2600K was fine, now overheating after re mount

[Bun-Bun]

Before at 4.8GHz 1.5v I was idling at 33°C and loading 74-83°C with LinX AVX. I decided to remount and try MX-4 instead of the MX-3 I had on there.

Now idle temps go instantly to 55°C and the CPU thermals and drops to x16 multi if I start LinX AVX.

What the heck? System in sig. Pump is running, fans are running, everything is cool to the touch.

I tried remounting it again. But still the same thing.

[defect9]

did anything else change in the system? meticulously check to make sure something in the mount isn't catching and causing an offset in the mount pressure from one side of the CPU block to the other. Even try to wiggle it around to make damn sure it's snug.

Did you drain and refill the system (if so, air in the radiator is a possibility).

Other than that, without coming over and poking it with a stick I can't narrow it down any farther.

[Sparky]

put mx3 back, remove only thign you changed

[Bun-Bun]

Nothing else changed.

I went to redo the mount and was going to do just as you suggested (put the mx3 back) but the damn swiftech backplate stripped. I had to pull the mobo. I have the stock intel cooler on it right now. I put it back to stock before I pulled the water off and it was still LinX at 70-80°C which is what it is doing with the stock cooler right now.

[Sparky]

Backplate stripped? dang tats crap.

[Bun-Bun]

What to do now. Trying to think if I can get the backplate back on good enough...


I still have the loop connect and when I started it up this time some air bubbles came out of the block. Maybe that's what it was...

Trying to repair the backplate right now... will get a new one from swifty once I have this back up and running. Really need to get around to cutting that hole in the mobo tray...

[Sparky]

All the blocks I have used have scres that go into the plate and use thumbnuts to tighten teh block. th e swifty screws into the plate? how did it strip? maybe they need ot use better metal.

can you use a nut on the back for now/

[Bun-Bun]

Yes the swifty threads into the backplate. The threads themselves are fine. But its like a nut thats fabed into the plate and its spinning freely. I am trying some industrial adhesive to get it to stay in place.

[lowfat]

You can use pretty much any LGA1155 backplate. They generally all use the same thread. I know that Thermalright ones will fit from experience.

[Sparky]

ohhh ok i thought the threads were screwed... haha see wat i did thar

[Bun-Bun]

You can use pretty much any LGA1155 backplate. They generally all use the same thread. I know that Thermalright ones will fit from experience.

Lets test your theory...

EDIT:

You are correct. The backplate from my Zalman CNPS10X heatsinks will work.

[Sparky]

Win.

[Bun-Bun]

And this backplate is a better design anyway... I think my main rig is going to just steal it and I will find a new one to replace this zalman one for my cruncher.

[Bun-Bun]

Done. everything back together and temps back to normal. Must have been air in the wb somehow.

[meanmoe]

you really should bleed and leak test before going back live... This is stated from painful experience (I will never use worm clamps again )

[Bun-Bun]

This loop has been running for years without issue. All i did was remount. I leak tested and bled it 6 months ago when I flushed it and re filled.

[meanmoe]

where did the air come from then?

[Bun-Bun]

Exactly.

[PatRaceTin]

your temp is too high

[Bun-Bun]

Elaborate?

[meanmoe]

Bun-Bun, did you ever resolve where the air came from?

My guess is that the system was never completely bled, e.g. air remained in the radiator or what not, but...
I thought about this a little more and came up with this scenario:
Suppose that you were completely bled and you had a fill port on the back side of the loop you had a fillport that wasn't completely closed or a barb that wasn't completely clamped down. If you have a very high flow rate then the dynamic pressure will be high causing the static pressure to drop. This could cause the sucking of air into the loop. When you turned off the pump though it would likely result in a leak though.

[Bun-Bun]

Never did figure out where the air came from. System has been fine ever since.

Fill port is at the top of the case with the line going to the top of the res and I fill liquid up into that line as well. No flow there anywhere near the air. I do not see how air could have gotten in.