Stock i7 930 w/ Megahalems: cannot get below 42C IDLE. Opinions?

[fireflame]

Hi everyone,
First I wanted to say that I've learned a bunch just by lurking around and reading the OC guides. The community seems pretty helpful here, and I've been involved with my share of forums.

To get to the point, let me start off with my components:

- i7 930 Bloomfield (stock, not OC)
- ASUS P6X58D Premium (on X.M.P Auto Setting)
- Megahalems revB w/ supplied TIM (PK-1) and single push Nexus RealSilent 120mm fan
- Dual 120mm case fans (Scythe S-flex D, Nexus Real silent)
- P183 case, clean cable mgmt
------------------------------------------------------------------------

For each seating of the HSF, I took the necessary precautions such as no finger touching the surfaces, used IPA70%, tightened mounting screws alternating one turn at a time, seated spring loaded bolts all the way down.

The following temps were recorded initially using Asus FanXpert readings, but I have interpolated them to fall in line with the ave temps of Real Temp software.

AMBIENT TEMP ranged from 21-24C (corrected from 18-20C)

Stock HSF: (fully seated plastic clips)
- 41-44C idle (for several weeks)
- sometimes it would hit 50-54C on nothing too intensive (internet browsing, playing some video/music), but loaded nonetheless, and it would take several minutes to cool back to the 40's. Like it couldn't keep up...

1st seating of Megahalems: (AC5's vertical line method, PK-1)
- 40-42C for an hour of idle (wasn't satisfied, so tried reseating again)

2nd seating of Megahalems: (5mm circular blob in the center of the IHS. I used this because the test spread was very nice)
- 43-45C all day idle
- Prime95 for 20 min (small FFT) yielded ave 69-71C. It stayed very constant at 67C for 10 min, then slowly climbed from that point.

3rd seating of Megahalems: (tinting both IHS and Mega, meticuous cleaning, and very thin vertical line 1.5mm
per AC5 instructions)

- stayed at 41-42C for 20 min upon bootup
- After running Prime95 for 3 minutes, RealTemp cores showed 70/68/71/67
- After Prime95, cores INSTANTLY cooled to 45C, but never got any lower
- idle at 45C
** EDIT: Just finished Prime95 for 1 hr: 71/69/72/67 were the maximum values. It hovered -2 below those for the most part. I expect these #s for a mild OC, but not at stock speed?
**EDIT2: Prime95 for 1 hr w/ tricool on high as HSF were 70/68/71/66. ambient was probably 1-2C higher than the original attempt)

For reference, here was my latest (3rd seating) install of the megahalems. I also determined that my megahalems has a convex surface (higher at the top), as indicated by the initial unseating, and confirmed using a SS straight edge ruler. It would just barely rock back and forth in one direction, but not in the perpendicular direction. The difference was probably on the scale of .2mm or something. Nothing obvious at all.


i7 930 after unseating the center blob method


the other half, megahalems after unseating


Seating#3: tinting, using clean credit card (later wiped gently and smoothened out with a completely lint free wipe)

Seating#3: the other half tinted ((later wiped gently and smoothened out with a completely lint free wipe)


Seating#3: my application method, you can see the tinted/finished IHS too


Installed (right before mounting the Nexus fan in push back direction)

In Summary:
- I didn't see any real noticeable difference between my 3 seating attempts. And sadly no idle improvement from the stock HSF except the rate at which it returns to near idle temps after some loading.
- I am concerned with the convex shape of the Megahalems. Is it normal? (I contacted Prolimatech)
- I checked the levelness of the stock HSF (intel) and the actual IHS chip and they were completely flat


For anyone who took the time to read through this whole thing, THANK YOU. I understand that monitoring idle temperatures doesn't tells the whole story, but even my Prime95 numbers seem a bit too high.

Any input would be greatly appreciated.

 

[CgS Drone]

I don't think your temps are that far off of the mark, seen where you said 18-20C ambient. With those temps and in a mid sized case and only using a single fan I think its doing fine. As far as the idle temps they look good.

Check out this thread also. http://www.overclockers.com/forums/showthread.php?t=652441

Your Nexus fan specs are only: 1000RPM and 37CFM, with the fact you only have two case fans and the other is even a lower CFM fan I think its doing REALLY well.

 

[fireflame]

I don't think your temps are that far off of the mark, seen where you said 18-20C ambient. With those temps and in a mid sized case and only using a single fan I think its doing fine. As far as the idle temps they look good.

Check out this thread also. http://www.overclockers.com/forums/showthread.php?t=652441

Yeah, I realized 18-20 is not correct.

It's just funny though, I read this thread here:
http://www.overclock.net/air-cooling/605008-post-your-megahalems-fan-configuration-temps.html

and they seem to get so much lower values. Is there a major difference in idle/load temps between the 920 and 930?

