P180 temperature observations

[keldog7]

Hi all,
I had planned to do some more benchmarking, but I'm out of time, moving PC to another building, and am SURE to lose the data I already have. Accordingly, here's a post for anyone interested about case temps and airflow in the P180.

First off, my sig should detail roughly what's in my rig. Keep in mind, that during the build, my main emphasis was QUIET performance. Overall, I'm pretty happy with the result. My sig doesn't explain the case fan setup, so I'll outline it:

My reason for doing this was that I found my PC to be suddenly unstable in its overclock, and all I had done was change the location of the rig. Specifically, it was moved much closer to a wall, with the back of the case about 4 inches from the wall. Also, the new room was 2-3 degrees warmer than the original. As a result, I was amazed to see my CPU / chipset and ambient temps all go up significantly - and likely causing the instability.

I'll post my temperature tests below, but first my case setup:

- Scythe Ninja HS has a Nexus 120 (orange) set to 12V on the "front facing" (towards the case front door) side of the heatsink. Airflow is obviously directed through the HS towards the back of the case. Because of the size of the RAM used, the fan had to be mounted "high" on the heatsink - as a result, only the upper 2/3 of the heatsink get meaningful airflow from this fan. Also, becasue the fan blows *sideways* (instead of downward like the stock HSF), the chipset surrounding the CPU doesn't get much airflow on my A8NSLI32

- The rear case fan is a Nexus 120 orange at 12V, blowing outward.
- The upper (top) blowhole fan is disconnected, allowing air to be pulled in the top, helping supply cooler air to the HS (see the www.silentpcreview.com review of the case for an explanation of why I did this)

- The HDA X-plosion is mounted low in the case, near the bottom-most PCI slot, and is unikely to affect temps / airflow. The 6600GT is in the top most PCI-E slot, with stock fan etc.

- The lower drive bay has both of the 2 hard drives, and the upper and lower bays have been sealed off from one another with the supplied accessories.
- Cabling is a bit untidy, but efforts were made to use rounded cables only, and I did my best not to impede airflow from the front of the case to the CPU area.

My sig has the original (cooler room) idle and load temps listed in it.
Ambient temps were taken from the video card's sensor.
LOAD temps are done with linux, using 2 instances of mprime, left running for at least 30 minutes. Unless specified, all temps are under full (100%) CPU load, in the new (warmer) room.
Temps are listed as follows:
Opteron CPU / Chipset / Ambient / GPU

1) New room and location, side panel attached, front door open
IDLE: 41/40/32/38
LOAD: 51/43/33/39

2) SilenX 120@12V (14dBA (hah!)/58 CFM) in front fan bay, side panel attached
IDLE: 39/39/30/37
LOAD: 49/41/32/39 (front door open)
LOAD: 51/42/33/39 (closed the front door)
LOAD: 47/35/28/34 (front door open, air filter removed, and covering grille open)
LOAD: 48/39/30/36 (front door open, air filter removed, and covering grille CLOSED)
LOAD: 49/39/31/37 (front door, front grill closed; air filter removed only)
LOAD 48/38/29/36 (front door and grille open, air filter installed)

I'll let folks draw their own conclusions - keep in mind that this was all done fairly informally. Still, I was very surprised to find the front grille had such an effect on the ambient temp in the case - its a very open grille, but changes the internal temp by 2 degrees! That pretty much translates to another 2-4 degrees added to the temp of my processor.

Anyway, this box is soon going to a room where the room temperature stays at about 17 degrees C, where I expect it to perform wonderfully.

Comments welcome as always.
-A

 

[Pens1566]

Idle temp seems a little high for an opty w/a ninja. My expert idles @24/25 with a 120. Even with the +10 adjustment for not knowing whether my DFI sensor is accurate, 7 degrees is a big difference.

 

[pulsedrive]

It is a given that with the air filters out there will be more air flow and hence, lower temps. But with out the filters, you get lots of nasty dust in the case, which you are actually aggravating because you are letting air in through the top of the case without a filter for the dust. The p180 runs very well, if you tape off the VGA duct holes, and either tape off, or use the top exhaust. In the end, the dust, will cause more problems than a few degrees of perfectly safe temps.

 

[keldog7]

Dust bunnies eh?... I challenge you to find some evidence to back up your claim. Don't take my suggestion the wrong way - I'm not trying to be rude, but I think the widely held belief that "dust will be pulled into the case, and cause [mayhem]..." is pure BS. Lets face it - most of us don't have our rigs in a woodshop... Just how dusty is your room right now? Its not like the air intake is at floor-level, pulling in stuff which has settled to the floor, either.
Personally, I think taping up the ducts is a bad idea...because it will force some air to be sucked in the small cracks - like the dvd burner slots etc...where even small amounts of dust CAN cause problems. I've vacuumed alot of dust out of *alot* of machines, and never found it to be the cause of many problems. In a few cases, where the environment was very dusty, dust *had* accumulated all over the CPU heatsink, inhibiting airflow - but this was the exception, not the rule, and took well over a year to occur, with 24/7 operation.

Finally, I think you may have missed the point of my post - the air filter *isn't* the big culprit here... The grille covering it seems to have had more of an effect on case temps than the filter!
-A