ok, maybe I'm being foolishly optimistic, but why have a closed case at all?

Wouldn't it reduce the temperature of everything involved if, for instance, the motherboard, ps, and hard drive were all mounted at the bottom of an open mesh cage? Don't computer run cooler without their side door on?

Yes, the case is a major reducer of the noise the computer puts out... but from the perspective of making something entirely utilizing passive cooling, isn't the case just keeping hot air in and cooler room air out?

Before I go drilling more holes in my poor experimental system, what am I missing that makes this a foolish idea?

-Chris

A well ventilated case will perform better with the case sides on. By well ventilated i mean intake and exhaust fans. This is because an even airflow is maintained by cold air being pulled in and pushing the hot air out. However if you take the sides off the fans dont work anywhere near as efficiently 'cos air just 'leaks' out of the sides.

However if you are aiming for a passively cooled solution, then what you say is correct, an open case would work much better.

Well, not necessarily. In fact, probably not. Most really effective air convection cooling systems use some kind of tunnel-chimney scheme to direct the airflow over specific hot spots. Rising hot air can create a negative pressure so that cooler air is drawn in at the bottom of the "tunnel-chimney", and the concentration of airflow allows for higher velocity -- and cooling -- where it is needed most. In contrast, having all the components out in the open means that you can't direct the convection airflow in any way, and there is no way to accelerate the airflow at all.

Examine the Mac Cube and you'll see this is precisely the idea they used. Ditto the new iMac -- even though it has a fan. There is a row of intake holes just at the rim of the half-sphere, and a single exhaust hole at the top center, around the post fo the monitor. The one fan seems to be directly below the hole. (It's not THAT quiet but not bad.)

Yeah ok A chimney setup is the best, but when i said an open case is better, i meant better than a normal case, i.e one with a top and a bottom.

So, seishino, maybe try and get a cage effect at the top and bottom, but keep the sides fully in tact.

So that you won't drop a metal coathanger onto your motherboard where it accidentally shorts two contacts together that carry moderately high power and makes a little 'pop' noise and creates a little blackened burn mark thereupon. Not that I would know from personal experience or anything.

In a similar problem, note that some old power supplies (anything before ATX) had somewhat dangerous on/off switch arrangements. The line would come in from the wall, then be connected directly to 2 wires that went to a powerswitch on the front of the case. Often the power wires from the PSU to the switch on the front of the case were a bit carelesssly routed and flimsy and had easily loosened covering where they connected to the switch. They were in close proximity to sharp pointy sheetmetal, and to people futzing around with screwdrivers and god knows what else. They carried the full 120VAC line voltage and current, so if you , say, jammed a pair of scissors in your open case just wrong on accident, or had loose insulation near the switch and dropped liquid on it, then you would be getting the full shocking effects just as if you had done the same thing to an electrical socket. A sturdy solid enclosed case would prevent many such accidents, so there is another reason to keep it closed. Probably less of an issue with ATX.

Also there is the 'faraday cage' effect of metal around your box to prevent leakage of all sorts of EM signals going off interfering with other electronics components. I believe metal mesh will work for this, though I am no expert.

I recently undervolted the cooling fans on one of my systems, at which point it became slightly unstable. CPU temps were ok, but the memory (1066 rimms) were blazing hot.

I removed the side panels, which dropped the CPU temp by 1 degree C....but raised the ram temps by 2 degrees. The system was actually far less stable with the sides off (!)

After some further testing and probing around with a thermister, I came up with the following **general** conclusions. Any component cooled with a dedicated fan _and_ not in close proximity to a greater heat source than itself will be cooler in a caseless system. Passively cooled components (ram, chipset, hard drives, etc) and fan-cooled components close to a high heat source will likely run hotter though.

> "...I believe metal mesh will work for this, though I am no expert...."

Yes, a mesh or screen will work fine for EMI shielding.

What you want is one of these.

