[lordz]

Hey guys,

I have a Q6600 G0 stepping with the stock HSF running orthos blend @ 63-65 degrees C (Core Temp 0.95b). I've reseated the heatsink and even used AS5 following the website's instructions. Edit: It idles at 45C. I've now seen it push 71C; I got scared and stopped the burn-in testing.

My motherboard is the crappy ECS 671T-M that came bundled with the CPU. Vcore is 1.248....I first noticed this when I was gaming and the board started beeping then eventually shutdown my computer. All fans are operational and my room is pretty chilly 65 degrees F.

Could somebody spare some insight into my problem?

Thank You.

[blackersabbath]

I'm thinking that your motherboard is reading the temps wrong. They seem a little high. Did you seat the heatsink correctly?

[lordz]

Honestly, I really dislike the snap-on stock HSF. I am certain that the HSF is seated correctly...I heard satisfying clicks for each leg I pushed in. I have not gone as far as to remove the motherboard and visually examine the underside though. I did a finger check + visual inspection to make sure the legs were flush to the board. I'm about to remove my old scythe ninja plus from my old rig and stick it on to see how that will work out.

[donuts]

It should work out much better. I've got one on a B3 quad and under load I'll see low 60's. I never did complete load testing with the stock HSF because of the temps.

[xilix]

I never saw anything above 58c with the stock cooler at stock speeds. Your ambient temps must be high. Contrary to common belief, the stock cooler is more than enough to do the job, and even handle mild OC's. If temps are high then something else is playing into the equation.

[lordz]

Originally Posted by xilix

I never saw anything above 58c with the stock cooler at stock speeds. Your ambient temps must be high. Contrary to common belief, the stock cooler is more than enough to do the job, and even handle mild OC's. If temps are high then something else is playing into the equation.

If that is the case with all four cores fired up at 100%, then I probably seated the heatsink incorrectly. Hopefully, that's the only thing that's wrong with my setup. Thanks for that information!

[jason4207]

How much TIM did you use (or did you just use the stock stuff)? That could possibly be an issue. I doubt the ECS has the option, but the Gigabyte board should allow you to lower the Vcore which can help temps significantly if you're still on the stock cooler.

[lordz]

Originally Posted by jason4207

How much TIM did you use (or did you just use the stock stuff)? That could possibly be an issue. I doubt the ECS has the option, but the Gigabyte board should allow you to lower the Vcore which can help temps significantly if you're still on the stock cooler.

I used AS5 sparingly. ECS board does not let you set voltages of any sort unfortunately.

[AlabamaCajun]

Temps may be off 10C. If it's that hot then you should be able to feel that feat just by putting your hand nead it.

Some notes on that board. I have the AMD versions but they all seem to follow one pattern.
The quality of components is good not what High end boards are using but good caps and other parts.
You won't get much OC from a high end chip, the VRegs are the standard 3 phase but it may handle enough to get to 3.0.
BIOS's are not that great, you should have a limited but usable VCore range but may not get any VDimm.
Ram settings, you may have to let the board decide about settings above a certain FSB depending on the dividers. On auto they do a descent job of relaxing lats as speed increases but be sure set the DRQ learning on (Centers DDR signaling for optimal stability).

Later if this board accepts a Penryn then it will make a nice HTPC of Folding farm.

[AlabamaCajun]

Originally Posted by lordz

I used AS5 sparingly. ECS board does not let you set voltages of any sort unfortunately.

Look for VID in the advanced section.
Edit; saw that's the sis(sy) chipset, little chance of a VCore setting.
The Gigabyte you get will own this ECS.

[Liquid3D]

Originally Posted by lordz

I used AS5 sparingly. ECS board does not let you set voltages of any sort unfortunately.

Arctic Silver has a very specific application method for Quad Cores with AS5.

You simply apply a thin line running from left to right which effectly covers the cores, then use the mounting pressure to compress the paste into both surfaces.

The push pins on the stock Intel heatsinks are the worst. You cannot even say their ok solong as you don't replace your CPU often because they are difficult to mount properly. Eevn trying to distribute the pressure evenly by depressing pins perpindicular of each other still pops another up. I've yet to install a stock heatsinks without problems.

For my TIM round-up I removed all four pins and dropped the heatsink on my water-block through mount system. It wqas the only way I could ensure even pressure and remove the heatsink as many times as I did (3 mounts for each TIM taking the lowest temp from each).

If you don't have the money to get a new heatsink right now, I suggest going to the hardware store and getting four phillips head bolts and 8 nuts. Get some plastic washers, and there's a way to remove the pins from the heatsink without breaking them. I mention this because I belaive warranty requres sending back CPU in original box with heatsink.

EDIT

[AlabamaCajun]

I just got my 6600 today and used the metal brackets with the Ninja that bolt through to a x bracket on the back. It makes a solid mount on the intel boards.

[Liquid3D]

Very smart, contact is everything.

[lordz]

Well, I got my DS3R and boy does it run rock solid. Runs my hardware configuration much smoother than the P5N-E SLi or the ECS board. My temps have also dropped to about 54C load on a stock HSF. Really odd, but it probably was a contact problem.

[jason4207]

Liquid3D - TAT only loads up 2 cores, correct? Shouldn't you be using P95 or something to stress? I see 2 of your cores around 50*, and the other 2 in the mid-upper 60's.

[Liquid3D]

Yes your right I had to change my test results mid-stride in a TIM round-up. The wird thing was when I first began testing a few months ago I found TAT was seemingly affecting all four cores.

I had switched to Prime95 v.25.5a when I found anomalies half way through. This cost me a month of ruined data. But even with Prime95 cores will run at slightly different temps.

[Liquid3D]

I know a 4C differnce isn't much compared to a 10C differnce but in many of the tests the results aren't so far apart running TAT on Quad? While it may not stress all four it does register them in it's scrolling results? Nonetheless your absolutely correct and its better safe then sorry. In fact I had began a TIM test about three months ago and if it weren't for you I wouldn't have gone back to check the results. I began testing with TAT and switched, but left the TAT info in there