Is the case the noisy component?

[WipeOut]

Ok, let me explain..

We are on a mission to minimise noise from our PC's but we put these heat generating components into closed boxes and then try an work out ways to silently pass air through them..

Now I know the case is there not only to house the components but also to block/reduce/minimise EMI.. How much EMI is there anyway and what does it actually do?? I know IT guys who NEVER have their cases closed up and as far as I know they don't have any issues..

So last night I was fiddling because it was a warm evening.. The ambient temperature was about 26C and using a food probe through the back of my case it was 33C inside the case at the top.. Thats a 7C difference.. So I got to thinking about how to get the inside of the case closer to ambient without making the fans run faster and making more noise..

So here is what came to mind.. What about a case made from mesh?? Something like the honeycomb design of the antec fan grills..

I figure I must be missing something, and its probably down to some regulation or other, because that would surely be the best way to silent computing.. The fan on the CPU cooler (if one was needed) would only have to spin really slowly if at all at idle and could spin a little more under load but it would be nothing like whats needed at the moment..

Does something like this exist?

[relitz]

A case made of something like mesh would let noise out so it's really only the better solution if you don't have any fans at all. In closed up cases we are trying to direct a little airflow through heatsinks, thus cooling the components.

[WipeOut]

Thats really the point.. If the PC was running at ambient temperature inside the fans, if any, would have to spin so slowly, if at all, that there wouldn't be any noise anyway.. If anything it would be fan bearing noise and so with the right fans even that wouldn't be an issue..

Its just a theory anyway, just wanted to see what others thought..

[pcy]

Hi,


7C above ambient isn't bad at all - if you try for anything under 5C the fan noise will be horrendous.


But braodly, the idea does not work. You need the fans even if there is no case at all.

The PC components heat up the air they are in contactt with; and if you have no fans, convection current from the temperature difference is simply too small to replace the air close to the PC components with cooler air (i.e air at room temperature).

If you forget the case, and simply build the entire PC onto a flat "tray", but with all it's fans attached it certainly does run cooler because the air has better access, so you can run the fans slower than if it was all closed up in a case. The fans are totally un-muffled, then, of course: and it's not clear whether the overall effect would be more noise or less.

Based on my own experience and experiments with the PaQ case I'd say that the noise level at the front would be higher at the front without the case, but lower at the back. So... sitting in the middle of anacoustic chamber the result would be a toss up, but in a typical installation, where noise from the front is much more noticable than noise form the back, the PC with a (good) case would be quieter.



As to EMI - if the PC is going to be used in a research lab or a hospital this might be a serious issue - in an ordinary home or office I doubt it matters at all.



Peter

[qviri]

There are generally two models for cases:

a) as open as possible, air comes in and leaves as convenient, fans are used to direct air over specific components
b) mostly closed, with airflow tightly managed, fans are used to control intake and/or exhaust, fans might be used for specific components

Historically, SPCR has gone for the second option, perhaps best exemplified in the MikeC-assisted-design Antec P180 and P150 cases.

Both options have their advantages and disadvantages. The closed model lets you do more with less fans; an exhaust fan can be ducted to a tower heatsink and also cool a hard drive placed in the intake path. On the other hand, if you have only one hot component, a single fan cooling just that might end up quieter. Closed cases can, if properly constructed, help with blocking/damping non-fan noise, of which hard drives are the most common example. Limited intake paths also make it easier to control dust.

It is very possible to mess up, from a quiet computing point of view, both models.

The open concept got a boost with the rise of picoPSU and the SSDs. If you choose your components carefully, your CPU might indeed be the only component requiring active cooling or otherwise creating noise.

[WipeOut]

Thanks for the interesting input..

I think a valid point is the selection of the components.. For example if running one of the extreme graphics cards then I see how the case would help massively to reduce the noise..

Maybe a bad example since gamers aren't "usually" too fussed with the noise the PC is making because the game is making more..

Perhaps the answer as many have demonstrated is a combination.. The way closed cases (eg. P180) are modified to improve the airflow by cutting out all the grills etc. is probably how you get the best of both worlds..

