General Advice for Custom Case

Got a shopping cart of parts that you want opinions on? Get advice from members on your planned or existing system (or upgrade).

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

dj.pihat
Posts: 3
Joined: Sun May 14, 2006 6:08 am

General Advice for Custom Case

Post by dj.pihat » Sun May 14, 2006 6:46 am

Ok, I've got a few questions for the masses, but I figure I might as well preface them with a little information. First, I'm looking to build a silent computer that will be used as a print, web, and ftp server (as well as possibly gathering network stats or other little widgets). I have an old MS-6368 motherboard (Micro-ATX with onboard video, sound, network...pretty much all I need) with a Celeron 1GHz (socket 370) and 512 MB of ram that I will be using, since this should be more than enough power for my applications. Next, I will not need a CD-ROM drive (I will use one when installing Windows, but will not need one all the time), nor will I need any add-in cards. So, I'm basically looking at just including the motherboard/cpu/ram, power supply, and a hard drive. Lastly, I should let you know that I just graduated with a degree in Mechanical Engineering, so plenty of thought has gone into the airflow/heat transfer within the case layout that I'm thinking of.

Anyways...to the questions:

1. Very simple one, I have often read that the motherboard mounting screws actually ground the motherboard to the case. Now, I'm planning on building my case out of wood (probably MDF), in which case it wouldn't really be much of a ground. So, if these screws do serve this purpose, where should I ground the MB to?

2. Second, I must say, that after seeing the picoPSU review, I'm leaning heavily in that direction for my power supply. So, my next question is: do you think the configuration I have laid out would use less than 80W of power? If not, do you think using a 2.5" drive would help...in which case:

3. Do you have any brand suggestions for 2.5" drives? I've seen a lot of reviews for full size hdds, but I haven't seen too many for laptop drives, especially with regards to silent computing. I'll most likely be going with a 40-80GB drive since space really isn't that big of an issue, just looking for quiet.

4. And lastly, I'm planning on going with two Panaflo 80mm fans (one on the cpu and one for general duct airflow), my question is how quiet will these really be? This computer will be on all the time and it is pretty close to my bed, so I really want it silent (or as close to it as humanly possible). I'm pretty sure the MDF will help a bit with sound dampening, and after reading Doug's attempt at a wood case, I think I have all the potential problem locations identified (like isolating the CPU fan from the case), I've also designed the case to accommodate a material such as Dynomat to try and remove some more noise, but I'm going to test it first before dropping the extra cash.

Alright, I think that just about does it for now. I was planning on documenting the whole process and posting it on SPCR so that others might be able to learn from my mistakes (as I'm sure there will be plenty) so any help would be greatly appreciated. Thanks in advance and I look forward to hearing your advice.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun May 14, 2006 7:37 am

Welcome to SPCR.....I might help a little.

Actually the MB is grounded when you plug in the PSU cable, either with a DC/DC converter or a regular PSU. You can always use the black molex wires to ground any other components, but probably not necessary. If you have any wire mesh inside your wood case for EMI concerns, that needs to be grounded.

I'm using laptop drives in my recent projects.....A Samsung 80gb SATA 5400rpm model (inaudible to me), a Fujitsu 50GB 5400rpm drive in my Thinkpad (also inaudible), a 50GB Hitachi in another SFF I use (terrible seek noises).

You should get by with an 80W power supply for the converter, but that leaves little room for expansion.

Unless you have vibration problems with the case, dynamat seems un-necessary to me. Foam dampening would work better.

I can't recommend Panaflo fans....but I can recommend the Sanyo 80mm Petit Ace. Two of these slim 80mm fans might do the trick. They are really quiet, even at 12V.

