Am I pushing the H50 to far?

Biffa · 6 Aug 2010 at 17:55

I've got a i7 920 D0 @ 4.2Ghz 1.4v 200x21 on an Asus P6T

With the H50 and the system under 100% load on all core with HT on I am hitting 100°C and the system is throttling back

After boot up it idles at 55°C

This is outside a case with fans in push/pull on the H50 radiator.

What I am not sure of is if the kit is faulty or if I am just asking too much of it.

Why am I concerned its faulty? Well the block/pump doesn't even get hot and the radiator is cool.

Fans all at 100% bios settings all switched off for fan profiles, even have the pump running on a 12v molex adapter from the PSU rather than the mobo headers, although I have tried it on every mobo header as well.

When I turn it on I can hear the pump start first.

I've stripped it off and fitted it again a couple of times, the notches are all in the right place in the rings and its tightened down as far as it will go. Could there be a problem with the mounting ring coming loose?

I'll try and get some pics when I get home but I heard such good things about this and I bought it to replace my swiftech H20-120 kit ready for my q6600 to i7 upgrade but at this rate I think I'll have to dig through my spare parts box and throw a custom rig together.

momo56 · 6 Aug 2010 at 18:02

MY h20-220 kit sharing a reservoir with a 4870x2, single rad and D5 pump hits 80 degrees on a core at 1.31v 200x21 w/ HT before it BSOD's using prime 95 so I think your h50 is fine.

momo56 · 6 Aug 2010 at 18:04

both rads are also push pull and the double rad is hot to the touch. I ordered a 2nd hand tfc 360 so I might add this into the loop and see how it goes.

setter · 6 Aug 2010 at 18:06

Biffa said:
pump running from a fan header from the CPU rather than the mobo.

When I turn it on I can hear the pump start first.

Should the pump not be connected to a 3 pin fan header or to molex, to ensure that it runs full speed?

Serenityxo · 6 Aug 2010 at 18:35

setter said:
Should the pump not be connected to a 3 pin fan header or to molex, to ensure that it runs full speed?

Correct, the CPU header wont provide the full 12v.

On a 12v header mine runs at ~1507RPM.

Yellowbeard · 6 Aug 2010 at 18:56

How are you measuring the temps? I have seen several ASUS users having trouble with ASUS Probe reading very high temps. Compare your BIOS idle temps to whatever program you are using in the OS.

PhillyDee · 6 Aug 2010 at 18:58

Surely it should be hot though? There should be SOME heat coming from the block!

bifday2k · 6 Aug 2010 at 19:00

Biffa the rad on the H50 just won't handle that vcore sad to say but the H50 is only as good as some air coolers, even with push/pull.

So the kit is'nt faulty ur just asking to much of it really with that vcore, its a good cooler, but won't be able to handle what your asking of it.

Biffa · 6 Aug 2010 at 19:01

momo56 said:
MY h20-220 kit sharing a reservoir with a 4870x2, single rad and D5 pump hits 80 degrees on a core at 1.31v 200x21 w/ HT before it BSOD's using prime 95 so I think your h50 is fine.

momo56 said:
both rads are also push pull and the double rad is hot to the touch. I ordered a 2nd hand tfc 360 so I might add this into the loop and see how it goes.

Thanks for the input but I'm not sure how that is relevant.

Biffa · 6 Aug 2010 at 19:10

Yellowbeard said:
How are you measuring the temps? I have seen several ASUS users having trouble with ASUS Probe reading very high temps. Compare your BIOS idle temps to whatever program you are using in the OS.

Coretemp and realtemp. Bios confirms idle temps.

PhillyDee said:
Surely it should be hot though? There should be SOME heat coming from the block!

My experience with watercooling would confirm that.

bifday2k said:
Biffa the rad on the H50 just won't handle that vcore sad to say but the H50 is only as good as some air coolers, even with push/pull.

So the kit is'nt faulty ur just asking to much of it really with that vcore, its a good cooler, but won't be able to handle what your asking of it.