I ran all my computers with cases off for the longest time until I became annoyed by the noise. Now that I'm on a quest for silence, I've been monitoring key temperatures for a few months with various configurations. With the case cover off, the motherboard (as indicated by the motherboard temp sensor on my asus mobo) runs hotter than with the case cover on. I have to believe this is due to the fact that, with the cover on, the air is forced to come in through the front, and out through the back, as directed by fans.

With the cover off, the rear fan (for example) can suck it's air straight from the already cool outside air, and not draw the warmer air that resides around the various components into it.

I've also found that adding fans doesn't seem to help, in some cases. My overall conclusion is that this is a very complex issue, and you'd need heat-sensitive cameras and a lot of time to really come up with an ideal solution.

Placing a system on it's side, with mobo at the bottom, and all heat-generating components at the same level, would probably be more predictable in a case-off scenario. But you'd have to have passive cooling for this to be a silent solution - otherwise you don't get the benefit of the silencing effect of the case.

I would like to reply to DBRI & MASHER:

The case does offer EMI (Electro-Magnetic Interference) & RFI (Radio Frequency Interference) protection to ensure any extraneous EMI / RFI does not affect the computer internal components.

The case also serves to protect other sensistive electronics adjacent to the computer from the internal components that can generate EMI / RFI "noise" like the switcher PSU, CPU, etc.

Yes, mesh will do quite fine in lieu of the steel cover. Provided the holes are quite small @ around 1/8-inches square. Small holes will provide a "Faraday" sheild, very similar to those used in radio / TV transmitters, receivers and the like. The high frequency (over 200 MHz) will not "pass" through the small holes and will be "shunted" to ground.

Another item that will help against internal / extraneous EMI / RFI is a good line filter and/or UPS (some UPS have good internal line filtering).

Personally, I use a 10-ampere filter made by Tycor Electronics in Calgary, Alberta. Tycor guarantees their product will protect the computer from damages and Tycor will pay up to $5,000.00 in the event the filter fails to perform its task & a computer fails.

My Tycor filter protects my computer, my monitor, ADSL MODEM and my Linksys 4-port router. Other Tycor models protect the phone / modem lines. As the printer can generate noise, I do not plug the printer into the Tycor filter.

I have had only one Tycor filter fail in 10-years (severe lightning storm) and my computer was saved. I would say that I had good luck so far.

A little off the topic, but I thought you might want to know.

Good luck and good computing.

TerryW

Oops! Some other questions came to mind!

If there is no computer case or side panels, will there still be air currents around the temperature sensitive or heat generating components? Will there be the cooling "tunnel-chimney" effect that MikeC suggests?

The answers do follow, but for now a breif background on my related experiences:

I am employed by an electronics company in Western Canada and we had orders to deliver computer-related products into South America. We needed to check the relative temperatures of our electronics in a sealed enclosure in direct sunlight for 16-hours per day (minimum).

The outside temperatures could reach +40C in the direct sunlight, our product is in a sealed enclosure, the average temperature rise (over ambient) was around 35C to 40C, and the 2 fans operate 24/7.

We found that in a sealed enclosure, the temperature was quite high @ around 75-80C and the fans provided substantial air currents around the electronics. Even though the air currents are hot, they did still provide a cooling effect on the components (air was cooler than components).

Answer follows:

Therefore, I would hazard to guess that computer cases are designed to provide the cooling tunnel-chimney effect that MikeC discusses. Without the benefit of the case and/or side panels, there will be little to no cooling tunnel-chimney effect and the temperature sensitive or heat generating components could be damaged.

TerryW

"Yes, mesh will do quite fine in lieu of the steel cover. Provided the holes are quite small @ around 1/8-inches square. Small holes will provide a "Faraday" sheild, very similar to those used in radio / TV transmitters, receivers and the like. The high frequency (over 200 MHz) will not "pass" through the small holes and will be "shunted" to ground."

Cool, thanks!