I can see I am getting to the point with my set-up now where some case modification will be the name of the game.. Don't think it will ever stop..

[jessekopelman]

In terms of just concern over EMI, mesh would be fine as long as it is metal. Another thing to consider is that EMI is clearly not that big a deal for most people as you see open cases like Antec Skeleton, as well as cases with whole sides made out of plastic/resin (neither of which do much to block EMI), and plenty of homemade cases made from wood (also does nothing much to block EMI).

[mbf]

As far as I understand it the case is supposed to block EMI coming from the inside of the case and not vice versa. I also remember having heard that open cases or even cases with windows are illegal in several countries because they do not properly contain EMI from the components (most of it comes from the CPU and nowadays the GPU).

[EsaT]

It is very possible to mess up, from a quiet computing point of view, both models.

For anything producing any airborne noise magpie's nest designs are at big disadvantage from the start no matter what you do: Closed case always muffles noise of components while "appartment house without walls between neighbours" is as quiet as the noisiest neighbour.
Because of muffling effect you could even run those fans at higher speed in closed case for same effective noise level so even in cooling efficiency per noise there isn't any guaranteed advantages for open/free airflow designs.

But there's one thing which can really spoil acoustics of any kind case regardless of cooling noises or lack of them: vibration.
HDD itself doesn't produce that much airborne noise but hard mounting makes case acts as loudspeaker amplifying noise greatly so that's why at least rubber grommets should be used in mounting for decreasing vibration conductance to case and case "hardened" against vibration by mass loading (which also helps to block airborne noise) if structure isn't already vibration resisting.

How much EMI is there anyway and what does it actually do??

Ever heard that beeping from loudspeaker just before cell phone starts ringing?
That's very good example of EMI at work... GSM transmitter itself operates at GHz range but there's still other sources like power regulating circuitry.

So good thing that most computer's components use various error detection/correction methods with EMI/RFI protection being very important design concern.

I also remember having heard that open cases or even cases with windows are illegal in several countries because they do not properly contain EMI from the components (most of it comes from the CPU and nowadays the GPU).

EMC compliancy is requirement of EU laws for CE mark... but as it apparently talks only about "electronic devices" makers on plain cases probably use that loophole... if they actually do any testing!
But as whole computers are required to comply which would no doubt cause lot of bans if actually tested. Here in Finland PC using Akasa Eclipse got such ban because of leaking out EMI.

[jessekopelman]

Ever heard that beeping from loudspeaker just before cell phone starts ringing?
That's very good example of EMI at work... GSM transmitter itself operates at GHz range but there's still other sources like power regulating circuitry.

I find it is GSM phones operating in the sub GHz range (eg 800/900MHz) that tend to interfere the most with speakers. These phones will often distort CRT displays as well (when transmitting). It is interesting to note that in the US, all radios (and almost all electronics) must be certified as not interfering, by the FCC, before they can be sold, yet pretty much every GSM phone I've ever seen exhibits this behavior at 800MHz . It is also worth noting that this issue is very specific to GSM. That it doesn't happen with CDMA is not surprising (since that technology is spread spectrum), but it also doesn't happen with Analog or IS-136 (aka TDMA, very similar to GSM but with smaller channel size). Because the devices pass FCC testing and the interference is isolated to very specific types of devices, my guess is that the interference is a result of intermodulation between the emissions of the phone and something in the interfered with devices (probably related to the powering of the devices). Intermodulation is not something you can generically test for, since it depends not only on specific devices but also their relative positioning. While intermodulation is certainly the result of EMI, but it is usually considered as a separate issue in terms of whether something is adequately shielded.

[sonic6k]

While on subject of this "beeping"... yes I've heard it million times. Can this cause any harm to the speakers themselves other than just make an audible beep beep beep?

[jessekopelman]

While on subject of this "beeping"... yes I've heard it million times. Can this cause any harm to the speakers themselves other than just make an audible beep beep beep?

Probably not. I'd only be concerned if you have the volume cranked up near max (prolonged usage of which might well break poorly engineered systems in and of itself).