Good luck.

justblair
Posts: 545
Joined: Fri Jul 15, 2005 12:33 pm
Location: GLASGOW, SCOTLAND, UK
Contact:

Post by justblair » Sun May 14, 2006 8:45 am

I have built a case from MDF and can chip in my two pence worth.
1. Very simple one, I have often read that the motherboard mounting screws actually ground the motherboard to the case. Now, I'm planning on building my case out of wood (probably MDF), in which case it wouldn't really be much of a ground. So, if these screws do serve this purpose, where should I ground the MB to?
Bluefront is right on this one. I have seen plenty of cases where the mobo is held in with plastic risers. So obviously this cant be an issue. On thing I would be concerned about is that mobos are built with an expected gap between them and the mounting plate. They have prodruding componenets on the underside. Might be better building risers in. Alternatively butcher an old case for its back plate and mounting tray. That way you have ready built mounts for the pci cards as well. I chose this approach and frankly I'm glad I did.
4. And lastly, I'm planning on going with two Panaflo 80mm fans (one on the cpu and one for general duct airflow), my question is how quiet will these really be? This computer will be on all the time and it is pretty close to my bed, so I really want it silent (or as close to it as humanly possible). I'm pretty sure the MDF will help a bit with sound dampening, and after reading Doug's attempt at a wood case, I think I have all the potential problem locations identified (like isolating the CPU fan from the case), I've also designed the case to accommodate a material such as Dynomat to try and remove some more noise, but I'm going to test it first before dropping the extra cash.
If you have not bought them yet, why use 80mm fans when 120mm loonies are not much more? My setup uses 120mm fans and I think its pretty quiet. I had 80mm in when I first built it. The larger fans do drop the temps for less noise.

MDF in my experiance doesn't help with sound dampening. It produces a live sound. Decoupling will certainly help but I would look at using the dynamat to dampen the sound. I used foam dampening. I think (though I haven't tried it) that heavier dampening or a combination of the two would be a good idea. But you can always try without and then add them later. I uses the paxmate foam. Can be bothered removing it all to try dynamat. The 120 mm fan swap made a difference but I cant help think I couls have got it quieter.

My Rig

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Re: General Advice for Custom Case

Post by IsaacKuo » Sun May 14, 2006 9:33 am

dj.pihat wrote:2. Second, I must say, that after seeing the picoPSU review, I'm leaning heavily in that direction for my power supply. So, my next question is: do you think the configuration I have laid out would use less than 80W of power? If not, do you think using a 2.5" drive would help...in which case:
IMHO, you'd do just fine with a plain old 80mm fan ATX PSU. It'll cost less, and you won't have to worry about power limits.

My computer which is next to my bed uses an 80mm fan ATX PSU and it's the only fan in the system. The only other noisemaker is a fully 2.5" drive in a heavily noise reducing enclosure.
3. Do you have any brand suggestions for 2.5" drives?
Western Digital or Samsung. But I find that even the quietest of 2.5" drives still need to be entombed in a noise reducing full enclosure to be silent. Here's HOW-TO for my homebrew enclosure.
4. And lastly, I'm planning on going with two Panaflo 80mm fans (one on the cpu and one for general duct airflow), my question is how quiet will these really be? This computer will be on all the time and it is pretty close to my bed, so I really want it silent (or as close to it as humanly possible).
I don't have any recommendations for specific fan brands. I just select my fans from many fans which I pulled from old equipment. I do find that the quietest of those fans is nearly silent when undervolted to 5v.
I'm pretty sure the MDF will help a bit with sound dampening,
Not really. Wood-like materials don't dampen sound significantly. It's what they DON'T do which is helpful--they DON'T resonate as much as metal.

However, the most important thing IMHO is to not produce sounds and vibrations in the first place. I have consistently had the best results with any sorts of dampening or muffling removed entirely--so airflow is as unrestricted as possible.

In particular, my bedroom computer (which is on 24/7) is in a plain steel eMachines case.

Anyway, you can cool your computer with just a single low airflow 80mm fan, given an appropriate airflow plan. My most common scheme is to flip the PSU fan and direct airflow like this:

intake -> fan -> PSU -> CPU -> case -> exhausts

This scheme is only appropriate for relatively low power systems like yours, because air heated by the PSU and CPU then enters the case as a whole. I do this simply because it's pretty easy to do with a stock case and I can get away with it. With a scratchbuilt case, you can do better. For example, my main workstation is a custom case with the following airflow plan:

intake -> fan -> PSU -> case -> CPU -> exhaust

If you are stuck on the idea of using at least two 80mm fans, though, you might want to consider using a single 120mm fan rather than two 80mm fans.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sun May 14, 2006 10:24 am