I guess by comparison, and some general rule I read about keeping it under 70 to be 'safe', I find myself unsatisfied...because I am not even OCed yet...

 

[trents]

First of all, your temp reports: Are you referring to core temp or cpu temp? What software are you measuring them with?

Maybe its just the lighting but it looks like the only place the heatsink base and the CPU face are making good contact is a narrow strip down the center from front to back. does your heatsink base have a curve to it?

 

[SecrtAgentMan]

Do you have a fan on the heatsink itself?

 

[jevery]

I don't think you have enough fan. Prolimateck suggests 57 CFM, (as minimum I assume). I just ordered a Megahalems with a GELID FN-FW12BPL-18 PWM @ 76 CFM.

 

[CgS Drone]

Yeah, I realized 18-20 is not correct.

It's just funny though, I read this thread here:
http://www.overclock.net/air-cooling/605008-post-your-megahalems-fan-configuration-temps.html

and they seem to get so much lower values. Is there a major difference in idle/load temps between the 920 and 930?

I guess by comparison, and some general rule I read about keeping it under 70 to be 'safe', I find myself unsatisfied...because I am not even OCed yet...

I went to the link you have shown, If you look at some of those numbers you will notice that they have fans with more CFM, also larger cases which are helping them. Some of the number I just flat out don't believe. When I posted last night after digging up the info on your fans all I could think is they are trying to built a silent system but they are doing it at the expense of heat build up in the system.

As far as the temps between the 920 and 930, the 930 turbo's to about 200mhz more than the 920, idle temps don't mean anything. So the 930 will use more volts-higher turbo (freq)-mid size case-low flow fans. Its not fair to compare that against those 920's and 930's posted. Take my 920 for example.

Comparing:

Yours-------------------------------------Mine
i7-930------------------------------------i7-920
Mid tower--------------------------------Full tower
2 case fans low flow-------------------6 case fans mid flow
cpu fan low flow (single)--------------cpu fan mid flow (dual)
69C average Load----------------------57 average Load
2990mhz Load--------------------------2806 mhz Load
41C idle----------------------------------38C idle
1.26Vcore???----------------------------1.22Vcore
CPU fan speed????--------------------CPU fan (continues full)

There are just to many things different between them to compare them fairly. But even if you do compare them I think with the fact yours is only about 10C higher than mine tells me with all the differences I think yours is doing good. Better fans I think would make a big difference in yours.

920 Load 24C ambient


920 Idle 24C ambient


In the end I look at computer cooling like golf, you can't compare your game/equipment against someone else's. Lower scores are better and you are competing against yourself for the best score, some just lie about their score.

 

[m0r7if3r]

i'd say it looks like a lapping would do you good unless you pulled the heatsink off weird or something and it gave you that weird spread on the paste...that said, your temps aren't too bad for the fan you have, i wouldn't worry personally. What do your load temps look like?

 

[CgS Drone]

Looking at your case manual, it says it came with tri-cool fans (Top/Rear). Assuming? you changed these out for a quieter system? It also says you can add up to 3 additional fans in the front of the case, you say you only have 2 case fans (pics show top and rear). Have you considered putting some fans in the front of the case to help air flow?

 

[fireflame]

First of all, your temp reports: Are you referring to core temp or cpu temp? What software are you measuring them with?

Maybe its just the lighting but it looks like the only place the heatsink base and the CPU face are making good contact is a narrow strip down the center from front to back. does your heatsink base have a curve to it?

Well, the single numbers were 'CPU temp' adjusted from Asus FanXpert, but adjusted to fit the average value core temp RealTemp would show on the sys tray.

I agree, I think the only good place of contact was the center. Maybe about 1/2-2/3 of the total area of the IHS. I want to know if it's normal and I am awaiting Prolimatech to see if this is designed to be this way. In my original post I did say that it was concave.

I don't think you have enough fan. Prolimateck suggests 57 CFM, (as minimum I assume). I just ordered a Megahalems with a GELID FN-FW12BPL-18 PWM @ 76 CFM.

I agree. Probably for optimal quiet performance, 57 CFM is the minimum. My Nexus RealSilent (YL) is only rated at about 37 CFM.

I went to the link you have shown, If you look at some of those numbers you will notice that they have fans with more CFM, also larger cases which are helping them. Some of the number I just flat out don't believe. When I posted last night after digging up the info on your fans all I could think is they are trying to built a silent system but they are doing it at the expense of heat build up in the system.

Right now I tried using a Tri-Cool fan as my Megahalems HSF, its rated at 79 CFM on HIGH. I am maintaining 39/36/41/35 at an ambient of around 20-22C. Sooo loud for my silent-minded system....but its making a noticeable idle difference (maybe around 4C based on my observations using the Nexus fan)

Are you saying that my temps are ok to do some mild OCing? maybe around 3.4-3.6 or so?

