Anyone tested CPU temps with passive vs. active cooling?

[Ken90630]

There's a new thread today here on this forum about how quiet some retail PCs are (Dell, et al.). The issue of noise vis-a-vis active and passive CPU cooling came up as a result. I posted a semi-related question several days ago about CPU cooling ducts. Unfortunately, that thread got sidetracked and I never did really get an answer. :-(

Anyway, I was wondering if anyone here has ever tested CPU temps with active vs. passive cooling. In other words, active as in a heatsink/fan combo cooling the CPU, and passive as in one of those plastic ducts over the CPU/heatsink [w/no HS fan] that acts like a vacuum and merely sucks hot air away from the CPU and channels it to an exit fan (sometimes via a sealed tube).

Member Cerb mentioned something to the effect of passive schemes like this resulting in temps running "20 C" hotter than active schemes with a heatsink fan. As far as I know though, I've never heard of Dell having any problems with machines dying early due to constant excessive heat (in fact, system longevity is one of Dell's known attributes, isn't it?).

Has anyone ever done a test with both methods on a particular CPU to see which really results in lower CPU temps?

Just curious.

Ken

PS: Before you ask, I don't work in the computer biz and don't have test equipment. If I did, yes, I'd do the testing myself!

 

[Connoisseur]

While I haven't done this testing myself, I have read several articles a while back testing these methods. The thing is, an active cooling solution where are is blown directly onto the heatsink will generally result in much lower temps than the sleeve solution. The CPU fan that's directly on top of the heatsink will blow much more air onto the HS and thus increase thermal dissipation as opposed to a case fan that's several inches away from the actual HS. Furthermore, blowing air onto the heat sink is apparently more efficient than sucking air off of it (I'm no thermodynamics expert so don't ask for an explanation). I forgot who, but a review did test the cooling effectiveness of a CPU fan sucking air away from the CPU. The reviewer first made a custom skirt for the HS so that the CPU fan sucked air directly off of the HS and not just the ambient air. Without the skirt, vacuum effect resulted in much higher temps. With the skirt, the cooling effectiveness was about the same as the blowing solution.

 

[Cerb]

For one thing, they are not entirely passive: there's just one fan cooling it all, but with the shroud, it's the CPU fan. I don't know of any ducted tests right now, but am looking forward to seeing some using the Hyper6 or HT-101.

Dying of heat. You don't have a profile, but I take it you don't live in the south. It happens. HP and Compaq PCs have always been worse about it, but it happens. One thing to remember, however, is that OEMs also have the option of hundreds of hours of testing and tweaking, with budgets we don't have. That is wuite simply their biggest advantage--until it happens to be in a 2nd-story bedroom in the summer .

If I had a A64, I'd be seriously tempted (oh hell, I'd spend the money in a heartbeart!) to get a heatpipe cooler or the new Alpha and try it out--PSU fan for case, CPU ducted out the back. Given the current running of AMD CPUs and their future, I definitely plan on using such a setup in the future (though may blow in through the CPU heatsink), but unless somebody else tries and finds what works, it'll be blind trial and error.

 

[Ken90630]

Thanks, guys. Great info. I'd really like to see that test you refer to, Connoiseur. If you ever run across it again, PM me or something with a link. What the reviewer found sounds really interesting. Do you happen to remember what CPU was being used for the test?

Cerb: I live in Southern California (not far from Disneyland), so I can't really relate to the heat/humidity you sound like you have to deal with. It gets hot & humid here a few months of the year (mainly June thru August), but the weather isn't really extreme enough to significantly affect computer temps one way or the other. Of course, I have very primitive equipment right now (a Celeron 333 system w/192 MB of RAM -- stop laughing) and don't get high temps in my case anyway, so I'm not the best judge of this. Even once I upgrade this system and maybe get another as well, I'm still gonna have modest stuff compared to real "power users."

What I'm trying to figure out with these duct/no HS fan cooling schemes is whether they're done 'cuz they cool better or simply because they're cheaper to manufacture/build than a HS-fan-only cooling scheme. I'd think a simple $5 HS fan and no duct-to-exit-fan apparatus would be cheaper and easier to build, so it only stands to reason then that these retailers like Dell, Compaq, Gateway, etc., would use the duct set-up 'cuz it cools everything better and thus allows the PC to last longer without failure, in-warranty repairs, and customer calls to tech support. If it doesn't cool the CPU & case better, then why would they use it if it isn't cheaper? Then consider what Connoiseur said re the review he saw. I admit I see no clear-cut answer.

You kinda hit the nail on the head when you say, "unless somebody else tries and finds what works, it'll be blind trial and error." It sounds like that reviewer did, but I'd like to see the full review. Maybe one of these days I'll buy some temperature-reading equipment, a duct apparatus, and try both. Yeah, right ... like I've got time for that.

Interesting topic, regardless.

 

[Cerb]

Cheaper, and more reliable--for relatively low-powered systems. The exhausting duct is absolutely great for IEG2 Celeron boxes. In fact, if you design your new box around low powered parts (passive GPU, Barton, Duron, or whatever the next one is), you could easily use something like the Alpha HSs and see how that works. Worst case there I'd bet an exhaust duct for the CPU and then the PSU exhausting heat would work fine.

But more power = more heat = more air to move.
If you have one HDD, a passive video card, and one big CPU HS...it's not big deal.
But when you have a 9800 Pro, some HDDs blocking the front (or in a place where they need extra cooling help), OC'd CPU...well, big heatsinks and many slow fans beat out one exhaust fan easily.

Hopefully after this semester I can manage to get the big J.O.B. so I'll have the money to spend on this stuff later. I'm excited about the possibilities offered by some of these new coolers, but will have to have fall-back plans, as I want to impliment things that aren't tried and true (cut can't exactly go w/o a PC for long ), and am willing to do a custom case if need be (I have one in the works now that I should be able to finish right after finals).

 

[Ken90630]

Yeah, I'm with you. The build I'm designing/planning right now will be centered around an Athlon 2500 or 2600 (Barton), with a 40 GB hard drive (just one) and probably a mid-level video card. No 9800 or OC'ing needed for this machine! This particular machine is primarily gonna be used just for Web surfing, e-mail, and some occasional, lite MS Office stuff. But I do want it to have enough oomph in case I decide to increase the demands on it a bit in the future. I'm conceptualizing an additional build for a bigger/better machine, but that's for after this one is done.

I may do some experimenting with cooling too, if I can find the time. Am hoping to be able to go back to school & finish my degree in the fall, so I'll need to do it before then. If you get a chance to do any cooling experiments yourself, by all means post the results here so the rest of us (particularly me) can learn from you.

Good luck with school and the career search.

Ken