Bluefront wrote:I can recommend the Sanyo 80mm Petit Ace. They are really quiet, even at 12V.
You're a great kidder, Bluefront! I googled "Sanyo 80mm Petit Ace" and discovered it's a 32mm thick 4,600RPM(!!) model. Yeh, that oughta be real quiet! :D

Gorsnak
Posts: 190
Joined: Sun Nov 23, 2003 6:27 pm
Location: Saskatoon, SK

Post by Gorsnak » Sun May 14, 2006 10:34 am

I'm going to disagree with Isaac about wood and resonance. Wood does resonate as much as steel or aluminum (probably more, actually), for any given thickness. A wood case with walls as thick as common metal cases would be like sticking a computer into a cello. I'll let you imagine what that would sound like. Of course, any wood you'd make a case out of is much thicker, but not enough to eliminate the issue. If you don't believe me, take a case fan and run it in free air holding in in your hand. Now set it on a wood surface - a shelf or tabletop or whatever. That hum that you suddenly hear is the noise the fan produces at the resonant frequency of the wood being amplified by the wood's resonance.

Trust me on this one - resonance IS your absolute #1 concern with a wood (or MDF) case. Everything that vibrates must be isolated.

As for dampening, wood isn't going to be much better than steel. What it IS better at is blocking sound entirely - that is, when sound hits the wall of a case, some of it is reflected, and some is transmitted through to the other side (and some is dampened, but when we're talking about hard solids there will be very little dampening). The greater the proportion that is reflected, the better the sound blocking. This is largely a function of mass, and the sort of wood one might build a case out of is generally more massive than the very thin steel or aluminum found in most cases. This is a huge advantage if you exploit it to its full potential. You can trap nearly all the sound produced inside the case, and devise the necessary ventilation ports to allow sound out only in directions where the sound can then be trapped by a reasonably effective sound absorber. Some dampening materials inside the case are also desirable, as the sound that is reflected by the case walls can then be dissipated before it finds its way out through a fan port, but realistically the room inside a computer case isn't sufficient for really effective dampening. Thin dampening materials, such as those used by most people on these forums, are only effective with high frequency sound - probably only 2kHz and up. That's not entirely a bad thing, as broadband noise is less unpleasant with the higher frequencies gone.

With the proposed system starting with very low noise components, in a well-designed case I think that "completely inaudible" in any ordinary room is a realistic goal.

Doug

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun May 14, 2006 10:35 am

ACK....do I have to link to everything? :lol:

This Sanyo thin 80mm fan is the one to which I referred. They have a different name for the thing now. IMHO.....If you need an 80mm fan, this is the one.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sun May 14, 2006 11:36 am

This Sanyo thin 80mm fan is the one to which I referred. They have a different name for the thing now. IMHO.....If you need an 80mm fan, this is the one.
Hmm. Sanyo Denksomething... quiet... 11 blades... quiet... not just one but two ball bearings... quiet... $18.95... What??

Gee, here I am suffering with my GW NCB $5 80mm 1900RPM fans (which I run at 5-7V) from SVC.com. I'd sure switch to them real quiet ball bearing fans if only I could afford them. (Brightly:) I accept charitable gifts toward the purchase of those superior 11-blade dual ball bearing quiet fans. :D :D

Just kidding, Bluefront. You've made a rare find there.

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Sun May 14, 2006 12:58 pm

Gorsnak wrote:I'm going to disagree with Isaac about wood and resonance. Wood does resonate as much as steel or aluminum (probably more, actually), for any given thickness. A wood case with walls as thick as common metal cases would be like sticking a computer into a cello. I'll let you imagine what that would sound like. Of course, any wood you'd make a case out of is much thicker, but not enough to eliminate the issue.
What you say is true, but I never really thought of it in those terms. As you say, you'd never make a computer case out of wood as thin as a metal case. It just wouldn't be sturdy enough.
What it IS better at is blocking sound entirely [...] You can trap nearly all the sound produced inside the case, and devise the necessary ventilation ports to allow sound out only in directions where the sound can then be trapped by a reasonably effective sound absorber.
My experience is that the only way to "trap nearly all the sound" is with a full enclosure with no significant openings whatsoever. For computing purposes, the only place where this is really applicable IMHO is with silencing a 2.5" hard drive. Everything else requires some sort of ventilation that will let a large fraction of the noise escape.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun May 14, 2006 1:58 pm

Isaac....you could make a strong case out of thin wood. I have used 5/32 aircraft plywood in different projects (it's three ply). The stuff is very strong, and easy to work with. Glued together with proper bracing it would make a unique computer, as strong as most standard cases and no bigger. I get the stuff at craft stores.....fairly cheap.

