Vent holes and Electro Magnetic Interference (EMI)

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 18 Jun 2004 15:11:47 -0700, "Timothy Daniels"
<TDaniels@NoSpamDot.com> wrote:

>"Battleax" wrote:
>>
>> Why are you going on and on about something that is totally obvious?
>
>
> Because to some, it's not obvious. What I have been advising is
> to put more reliance on basic science - like the pros do - than on
> the minimum requirements of an industry form factor guideline.
>
>*TimDaniels*

To pretend that you are relying on basic science is an offense to anyone,
anywhere, that follows basic scientific methods.

Google search "scientific method" and see which important steps you've
left out.

----------------------------------------------------------

[Go ahead and do that now, it is central to why you don't have a valid
argument].

----------------------------------------------------------

What you fail to realize is that this adherance to Dell-worship, even
though you try to obfuscate it by occasionally mentioning "OEMs" instead,
does nothing but hurt your credibility in general. You have established
that you don't use reason to solve problems, rather jumping to conclusions
then failing to re-evaluate even after being given more information than
used to formulate your original theory.

Based on your Dell system, you have ZERO evidence of anything, except that
it keeps the components cool enough to contiune running up until this
point. Is it not a clue to you that even Dell uses more aggressive
cooling measures on their more mission-critical oriented systems, that
just maybe in their infinite wisdom, they KNOW there is a difference,
improvements to be made over your system, but at a higher cost?
Hint- I'm not really asking a question, it is a question for you to
ponder, how it affects your perception of Dell's cooling strategy.

Digging a bit deeper, previously you felt it necessary to cut out a
grating to improve the Dell's cooling... are you now claiming it is great,
yet the user still needed to alter it, void warranty if any remained?
Seems like a great cooling design wouldn't leave a user thinking about
cutting out holes, no?

So tell us, what was the purpose behind your desire to cut out the
grating? It wouldn't have happened to be for an increase in FLOW RATE,
would it?

Until you start accumulating cases and doing your own mods to them, not
just cutting out a fan hole where there already was on,e, and can compare
before and after temps, you have no evidence that ANY of what you've
written, is correct. You repeat over and over what SOMEONE ELSE wrote, in
a DISSIMILAR SITUATION, but never anything relative. Since you don't have
any direct evidence, it is IMPOSSIBLE to make the conclusion you have and
consider it scientific.

I fully expect you'll repeat the same things all over again, because you
still can't grasp that it's not the information you have that's the
problem, it's the information you DON'T HAVE that makes your theory
incomplete, unreliable, and unscientific.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"kony" wrote:
> What you fail to realize is that this adherance to Dell-worship,
> even though you try to obfuscate it by occasionally mentioning
> "OEMs" instead, does nothing but hurt your credibility in general.


I have never used the term "OEMs". And my admiration for
the efforts of cooling engineers employed by big name vendors
is certainly not worship.


> Based on your Dell system, you have ZERO evidence of anything,


Nothing except my bachelor degree in physics and my common
knowledge of watching smoke swirls and stirring coffee and sitting
in jacuzzis - all the stuff you ignore.


> Digging a bit deeper, previously you felt it necessary to cut out a
> grating to improve the Dell's cooling...


I cut out an exit grating not because I felt it was necessary,
but because I wanted to maximize cooling. There can be
nothing wrong with getting components down to near
room temperature.


> So tell us, what was the purpose behind your desire to cut out the
> grating? It wouldn't have happened to be for an increase in
> FLOW RATE, would it?


Of course it would! It would also increase turbulence, because
turbulence rises approximately as the square of the velocity of
a fluid relative to obstructions.


> Until you start accumulating cases and doing your own mods
> to them, not just cutting out a fan hole where there already was
> on,e, and can compare before and after temps, you have no
> evidence that ANY of what you've written, is correct.


And neither have you any evidence that turbulence *doesn't*
aid heat transfer. You're just subscribing to the propaganda
(and oversight) of fan and heatsink manufacturers. I, on the
other hand, am relying on the proven scientific principles of
the effects of turbulent flow. If you want to say that scientific
research doesn't apply to your situation, say it. Say that
gravity doesn't apply as well. What you do to cool or not
cool your computer is of no concern to me in the least.

On the other hand, if a homebuilder wants to trust in scientific
research - the results of which have been published - he can
easily use it in his case set up.


> ...it's the information you DON'T HAVE that makes your theory
> incomplete, unreliable, and unscientific.


The effects of turbulence on heat transfer has gone *way*
beyond theory and they are quite reliable - like the "theory"
of gravity.

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "kony" wrote:
>
>>What you fail to realize is that this adherance to Dell-worship,
>>even though you try to obfuscate it by occasionally mentioning
>>"OEMs" instead, does nothing but hurt your credibility in general.
>
>
>
> I have never used the term "OEMs". And my admiration for
> the efforts of cooling engineers employed by big name vendors
> is certainly not worship.

It is when you ascribe to them theories you have absolutely no evidence
for; just that 'the gods must have a reason' (yours, of course)

>>Based on your Dell system, you have ZERO evidence of anything,
>
>
> Nothing except my bachelor degree in physics and my common
> knowledge of watching smoke swirls and stirring coffee and sitting
> in jacuzzis - all the stuff you ignore.


You need to learn the difference between theoretical and application.

No one disputes the 'evidence' you present, in the rare instances you
bother to present any. What they dispute are your irrational generalizations.


