Micro ATX desktop cases

[Ed Coolidge]

You're quite right, I did.

You, of course, also knew that "the amount of heat that the heatsink
can dissipate" is dependent on not only ambient but the airflow across
it's dissipating surfaces.

Well, I'm glad that is finally out of the way. So yes, it can increase
dissipation, but the real question is how much does the heatsink really
require to properly function? Obviously, some heatsinks require the fan
to be in direct contact to meet the CPUs requirements. The heatsink in
question however has far greater surface area then the average fanned
heatsink.

BTW, outside of the context of CPUs there are in fact heatsinks that are
designed to function with no extra source of airflow, i.e. no fans!

Perhaps. What was case temp?

I was referring to my experiences with other PCs. This particular OEM
mainboard, which is a Gateway if you must know, doesn't seem to have any
temperate sensors. I tried loading Main Board Monitor to check, but it
couldn't detect any sensors. So if it does have any, I don't have any
way to read them. I don't have a probe that can test the temperature
inside the closed case either.

Then how do you explain the millions of them working just fine?

I was referring to the average cases you would find in the open market,
not the ones that OEMs use. Many of them seem to have very little air
flow through the decorative case front. While they may be merely
adequate for some PCs, others seem to mysteriously crash or fail when
using a cheap case and PSU, especially when overclocking.

It's closer to the top than the bottom but your whole 'retort' is a
good example of what I mean when I say you are more interested in
arguing than in a solution.

I was merely pointing out that you comment wasn't as relevant as you'd
believed.

 

[David Maynard]

Well, I'm glad that is finally out of the way. So yes, it can increase
dissipation,

Not just 'can', it does.

but the real question is how much does the heatsink really
require to properly function?

And you have a clue: it was originally ducted to a rear case fan. Or so you
told me.

Obviously, some heatsinks require the fan
to be in direct contact to meet the CPUs requirements.

Mounting the fan on the heatsink is simply one way of accomplishing it. It
has the virtue of being 'self contained' but it has the vice of being stuck
with a fan no larger than the heatsink form factor will bear.

Ducting has the advantage of being able to use a larger, quieter, fan and
more efficient use of the airflow since it's directed and can serve
multiple purposes when engineered properly.

The heatsink in
question however has far greater surface area then the average fanned
heatsink.

You calculated the surface area?

If it's anything like the one I had it's 'big' but also rather widely
spaced and open to reduce airflow resistance since an axial fan can't draw
much pressure.

None of which matters much because you haven't a clue what the heatsink
'needs' other than it was originally ducted to a rear case fan and common
sense says they didn't do it just for chuckles.

BTW, outside of the context of CPUs there are in fact heatsinks that are
designed to function with no extra source of airflow, i.e. no fans!

Of course there are. Don't you remember me talking about "fanless" PSUs and
"fanless" PCs? You know, where I pointed out that 'normal' people kinda
figure that "fanless" means no fans are needed.

Moot, though, since yours is obviously not one of them.

I was referring to my experiences with other PCs.

Then it's of little use here.

This particular OEM
mainboard, which is a Gateway if you must know, doesn't seem to have any
temperate sensors. I tried loading Main Board Monitor to check, but it
couldn't detect any sensors. So if it does have any, I don't have any
way to read them. I don't have a probe that can test the temperature
inside the closed case either.

Then you're shooting in the dark. Get a thermometer.

Since you don't have any numbers you have no basis on which to presume the
problem is 'high case temp' and drawing on my experience with a myriad of
computers putting out 20 watts, and more, there's not normally a problem
cooling such a case with simply the PSU fan, or an added front at most,
even with a fanned heatsink dumping the whole 20+ watts directly into the
case, as was the typical situation for most PCs of the era using the stock
Intel heatsink/fan.

I was referring to the average cases you would find in the open market,

'Average' when and intended for what?

not the ones that OEMs use. Many of them seem to have very little air
flow through the decorative case front.

That's because it didn't take much for the typical systems of the day that
were built with 20 watt CPUs.

While they may be merely
adequate for some PCs, others seem to mysteriously crash or fail when
using a cheap case and PSU,

Now you've changed from 'average' to 'cheap' and without identifying what
caused them to "mysteriously crash."