Cerb
Posts: 391
Joined: Tue Apr 13, 2004 6:36 pm
Location: GA (US)

Post by Cerb » Sun May 14, 2006 2:29 pm

1. You might be able to find a case to rob the tray out of. This could also make cutting the case easier. Else (like if you want it low profile), just space it from the MDF. If you're concerned with case grounding, space it with standoffs, and attach wires to the standoffs that all connect to the PSU chassis.

2. There's no way it's anywhere near 80w, especially with onboard video.

4. If the Celeron is all that really needs cooling, use negative pressure and have a single exhaust fan, ducted from the CPU HSF. You could probably get away easier, and maybe better, with another fan. If you're going for small size, another (Bluefront's?) 80mm, otherwise a nice quiet 120mm (Nexus?), decoupled.

In fact, one idea might be to use a standard PSU, replace the fan with a really quiet one (or remove it and duct the PSU, too) and duct the CPU HSF through the PSU. While the PSU as exhaust may be bad in general, you're going to be dealing with negligible power to handle--just use a quality PSU, and it should be OK (especially if you get a new one, like one of the 80+, or a SFF PSU).

Good luck, and yes, we want pics! :)

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Sun May 14, 2006 3:56 pm

Bluefront wrote:Isaac....you could make a strong case out of thin wood. I have used 5/32 aircraft plywood in different projects (it's three ply).
But as thin as a typical computer case's metal panels? 5/32 is about 4mm, which is much thicker than a typical metal case panel.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Sun May 14, 2006 4:04 pm

5/32 is about 4mm, which is much thicker than a typical metal case panel.
Tru dat. Most PC cases use 0.4-0.8mm thick SECC steel.

dj.pihat
Posts: 3
Joined: Sun May 14, 2006 6:08 am

Post by dj.pihat » Sun May 14, 2006 6:17 pm

Wow guys, thanks for all the input on this project. I've got a few points to bring up in response to what you have said...and a few questions to go along with it.

First of all, I was indeed planning on spacing the motherboard from the floor of the case with risers or something. I was actually thinking small blocks of plexiglass (I have some 1" thick x 1" wide strips laying around) just because of how easy it would be to tap holes in, then I'll just use standard motherboard standoffs.

Second, when I said MDF will be a decent sound dampener, I think I was thinking sound blocker like Gorsnak pointed out. I'm decent with a welder...and believe me I considered this option...but it just seems to me that going with 3/4" or 1" MDF would be far better then 1/16" to 1/8" steel or aluminum (and far cheaper). But I definitely agree that everything that vibrates needs to be completely isolated from the wood.

Lastly, the reason I wanted to go with the picoPSU is two-fold. First, it will put off far less heat then a standard PSU because of the fact that the AC-DC converter will not be located inside of the case (which is the primary source of heat dissipation). And second, currently I have 3 PSU's lying around that I'm not using but they are all as loud as a jet engine, which just isn't going to fly (no pun intended :roll: ). So, if I was going to spend the cash to get a new PSU anyways, why spend $100-$150 getting a quiet or fanless 200-300W power supply when I can spend $80 and get a completely silent one that supplies enough power, is much smaller, and helps keep the ambient heat down.

So yeah, that’s some of my thoughts on the matter.

Now IsaacKuo brought up an interesting point...do you really think two fans are necessary? I'm all for dropping down to just one if you think I can get away with it. In which case, my question is, where would it go? Should I try to duct in air past the cpu and out of the case? Or should I just direct it right on the cpu and screw trying to get it out of the case...just deal with the ambient heat (I think this is a bad idea by the way). Keep in mind that the motherboard chipset didn't even come with a heatsink attached...so that tells you what kind of massive heat this thing is putting out 8) . If I did go with a single 80 or 120 mm fan and duct everything through, would I need to include something like the hard drive? Or can I basically just pass it right over the cpu heatsink and be done with it, seeing as though there will be no pci cards, and no real power supply to speak of.