>>Digging a bit deeper, previously you felt it necessary to cut out a
>>grating to improve the Dell's cooling...
>
> I cut out an exit grating not because I felt it was necessary,
> but because I wanted to maximize cooling. There can be
> nothing wrong with getting components down to near
> room temperature.

As you are so want to claim, surely they did that for a reason; for the
best cooling. It MUST BE to add turbulence. It just MUST, because you
believe it to be so. If not then why did they do it? HUH? HUH? HUH?

And don't bother with an explanation because we've all learned from you the
'Tim method' of dealing with explanations: ignore them, discard all
evidence to the contrary, and repeat the same stupid question over and over.

>>So tell us, what was the purpose behind your desire to cut out the
>>grating? It wouldn't have happened to be for an increase in
>>FLOW RATE, would it?
>
> Of course it would! It would also increase turbulence, because
> turbulence rises approximately as the square of the velocity of
> a fluid relative to obstructions.
>

But they PUT it there for 'the best cooling'. They MUST have! The GRILL
does the turbulence. It MUST be. It MUST be.


>>Until you start accumulating cases and doing your own mods
>>to them, not just cutting out a fan hole where there already was
>>on,e, and can compare before and after temps, you have no
>>evidence that ANY of what you've written, is correct.
>
> And neither have you any evidence that turbulence *doesn't*
> aid heat transfer.

The item under dispute, well, only by you, is the 'intent' of cheap punched
vent holes and not the 'science' of turbulent cooling.

> You're just subscribing to the propaganda
> (and oversight) of fan and heatsink manufacturers.

Wrong. We're looking at every knowledgeable source on the matter of case
cooling. You, on the other hand, can't find ONE source to suggest your
speculation has any relationship whatsoever to why manufacturers use cheap
punched hole vents.


> I, on the
> other hand, am relying on the proven scientific principles of
> the effects of turbulent flow. If you want to say that scientific
> research doesn't apply to your situation, say it. Say that
> gravity doesn't apply as well. What you do to cool or not
> cool your computer is of no concern to me in the least.
>
> On the other hand, if a homebuilder wants to trust in scientific
> research - the results of which have been published - he can
> easily use it in his case set up.

Then provide just ONE source that talks of the need to create 'useful vent
hole turbulence' for PC case cooling, how to do it properly, and what the
benefits are.

>>...it's the information you DON'T HAVE that makes your theory
>>incomplete, unreliable, and unscientific.
>
> The effects of turbulence on heat transfer has gone *way*
> beyond theory and they are quite reliable - like the "theory"
> of gravity.

But what hasn't gone "way beyond theory" is your purely fanciful
speculation of why case manufacturers punch cheap holes in the case for vents.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "David Maynard" scoffed:
>
>>Timothy Daniels wrote:
>>
>>> The problem seems not to be with what's on the PCBs, but rather
>>> with the long, sometimes unshielded, cables - which act as
>>> radiating antennas. Consider the continuous square wave clock
>>> pulses that travel down most of the cables inside and outside the
>>> case. The spectrum of emissions from those puppies must look
>>> like a corn field.
>>>
>>>*TimDaniels*
>>
>>Ya know, Plato got just about everything wrong too by trying to simply
>>'guess' how things worked.
>
>
>
> The Amateur Radio Relay League (ARRL - the ham radio guys)
> have something to say about computer EMI on their web site at
> http://www.arrl.org/tis/info/rficomp.html :
>
> "Q: How do all of these signals get out of the computer
> system and into my station receiver?
>
> "A: Most of them are radiated, and then picked up by
> your antenna. This suggests one easy cure -- put
> as much distance between the computer system and
> your receiving antenna as you are able to.
> Interference that is intolerable with an indoor
> antenna might go away completely if you erect a
> good, outdoor antenna. In addition to reducing the
> amount of signal picked up from the computer, this
> has the benefit of increasing the amount of desired
> signal. Most computer systems have shielding to
> minimize the radiated signals. The shielding can be
> external, provided by a good metal case and shielded
> cables, or internal, provided by ground planes or
> metal partitions on critical areas on the printed
> circuit boards. This shielding is not perfect, however;
> a surprising amount of energy can be radiated from
> the slots and seams that are part of most shielding
> systems. All other things being equal, a computer with
> a metal case is usually much quieter than one with a
> plastic case. The expression "Components don't radiate
> wires do" is well-known in the EMC field. The wires
> and cables used to interconnect the parts of your
> computer system are much better antennas than are the
> printed-circuit boards used inside, so most radiation
> takes place from the wires. The video, keyboard, mouse
> and printer cables are prime suspects. The interference
> could also be conducted -- transmitted directly by wires
> connected to both the computer and your receiver. If
> your computer is plugged into the same AC circuit as is
> your receiver, you are asking for trouble. If so, you
> may be able to use an AC-line filter to filter the
> computer, your receiver or both. "
>
> On the same webpage is found:
>
> "Q: How is this noise generated by a computer?
>
> "A: There are several things in a computer that can
> generate noise. All computers use digital signals --
> square waves rich in harmonics. These signals can be
> generated by the several oscillators found in most
> computer systems. Signals from the oscillators can
> interfere with the signals we want to receive. In
> addition to the oscillators, all computer circuits
> sub-divide these oscillators into signals that are
> sub-multiples of the oscillator frequencies.
> Additional digital noise can be generated by the
> video monitor circuits. Computers also use switching
> power supplies. "Switchers" can also be prolific
> generators of RF noise. The monitor has a separate
> power supply, plus sweep and high-voltage circuits
> that can also generate noise. When you put them all
> together, a computer system can generate RF signals
> from below the HF band well into VHF!"
>
>
> So not only are square waves rich in harmonics (which
> any engineer knows who has taken a course in Fourier
> Analysis), a computer, by its frequency dividers, also
> produces sub-harmonics, i.e. frequencies *lower* than
> the fundamental frequency. That is why a display of the
> emitted spectrum of a computer is rich in spikes and
> "looks like a cornfield" or like a Roman army.
>
> *TimDaniels*

I am well aware of EMI and harmonics.