Not to mention I ran many systems with the cheapest case/PSU junk combos
for years with no problems at all. My favorite was a 2.4Vcore, 29.5W,
K6-III 450, mini-tower AT case with no front fan or anything else. Just PSU
and heatsink fans.

especially when overclocking.

Overclocking is another whole game.

I was merely pointing out that you comment wasn't as relevant as you'd
believed.

It was a heck of a lot more relevant than you trying to argue that a P-II
350 case/thermal solution was designed for a 'high end P-III' that didn't
even exist when the case was designed.

Like I said, you just want to argue but what you think you're accomplishing
by it I have no idea.

 

[Ed Coolidge]

Mounting the fan on the heatsink is simply one way of accomplishing
it. It has the virtue of being 'self contained' but it has the vice of
being stuck with a fan no larger than the heatsink form factor will bear.

Ducting has the advantage of being able to use a larger, quieter, fan
and more efficient use of the airflow since it's directed and can
serve multiple purposes when engineered properly.

I never said the duct was a bad idea. I did say that it just doesn't
help in this case, because there is no place to add a rear fan and the a
duct on the PSU would be almost pointless due to its position relative
to the CPU.

You calculated the surface area?

No, but I do have other PII and PIII slot CPUs with fanned heatsinks.
The fin spacing is about the same, but the length of the fins is at
least double. I really don't think you need a calculator to figure out
which as the larger surface area.

If it's anything like the one I had it's 'big' but also rather widely
spaced and open to reduce airflow resistance since an axial fan can't
draw much pressure.

None of which matters much because you haven't a clue what the
heatsink 'needs' other than it was originally ducted to a rear case
fan and common sense says they didn't do it just for chuckles.

And yet somehow you are under the impression that the CPU would fry
without it.

Of course there are. Don't you remember me talking about "fanless"
PSUs and "fanless" PCs? You know, where I pointed out that 'normal'
people kinda figure that "fanless" means no fans are needed.

Moot, though, since yours is obviously not one of them.

*sign* The heatsinks I was referring to have nothing to do with the PSU
or the PC in general. Somehow you seem to be under the impression that
no heatsink could ever operate without some fan blowing on it.

Then it's of little use here.

Not really. Do you throw out all the knowledge of you experiences
simply because you start a new project? I seriously doubt it.

Then you're shooting in the dark. Get a thermometer.

I have a thermometer, just not one that can take the inside temp while
the case is closed. I usually buy retail mainboards that have sensors
on board.

'Average' when and intended for what?

Average as in the average case you would find on the market. They have
been slowly improving though the air flow though.

That's because it didn't take much for the typical systems of the day
that were built with 20 watt CPUs.

Yes, but todays CPU can put out about 5x that. I talking about the
cases that are on the market today, not 5 years ago.

Now you've changed from 'average' to 'cheap' and without identifying
what caused them to "mysteriously crash."

The market is flooded with cheap cases, so the average would tip in that
direction.

Not to mention I ran many systems with the cheapest case/PSU junk
combos for years with no problems at all. My favorite was a 2.4Vcore,
29.5W, K6-III 450, mini-tower AT case with no front fan or anything
else. Just PSU and heatsink fans.

Did you measure it's CPU and case temps?

It was a heck of a lot more relevant than you trying to argue that a
P-II 350 case/thermal solution was designed for a 'high end P-III'
that didn't even exist when the case was designed.

Like I said, you just want to argue but what you think you're
accomplishing by it I have no idea.

OEM sometimes reuse the same cases and have take into account the
requirements of future components, even if they are not on the market yet.

 

[David Maynard]

Ed Coolidge wrote:

<snip of Ed's incessant arguing about nothing>

I never said all heatsinks need fans nor any of the other fantasies you
invented but I tell you what, you just go on and enjoy staring at the case
mumbling to yourself about ducts with no purpose and heatsinks made for non
existent processors, or whatever other nonsense strikes your fancy, because
I'm finished with you.

 

[Ed Coolidge]

Ed Coolidge wrote:

<snip of Ed's incessant arguing about nothing>

I never said all heatsinks need fans nor any of the other fantasies
you invented but I tell you what, you just go on and enjoy staring at
the case mumbling to yourself about ducts with no purpose and
heatsinks made for non existent processors, or whatever other nonsense
strikes your fancy, because I'm finished with you.