I think that pretty much does it, let me know what you guys think because I'd like to finalize the layout over the course of this week and hopefully post a few 3D models for you to take a look at before I start actually building. Look forward to hearing from your replies.

...and thanks again.

Gorsnak
Posts: 190
Joined: Sun Nov 23, 2003 6:27 pm
Location: Saskatoon, SK

Post by Gorsnak » Sun May 14, 2006 6:59 pm

IsaacKuo wrote:My experience is that the only way to "trap nearly all the sound" is with a full enclosure with no significant openings whatsoever. For computing purposes, the only place where this is really applicable IMHO is with silencing a 2.5" hard drive. Everything else requires some sort of ventilation that will let a large fraction of the noise escape.
Then you need to go back and re-read the article about my case. :D In particular, note the sound absorbers at the intake /exhaust ports (carpet and rear acoustic panel respectively). Once you're out of the case, you have room to use effective sound absorption, and if you've set up the design right, you won't need much of it. My computer is loud. I mean, that 80 gig WD hdd just screams. But I can almost never hear it.

theyangster
Posts: 424
Joined: Sat Oct 01, 2005 9:08 pm
Location: Somewhere over the rainbow....

Post by theyangster » Sun May 14, 2006 8:57 pm

dj.pihat - Welcome to SPCR!

Great Timing on your part, I'm planning to the same thing with the similar components (intel 370 1.3 celeron with intergrated everything [intel 810], Laptop HD, hoping for a DVD slim drive, but those are expensive) and build a wood case

Perhaps we can share some ideas ;)

Some great suggestions by many of the SPCR greats :)

I'll add some of my own-

Now, I think we/I can get away with a 60w DC to DC kit
http://www.logicsupply.com/product_info ... ucts_id/90
I talked to Frostedflakes about this and he agreed that that's about all we need. Plus for $56 dollars + shipping, i think it is an affordable PSU

I myself will using this a secondary general usage PC for Internet surfing, email and word.

If I want to game, that's for my other PC

I'm glad that the "wise ones" :D brought up the issue of Resonance, I didn't think of it until I read it here

I'm planning for my case to have
Wood panels -of course much easier to work with for the average joe
Vinyl floor panels - for the base layer for additional vibration and sound control (the ones from the dollar store)
Melamine Foam- still have a lot laying around, this will serve as the last layer
Suspended 120 fan and HDD - for the maximum vibration control

I am looking for my case to be a slim as possible and of course as quiet as possible

As far as intake and exhaust, I haven't figured out where to put them

Are you going to use sketchup? (send me copy :) )

EDIT, which CPU heatsink are you going to use? I was planning to get one off ebay (copper base aluminum fins)
Last edited by theyangster on Sun May 14, 2006 9:13 pm, edited 1 time in total.

Cerb
Posts: 391
Joined: Tue Apr 13, 2004 6:36 pm
Location: GA (US)

Post by Cerb » Sun May 14, 2006 9:04 pm

Include the HDD, but don't be over-zealous. Unlike the CPU, it just needs non-stagnant air.

You don't necessarily need to duct any of it--it depends on the size deisgn. Just make sure you have the air pathway(s) as unobstructed as possible, except by things that need the cooling.

Also, you might wish to try a nice S462 cooler, and see if it will fit well, and work passively. It might just work...then have a single very low noise/slow case fan (partly for assurance :)).

theyangster
Posts: 424
Joined: Sat Oct 01, 2005 9:08 pm
Location: Somewhere over the rainbow....

Post by theyangster » Sun May 14, 2006 9:24 pm

true, the HD will need a bit of air

Cerb - you had me confused as to what a S462 heatsink was, I googled it, got a watercooling kit as on of the hits and was like ??? until I figured out you meant socket A :)

Cerb
Posts: 391
Joined: Tue Apr 13, 2004 6:36 pm
Location: GA (US)

Post by Cerb » Sun May 14, 2006 9:41 pm

Hehe, yeah. Most fit on S370, but some are odd, and the clip twists a little (two flanges, more pressure on the chip?). If you build other computers, a SI-97A would be worth a shot--if it fails, it can go into any AMD system (sockets 462, 754, 939, not sure about AM2).