And just about every signal on the motherboard you discounted is a 'square
wave' too and not just in the cables you speculated about.

And, contrary to your speculation, there are no unshielded cables carrying
high frequency signals outside the box. And even what you apparently
consider an 'unshielded cable' inside the box, e.g. the IDE ribbon cable,
has a ground in-between every signal wire that acts as a shield. And every
low frequency signal that penetrates the box goes through low pass
filtering to suppress high frequency EMI.

And, of course, there is the shielding of the metal chassis with it's cheap
punched vent holes.


Your article, trying to explain how a 'computer' of unspecified manufacture
and age could cause interference with a ham radio, is good enough for
explaining the one, limited, thing it's explaining but it isn't terribly
useful as a tutorial on modern EMI suppression techniques in FCC compliant
computer systems, other than hinting at a few of the things the design
engineer has to deal with in achieving it.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"David Maynard" wrote:
>
> I am well aware of EMI and harmonics.
>
> And just about every signal on the motherboard you
> discounted is a 'square wave' too and not just in the
> cables you speculated about.
>
> And, contrary to your speculation, there are no unshielded
> cables carrying high frequency signals outside the box.
> And even what you apparently consider an 'unshielded cable'
> inside the box, e.g. the IDE ribbon cable, has a ground
> in-between every signal wire that acts as a shield.


A shield to suppress crosstalk between signal-carrying
wires within the cable, not so much as to suppress EMI
radiating from the cable. Indeed, if you can see the
wire along its entire length, it ain't shielded.


> And every low frequency signal that penetrates the box
> goes through low pass filtering to suppress high frequency EMI.


It's good to hear that EMI has been so successfully suppressed.


> And, of course, there is the shielding of the metal chassis
> with it's cheap punched vent holes.


And its leaky slits in the back which leak EMI like a
sieve - if your contention about hole shape is true,
and which make the size and shape of the intake holes
totally moot for EMI suppression.


> Your article,


No, not *my* article, the Amateur Radio Relay League's
article.


> trying to explain


*successfully* explaining, based on many decades of RFI and
EMI investigations and solutions.


> .....how a 'computer' of unspecified manufacture and age could
> cause interference with a ham radio, is good enough for
> explaining the one, limited, thing it's explaining


Yes - EMI.


> ....but it isn't terribly useful as a tutorial on modern EMI
> suppression techniques in FCC compliant computer systems,


I see. ARRL's computers aren't modern. LOL.


> ...other than hinting at a few of the things the design
> engineer has to deal with in achieving it.


What is it that ARRL is wrong about?.

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "David Maynard" wrote:
>
>>I am well aware of EMI and harmonics.
>>
>>And just about every signal on the motherboard you
>>discounted is a 'square wave' too and not just in the
>>cables you speculated about.
>>
>>And, contrary to your speculation, there are no unshielded
>>cables carrying high frequency signals outside the box.
>>And even what you apparently consider an 'unshielded cable'
>>inside the box, e.g. the IDE ribbon cable, has a ground
>>in-between every signal wire that acts as a shield.
>
>
>
> A shield to suppress crosstalk between signal-carrying
> wires within the cable, not so much as to suppress EMI
> radiating from the cable. Indeed, if you can see the
> wire along its entire length, it ain't shielded.

It does both.

>>And every low frequency signal that penetrates the box
>>goes through low pass filtering to suppress high frequency EMI.
>
> It's good to hear that EMI has been so successfully suppressed.

To the point they pass FCC regs, yes.

>
> > And, of course, there is the shielding of the metal chassis
>
>>with it's cheap punched vent holes.
>
> And its leaky slits in the back which leak EMI like a
> sieve - if your contention about hole shape is true,
> and which make the size and shape of the intake holes
> totally moot for EMI suppression.

And who's case with "leaky slits in the back which leak EMI like a sieve"
are you talking about? Oh wait. never mind. It's another one of your
fanciful 'imaginations'.

>
>>Your article,
>
> No, not *my* article, the Amateur Radio Relay League's
> article.
>
>
>
>>trying to explain
>
> *successfully* explaining, based on many decades of RFI and
> EMI investigations and solutions.

Meaningless hand waving.

>
>>.....how a 'computer' of unspecified manufacture and age could
>>cause interference with a ham radio, is good enough for
>>explaining the one, limited, thing it's explaining
>
>
> Yes - EMI.

In a limited generalized context.

>
>>....but it isn't terribly useful as a tutorial on modern EMI
>>suppression techniques in FCC compliant computer systems,
>
>
> I see. ARRL's computers aren't modern. LOL.
>

You have no idea what their computers are or what computers they were
talking about. Nor do THEY when telling people what to 'look out for' with
whatever the hell 'computer' they've shoved next to their ham radio.