First you start the whole pointless rant about the word "fanless", then
deviate to a "normal" persons concepts of a fanless PC, only to finally
get stuck on the notion that I have to put a duct (which I don't have)
on the rear exhaust fan (which doesn't even exist in my PC case). Now
you accuse me of incessant rambling about pointless fantasies. Thanks
for wasting my time.

 

[David Maynard]

First you start the whole pointless rant about the word "fanless",

I made no rant whatsoever. I merely said the term "fanless," in that
context, was "a bit of a misnomer" since the supposed 'fanless heatsink'
needs sufficient airflow from a fan somewhere.

I was merely giving you a heads up that taking a 'fanless heatsink' out of
it's engineered 'whole system' thermal solution and simply shoving it into
a typical case was likely to be a problem and guess what, it was.

And you've been trying to argue it isn't a problem, even though you got a
problem, ever since.

then
deviate to a "normal" persons concepts of a fanless PC, only to finally
get stuck on the notion that I have to put a duct (which I don't have)
on the rear exhaust fan (which doesn't even exist in my PC case).

And not one bit of that is true either but then you're so busy trying to
find something to argue about you've never been able to grasp a single
thing said anyway.

Now
you accuse me of incessant rambling about pointless fantasies.

And you haven't stopped.

Thanks
for wasting my time.

The waste is entirely one of your own making.

 

[Ed Coolidge]

I made no rant whatsoever. I merely said the term "fanless," in that
context, was "a bit of a misnomer" since the supposed 'fanless
heatsink' needs sufficient airflow from a fan somewhere.

I was merely giving you a heads up that taking a 'fanless heatsink'
out of it's engineered 'whole system' thermal solution and simply
shoving it into a typical case was likely to be a problem and guess
what, it was.

And you've been trying to argue it isn't a problem, even though you
got a problem, ever since.

I never said that there wasn't a problem. I even stated in the opening
post "I don't think that the little fan in the PSU is going to give
enough airflow". The disagreement has been about how to fix the
situation. Again, I never argued that the CPU wasn't from a Dell,
because I really have no idea what it was out off and it does look like
the heatsink I've seen used in some Dell PCs. The problem is that since
my case doesn't have a rear exhaust fan, Dell's solution is fairly
irrelevant. You however been arguing that it's essential. Since it was
working "just fine" for over a year in a generic mid tower case, it
obviously isn't true. I even questioned how much the duct in the Dell
improves the heatsinks performance and you practically flipped. I
however doubt that you've ever tested the CPU temps of the Dell with and
without the duct so any farther discussion about Dell's duct is moot.

Now without ridiculous comments, how about the real problem. The
original concern was that the new case has even less airflow than the
old one. Even though the CPU still seems to be operating within it's
thermal limits, even under full load, I think the current situation
could be improved. Farther investigation into the case design hasn't
proved to be promising. Basically there is no way to add an exhaust or
intake fan to the case without cutting holes. Since the case was free I
would have no problems with that if it wasn't for the fact that I really
don't have any metal cutting tools and purchasing them would be more
than I would care to do for this "cheap" project. The only other
solution that I could think of would be to go the "flower" route and
place another fan closer to the heatsink since it's the largest heat
source anyway. Instead of just blowing the heat from the heatsink all
over the case however, I'm going to direct the air flow from the
heatsink up into the nearby PSU, which is the only viable exhaust point
anyway.

Unless you have anything useful to add, I consider this discussion closed.

 

[David Maynard]

I never said that there wasn't a problem. I even stated in the opening
post "I don't think that the little fan in the PSU is going to give
enough airflow".

Which I've been trying to explain to you is part of the problem: laying
everything on 'case ventilation'.

The disagreement has been about how to fix the
situation. Again, I never argued that the CPU wasn't from a Dell,
because I really have no idea what it was out off and it does look like
the heatsink I've seen used in some Dell PCs. The problem is that since
my case doesn't have a rear exhaust fan, Dell's solution is fairly
irrelevant.

It's relevant in that it demonstrates the thing needs airflow as they
didn't duct it to a rear fan for no reason nor did they do it for a
non-existant 'future processor'.

You however been arguing that it's essential. Since it was
working "just fine" for over a year in a generic mid tower case, it
obviously isn't true.