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Mon May 15, 2006 2:19 am

Since the case for this project is still on the drawing board, and can be constructed in a number of ways, and can be optimised for silence, may I suggest the "Absolute Ultimate Airflow Technique". I already did the initial testing, including the DC/DC converter power supply. The only problems I'm having right now involve fan control. I can run at a lower rpm, but my current controller won't allow it (that's soon to be fixed).

Since you are not constrained by an existing case design.......here's what I suggest. Make the case with the board sitting horizontal. The width should be at least 4" wider than the MB. That gives you a 3/4" wood side with room for 1" of dampening foam.

Make the case deep enough so the optical drive with it's cables does not overlap the MB. This leaves room for optimal passive exhaust flow, not impeded by the optical drive.

The height is optional. I'd make it tall enough to hold an least two 5 1/4" bays. One for the optical drive, another for a fan controller or whatever. The height is also determined by the CPU cooler. I recommend an XP-120 or some other wide, relatively short CPU heatsink. This makes optimal use of this positive airflow technique.

Use a Laptop drive and a DC/DC converter. This eliminates two major heat sources from the case. Since you intend to use onboard graphics, you can get by with a narrower intake duct, one that doesn't need to overlap the AGP slot. Maybe you could use only one 92mm fan, maybe even one 80mm fan. With the right fan controller, you could have a virtually silent computer......without any heat concerns.

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Mon May 15, 2006 4:48 am

Gorsnak wrote:
IsaacKuo wrote:My experience is that the only way to "trap nearly all the sound" is with a full enclosure with no significant openings whatsoever. For computing purposes, the only place where this is really applicable IMHO is with silencing a 2.5" hard drive. Everything else requires some sort of ventilation that will let a large fraction of the noise escape.
Then you need to go back and re-read the article about my case. :D In particular, note the sound absorbers at the intake /exhaust ports (carpet and rear acoustic panel respectively). Once you're out of the case, you have room to use effective sound absorption, and if you've set up the design right, you won't need much of it. My computer is loud. I mean, that 80 gig WD hdd just screams. But I can almost never hear it.
I read it right the first time, and I simply absolutely disagree.

My suspicion is that the noise floor you're working with must be rather high if you've had such success with such noise absorption.

[added:]

To clarify, I am NOT saying that such sound absorbers don't reduce noise significantly. They do. But in my experience they never eliminate "nearly all" of the noise. It's a certain amount of noise reduction, but nothing like the "nearly all" reduction a full enclosure can acheive.
Last edited by IsaacKuo on Mon May 15, 2006 5:06 am, edited 1 time in total.

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Mon May 15, 2006 4:49 am

[duplicate post]
Last edited by IsaacKuo on Mon May 15, 2006 5:04 am, edited 1 time in total.

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Mon May 15, 2006 4:56 am

dj.pihat wrote:First of all, I was indeed planning on spacing the motherboard from the floor of the case with risers or something. I was actually thinking small blocks of plexiglass (I have some 1" thick x 1" wide strips laying around) just because of how easy it would be to tap holes in, then I'll just use standard motherboard standoffs.
It doesn't matter if you do this.
And second, currently I have 3 PSU's lying around that I'm not using but they are all as loud as a jet engine, which just isn't going to fly (no pun intended :roll: ).
You already have some PSU's? Then you should definitely try modifying one or all of them before going out and spending money on others. I've had great success with old loud PSUs by undervolting the fan to 5v (replacing it with whichever 80mm fan was quietest). I typically cut out the fan grill also, if necessary, as well as flip the fan (this usually fits in with my airflow plans).
Or should I just direct it right on the cpu and screw trying to get it out of the case...just deal with the ambient heat (I think this is a bad idea by the way).
With a relatively low power CPU like what you have, this ambient heat won't be a big deal. Personally, I like ducting air from the PSU at the CPU, and then let the rest of the case simply live with the heat. Really, only the PSU and hard drive would suffer from the extra heat--but in my case the PSU is receiving fresh air and the hard drive is a 2.5" drive.
duct everything through, would I need to include something like the hard drive?
If you're using a 3.5" drive, then include it at the start of any airflow plan. If you're using a 2.5" drive, it doesn't matter.