In the end I look at computer cooling like golf, you can't compare your game/equipment against someone else's. Lower scores are better and you are competing against yourself for the best score, some just lie about their score.

well said, well said. I guess though, some people would be unsatisfied if they had to question the consistency of manufacturing of their clubs

i'd say it looks like a lapping would do you good unless you pulled the heatsink off weird or something and it gave you that weird spread on the paste...that said, your temps aren't too bad for the fan you have, i wouldn't worry personally. What do your load temps look like?

Well I'm still wondering if this is a defect on the Megahalems? that is key.... if it is, I can simply return it and get a new one.

Load temps, as shown in my first post were max 71/69/72/67, but stayed -2C below that for the most part.

Looking at your case manual, it says it came with tri-cool fans (Top/Rear). Assuming? you changed these out for a quieter system? It also says you can add up to 3 additional fans in the front of the case, you say you only have 2 case fans (pics show top and rear). Have you considered putting some fans in the front of the case to help air flow?

Yes, I did change them out. I have considered putting fans in the front, but have not tried it yet.


I guess so far in summary, I've learned

There will always be a trade off between good cooling and a quiet system.

All my fans dont exceed 40CFM.

Rear case fan: Nexus Realsilent 36.8 CFM
Top case fan: Scythe S-Flex SFF21D 33.5CFM
HSF for MH: Nexus Realsilent 36.8 CFM

I guess just not enough air flow/movement?

Open concern: Prolimatech Megahalems heatsink base, is it designed to have a slight convex shape for the 1366/1156 version?
EDIT: This website also noticed the very slight convex shape: http://www.frostytech.com/articleview.cfm?articleid=2481&page=3
"The nickel plated copper base of the Prolimatech Megahalems heatsink is perfectly flat in one axis and very slightly convex in the opposite (it's nearly flat). The metal has been machined very smooth, surface roughness is ~4 microinches or better, which is excellent."

 

[CgS Drone]

Some believe one way some believe the other way about air in the case. Personally I believe the exhaust fans should exceed the fans putting air into the case in the manner of CFM. My belief is this way it will pull the heat out of the case better.

There will be a trade off for the sound vs CFM/RPM, Some fans are quieter than others and I'm sure you will get some good advice about what fans to consider, as well as people saying to use a higher static fan for the CPU cooler but I will leave that to others with more experience.

 

[m0r7if3r]

Some believe one way some believe the other way about air in the case. Personally I believe the exhaust fans should exceed the fans putting air into the case in the manner of CFM. My belief is this way it will pull the heat out of the case better.

There will be a trade off for the sound vs CFM/RPM, Some fans are quieter than others and I'm sure you will get some good advice about what fans to consider, as well as people saying to use a higher static fan for the CPU cooler but I will leave that to others with more experience.

Well, the yate loon (which is what that nexus is, it's just rebranded) he has is pretty good in terms of quiet performance on a budget. I'd say put the sflex on the cooler though and see how that works, it might perform better.

as for the concavity, what you really need to do is compare it to your cpu, cpus are often a little convex, so I think that hsf mfg's set their spec to have a flat-concave base rather than convex-flat if that' makes sense. The problem with this is that often they do not match up with the cpu, it looks like your cpu is more convex than your heatsink is concave. The way that you can correct this is with lapping. Check your cpu, if it's not flat then a new cooler won't help you at all. I'd say that's where part of your issue lies. You can try contacting prolimatech, but I bet they'll just tell you that it's within spec and they can't do anything...if you're not too attached to the writing on your cpu, lap it and the hsf (if you are, take a pic first )

 

[CgS Drone]

Right now I tried using a Tri-Cool fan as my Megahalems HSF, its rated at 79 CFM on HIGH. I am maintaining 39/36/41/35 at an ambient of around 20-22C. Sooo loud for my silent-minded system....but its making a noticeable idle difference (maybe around 4C based on my observations using the Nexus fan)

A top end cooler is going to perform its best when the cpu is being pushed, one way to check and see if you only need to go to a better cpu fan would be to load test the system with the Tri-Cool and see what your temps are. The idle's I really don't consider important and stock coolers can stay close to what a high end cooler will do at idle, its only when the pressure is put on that they shine. If you test it and the load temps drop like you want then you would be looking at only having to change that fan. It wont cost you anything to try it.

 

[burebista]

Open concern: Prolimatech Megahalems heatsink base, is it designed to have a slight convex shape for the 1366/1156 version?

Yep, it is by design.
TBH you're OK. 70C on load with a 1000 RPM Nexus fan looks perfect plausible.
You know silent and high-end rigs are not so good friends. At least in load and with some OC.
Anyway try an OC but keep Vcore as low as possible. Also keep distance to TJMax>20 in daily use and forget about core temperatures and enjoy your rig.