>
>>...other than hinting at a few of the things the design
>>engineer has to deal with in achieving it.
>
>
> What is it that ARRL is wrong about?.

I didn't say they were 'wrong' about anything. It's your irrational
bounding leaps of illogic that are wrong.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"David Maynard" srote:
> Timothy Daniels wrote:
>
> > "David Maynard" wrote:
> >>.....there are no unshielded
> >>cables carrying high frequency signals outside the box.
> >>And even what you apparently consider an 'unshielded cable'
> >>inside the box, e.g. the IDE ribbon cable, has a ground
> >>in-between every signal wire that acts as a shield.
> >
> > A shield to suppress crosstalk between signal-carrying
> > wires within the cable, not so much as to suppress EMI
> > radiating from the cable. Indeed, if you can see the
> > wire along its entire length, it ain't shielded.
>
> It does both.


Poorly.


> > And its leaky slits in the back which leak EMI like a
> > sieve - if your contention about hole shape is true,
> > and which make the size and shape of the intake holes
> > totally moot for EMI suppression.
>
> And who's case with "leaky slits in the back which leak
> EMI like a sieve" are you talking about?


From your vaunted Form Factors link,
http://www.formfactors.org/developer/specs/microatx_thermal_dg01.pdf :

"2.5.2 Apertures
Keep maximum linear dimensions of ventilation apertures,
I/O ports, and open areas along chassis seams less than
1/20th of a wavelength ( l ) of the highest harmonic frequency
( f ) of concern (1/20th rule, see also Figure 2). Absorption
and shield thickness may contain low frequency magnetic fields,
but high frequency electric field radiation out of slots becomes
the next concern. Apertures (or slots) can be viewed as
half-wave dipole antennas and are thus able to radiate
maximum energy at dimensions of 1/2 a wavelength. In fact,
slots longer than 1/100th of a wavelength can cause
considerable leakage. Therefore, it is necessary to keep slot
lengths as short as possible to minimize leakage. Currently,
the FCC has requirements up to 2 GHz, which, as derived below,
correspond to a recommended maximum aperture size
(vertical or horizontal) of about .75 centimeters;"


In other words, those leaky cracks in the back of a computer
case, the ones between the PCI board brackets, the ones
that collect dust from air leaking in - those are long apertures,
i.e. "half-wave dipole antennas", which leak EMI out the kazoo.
So are the cracks in the front of the case, the ones between
the ATAPI devices and the case, the cracks that everyone
sees filled with dust just like the ones in the back. Those, too,
act as half-wave dipole antennas, leaking EMI like a sieve.
To think that the vent holes in the front are designed primarily
with EMI in mind is to ignore what they obviously do by the
very nature of aerodynamics and the principles of fluid flow - they
generate turbulence, hated turbulence, flow-restricting turbulence,
bad bad bad turbulence, COOLING turbulence.

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"Timothy Daniels" <TDaniels@NoSpamDot.com> wrote in message
newsaudnfYz54kC9U7dRVn-uw@comcast.com...
> "Battleax" wrote:
> >
> > Why are you going on and on about something that is totally obvious?
>
>
> Because to some, it's not obvious. What I have been advising is
> to put more reliance on basic science - like the pros do - than on
> the minimum requirements of an industry form factor guideline.
>
> *TimDaniels*

The fact is that turbulence is inherent in a PC. All I think about is just
getting the hot air out and the cool air in and creating a clear path for
airflow. Turbulence takes care of itself with all the crooks and crannies
and cabling present.

Ed

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"Ed Medlin" wrote:
>
> The fact is that turbulence is inherent in a PC. All I
> think about is just getting the hot air out and the
> cool air in and creating a clear path for airflow.
> Turbulence takes care of itself with all the crooks
> and crannies and cabling present.


Yes. That is characteristic of downstream air flow
after it has encountered the cabling and the various
components on the PCBs. But upstream, near the
head of the air flow where such components as the
hard drives reside, the turbulence greatly aids the
cooling. Indeed, tests have shown that cooling
hard drives by blowing along their length does not
help to cool them as much as a fan having blades
which sweep over the hard drive's flat side,
sweeping away the boundary layer and sweeping
cool air in to contact the hard drive's case. Such
components don't generate their own turbulence
as well as the surface of a printed circuit board,
and turbulating vent holes do that for them.

Furthermore, the swirling and buffeting of turbulent
flow helps air to get into those nooks and crannies
because turbulent flow includes air having a
constantly changing direction. If there is a little
nook where the steady-state flow does not reach,
it will be reached by the turbulent flow.

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"David Maynard" wrote:
> Not to mention that there ARE cases that use 'large holes'
> with wire grills, but then the only damn thing you apparently
> know about is some Dell case so you dream how you've
> divined the secret to eternal life by 'guessing' what the
> designers were thinking when they made it, and to hell
> with all the literature which uniformly and without exception
> contradicts your 'faith'.


My faith begins with knowing how air flows, and by the
writing of others who are more informed:
http://www.overclockers.com/tips90/
http://www.begellhouse.com/books/497d60632054f587,6ddfe1a32b58c789.html
http://www.cougarlabs.com/cool2.html
http://www.ceere.org/beep/docs/FY2002/Turbulent_Flow_in_Enclosure.pdf



> And THAT you have the audacity to call 'science'.