BillyBuerger
Patron of SPCR
Posts: 857
Joined: Fri Dec 27, 2002 1:49 pm
Location: Somerset, WI - USA
Contact:

Post by BillyBuerger » Mon May 15, 2006 6:05 am

Socket 370 CPUs draw power from the 5V line don't they? Could that be a problem for the PicoPSU? It only has 6A on the 5V line for a total of 30W. The CPU could very well be within 30W. But the motherboard and other devices might push that. I don't have any experience with the PicoPSU in an old system. It was just a concern I had.

I recently built a wood case for an old P133. The only real noise from the system was the old hard drive. Personally, I think it did a good job of "sealing" in the noise. With the side off you can hear the drive pretty well. With the side on, I can only hear a low hum that resonates through. I didn't even try to decouple the fans. At 5V the 80mm fan doesn't make enough noise/vibration to be a concern.

theyangster
Posts: 424
Joined: Sat Oct 01, 2005 9:08 pm
Location: Somewhere over the rainbow....

Post by theyangster » Mon May 15, 2006 6:23 am

The Pico PSU is only a consideration at this point, but I hope you are wrong :)
but, the specs of the Morex 60w have me worried now that you mention it
Specifications:
Dimensions (mm) 160 x 45
Input Voltage Required (V) 11.4V~12.6V
Maximum Output (W) 62.4
Output Voltage V 5VSB +3.3V +5V +12V -12V
Current A 1.5 5.0 5.0 2.5 0.1
Power Output W 7.5 16.5 25.0 30.0 1.2

Bluefront- I was thinking of doing the MB horizontal and put the 120 fan directly above the heatsink ala your technique

IsaacKuo- do you think that an old atx PSU would be able to run only with one 120 fan running at 5-7v? The idea here is to run fanless and eliminate all the fans except one. The old PSU's are inefficient and would add excess heat to the design.

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Mon May 15, 2006 6:39 am

theyangster wrote:IsaacKuo- do you think that an old atx PSU would be able to run only with one 120 fan running at 5-7v?
Trivially. Most of my systems use cheap old leftover ATX PSUs which are NOT very efficient (I can tell just from the amount they heat the air). I run my 2.4Ghz Celeron system with an old Enlight 300 with just one very low airflow 80mm fan at 5v.

That particular PSU has been severely abused by me--for a while I was using it to fanlessly power a 550mhz P3 system. (The components were in open air, operating completely fanlessly for true passive operation. I wasn't satisfied with the results and I was afraid of the long term effects on the PSU. The air above it was VERY warm!)
The idea here is to run fanless and eliminate all the fans except one. The old PSU's are inefficient and would add excess heat to the design.
The inefficiency of old PSUs is certainly something to be aware of. On my midfan file server, I was shocked at how much cooler the exhaust air was when I switched from an old Fortron 300 to the ePower Lion. With the Fortron, the exhaust air was rather warm even at idle. With the Lion, the exhaust air was not much warmer than ambient at idle. (Based on the SPCR review of this PSU, my file server is hitting its "sweet spot" of maximum efficiency.)

And a Fortron 300 is perhaps more efficient than some no-name el-cheapo old power supply.

Gorsnak
Posts: 190
Joined: Sun Nov 23, 2003 6:27 pm
Location: Saskatoon, SK

Post by Gorsnak » Mon May 15, 2006 7:07 am

IsaacKuo wrote:I read it right the first time, and I simply absolutely disagree.

My suspicion is that the noise floor you're working with must be rather high if you've had such success with such noise absorption.

[added:]

To clarify, I am NOT saying that such sound absorbers don't reduce noise significantly. They do. But in my experience they never eliminate "nearly all" of the noise. It's a certain amount of noise reduction, but nothing like the "nearly all" reduction a full enclosure can acheive.
And have you ever worked with professional sound absorption panels made of hi-density fibreglass?

1) The noise floor I'm working with is below 30dBA, as measured by a very good B&K spl meter. Okay, that was the middle of the night with the fridge unplugged, but it took very significant audible sound above ambient to get a reading. By my estimate the ambient at that time was 20ish. Ordinary daytime ambient is probably 25ish, maybe a tad higher.

