New H80i, possible pump issue - CPU immediately overheating on boot

[SpecMode]

I purchased a new H80i yesterday with the intention of replacing a Cooler Master Hyper 212 Plus air cooler for overclocking and better general system cooling. However, after installing the cooler (and reseating the pump module a few times to ensure solid contact with the CPU), upon powering up the system, the CPU temperature rapidly rises to its thermal limit, causing the system to first throttle CPU speed, then eventually power itself down.

 

I have removed and reinstalled the entire cooler assembly multiple times, checking all connections each time, and everything seems to be hooked up properly. All fans connected to the pump module are running full blast, and I can see a speed indicator of 2000+ RPM in the BIOS real-time monitoring display. When I could get the system briefly booted into Windows to check Corsair Link, however, I could not get the application to respond properly due to CPU throttling.

 

I've connected the pump power cable to both SATA headers on the power supply, with no difference in behavior noted. I also checked for obstructions on the motherboard, and it does not appear that the pump assembly is making contact with anything other than the CPU itself; also, the plate seems to be making even contact across the CPU surface as evidenced by an even thermal compund layer. I also rotated the pump module such that it should be avoiding the row of capacitors adjacent to the CPU socket.

 

I finally ended up putting the original CM air cooler back in, and the system is back to normal (no CPU overheating, normal temperatures observed).

 

I would appreciate any suggestions on what I can do to fix the problem!

[Toasted]

Was the H80i 3-pin plugged into the CPU_FAN header and it was reading ~2000RPM?

[SpecMode]

Was the H80i 3-pin plugged into the CPU_FAN header and it was reading ~2000RPM?

 

Affirmative.

[RAM GUY]

Since pump is working, my best bet would be the contact between your CPU and the cooler's cold plate. I know that you have made sure that they are making contact but it does not mean that you have made proper contact.

 

BTW, how were you monitoring the CPU temp? BIOS?

[SpecMode]

I monitored BIOS temperature to thermal margin and individual core temps in HWMonitor (which were reading 95-104 C before I shut the system down to prevent damage). The last time I got Corsair Link running, I was able to check the cooler temperature itself, which was indicating 80+ C and rising before I had to shutdown again.

 

As far as checking proper contact between the CPU and cold plate - besides the visual indication of thermal compound transfer between the plate and CPU surface, how would I verify that proper contact is being made? Manually pressing on the pump module surface while running had no effect on CPU temperature, so I'm uncertain as to what more I can do at this point.

[wytnyt]

the best way to check the contact of thermal past is to remove the hydro and visually look at the paste contact,if done then clean well and re-apply new paste,it certainly shouldnt heat up fast but the computer would safely shut down before it overheated.go into bios and check the speed of the pump,it should be @2200 rpm

[SpecMode]

The last couple times I removed the cooler, I completely removed the thermal compound from both surfaces and reapplied a fresh thin layer to the cold plate before reinstallation; when the cooler was subsequently removed, there was an even layer of compound remaining on the CPU surface, in theory indicating even contact.