What do *you* call science?


> A flat lie on your part. And no, I'm not going to repost the link to the
> waveguide filter design, the 'airflow straightener' filter, or any of the
> others.


Of course not. It was written by a maker and seller of
collimating filters! He *certainly* isn't going to point out
that his filters do the opposite of what is desired.



> Oh sure. I see LOTs of Dell ads talking about their 'superior
> turbulent vent hole' designs. LOL


Of course not. It would go against the Common Wisdom
of case ventilation and confuse buyers. Better to go along
with the published (and misleading, and simpler to "understand")
guidelines which say "maximize air flow". Too bad that people
don't understand that "maximize air flow" means "maximize
velocity along the surface undergoing heat exchange", because
that is what turbulence does.

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Sun, 20 Jun 2004 11:46:21 -0700, "Timothy Daniels"
<TDaniels@NoSpamDot.com> wrote:

<the same guesses posted a dozen times already>

Those guesses are still wrong.
We told you why, yet you seem to think reposting same thing over and over
makes a difference. If I tell you the moon is made of chesse, 1,000 times
in a row, it's no more true than the first time.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"kony" wrote:
> Those guesses are still wrong.


If you want to think that turbulence doesn't aid in
heat transfer, go ahead. The fan manufacturers
and the fan sellers depend on your ignorance.

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:
> "David Maynard" srote:
>
>>Timothy Daniels wrote:
>>
>>
>>>"David Maynard" wrote:
>>>
>>>>.....there are no unshielded
>>>>cables carrying high frequency signals outside the box.
>>>>And even what you apparently consider an 'unshielded cable'
>>>>inside the box, e.g. the IDE ribbon cable, has a ground
>>>>in-between every signal wire that acts as a shield.
>>>
>>> A shield to suppress crosstalk between signal-carrying
>>> wires within the cable, not so much as to suppress EMI
>>> radiating from the cable. Indeed, if you can see the
>>> wire along its entire length, it ain't shielded.
>>
>>It does both.
>
>
>
> Poorly.
>

That is, no doubt, the 'scientific' usage of 'poorly' you mean.

>
>>> And its leaky slits in the back which leak EMI like a
>>> sieve - if your contention about hole shape is true,
>>> and which make the size and shape of the intake holes
>>> totally moot for EMI suppression.
>>
>>And who's case with "leaky slits in the back which leak
>>EMI like a sieve" are you talking about?
>
>
>
> From your vaunted Form Factors link,
> http://www.formfactors.org/developer/specs/microatx_thermal_dg01.pdf :
>
> "2.5.2 Apertures
> Keep maximum linear dimensions of ventilation apertures,
> I/O ports, and open areas along chassis seams less than
> 1/20th of a wavelength ( l ) of the highest harmonic frequency
> ( f ) of concern (1/20th rule, see also Figure 2). Absorption
> and shield thickness may contain low frequency magnetic fields,
> but high frequency electric field radiation out of slots becomes
> the next concern. Apertures (or slots) can be viewed as
> half-wave dipole antennas and are thus able to radiate
> maximum energy at dimensions of 1/2 a wavelength. In fact,
> slots longer than 1/100th of a wavelength can cause
> considerable leakage. Therefore, it is necessary to keep slot
> lengths as short as possible to minimize leakage. Currently,
> the FCC has requirements up to 2 GHz, which, as derived below,
> correspond to a recommended maximum aperture size
> (vertical or horizontal) of about .75 centimeters;"
>
>
> In other words, those leaky cracks in the back of a computer
> case, the ones between the PCI board brackets, the ones
> that collect dust from air leaking in - those are long apertures,
> i.e. "half-wave dipole antennas", which leak EMI out the kazoo.
> So are the cracks in the front of the case, the ones between
> the ATAPI devices and the case, the cracks that everyone
> sees filled with dust just like the ones in the back. Those, too,
> act as half-wave dipole antennas, leaking EMI like a sieve.
> To think that the vent holes in the front are designed primarily
> with EMI in mind is to ignore what they obviously do by the
> very nature of aerodynamics and the principles of fluid flow - they
> generate turbulence, hated turbulence, flow-restricting turbulence,
> bad bad bad turbulence, COOLING turbulence.

In the first place, the joint between mounting plates and chassis aren't
"ventilation apertures" nor are they "slots.

Hell, forget it. All you do is 'speculate' and 'guess' about things you
haven't a clue about and then ignore reality in favor of your bogus
speculation and guesses.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "David Maynard" wrote:
>
>>Not to mention that there ARE cases that use 'large holes'
>>with wire grills, but then the only damn thing you apparently
>>know about is some Dell case so you dream how you've
>>divined the secret to eternal life by 'guessing' what the
>>designers were thinking when they made it, and to hell
>>with all the literature which uniformly and without exception
>>contradicts your 'faith'.
>
>
>
> My faith begins with knowing how air flows, and by the
> writing of others who are more informed:
> http://www.overclockers.com/tips90/
> http://www.begellhouse.com/books/497d60632054f587,6ddfe1a32b58c789.html
> http://www.cougarlabs.com/cool2.html
> http://www.ceere.org/beep/docs/FY2002/Turbulent_Flow_in_Enclosure.pdf

Yes, it 'begins' there and then you layer speculation, guess, and fantasy
on top of it, as is done with any myth.