Thats what I am concerned about, hence the title, but I'm sure I've seen reviews where others have been able to maintain somewhat more sane temps without their CPU's throttling at similar speeds/vcores. :confused:

Thanks for the comments guys. I'm not new to watercooling and on this rig wanted something simple to just chuck on and not bother with it. Guess I am not going to be able to do that this time

momo56 · 6 Aug 2010 at 19:23

Biffa said:
Thanks for the input but I'm not sure how that is relevant.

Was simply attempting to assure you that there was nothing wrong with your h50 and used my setup as an example of the kind of temps from a 920 D0...

rebel rebel · 6 Aug 2010 at 20:43

It sounds about right, if I was you I would drop it down to 4.0 because 100c is the shut off temp. I've got a i7 920 D0 @ 4.0 1.3v qpi/dram 191x21 on an Asus P6T delux oc palm edition, with 2 vipers on the h50. It hits the mid 80's on full load in prime or intel burn.

rebel rebel · 6 Aug 2010 at 20:50

Serenityxo said:
Correct, the CPU header wont provide the full 12v.

On a 12v header mine runs at ~1507RPM.

Thanks for the tip, I have mine connected to the pwr header and it runs at 1360rpm, so its good to know I can get an extra 147rpm.

minicle · 6 Aug 2010 at 20:52

1.4v seems more than necessary to me, it would also explain the temps.

Biffa · 7 Aug 2010 at 15:24

Dropped it down to 4Ghz at 1.3V and temps are just sub 90°C

JonnyG1982 · 7 Aug 2010 at 16:10

Biffa said:
Dropped it down to 4Ghz at 1.3V and temps are just sub 90°C

How hots your room? THink when i had a H50 I was getting around 72-76c depending on room temp at that speed/vcore. But i have decent case cooling.

PhillyDee · 7 Aug 2010 at 17:34

Just been running IntelBurnTest on my 4.2Ghz i7 (Ocuk bundle) with my new H50.
Tops out on maximum 90 Degrees. Thats using my top 140mm fan (900RPM) as in intake into the case. I am currently using positive air pressure, with 2 intakes, gfx exhaust and psu exhaust.
The H50 gets red hot! So proves its working hard.

JonJ678 · 7 Aug 2010 at 17:57

Assuming you've done the standard remounting / checking you've installed it correctly, conclude that the cooler is faulty. A processor running hot enough to throttle / switch off should leave the radiator very obviously hot to the touch.

I suppose it's possible that the H50 is not capable of coping with this heat load, but I can't imagine it reacts to excessive heat by remaining cool. Perhaps too much heat killed the pump, but it feels like a long shot compared to a manufacturing fault.

^Is that hyperbole PhillyDee? If not I'm genuinely curious as to which part is glowing.

PhillyDee · 7 Aug 2010 at 19:02

JonJ678 said:
^Is that hyperbole PhillyDee? If not I'm genuinely curious as to which part is glowing.

Huh?

You know what I meant

.

Its very hot to the touch. I could get you an exact temp if you really wanted me too.

Biffa · 8 Aug 2010 at 02:05

JonJ678 said:
Assuming you've done the standard remounting / checking you've installed it correctly, conclude that the cooler is faulty. A processor running hot enough to throttle / switch off should leave the radiator very obviously hot to the touch.

I suppose it's possible that the H50 is not capable of coping with this heat load, but I can't imagine it reacts to excessive heat by remaining cool. Perhaps too much heat killed the pump, but it feels like a long shot compared to a manufacturing fault.

^Is that hyperbole PhillyDee? If not I'm genuinely curious as to which part is glowing.

Its like its not pulling down hard enough on the cpu, or the pumps not running or there is some heavy restriction between the block/pump and the radiator.

I know how hot full WC rig components can get just at 70/80°C, this little rad is ambient, the pipes are ambient and the top and sides of the pump/block are ambient, but the CPU is at 80/90/100

To the JohnnyG: Its sitting on a bench outside of the case. Room is 15-25 degrees depending on the time of day.