Sager NP8150 Overheating, unusual cyclical GPU activity, cleaned

All,

Purchased a Sager NP8150 back in March 2012, to basically be my HTPC. It sits under my plasma, and on top of cooling stand with fans. It's been an excellent workhorse for the past year, doing everything I could ask of it. I'm primarily a gamer, and it's been able to give me long hours of that.

Two days ago, while playing some Skyrim it suddenly shut down. I automatically assumed a GPU overheat, even though I couldn't hear my fans working that hard. So I ran catalyst control center and saw that my idle temps were 70 C. I was hitting around 90 C during activity. I figured I would need to clean it out. At the time, I realized that I hadn't updated my drivers in a while, so I updated them as well as CCC. After updating the drivers and CCC, I reran it, and saw no change. So I pulled it down, removed the fans and heat sinks, and gave them a good clean. (dust cloth, q tips, and a bit of rubbing alcohol)

Put everything back together, and ran it. My fans immediately started running on overdrive. Full out with no stop, then once it had booted, it calmed down. Opened up CCC, and everything started running overdrive again. In an almost cyclical fashion, my GPU would run full out, my temps would hit 90-100C, and my fan would be at 100%. Then it would slow down for a minute or two, and go at it again. Cycles started out at a frequency of a couple of minutes apart..then every minute...then full out until I shut down for fear of damage.

I immediately assumed that one of my fans got disconnected, or I had tightened the screws in the aluminum fan covers (overlays? not sure what they are called) too much and restricted movement. I pulled off the back and ran it. Nope, fans are running fine, no slowdown. However, I noticed the CPU heatsink is MUCH hotter to the touch than the GPU. I'm talking a factor of 2-3 times. I'm just not knowledgeable enough to know if that is an issue.

Not knowing what else to do, I rolled back the drivers, and I'm posting on here. Please help! I know enough to get into trouble, but not enough to know what my next step is.

I'm running an I7 with a Radeon 6990M, if that information helps.

 

Sounds like you've done everything correctly. What kind of thermal paste did you reapply to the CPU/GPU and their heatsinks. It could be a bad paste job causing the higher temps.

 

Arctic MX-2. And that could be, I followed a youtube video: it's not like I've done this often.

But would that have anything to do with it suddenly cycling from idle to full load for no reason?

 

If not done properly your temps could go right to the the threshold which tells the fans to kick on. I'd reapply it to see what happens. Here are application instructions for IC Diamond, I know you said you have MX-2 but the dot it the middle method described here is preferred Application

 

Thank you, I'll give it a try. What's the preferred method for removing the current thermal paste?

 

There is special cleaner you can purchase, but honestly, you can clean it with acetone nail polish cleaner, or some use isopropyl rubbing alcohol. Use a lint free cloth along with one of said cleaners, and make sure you don't some off onto the board.

You can see some application methods here. Feel free to read further for data collected when testing various compounds.

 

Hey, sorry for the delay in follow up, I had to wait for more paste to arrive. I removed the heatsinks, and cleaned off the gunk. Reapplied. idle temps are 45-47 C now. No cycling. Netflix viewing shows up at 57 C max, and Far Cry 3, maxed out, while blowing up is 83 C. The only thing I found odd was that at 83 C, fan was still only running at 30%. I'll keep an eye on it, but it seems fixed for now. Thanks for the assistance.

 

how did u monitor the fan speed? i guess the tool just gave u bogus numbers....

Sent from my Galaxy Nexus using Tapatalk 2

 

Most programs are pretty good about it, but it's possible.