>
>>And THAT you have the audacity to call 'science'.
>
> What do *you* call science?

Confirmation of hypothesis by direct observation.

It is certainly not "I imagine the designer's thought, ergo they did."


>>A flat lie on your part. And no, I'm not going to repost the link to the
>>waveguide filter design, the 'airflow straightener' filter, or any of the
>>others.
>
> Of course not. It was written by a maker and seller of
> collimating filters! He *certainly* isn't going to point out
> that his filters do the opposite of what is desired.
>

Only one was written by the 'maker and seller' and that you've decided to
invent a conspiracy theory to discount facts which contradict your fanciful
mythology is just so typical of fanatics.


>>Oh sure. I see LOTs of Dell ads talking about their 'superior
>>turbulent vent hole' designs. LOL
>
>
>
> Of course not. It would go against the Common Wisdom
> of case ventilation and confuse buyers.

ROTFLOL. And now the 'selling advantage' is kept 'secret' lest if confuse
the buyers. "Buy our stuff. Why? Oh, that's a secret."

> Better to go along
> with the published (and misleading, and simpler to "understand")
> guidelines which say "maximize air flow". Too bad that people
> don't understand that "maximize air flow" means "maximize
> velocity along the surface undergoing heat exchange", because
> that is what turbulence does.

Waaaaaahahahahaha

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "kony" wrote:
>
>>Those guesses are still wrong.
>
>
>
> If you want to think that turbulence doesn't aid in
> heat transfer, go ahead. The fan manufacturers
> and the fan sellers depend on your ignorance.

And this is another all too common 'tactic' I despise, because it is
dishonest: misrepresenting what someone says and then 'arguing' against the
false representation

> *TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"David Maynard" wrote:
> In the first place, the joint between mounting plates and chassis aren't
> "ventilation apertures" nor are they "slots.


Your notion of a "slot" is very limited to that of a low aspect
ratio rectangular hole.


> Hell, forget it. All you do is 'speculate' and 'guess' about things you
> haven't a clue about and then ignore reality in favor of your bogus
> speculation and guesses.


Oh, I have a clue. And it's derived from basic science, not
dependence on industry standard "guidelines" to tell me what
and how to think.

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"David Maynard" wrote:
> Only one was written by the 'maker and seller' and
> that you've decided to invent a conspiracy theory
> to discount facts which contradict your fanciful
> mythology is just so typical of fanatics.


A conspiracy theory is not necessary to see that
the maker of collimating filters has some desire
to have you believe that collimated air flow is best
for cooling.


> > Of course not. It would go against the Common Wisdom
> > of case ventilation and confuse buyers.
>
> ROTFLOL. And now the 'selling advantage' is kept 'secret' lest if confuse
> the buyers. "Buy our stuff. Why? Oh, that's a secret."


Odd as it may seem, a boast of "turbulent intake" would
certainly turn *you* away, wouldn't it? And, odder still,
Dell considers your money as good as anyone else's. So
Dell wouldn't tell you what techniques it uses to do its
cooling.

Too bad that people don't understand that "maximize air flow"
means "maximize velocity along the surface undergoing heat
exchange", because that is what turbulence does - it super-
positions a random distribution of velocities on top of the
velocity of the steady-state flow. The square *root* of the
time-average of the square of those added velocities is a
scalar value, a speed - which adds to the speed of the steady-
state flow. And it is that sum of average speeds across the
surface that cools (since cooling doesn't depend on the
direction of flow), an average speed that has been made
greater than that of the steady-state flow by the addition
of turbulence. Now go "Waaaaaahahahahaha", like an
imbecile.

>Waaaaaahahahahaha

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Sun, 20 Jun 2004 17:42:41 -0700, "Timothy Daniels"
<TDaniels@NoSpamDot.com> wrote:

>"kony" wrote:
>> Those guesses are still wrong.
>
>
> If you want to think that turbulence doesn't aid in
> heat transfer, go ahead.

Ignoring the facts will not help you prove anything.
I never claimed turbulence doesn't aid in heat transfer. I claimed it
will not help when it is created at the expense of flow rate, far from the
component being cooled. Your examples coincide with this too, but still
you choose to ignore it.

> The fan manufacturers
> and the fan sellers depend on your ignorance.
>
>*TimDaniels*

Go ahead and create intake turbulence if you want, then you'll be the one
needing more fans or more powerful fans to overcome the problem you
created, to achieve the needed flow rate. Given a chassis with components
creating a constant amount of heat, turbulence MUST remove that heat from
the chassis, it can't just keep traveling around though the case, getting
hotter and hotter. Of course it WILL exit the case still, but it was
hotter all along it's travel, reducing it's effectiveness by the time it
reached the upper area of the chassis. You keep ignoring this mitigating
factor and so you never had a chance for a valid argument. Instead you
like a plan that allows only the hard drives to receive enough cooling,
only if the exhaust fan is powerful enough.

Several times I"ve mentioned, several times you've ignored, that for your
grand turbulent hole theory to be EVEN CONSIDERED, it would have to be
observed ONE SINGLE TIME, but you haven't observed it, never did a single
test. Cut out that grill in your Dell , below the hard drive, then tell
us what happened to chassis temps. hint - they go down, I've done it
myself several times with several different OEM cases.

