I7 920 temps too high?

[mcj3871]

CPU i7 920 4.2GHz @ 1.26v lapped
MB Rampage II Extreme
Ram Corsair XMS3 1600 1.56v
PSU Corsair HX1000
Pump swiftech MCP655
Rads 2 swiftech MCR220. 1 Danger Den X-FLow Black Ice Pro II 220
Cpu block Danger Den MC-TDX
Res Danger Den RAD-Reservoir
1/2" Tubing
Thermal compound Antec formula 5. Same as asc 5 I was told.
Case HAF 932
Fans pulling air in to the case push pull on 2 internal rads. Fans pushing air on back rad.
Room temp 21c
Idle temp 41,38,39,35
Max load temp 81,79,75,73


These temps just don't seem right to me. I have reseated the cpu block 20 or more times over the last 2 weeks and the temps change maybe 1 or 2c at idle and 3 to 4c full load. I have a heat killer 3.0 on the way but would like to know if there is any thing I can do to get it colder. I mean I have more than enough rad for an I7 and a Video loop. My pump is good and my flow is fine. One thing I did find odd is that none of the rads get even warm when its pushing 80c out of the cpu. Is this an indicator that my cpu block just isn't picking up the heat or just not enough?
Here is a pic of my rig. You can see that I didn't make any really long runs with the tubing and only used 1 T fitting and 1 90.
Any help would be great. If you need more info just let me know and ill post it.
http://hardwaredemon.com/pix/109_0837.jpg

[masska]

what are you measuring your temps with?
And btw those temps dont look that bad to me. I'm running two loops in mine (with a 920 @ 4.2) and my temps are almost identical to yours.
Your cpu is OC'd so you must expect slightly higher readings :up:

[Kurz]

The new block would probably help a bit.

[snoro]

do you have been running your loop for aalong time with that uv red coolant ? people that have runned their wit dyed coolant for six month and more had some dye clogging up in the block and the rad.it could explain the high temp.

[zalbard]

The new block would probably help a bit.
Yeah, most likely the case.

do you have been running your loop for aalong time with that uv red coolant ? people that have runned their wit dyed coolant for six month and more had some dye clogging up in the block and the rad.it could explain the high temp.
Possible also...

And not sure about your thermal compound. But that shouldn't make a dramatic difference.

[scamps]

Without water temp no funded assistance possible ;)

73° is a middle value, 81° is far to much for this triple-dual setup.

Is the big difference between your cores already known without other blocks? Run Realtemp´s sensor test and check. Realtemp will show you whether the sensors are only bad calibrated (and the highest values are crap and you can ignore them) or not. If not there is a problem with the mounting of the block or the block itself.

For comparison only: My D0 runs 24/7 with 4,2 @1,32 V load. Delta between water and cores is about 35°. So I would need a 40° water temp to have your core temps :eek: (my external PA140.3 won´t allow that :D)

[mcj3871]

I'm measuring temps with real temp.
I'm hoping for a good 5c drop in temp and if that's not the case I'm going to have to think maybe I messed up the lapping of the cpu some how.
The loop has been running for less than a month now. The water has been changed many times and there have been no clogs. I'm going to be running just plane old distilled water when i put the new block in so ill be flushing the hole system again.
I have no way right now to check the water temp. Whats the best way to do this with out spending a bunch of cash.
Im running the sensor test right now and ill post the out come.
By the way the first set of temps where taken using real temp and intel burn test set to very high.
http://hardwaredemon.com/pix/sensortest.jpg

[Kurz]

The fact you lapped your CPU... Hmmm there might be a problem with that. Since most newer blocks are bowed. Still give it a try.

(DONT LAP THE Waterblock when you get it you probably damage the bow).

[scamps]

I agree with Kurz. Though a comparison between two cpus in two different systems is not as simple as it seems to be your temps are absolutely very very high. My temps in idle are much lower just because I have EIST active, but also all the way to 100% load they are very much better indeed.

Perhaps you are able to measure your water temp with a clinical thermometer on load? In my loop it would not be possible, because the water temp does not exceed 32°.

http://www.abload.de/img/rthbyy.jpg (http://www.abload.de/image.php?img=rthbyy.jpg)

[rge]

Fan speeds?

2 things, first is when I changed my GTZ block to HK3, I got ~3C better load temps, and looking at internals of the block your using, I bet you will get at least that much improvement at load.

Second issue, look at Scamps minimum temps with Realtemp then at yours, yours avg 40C his 28C, so 12C higher. Scamps idle vcore is 1.07 with reduced MHZ, so his ambient temp is probably ~20-21C (about a 7-8C gradient from core temps to ambient at low vcore/mhz <1.1).

MCJ, your idle vcore is 1.26, idle temps should only be 9-12C above ambient, ie few C higher than scamps not 12C higher, since you state your ambient is 21C, ie same approximate as Scamps . If your ambient is also ~21C, your idle temps should avg ~32C. Idle temps you can compare cpu to cpu, since leakage/differences are negligible at low idle power output and i7s have accurate temps across range.

You need to at least measure air temp in front of rads with any temp measuring device you have, any thermometer or gauge. My guess is air temp is higher than you think, also you are pushing hot air coming out of case through your rear rad, probably not helping either.

But changing the block/tim will also probably get you close to the improvement you want as well. Also putting thermometer tip on rad should get you temps within 1-2C of water temps, like Scamps said without measuring air/water temps, it is flying blind trying to figure out what is wrong.

[mcj3871]

here are my idle temps today. Its a little bit cooler since the tmi has had a chance to cure a bit. 33, 40, 36, 37 average of 36.5c and scamps 32.5c. I dont have time right now to check load water temps but ill do that later tonight.

[franklin5252]

Man i love that case you have.

[mcj3871]

May have got it figured out. i made the mistake of using my finger to spread the tmi on the cpu. As well I changed to IC diamond 7 carat and this is what i ended up with. A 9.7 drop in max load temp. So it was bad tmi job. Im going to have to check and see if my heat killer has shipped yet or not. If it hasn't shipped yet I may call it off. As these temps don't look too bad at all. The temps are just after putting on the new tmi and i didn't let it cure before starting so they should get a little bit better.
http://hardwaredemon.com/pix/test3.jpg

[masska]

good work! now I need to remount mine :up:

[mcj3871]

By the way i am running at 4.3 now with 1.312V and my temps didn't change at all from 4.2 and its still rock stable.

[masska]

way to go! I'm happy its all sorted for you :up:

[mcj3871]

Well my order showed up today. Problem was it wasn't my stuff in the box. They sent me 2 fans and sent my block and back plate to Texas for some reason. They told me they intercepted my order and it should be here on Friday the 12th. I just hope they get their head out of their ass and get me my parts. I was pissed when I cut the box open to find just 2 fans in it. I guess this is why I don't like to order stuff off the internet.

[Vinas]

Who did you order from?

I was also wondering if you had LLC enabled or not? I noticed my CPU burns like the surface of the sun with LLC on, but hey that's why I got 2 rads.

[mcj3871]

I ordered from sidewinder and yes I have LLC on. Ill turn it off and see if changes at all.

[Vinas]

Any progress with LLC off? I take it you have to bump significantly higher to adjust for the vdroop. Another thing you can try is just tweak the ratio a little so instead of 200x21 try 21x199 etc just to see how low you can go...

[mcj3871]

temps are fine now.