3770k 4.8 ghz temps

Luke Phillips

Honorable
May 27, 2013
125
0
10,690
Hey guys i recently got an oc of 4.8ghz using 3.8v on my i7 3770k. It is stable so far, but when i run IBT my temps get a little warm, the hottest core avg is 87-88c under full load. However, i NEVER put my CPU under full load, the most demanding thing i do on my PC is run bf3 on ultra on nvidia surround. When i run that, my max temp on the hottest core is 72c. Do you think this OC with these temps is safe because under a daily usage, i am never going under full load. Thanks in advance:) Also i am using an XSPC Rx 240 kit to cool with an additional NexXxos 120mm radiator.
 
Solution
What case is this? Sounds like you are running air intake thru rad and exhausting into case. For the large majority of cases this doesn't work well.

Ex of good case flow with rads: intakes bottom and front. Rads mounted top and/or back and exhausting out case. This utilizes convection currents to increase total case cooling potential and maximizes airflow potential which provides more than sufficient cool air to rads. Also gpu, vrm, and chipset cooling is improved due to increased proper flow direction, and not having warm air pumping to them

The x20 750 is an... ok... pump and very likely part of your problem. At near 2/3 capabilities of a D5 pump, it isn't really designed for addons - well very well anyway. It works well in...
pvts.png


Your voltage is acceptable for 4.8GHz. However, your load temps in IBT are high and you should consider dropping down to 4.7 or 4.6 until you have max temps of 70-80 under load on IBT.
 
Gonna agree with intel. But with 3x120 rad space and knowing what tdw those rads are capable of, something isn't tweaked/set right in your cooling I suspect. Flow rate, fan setup, TIM type and application method, block contact - something is off. What pump are you using? Fans model and arrangements? Case airflow?
 
I have 2 200mm intake fans as well as a 120 mm right in front of the rads, on the 240 rad i have 2 corsair af120 quiet edition fans and on the 120mm rad i have one corsair af 120 quiet edition. My block is on very well and covers the entire cpu, and is on tightly. I am using the XSPC X20 750 dual bay res/pump combo. Not sure what the flow rate is though, its at whatever it comes set to stock.
 
just looked it up flow rate is 750lph on my pump

 
What case is this? Sounds like you are running air intake thru rad and exhausting into case. For the large majority of cases this doesn't work well.

Ex of good case flow with rads: intakes bottom and front. Rads mounted top and/or back and exhausting out case. This utilizes convection currents to increase total case cooling potential and maximizes airflow potential which provides more than sufficient cool air to rads. Also gpu, vrm, and chipset cooling is improved due to increased proper flow direction, and not having warm air pumping to them

The x20 750 is an... ok... pump and very likely part of your problem. At near 2/3 capabilities of a D5 pump, it isn't really designed for addons - well very well anyway. It works well in basic situations, but when adding exyra rads or blocks, which increases restrictions, just doesn't have the power to compensate.

At least re config fans as necessary and see where you sit. You may need to upgrade yhat pump for your setup. Sadly, this also means upgrading to a different resevoir as the pump is built into it. XSPC does offer that kit with a d5 pump instead....

Make sure your tubing is short as possible, and straight as possible. The less bending and corners the better. If you can upload a pic of your setup, mightbhelp us help you improve those temps
 
Solution
Its the thermaltake chaser mk1 case. I have my rads and intake fans set up just as you suggested. 200mm and 120mm intake in front, 200mm on side, 120mm rad is in the rear, the fan on that is pushing air our of the case, and the fans on the 240 are also pushing air out of the case on the top. I got my runs as straight as possible for me without kinks, unfortunately the first run from the res/pump has to go straight up to the 240mm rad, then slightly down to the second rad on the rear, then it goes down then up from the 2nd rad to the cpu, then straight back from the cpu to the res. Will upload pic of rig asap, you can also see it in my avatar photo on my account. Thanks for the help, i would give you best answer if i hadn't selected it already.
 
Handle... sure - effectively with oc conditions... prob not. Rads arent a huge restriction factor themselves, but tubing and fittings are. In addition, with oc conditions you need high flow to the tune of 4 Lph. That pump doesn't support that. With no restrictions a pure flow test provides 2.5-.75 Lph. So you can see the issue.

Mk1 is a nice case and great for airflow. It is a highly vented case, so it does benefit from positive case pressure - meaning more intake than exhaust cfm. Not a major thing but helpful for tweaking and getting max out of your cooling potential.

Granted flow direction between components is not imperative given thermal mass equilibrium of liquid. But especially with a weak pump, if you can snake it so res/pump> rad> rad> res you will reach equilibrium at a longer interval, giving the system time before it reaches max temp- just a thought not a must
 
I know, the errors today are ridiculous. I have the opposite setup that you recommended i think, mine goes res/pump>rad>rad>cpu>res/pump. Do you think it would really make a big difference in temperatures though? Because i have read in many places that the order of the loop will not effect temperature that much. The thermal paste i used was XSPC k2 thermal compound which isn't bad from what i hear. I placed a decent sized circle of it in the middle and pressed down the cpu block on top, but made sure not too use too much because you know what they say, less is more. I tightened all the screws on the water-block equally as well.
 
Generally no won't make a difference. But again we have an extreme clock and weak pump. So yes a few degrees gained will help.

Not sure what a decent size is in my idea vs yours lol. More than half a pea, or about bb size is usually about right. So if you used more than that, consider unmount, rubbing alcohol n qtip to clean, and re seat. Iirc, xspc tim is a bit thicker so you probably had height on that circle too, and that increase volume used
 


 
One final option - try stepping down PLL voltage to lower temp. Of course retest stability at each stepping.

Every degree from every tweak adds up. You don't have any one thing to cure. Just a bunch of little tweaks that add up
 
One final option - try stepping down PLL voltage to lower temp. Of course retest stability at each stepping.

Every degree from every tweak adds up. You don't have any one thing to cure. Just a bunch of little tweaks that add up
 
I was going to do that, but i could not find it in my bios. My mobo is an MSI Z77A-G45. Do you think it would be called something else? Also i found vdroop offset control, not sure if it would help though. Also sorry i haven't uploaded any pics yet, my internet is being very slow when its usually fast, not sure why. Working on uploading one now...
 
Ok - flipping hose on input output is all you need to change flow. Far as back rad - not a fat of 90 fittings, but it looks like a 60 degree fitting pointed at left would allow a shorter run and clear the gpu. Cant see well enough to tell if you have clearance but flipping the rad up and 2 60 fittings would shorten even more

Flow on a loop is consistent from start to finish. Pressure degrades as you go further from the pump. The micro nozzles on a cpu block work best with high pressure. Normally not an issue with a d5 pump. But your x20 will help cpu block perform better if closer