On the other hand, EM emissions will go up. That has been established,
has not been argued against. The issue is then whether this causes any
problem. If you have a HAM radio next to your case and suddenly have
problems, then of course the EMI has to be dealt with.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

"kony" wrote:
> Cut out that grill in your Dell , below the hard drive, then tell
> us what happened to chassis temps. hint - they go down,
> I've done it myself several times with several different OEM cases.


You cut out the grill, then measured chassis temps.
Why didn't you measure the temp of the hard drive?

And that's not a measure of the value of turbulence
because the steady-state air flow changed. You
should have slowed the exit fan so that the air flow
remained the same, *then* measured the temp of
the hard drive. Only by varying just the turbulence
could you have seen what it added to cooling.
Instead, you measured the effect of increased
steady-state air flow.


> On the other hand, EM emissions will go up. That has been established,


Did you measure the EMI?

*TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "David Maynard" wrote:
>
>>In the first place, the joint between mounting plates and chassis aren't
>>"ventilation apertures" nor are they "slots.
>
>
>
> Your notion of a "slot" is very limited to that of a low aspect
> ratio rectangular hole.
>

More 'guesses', eh?

>
>>Hell, forget it. All you do is 'speculate' and 'guess' about things you
>>haven't a clue about and then ignore reality in favor of your bogus
>>speculation and guesses.
>
>
>
> Oh, I have a clue. And it's derived from basic science, not
> dependence on industry standard "guidelines" to tell me what
> and how to think.

A claim which proves you haven't a clue.

> *TimDaniels*

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "David Maynard" wrote:
>
>>Only one was written by the 'maker and seller' and
>>that you've decided to invent a conspiracy theory
>>to discount facts which contradict your fanciful
>>mythology is just so typical of fanatics.
>
>
>
> A conspiracy theory is not necessary to see that
> the maker of collimating filters has some desire
> to have you believe that collimated air flow is best
> for cooling.

Of course it is, because the unstated presumption in your conspiracy theory
is that everyone is inherently a self serving liar.

>
>>> Of course not. It would go against the Common Wisdom
>>> of case ventilation and confuse buyers.
>>
>>ROTFLOL. And now the 'selling advantage' is kept 'secret' lest if confuse
>>the buyers. "Buy our stuff. Why? Oh, that's a secret."
>
>
>
> Odd as it may seem, a boast of "turbulent intake" would
> certainly turn *you* away, wouldn't it? And, odder still,
> Dell considers your money as good as anyone else's. So
> Dell wouldn't tell you what techniques it uses to do its
> cooling.

More 'guesses'. And, as usual, incorrect ones.

> Too bad that people don't understand that "maximize air flow"
> means "maximize velocity along the surface undergoing heat
> exchange", because that is what turbulence does - it super-
> positions a random distribution of velocities on top of the
> velocity of the steady-state flow. The square *root* of the
> time-average of the square of those added velocities is a
> scalar value, a speed - which adds to the speed of the steady-
> state flow. And it is that sum of average speeds across the
> surface that cools (since cooling doesn't depend on the
> direction of flow), an average speed that has been made
> greater than that of the steady-state flow by the addition
> of turbulence. Now go "Waaaaaahahahahaha", like an
> imbecile.

Too bad YOU don't understand and prefer 'guessing'.

>
>>Waaaaaahahahahaha
>
>
> *TimDaniels*
>
>

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Timothy Daniels wrote:

> "kony" wrote:
>
>>Cut out that grill in your Dell , below the hard drive, then tell
>>us what happened to chassis temps. hint - they go down,
>>I've done it myself several times with several different OEM cases.
>
>
>
> You cut out the grill, then measured chassis temps.
> Why didn't you measure the temp of the hard drive?
>
> And that's not a measure of the value of turbulence

The purpose isn't to measure "the value of turbulence." The purpose is to
determine whether having small vent holes or a big hole provides better
cooling.

> because the steady-state air flow changed.

Which is what we've been telling you all along.

> You
> should have slowed the exit fan so that the air flow
> remained the same, *then* measured the temp of
> the hard drive.

Artificially forcing the air flow to a lower value isn't the 'real world'
and we are not concerned with your 'obsession' of 'turbulence'; the goal is
to efficiently cool a PC and the point of the experiment is to see if a
bunch of small holes or a large hole works best.

> Only by varying just the turbulence
> could you have seen what it added to cooling.

If it added anything at all it was less than the drop in airflow so it does
not 'help' with the purpose.

> Instead, you measured the effect of increased
> steady-state air flow.

We ended up with a cooler PC, which was the POINT.

>>On the other hand, EM emissions will go up. That has been established,
>
>
>
> Did you measure the EMI?

Are you disputing that gaping holes let more EMI leak than small ones?

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Sun, 20 Jun 2004 23:33:09 -0700, "Timothy Daniels"
<TDaniels@NoSpamDot.com> wrote:

<snip>

> Oh, I have a clue. And it's derived from basic science, not
> dependence on industry standard "guidelines" to tell me what
> and how to think.

Yes, your clue is based on science, but then you ignored everything but
that one single clue, where there were clearly other relevant clues.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Mon, 21 Jun 2004 00:26:41 -0700, "Timothy Daniels"
<TDaniels@NoSpamDot.com> wrote:

>"kony" wrote:
>> Cut out that grill in your Dell , below the hard drive, then tell
>> us what happened to chassis temps. hint - they go down,
>> I've done it myself several times with several different OEM cases.
>
>
> You cut out the grill, then measured chassis temps.
> Why didn't you measure the temp of the hard drive?

