Announcement

az · 18 January 2004, 07:23

I would blow cool air onto the sinks, that way you ensure the cool room air doesn't enter the cabinet at places where it's not needed (i.e. not where the heatsinks are).

AZ

**TransformX** · 18 January 2004, 08:34

I'd say, dismantle the heatsinks and put them by the window to keep them cool, and burry the noisy part under a blanket or something

**rugger** · 18 January 2004, 08:37

Would it be possible to use just put holes in the top of the cabinet and use convection currents to keep it cooler?

**rylan** · 18 January 2004, 11:38

Fan at the top of the cabinet exhausting the air with an intake vent near the borttom or near the equipment would be good, and you wouldn't need 2 fans then.

**Brian R.** · 18 January 2004, 14:47

what rylan said plus a duct from the intake vent to the tops of the heat sinks if possible. Duct the cool air from the intake duct directly onto the heat sinks. Make sure your exhaust fan(s) are blowing alot of air out the top directly up if possible.

**Wombat** · 18 January 2004, 15:07

Originally posted by Brian R.
what rylan said plus a duct from the intake vent to the tops of the heat sinks if possible. Duct the cool air from the intake duct directly onto the heat sinks. Make sure your exhaust fan(s) are blowing alot of air out the top directly up if possible.

If you do something like this, you really need to put some kind of air filter in place. A $5 car filter from the auto shop will do the trick.

**ZokesPro** · 18 January 2004, 19:15

Exhaust fan on the top and intake fan on the bottom insuring MAXIMUM cooling. (hot air rises and cold air descends) The air will circualte and the rest is up to the heatsinks. IMO, should be plenty sufficient.

**TnT** · 19 January 2004, 00:11

Re: Physics question!

Originally posted by RhinoZ
I thought the fan on a cpu heat sink blew air into the heat sink, I just found out it's the other way around. But I would think it would be more effeicent to blow cooler air into the heat sink then removing the heat? No?

From what I've read heatsinks with fins work best with air being blown onto them; heatsinks with pins work best with air being sucked from them.

**moreau** · 19 January 2004, 08:57

The curl of the electric field due to a stationary point charge is zero, always.

Oh, wait, I though you had a physics question...

What you really have is an engineering question... and unfortunatley for my social life I am not an engineer.

I can address, however, a few things you've read here...

From what I've read heatsinks with fins work best with air being blown onto them; heatsinks with pins work best with air being sucked from them.

From a physics standpoint, this shouldn't be true... (you are looking at time reversal of the same system). If there are measurements indicating that fan direction matters (which I have also seen around the net), then this is not only going to be heatsink dependent (pins vs. fins), but also environment dependent (where are the cold air intakes and hot air exhausts).

If you do something like this, you really need to put some kind of air filter in place. A $5 car filter from the auto shop will do the trick.

I would suggest that an auto filter would be a bit high in impedence to serve here. the fans we are talking about don't have that much oomph. Look in any electronics catalog, as well as many computer specialty shops, and you will find fan filters designed for this purpose.

Would it be possible to use just put holes in the top of the cabinet and use convection currents to keep it cooler?

Exhaust fan on the top and intake fan on the bottom insuring MAXIMUM cooling. (hot air rises and cold air descends) The air will circualte and the rest is up to the heatsinks. IMO, should be plenty sufficient.

In a living room or outside, this is certainly true. And a case with a whole on top and bottom and no fans, also true. But, in all of those cases there is one thing in common: lots of volume for the given heat load and also a return path. Your living room in the static case certainly does have thermal gradient from top to bottom (top is hottest, bottom is coldest). And, if you have a really hot thermal source (radiator, stove, etc), you can even set up convection currents in the air. But... if you have a sufficient heat load you can actually heat the air to the point that there cannot be a sufficient temperature gradient to set up any convective flow. What that heat load is, and what the steadystate temperature that produces, I cannot say, as it will depend on the specifics of the room in question. However, in the case of a completely closed equipment chassis, it doesn't take much heat load at all for all of the air inside to reach thermal equilibrium. Now, if you have a whole in the top and in the bottom with no fans, you will have a glorified chimney. In order for this to work, you must have a return air path. In this case, it is a line that circulates around the outside of the chassis, starting at the top whole and terminating at the bottom hole.

If in the above example, you wanted to put in fans to assist the natural convection, you certainly an. This is forced convection. But it is likely that once you put the fans in place, the forced convection would dominate, and you could put in and out as mentioned above or front to back with equal efficacy.

Anywho... I wasn't trying to be a jerk or anything. I was just having fun. I am a physicist and I don't play one on TV. People don't seek out physics answers as often as you might think. I've considered renting a storefront and hanging out a shingle, as they say, but I doubt I'd need a waiting room.

Maybe I'd get a lot of traffic if it was "Physics Solutions & Chainsaw Sharpening". Hmmm....

**Drizzt** · 19 January 2004, 09:21

If noise is your problem, I don't see how adding two fans to the cabinet could solve the problem...
Better, it should surely solve the warm problem, but then you will find that you have to put the first cabinet into a second cabinet because it's too noise, then you discover that you have to put two fans at the outer cabinet because the inner cabinet it's too warm, but then you have to put the outer cabinet into another cabinet and so on

So: if the noise is made by the fans of the "guy", try changing the fans with better and noiseless ones, maybe simply changing the alluminium heatsinks with something better and using slower fans.
If the noises are made by the "guy" himself...uhhmmm...noise insulation isn't an option? At home I used "egs boxes" to reduce the noise made by my PC...

**TnT** · 19 January 2004, 13:19

Originally posted by moreau

From a physics standpoint, this shouldn't be true... (you are looking at time reversal of the same system). If there are measurements indicating that fan direction matters (which I have also seen around the net), then this is not only going to be heatsink dependent (pins vs. fins), but also environment dependent (where are the cold air intakes and hot air exhausts).

Now that I thought about it again, the suck/blow issue is more based on how the HS is designed using fins or pins by the HS manufacturer.

**RhinoZ** · 20 January 2004, 20:55

Wow Thanks guys for the all the suggestions and theory.
Sorry I'm a bit late on the return reply, I never had a good chance to read up. So it's looks like I'll just add one fan to blow cooler air into the cabinet. I can't make a hole on the top of the cabinet as other stuff is on top of the cabinet, so it's a nice gentle breeze for the heatsinks.
Also I just find out I'm not the only one who thought the cpu fans blew air into the heat sink.
Thanks again.
R

Announcement

Physics question!

Physics question!

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment