (steam) on it. (+GTX 260 SP 216)
Core i-5 3570K CPU
Intel Z77GA-70K Extreme mobo
Corsair VengenceLP DDR3 1600 16GB kit (4x4GB) (CML16GX3M4A1600C9b) *certified w/mobo
H50 Cooler w/ x2 Couger Vortex HDB Fans (1200RPM , 60cfm, 1.73mmH2O, 17.7db)
EVGA FTW GTX 660 ti Superclock 3GB w/ Backplate
Seagate 1TB sata II + Intel 520 120GB (RST & READYBOOST Partitions)
Corsair TX850 PSU
Antec 902 Case
H50 is mounted push/pull intake from the rear fan mount. Rad is oriented tubes down (I'm a hydraulics specialist in the real world).
Room is 23c ambient. TIM is stock. It runs cold. I have an infrared thermometer to verify reported temps.
Prime95 CPU temp (multiple tests 10min -3 hrs)
idletemp: 22c +/-
torture @ stock 3.4ghz: 52c +/- (stable -no fail)
torture @ oc 4.5ghz: 62c +/- (stable -no fail)
The air exhausting from the H50 radiator is 3 degrees celcius warmer than intake air. The rad itself does not get hotter than ambient. The case is not warmed by the exhaust, in fact with the side door off, the fans blow cold air directly onto my face, it cools the mainboard components aswell. The Couger fans are silent, when i put my head into the case i can only hear the antec branded 120mm front fans and the BigBoy 200mm above. My GTX 660 ti, at idle runs +10 degrees above ambient, much cooler than my GTX 260. The TX 850 PSU is amazing compared to the TX 750's I have. Under full Prime95, the fan doesn't even spin up (85 F-101 F). The hottest spots on the mobo are the bridgechip, RAm and capacitors (30-35c)
I'm a noob at this oc'ing and tuning. Intel visual bios was simple but i read, underwhelming and conservative in oc'ing compared to other mobo's. As a noob, I'm ok with that. I was very impressed with the quality of my old DQ965GF board. It's been running 24/7/365 and constantly downloading/gaming/surfing/dataserving for 5 years with only a few reboots/year. I'm passing the old rig onto my daughter, who's six and gaming on my wifes casual steam games already.
_________ One Question ?___________
The ram is reporting as 0.74-0.75 volts in Intel Desktop Utilities. Bios specs are asking for 1.5v. I troubleshot the RAM in pairs both syncronous and asyncronous modes and swapped , looking for a bad one. I changed oc speed from 1333 to 1600 and back again, used automatic and XMP profiles... no change in voltage. I did update the bios when i installed the OS, which required a CMOS purge by removing the battery to get it to reboot after the bios update. Can anyone give me more insight into this ram voltage issue. Google is of no help, searching for undervoltage and underclocking only gives advice on intentional downclocking... not unintentional. Searching for an arbitrary voltage like 0.75v doenst find searchstrings for alternate undervoltages like 0.76v. I have a support ticket in with Corsair RAMguy. Maybe you guys can save me days on the phone talking to intel. Has passed Memtest86 v4.2 . [ Doing more research I realised the "LP" designator on the RAM is "low power".. but it should be 1.35v ... or is 0.75 v normal? Is the intel software giving me a bad reading? ... why does the bios report 1.5v instead of 1.35v LP?]
This is during torture test Prime95 with side doors installed, fans on minimum.
Just figured out how to use the fan control in the bios. Pump and fans (1200 RPM) now run at full speed. Is that the proper pump rate for the H50 or should i turn ot down?
[Images below of idle and full torture. NOW 10c cooler at full load. The pump had been running 800 RPM and inner fan @ 800 RPM retarding flow rates].
Here is a final torture test with CPU overclocked by bios slider to 4.5Ghz (Note: 20c cooler than 1st OC attempt with improper fan and pump setting)
________________ Last Thoughts_______
There are many variables when setting up the watercool option. Proper case venting, fan choice, TIM application, bracket tightness, fan speed , pump speed, ambient temps, CPU TDP, proper radiator orientation to prevent cavitation by entrained air within the fluid... all contribute to differing results. Your "milage may vary" in proportion to the good sense with which you drive.
To the "particular individuals" who continue to advocate aircoolers by the reasoning the do not want "all that hot air blowing onto my components" ... where the hell do you think all the heat dissapated by your precious aircooler is going? To the guys who are exhasting the air rather than intaking 1) get a modern case to fix your flow problems 2) Stop, ignoring the engineers advice and then having the to gall to complain. To the guys scraping the TIM off and lapping new products... You have no clue. A carpenters straight edge? Are you for real? I have 20 years experience with lapped sealing faces. I have ABLAP diamond lapps (1 lightband flatness), monochromatic light sources and optical flats for checking flatness. Do you? .. But why would I do that to factory new materials? The engineers at Intel & Corsair have produced a finished product to specification... and it works. Any defects will not be fixed by your heavyhanded and improper technique with materials less flat than what you started with. BTW, professionals do not use sandpaper of any grit for a mirror finish, they use an oiled "anti-corrosion" brown paper on a flat backing ( ie heavy fine glass).
This has been fun....
Friend me on steam
(dod:s division)Edited by opforwarrior - 1/7/13 at 1:42pm