 

[CgS Drone]

Fireflame, m0r7if3r put me onto these and I believe its a case of "Having your cake and eating it to" clicky--->Gentle Typhoon 1850<---clicky.

These fans are QUIET. I bought two for my wife's computer and could not believe how quiet they are. Don't let the 28db fool you or the 1850 rpm, Did I mention I cannot believe how quiet these things are?

I don't think you have enough fan. Prolimateck suggests 57 CFM, (as minimum I assume). I just ordered a Megahalems with a GELID FN-FW12BPL-18 PWM @ 76 CFM.

As you can see in the specs on the fans they are 58.3cfm, Oh and if I haven't mentioned it I would like to say they are VERY QUIET.

m0r7if3r posted a very good link over in another thread so you can go to video's to hear the difference between a bunch of fan. Fan noise video

 

[Raypundo]

To check to see if case air flow is an issue just leave the case open on the side, and maybe a house fan blowing on it too

 

[fireflame]

Fireflame, m0r7if3r put me onto these and I believe its a case of "Having your cake and eating it to" clicky--->Gentle Typhoon 1850<---clicky.

These fans are QUIET. I bought two for my wife's computer and could not believe how quiet they are. Don't let the 28db fool you or the 1850 rpm, Did I mention I cannot believe how quiet these things are?



As you can see in the specs on the fans they are 58.3cfm, Oh and if I haven't mentioned it I would like to say they are VERY QUIET.

m0r7if3r posted a very good link over in another thread so you can go to video's to hear the difference between a bunch of fan. Fan noise video

Did you say they were quiet? ...haha

You can consider me a lurker around the OC.com forums....my first and foremost goal was a quiet system, but my OCD self started upgrading/changing every component to as quiet as reasonably possible.

I will definately do reserach on these Scythe GT fans, but have you ever heard Nexus (YL) 1000RPM or Scythe D S-flex models? I guess I can head over the silentpcreview for that


Thanks for all the help, I will be posting my Prime95 # w/ high setting TriCool in a bit.

EDIT: Antec TriCool performance on the Megahalems: Prime95 for 1 hr w/ tricool on high were 70/68/71/66. ambient was probably 1-2C higher than the original attempt)

I am surprised to see no major change in temperature between a Antec Tri-Cool on High 2000RPM @ rated 79 CFM, vs. a Nexus (YL) 1000RPM rated at only 37 CFM. To make it completely fair (ambient temp is the only variable I am not sure about), I will run the test once more with the Nexus.

EDIT2: I tested the Nexus again on Prime95, didn't wait for an hour, but back to back against the TriCool on high, the numbers are very close 71/69/72/67.

For one degree of temperature rise, I would gladly stick with the Nexus!!!

Can anyone help explain the results?
79 CFM vs. 37 CFM. --- > 1 degree core temperature difference under full Prime95 load?

 

[CgS Drone]

I will definately do reserach on these Scythe GT fans, but have you ever heard Nexus (YL) 1000RPM or Scythe D S-flex models? I guess I can head over the silentpcreview for that

While I have not heard those fans and I am willing to bet with their low speed that they are quieter than the GT. I am looking at it this way, the GT will be hidden away inside the case and not at the openings like the other fans will be so this should hide its VERY low noise.

When you look at the video of the GT remember he is recording that sound 1" away from the fan and in the open. I had my fans out of the case and in the open and me and my wife could not hear them past 12" but some are more sensitive to sound. I'm betting that you will be happy with a single GT on the heat sink, plus once you know that it is moving enough air at full speed to keep the cpu cooled to the numbers you like you can always go back in and enable Smart fan on the mb to slow them down when the system isn't being pushed hard even though I don't think you will need to for noise.

I am basically trying to let you know of the quietest fan that I know of that will supply the cfm/static that you need for the cpu cooler to perform at its best.

 

[fireflame]

Well said. I may give the 1850RPM Gentle Typhoon a try, but on my previous post, I showed only a one degree improvement from the TriCool vs. the Nexus. That was very unexpected.

 

[jevery]

I ran an experiment to compare the effectiveness of increasing the airflow across the CPU heatsink to increasing the airflow through the case. My case has a 3 speed controller for five 140mm fans and a PWM fan on the CPU sink. Increasing the speed of the CPU fan from 1100 to 2200 RPM only made a 1C difference at idle, (case fans on low), but increasing the speed of the case fans, (with CPU fan at 1100), made a 4C difference. I'm surmising that, at least at idle, the airflow across the sink is sufficient so that the limiting factor is the mass and fin area of the sink. I'd imagine that under load CPU fan speed would affect more of a difference.