Trying to install Thermalright True

I still plan on testing at some point as two of my cases have top and rear exhaust fans (Antec 900 and 300) - and I swear the claims I read were also in such a case with a top and a rear fan...

Just a matter of finding the time

odd...and you think the top would pull more air though than the back would?

Hot Air Rises - and that has been my logic for using the "bottom-to-top" airflow in cases with top exhaust fans (like the Antec 300 and 900). Again, I haven't actually compared the two orientations in my PC's, but numerous people have suggested that front-to-back is more efficient with tower type heatpipe coolers in general (something to do with how the heatpipes work relative to gravity or gas to liquid states of matter).

The claims I've read on this very forum were on the order of 5*C or more - but I just haven't had any time to dedicate to testing this theory alone (been busy building PC's and recording music with said PC's )

Randyman... said:

Hot Air Rises - and that has been my logic for using the "bottom-to-top" airflow in cases with top exhaust fans (like the Antec 300 and 900). Again, I haven't actually compared the two orientations in my PC's, but numerous people have suggested that front-to-back is more efficient with tower type heatpipe coolers in general (something to do with how the heatpipes work relative to gravity or gas to liquid states of matter).

The claims I've read on this very forum were on the order of 5*C or more - but I just haven't had any time to dedicate to testing this theory alone (been busy building PC's and recording music with said PC's )

Click to expand...

scientifically speaking, cold air actually sinks, so when you're forcing air up, convection currents can't really form to let the hot air rise. Hot air also rises pretty slowly (well...cold air sinks slowly since we're being technical), since the differences in densities is incredibly minimal. Any forced airflow basically puts that out the window, or so i understand it with my very limited fluid dynamics...which leads me to believe that since the fan on the back is closer, it makes it easier to exhaust the hot air and helps pull some out of the sink. That said, some open-air testing would certainly resolve this, but i don't have a rig for that (nor do I have any TIM left, but that's an easy fix)

if you have a hot running graphics card and your hsf fans exhausting out the top i could see a potential issue w/ hot air off card being pulled through hsf.

j0rd said:

if you have a hot running graphics card and your hsf fans exhausting out the top i could see a potential issue w/ hot air off card being pulled through hsf.

Click to expand...

huh...that's an interesting point, would that heat be coming off the pcb or are we talking about the old, non-shrouded gfx cards and non-ref cards that come with the fans that don't exhaust out the back? I know on my 480 has a hole on the back of the pcb behind the fan that it uses to draw air in, but I'm not as familiar as i'd like to be with most other gfx cards...

i would think it depends on the card and air-flow or rig in question. I havnt done a huge amount of research on this, just remember reading about it in atomic pc mag. I guess if the card runs hot enough then heat radiating off back of pcb could cause issues or if you are using a hot running, non shrouded card perhaps. guess its a case by case thing but one to be aware of. i guess some rudimenterarry resilts coulkd be gotten by loading the gfx card hard for a while and measure temps w/ thermomitor or just touch it, if it feels hot then... If the hsf if mounted to pull air vertically then i guess a test w/ card loaded and one w/ it idle should at least give an idea if its an issue but its not somthing im in a position to do otherwise i would be all over it. or just stick hand between gfx card and hsf wilst gfx card is @ load, if air feels warm running through fingers then...

edit: atomic pc mag (australia) did a build that was featured in the last issue, dual socket evga clasified. The left most processor was running allot hotter due to the heat coming of the gfx cards (though there were 4 of them). Ultimatly it limited the max OC of cpu's. will see if i can find link to article.

edit2: short version of article here. If anyone wants i can copy relevent parts of print article that mention the gfx / cpu hgeat issues and numbers to match.

last edit: the article linked was pre-OC. The print version has post-OC results in it.

well, my 480 has a pcb temp sensor and the pcb temp definitely increases with loading, i can get some numbers on that if you're interested...i just wonder how much of that conducts through to the other side of the pcb to radiate off the back...

my bad. my apologies to iskatefast's.

back on topic...now

muddocktor said:

OK guys, if you want to go off on a tangent, start your own thread and quit hijacking iskatefast's thread here, which is on fitting his TRUE on his system.

mudd

Click to expand...

Mudd,
It's okay. Besides the fact that I am still learning and didn't know that I could spin the True...I was thinking that with 3 fans pulling hot air off the video card would help. For the breif time I was able to run Prime95 before getting the True, temp #3 on speedfan was very high. So it was interesting to read the added discussion!

So last night I did the spin thing. Looks like I should have plenty of room. I never thought of changing the orientation of the bracket...makes sense. I still have alot to learn.

And now I have another cooler on the way...Coolit Vantage..if it really was in stock.....what to do, what to do!! I guess I could test both. Or build another rig..

Glad to hear that you figured out how to mount the TRUE in a vertical position man. And yes, definitely try out the Coolit Vantage as that is a heatsink I haven't tested and the TRUE is still a leading performance sink, so that will be interesting to see how well it performs compared to it.

I'm just so glad that Thermalright came out with the revised AMD mount that will let you mount it in any orientation. Last time I tried a TRUE out on an AMD system was before they came out with the new rev mount and it would only let you mount in 1 orientation and if the socket was oriented wrong you were just screwed. Whatever you think of Intel (there are folks here who absolutely hate them), you got to hand it to their engineers for using a square mounting pattern instead of a rectangular one like AMD for heatsink mounting.

muddocktor said:

Whatever you think of Intel (there are folks here who absolutely hate them), you got to hand it to their engineers for using a square mounting pattern instead of a rectangular one like AMD for heatsink mounting.

Click to expand...

Or you can just call the heatsink manufacturers cheap for not making the AMD a front-back mount in the first place. Seems like that should be the "standard" with up-down as an add-on ...

FWIW orientation doesn't seem to have an effect on cooling (I had to stop by the Studio on the way home to pick up my DAW PC and did the test last night!). So go with which ever way works best for clearance. Glad to hear you can rotate on AMD as well!

Thanks to all that replied..got everything back together and started some testing..
I was able to get to 3.996 MGHZ by just using the Asus tuning software. Ran prime 95 for a about 2 hours and never saw a temp over 42c!! I was very surprised at the 3.9 by autotuning..wondering if I can push it higher manualy?

I'm sure you can, if you can do that with the autotuning software. I say go for it!

muddocktor said:

I'm sure you can, if you can do that with the autotuning software. I say go for it!

Click to expand...

I'll try and get some screen shots up later. I have been running CPUZ, speed fan, and hardware monitor just to get a look at temps..anything else you need to see shots of?

You might download and run OCCT and use the Linpack option and see what temps do and if she's stable or not. OCCT/Linpack is much more intensive than P95 in my experience. Just run a default 1 hour session at first.

muddocktor said:

You might download and run OCCT and use the Linpack option and see what temps do and if she's stable or not. OCCT/Linpack is much more intensive than P95 in my experience. Just run a default 1 hour session at first.

Click to expand...

my experience is that prime will yeild higher temps over time, but to jack the temps up as fast as you can to within 2C of the maxes, linx (IntelBurnTest, Linpack, w.e) is the best...linx is my quick and dirty, prime is my long term stabilty...just my 2c, nothing to back that up but my own experience

muddocktor said:

You might download and run OCCT and use the Linpack option and see what temps do and if she's stable or not. OCCT/Linpack is much more intensive than P95 in my experience. Just run a default 1 hour session at first.

Click to expand...

Finally got some time to test....Ran OCCT for an hour,got some errors...Which images would you like me to post...