Did I claim I didn't?
Why would I mention "below the hard drive" if I weren't planning flow to
keep drive cool too?

In case you hadn't noticed, a hard drive is not particularly hard to cool,
and no, I do not only measure chassis temp, also video card, CPU, the
integral temp sensor on the motherboard, and power supply exhaust temp.
They're all cooler.


>
> And that's not a measure of the value of turbulence
> because the steady-state air flow changed.

This is exactly what you kept ignoring all along.
Maybe you are ready to learn?
The creation of turbulence decreases flow rate, period. There is no way
around it, whatever hole size and number you start out with, the flow is
impeded at the point where the turbulence is created.


> You
> should have slowed the exit fan so that the air flow
> remained the same, *then* measured the temp of
> the hard drive.

That's exactly what we DON'T want to do. You still don't get it... the
goal is not to have the exact same flow rate, the goal is to determine if
creation of turbulence, at the holes, is a desired thing. If they (case
designers) had made an effort, it would've been to decrease turbulence so
that the flow rate went up.


> Only by varying just the turbulence
> could you have seen what it added to cooling.
> Instead, you measured the effect of increased
> steady-state air flow.

Again you ignore IMPLEMENTATION of your theory. Given exact same case,
same fan, same components, etc, etc, the system will become cooler the
more the intake turbulence is reduced, because the flow rate increases.

If you want to consider the exact same flow rate, then with a reduction in
turbulence a lower speed fan can be used to reduce noise and wear.
Either way, the turbulence at the intake point is a detriment. To
maintain a system at same temp with either intake "design", a
turbulent-intake design will cause the system to be louder.


>> On the other hand, EM emissions will go up. That has been established,
>
>
> Did you measure the EMI?

No. I did not need to because this is clearly specified, tested, not
just a theory in an isolated, different situation but rather the same
situation.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

In article <uLOdnW-dVN2EF0vdRVn-uw@comcast.com>,
Timothy Daniels <TDaniels@NoSpamDot.com> wrote:
>"David Maynard" wrote:
>> Only one was written by the 'maker and seller' and
>> that you've decided to invent a conspiracy theory
>> to discount facts which contradict your fanciful
>> mythology is just so typical of fanatics.
>
>
> A conspiracy theory is not necessary to see that
> the maker of collimating filters has some desire
> to have you believe that collimated air flow is best
> for cooling.
>
>
>> > Of course not. It would go against the Common Wisdom
>> > of case ventilation and confuse buyers.
>>
>> ROTFLOL. And now the 'selling advantage' is kept 'secret' lest if confuse
>> the buyers. "Buy our stuff. Why? Oh, that's a secret."
>
>
> Odd as it may seem, a boast of "turbulent intake" would
> certainly turn *you* away, wouldn't it? And, odder still,
> Dell considers your money as good as anyone else's. So
> Dell wouldn't tell you what techniques it uses to do its
> cooling.
>
> Too bad that people don't understand that "maximize air flow"
> means "maximize velocity along the surface undergoing heat
> exchange", because that is what turbulence does - it super-
> positions a random distribution of velocities on top of the
> velocity of the steady-state flow. The square *root* of the
> time-average of the square of those added velocities is a
> scalar value, a speed - which adds to the speed of the steady-
> state flow. And it is that sum of average speeds across the
> surface that cools (since cooling doesn't depend on the
> direction of flow), an average speed that has been made
> greater than that of the steady-state flow by the addition
> of turbulence. Now go "Waaaaaahahahahaha", like an
> imbecile.
>
>>Waaaaaahahahahaha
>
>*TimDaniels*
>
>


FYI: here's a nice paper on the issues for cooling a PC case.

http://www.electronics-cooling.com/Resources/EC_Articles/
MAY96/may96_01.htm

ISTM that we have confused "laminar airflow heat transfer" (which
applies only to maybe a few hundredths on an inch over a surface)
and eliminating hot spots inside a PC chassis.

Eliminating hotspots in a case isn't rocket science, and only vaguely
related to turbulance, or laminar flow in the aerodynamic sense. In
the engineering papers I've found with google it's a minor issue.

FWIW I have seen so many cost-is-no-object large electronics systems,
over (gaaad) 40 years and never seen a single one that was designed
either for laminar airflow, or for turbulance in any way that you
wouldn't recognize on an audiophlile 200 watt Class A amp, or kilowatt
solid state transnmitter. Two systems I worked with had power supplies
that need 5 people to swap (4 to lift, 1 to unbolt.) Mainframe power
supplies that predate switching were _huge_ and massively inefficient.

The hottest air-cooled mainframes I've seen simply has a huge number
of 8 inch fans which moved massive amunts of air. The fans made enough
noise that there were custmer complaints and maybe OSHA problems. For
the follow-on model the manufacturer hired thermdynamics consultants
from the automotive industry. They put in foam-padded ducts to direct
the airflow and dampen the radiated noise. Any laminar (or turbulence)
issuse were secondary to getting enugh air to the right places in the
cabinet. Anything larger than these systems gets water or freon
cooling.

Since laminar flow is _very hard_ to maintain, it's safe to say that
in the heat-transfer region, it's almost impossible to have laminar
flow unless the HS is convection cooled.






--
Al Dykes
-----------
adykes at p a n i x . c o m