I NEVER said it was 'essential' to have a duct and if you ever paid any
attention to what's said you'd know that as I've given you at least three
possible solutions of which only one involved a duct. And then only because
you said you were going to 'aim' an intake fan at the thing.

I even questioned how much the duct in the Dell
improves the heatsinks performance and you practically flipped. I
however doubt that you've ever tested the CPU temps of the Dell with and
without the duct so any farther discussion about Dell's duct is moot.

I didn't 'flip'. I just pointed out the obvious, that PC makers don't go to
the effort and expense of designing and installing ducts for no reason yet
that is precisely what you argue they did. Weeee... let's put a duct in for
chuckles.

Now without ridiculous comments, how about the real problem. The
original concern was that the new case has even less airflow than the
old one. Even though the CPU still seems to be operating within it's
thermal limits, even under full load, I think the current situation
could be improved.

If it ain't broke then why are you trying to 'fix' it?

Farther investigation into the case design hasn't
proved to be promising. Basically there is no way to add an exhaust or
intake fan to the case without cutting holes. Since the case was free I
would have no problems with that if it wasn't for the fact that I really
don't have any metal cutting tools and purchasing them would be more
than I would care to do for this "cheap" project. The only other
solution that I could think of would be to go the "flower" route and
place another fan closer to the heatsink since it's the largest heat
source anyway.

That's one of the solutions I suggested, even to using the the same Zalman
flower allegory I made.

Instead of just blowing the heat from the heatsink all
over the case however, I'm going to direct the air flow from the
heatsink up into the nearby PSU, which is the only viable exhaust point
anyway.

Just remember that if you're trying to 'pull' air with a fan it'll pull
from the area of least resistance, with is usually right there at it's rim
unless it's forced, like with a duct or heatsink shroud, to get the air
from someplace else. That's why the Zalman fan blows 'onto' the heatsink
and heatsinks with fans generally blow 'down' onto them. The expelled air
is 'impelled' in that direction and so goes that way, for at least short
bit, but the intake side isn't impelled. It just comes from where ever is
easiest.

Unless you have anything useful to add, I consider this discussion closed.

You could also consider replacing the PSU fan with a more powerful one.

 

[Ed Coolidge]

David Maynard wrote:

<snip>more pointless arguing</snip>

Since we agree that in the context of the Dell case, the duct is neither
essential or useless, can we leave it at that?

Just remember that if you're trying to 'pull' air with a fan it'll
pull from the area of least resistance, with is usually right there at
it's rim unless it's forced, like with a duct or heatsink shroud, to
get the air from someplace else. That's why the Zalman fan blows
'onto' the heatsink and heatsinks with fans generally blow 'down' onto
them. The expelled air is 'impelled' in that direction and so goes
that way, for at least short bit, but the intake side isn't impelled.
It just comes from where ever is easiest.


You could also consider replacing the PSU fan with a more powerful one.

I have several fans on hand, so I'll see what I can do. BTW, the whole
point is to make the best of a less than ideal situation. For the PC
itself I invested ~$25 in parts, the rest was just stuff I had or got
for free. If in the end it doesn't work I'm not out much.

 

[David Maynard]

David Maynard wrote:

<snip>more pointless arguing</snip>

Since we agree that in the context of the Dell case, the duct is neither
essential or useless, can we leave it at that?

If your 'summary' were, in fact, correct we'd have agreement but it isn't.

I have several fans on hand, so I'll see what I can do. BTW, the whole
point is to make the best of a less than ideal situation. For the PC
itself I invested ~$25 in parts, the rest was just stuff I had or got
for free. If in the end it doesn't work I'm not out much.

Why don't you post what the case is so I have some idea what it's like?

You're hampered by saying you can't cut holes. I had a super small desktop
mATX case (flex ATX actually) that had a ventilation problem and I put a
90mm fan on it's top side. Took some 'creativity' because the sides were
essentially a welded rectangular tube that the chassis slid into so you
couldn't mount the fan and then slide the chassis in. So I made a 'hook'
probe from a coat hanger, laid the fan (plugged in) on the motherboard in
roughly the right place, slid it into the case and then used the hook to
grab the fan, pull it up to the case wall, and hold it there while I screw
it down. Not exactly a convenient assembly process but it sure solved the
ventilation problem and looked great to boot.

