Watercooled Phenom II 965 running over 70 at 3.4GHz full load

Hey guys, I've been struggling with this problem for awhile now and it's finally gotten the better of me and I just dunno what else to try.

I'm running a 3.4GHx Phenom II x4 965 with a Corsair H60 cooler on an Asus M5A99X EVO AMD 990X motherboard. Idling the core temperature varies between 35 and 50 degress and under load climbs up to 55 almost instantly and hits 70+ within a few minutes.

I did think it might have been the motherboard before and recently replaced the motherboard and cleaned and repasted the CPU but it hasn't helped.

Does anyone know if there is a temperature at which this chip will shut down or anything as I haven't wanted to just let it do a CPU test until it crashes as it still keeps running right upto 75 degrees.

I don't think the problem is airflow related as nothing else in the system runs above 36 degrees. Graphics wise the card I have doesn't go above 65 degrees under load. And I can easily hit the 70 mark on the CPU with everything else under or around 35 though.

Thanks for any possible help guys.

Full system specs:
AMD Phenom II X4 965 3.4GHz,
Corsair H60 sealed liquid cooler,
Asus M5A99X EVO AMD 990X motherboard,
MSI AMD 6970 Lightning 2GB,
Thermaltake TR2 800W PSU,
12GB DDR3 (2 x 2GB sticks, 2 x 4GB sticks),

 

Did you apply thermal paste to the CPU when you installed the H60? If it's present and you've tried reapplication, then you may have a defective temperature sensor or H60. Try the stock cooler with a fresh paste application and see if the problem persists.

 

I can repaste it again but it'll be the 3rd repasting with the H60, aye I will give the stock cooler a go shortly. Think I've got enough paste left to do a few more experiments, I've been using Arctic Silver 5 paste and using Arctic Clean to remove each previous pasting.

One thing which I will mention is that if I touch the radiator on the H60 it's hot enough to burn when the CPU is under load.

Another thing which I did forget to mention is that I'm running this CPU completely stock in terms of voltage, clock speed FSB and everything. As I'm typing I just burnt my finger on the damned rad.

 

What fans do you have attached to the radiator? If the CPU really is generating that much heat, it might be defective.

 

I'm leaning towards the chip being defective personally as you can feel the airflow from behind the radiator. I'm using the fan which comes with the H60 on the outflow of the radiator, my case intakes are Akasa Viper PWM fans so there's a lot of airflow in the case towards the radiator.

 

Well try the stock cooler and see if you get of a load temp of around 60 tops (as it should be around there at stock). The pump on the H60 may be cycling extremely slowly.

 

Ok, thanks for the all the suggestions I will have to get some sleep now but will try them out tomorrow morning, according to Speedfan the H60 is running at 4410 rpm, but I'm afraid I didn't check it when I first bought it so I don't know what it was saying it was doing.

 

Right, I tried fitting the stock heatsink and the temps went straight upto 70 in about 30 seconds on the Intel Burnin test, then to 80 within 3 minutes. Interesting enough is that the chip hasn't shut itself down or crashed when running at full load at those temps.

 

when did you buy this cpu? is it the first time you run it or did it come from your previous rig?
did you check the voltage in the bios, it must be around 1.37v.

 

The CPU is about 5 months old. I got it as part of a bundle with a Foxxcon A88GMX motherboard and recently replaced the Foxxcon board as the temps on that board were just too high for my liking.
The temps on the CPU are the same on my new board as the old one, only real advantage is now I have a new motherboard with AM3+ socket, USB3, SLI/Xfire support, etc. so when this chip does die I won't have to buy a new board at the same time.
The CPU is running 1.33 volts peaking at 1.36 volts under load. The BIOS voltage settings are set to 1.37.

 

1.37 set to 1.33 real seem alot of vdrop, did enable you cool and quiet?
for your temp i think you cpu is dying now , the first sign of a bying chip (cpu or gpu) is the temp getting higher and higher, i think you might RMA it before it burn with you new mobo.

 

I've tried it with Cool and Quiet enabled and disabled, the temps were actually higher both idling and on load with it enabled.

As for RMA I did take the CPU back to the supplier as it's an OEM chip so AMD won't cover it on their warrenty and I was told that the chip was completely fine by Novatech.co.uk. I'm hoping the chip will last until Piledriver comes out as I'll probably buy one of those as soon as they come out and I would prefer not to have to buy Bulldozer for now and replace that again in 5-6 months.
Do you know what the chance of the CPU killing the board is?

I can try overvolting the CPU on the board to push it to 1.37 but I'm a bit worried that if I do that it might then push the voltage upto 1.37 natively and it end up at 1.41 with the overvolting.

edit: When I first put the H60 on it was idling at 20-30 degrees and going to 45-50 under load with with 1.5 volts and 4.2GHz.
These temps were on the Foxxcon A88GMX. However on Monday when I was still using that board the temps were the same as they are on the ASUS board, hence I changed the motherboard on Tuesday (yesterday)

What temps are you getting from you CPU Bonob if you don't mind me asking?

 

Are you sure that the waterblock have a good contact with the chip? when you take it of did you see the print of the chip on the block?
for the damage if the cpu die it can fry your socket or blow up your power phase.
i'm getting on low rpm fan 30°c CPU and 32°C core idle, and around 45°C both full load on a custom loop in a warm roon around 21-22°c.

 

When I repasted it yesterday there was a chipmark on the block, I can take the block off again and have a look but if it's lined up with the old chipmark I don't know if I'd be able to see a new one.

I've undervolted and underclocked the chip to 1.3V and 3GHz and the temps have dropped by about 6-7 degrees.

 

can you see if the coolant is flowing well in the loop?

 

Do you know if there's a way to check? The H60 is a sealed loop cooler so I don't know how I'd be able to find out really. you can feel the pump going if you touch the block and the pipes do feel like something is moving inside them but I'm afraid that's the best I can do without risking breaking the system.

 

some guys on this site had the same problem as you, and he as to RMA his h60 due to a faulty pump, check the pics of this site to see if you had the same mounting problem
http://forum.canardpc.com/threads/61964-Attention-Corsair-H60-Carte-mère-micro-ATX-!
you can also check the corsair forum too.

 

Well I don't have the same mounting problem the board has a good amount of clearance around the socket, closest thing is actually the memory slots. Aye I did send e-mails to both AMD and Corsair's technical support departments last night but haven't received replies from either yet. Thanks, I actually didn't think to check the Corsair forums, will have a look shortly.

 

Im gonna assume it had a bad pump or you have not mounted it correctly.

 

I'm fairly certain the H60 is mounted and if it was a bad pump I would expect that the stock heat sink would give lower temps. (post #8)

Also when I stop using the CPU the temps do drop straight back down again which also suggests that the pump is working.