2) 1" compressed fibreglass has a sound absorption coefficient of 1 or higher down to 1kHz, dropping to around .6 at 500Hz. What that means is that a panel made of the stuff will absorb everything bit of noise over 1kHz that hits it, and a good chunk of somewhat lower frequencies too. Sound over 1kHz doesn't difract much at all, so it cannot turn corners to sneak through winding gaps. Admittedly, the carpet isn't as effective an absorber as the panel, but the corners involved at the bottom are more substantial, meaning that most sound will have to bounce off the carpet several times before it could escape, effectively increasing the absorption coefficient.

I dunno what to say. Part of my job involves building and installing these things in a variety of venues. I've heard how just a couple panels in a large room can eliminate slapback from speaker output bouncing off of back walls. I've heard how, once substantial square footage is installed, the reverb time in a room drops dramatically. I just recently put some into a fancy home theatre, a strip around the room about 44" wide. The room went from being extremely live to extremely dead, so much so that you can practically feel the atmosphere change as you walk into the room. If you're basing your beliefs about sound absorption about the performance of stuff like Acoustipak, I suggest that you simply don't have experience with real sound absorbers.

The way my machine is set up, virtually no sound over about 800Hz can escape it, and very little sound under 800Hz is produced in the first place. Whether you believe it or not.

Sorry for the thread hijack, dj.pihat. :D

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Mon May 15, 2006 7:26 am

Gorsnak wrote:And have you ever worked with professional sound absorption panels made of hi-density fibreglass?
No. Perhaps I simply have never used really good sound absorbing materials. I'm only speaking from my experience, of course.

All I know for sure is that a full enclosure with even cheap homebrew materials can reduce noise by an incredible amount, but that even a small opening leaks out an unacceptable amount of noise.

I personally haven't had success silencing a 3.5" hard drive with convoluted air paths and sound absorption. Others here have had the same difficulties. Some of us have turned to using full enclosures on 3.5" drives despite the thermal risks. I've turned to using fully enclosed 2.5" hard drives.

If you've figured out a good way to silence a loud 3.5" hard drive while still providing airflow, then that's of great interest to us all!

[edit added:]

Oh, I should note that on theoretical grounds I agree with your ideas. Based on physics theory, the high pitched noises which I find most irritating SHOULD be pretty straightforward to contain. It SHOULD just be a matter of line-of-sight, forcing bouncing paths, and proper sound absorbing materials. I just haven't had personal success yet in my attempts to apply this theory in practice.

The best results I've come up with involve making all the intakes/exhausts point to the rear, and some sound absorption on the facing wall (with some space for attenuation). It works some...but it's never been something so dramatic that it would ever turn a loud computer into a quiet computer.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Mon May 15, 2006 7:42 am

If you're basing your beliefs about sound absorption from the performance of stuff like Acoustipak, I suggest that you simply don't have experience with real sound absorbers.
Unfortunately, "real" sound absorbers are too expensive for all but the most affluent silencers. So the logical consequence is to stop noise being generated in the first place, rather than going to the trouble and expense of trying to block it once it's been generated.

Gorsnak
Posts: 190
Joined: Sun Nov 23, 2003 6:27 pm
Location: Saskatoon, SK

Post by Gorsnak » Mon May 15, 2006 7:57 am

Well, I'm not affluent at all. But then stealing scraps of fibreglass from work doesn't cost me much, either, and the stuff we use isn't readily available on the retail market I'll admit. Still, you could make something comparable using regular fibreglass bat insulation. It's low density, so you need quite a bit of it. My wild-assed guess is that the R20 stuff for 2x6 walls would be sufficient, if bulky. You could compress it into a manageable size by making a sort of upholstered cushion using an open-weave fabric (i.e., something burlapy). This wouldn't be very attractive, but if your case only had openings at the rear, and it were sitting under a desk or something, you could have the fibreglass pillow in behind it without compromising airflow or looking ugly.

However, my initial point in this thread was somewhat different. If the OP is starting with a very low-noise machine to begin with, then with smart placement of intake/exhaust ports only a modicum of absorption would be necessary to achieve the desired effect. Carpet on the bottom, and heavy curtains hanging behind, for example. If there's only a little sound coming out, you only have to prevent it from being reflected directly at the user. I suspect at this point we aren't really in much disagreement.

Post Reply