But you have to cut a hole for it, of course.

 

[Ed Coolidge]

If your 'summary' were, in fact, correct we'd have agreement but it
isn't.

Well, the CPU can run without the duct so it's not essential, but it
surely provides some airflow over the heatsink so it's not useless either.

Why don't you post what the case is so I have some idea what it's like?

The case is from a NEC Powermate VT. It's basically a really short
tower with removable side panels. It even takes a standard ATX PSU.
The chassis is almost exactly 12" from front to back on the inside, the
same length as my Awe32 which was a bitch to shoehorn into it. The hard
drive unfortunately mounts sideways with the board against the front of
the chassis.

You're hampered by saying you can't cut holes. I had a super small
desktop mATX case (flex ATX actually) that had a ventilation problem
and I put a 90mm fan on it's top side. Took some 'creativity' because
the sides were essentially a welded rectangular tube that the chassis
slid into so you couldn't mount the fan and then slide the chassis in.
So I made a 'hook' probe from a coat hanger, laid the fan (plugged in)
on the motherboard in roughly the right place, slid it into the case
and then used the hook to grab the fan, pull it up to the case wall,
and hold it there while I screw it down. Not exactly a convenient
assembly process but it sure solved the ventilation problem and looked
great to boot.

But you have to cut a hole for it, of course.

I don't know about your particular case, but I guess that you could have
made a bracket that bolts to the chassis and holds the fan up to the
hole in the case. Either way, I've always hated those single piece
cases. They're even worse on towers.

My dad as a lot of tools and stuff so I'll just wait until I get the
time head over there and "mod" the case. One of the 120mm fans I have
setting around should be more than enough to settle the airflow issue
once I find the right place to put it.

 

[David Maynard]

Well, the CPU can run without the duct so it's not essential, but it
surely provides some airflow over the heatsink so it's not useless either.

I'm not going to argue about it because you want to play word games. Case
in point, I *can* drive my car with only one forward and one reverse gear
working properly so, using your criteria, the others are "not essential."

If, however, the issue at hand is how well the car drives (or how well the
heatsink cools the processor) then saying they're "not essential" is a word
game.

The case is from a NEC Powermate VT. It's basically a really short
tower with removable side panels. It even takes a standard ATX PSU.
The chassis is almost exactly 12" from front to back on the inside, the
same length as my Awe32 which was a bitch to shoehorn into it. The hard
drive unfortunately mounts sideways with the board against the front of
the chassis.

Sounds like a typical shallow depth micro-tower made by a myriad of
manufacturers.

I don't know about your particular case, but I guess that you could have
made a bracket that bolts to the chassis and holds the fan up to the
hole in the case.

I thought about it but there wasn't any good way to mount a bracket.

Either way, I've always hated those single piece
cases. They're even worse on towers.

I don't 'like' them either but the purpose behind that case was for
something about the size of a VCR. It fit the bill in that regard but,
unfortunately, the proprietary PSU was a piece of junk that self destructed
twice (and while OFF, no less) for no known reason.

My dad as a lot of tools and stuff so I'll just wait until I get the
time head over there and "mod" the case. One of the 120mm fans I have
setting around should be more than enough to settle the airflow issue
once I find the right place to put it.

If you're lucky the airflow might even go by the heatsink.

 

[Ed Coolidge]

I don't 'like' them either but the purpose behind that case was for
something about the size of a VCR. It fit the bill in that regard but,
unfortunately, the proprietary PSU was a piece of junk that self
destructed twice (and while OFF, no less) for no known reason.

Sounds like the typical win/lose situation. Find the right case, but
get saddled with a bad PSU.

If you're lucky the airflow might even go by the heatsink.

I would like to place the fan on the side panel over the heatsink, but
the only location I have for the PC is going to render it useless. The
only other options would be to place it farther way at the front or to
find a way to place it closer to the CPU inside the case.

 

[David Maynard]

Sounds like the typical win/lose situation. Find the right case, but
get saddled with a bad PSU.

Yep. I was not a happy camper.

I would like to place the fan on the side panel over the heatsink, but
the only location I have for the PC is going to render it useless.

Bummer. Because that sounds like the ideal spot.