a silent air cooled computer project

[Floyd L. Davidson]

My highest priority is noise actually, so pump noise a major concern.
But I agree - it's very cool and I might do it just to do it, although
playing with airflow can be fun too. I had so much fun I fried one
motherboard trying to mount a fan (accidently moved the cpu, I guess) - was
too lazy to turn the system off

You've got a great setup for having *fun* too. Being able to
actually watch the results, and adjust fan speed etc while
getting actual feedback on results is great.

I'm more into software than hardware these days, so the
Crystalfontz unit suits my needs; but for anyone who doesn't
want to do the software a manual controller like you are using
is the equivalent as far as fun goes.

Can you post a URL to a product info sheet or something similar
for the units you've got? Might be interesting...

 

[stormrider]

http://silentgamingreview.com/components.htm
This is a listing by component - there are full model names etc. No product
sheets but you could google it
I would do the software, but I could not find anything that can turn the
fans completely off.

 

[stormrider]

I thought about that but I think the recommended weight limit is way too
concervative. There is a metal plate behind the socket that is distributing
the stress.
I lug my computer around without mercy and tower is mounted horizontally
with additional weight of two aluminum fans and a heat sync (the heat synch
is very light on it's own) Check this out:
http://silentgamingreview.com/images/coolmaster fan2.jpg (computer lying
on its side)
http://silentgamingreview.com/images/coolmaster fan1.jpg
So far no problems. I think it's a non-issue in this case. If you really
want to be safe, just put the computer horizontally making the tower
vertical

if anyone is curious I documented all the steps and components that I
used to build a silent air cooled pc for games:
http://silentgamingreview.com/components.htm

The problem with the huge and heavy Thermaltake Sonic Tower CL-P0071, and
others like it, is that they far exceed the recommended weight limit for
the CPU socket lugs. This heatsink is more than twice that maximum. You
may end up with broken lugs after awhile (unless you orient the
motherboard horizontally in a desktop case rather than vertically in a
[mini-]tower case). You *definitely* need to remove the overweight
heatsink before transporting the computer.

 

[Floyd L. Davidson]

http://silentgamingreview.com/components.htm

This is a listing by component - there are full model names
etc. No product sheets but you could google it

Turns out that it isn't easy to find information on anything from
Vantec Nexus. Here's the best review that I could find, as far
as giving at least a few specs. But it is still very unclear as
to what it can actually do.

http://silentgamingreview.com/components.htm

I would do the software, but I could not find anything that can
turn the fans completely off.

The Crystalfontz CF633 can turn the fans off. It can also
measure their speed while not running at full throttle too,
which is tricky when they are controlled with PWM.

The CF633 is far more versatile, but it is *totally* under
software control. Hence it can either be used with available
software, to whatever degree that happens to fit needs, or
custom software has to be written. I use Linux and custom
software, so I'm not at all sure what the various Windows
software packages will do.

The NXP-301 and NXP-305 are totally manual, which is perfect
non-programmers who shut their computers off when they aren't
sitting at the console. They also include a lot of purely
snazzy stuff for eye appeal (light controllers!).

Hence a CF633 is much more likely to end up in a server or an
engineering model, while the NXP-305 is more appealing for a
"consumer" system.

 

[stormrider]

CF633 looks awesome! I am gonna get me on these!

 

[Floyd L. Davidson]

CF633 looks awesome! I am gonna get me on these!

I have to warn you though, the price is almost exactly *twice*
what they say it is!

Yep. That's cause as soon as you get one and put it into
service actually doing something... you can't play with it any
more and the only solution is to buy another one just to play
with.

 

[kony]

All of them?

How many parts do you expect to be water-cooled? It's
usually not but a handful if more than 2. Plus, once one
has a single one actively monitored, it is not such an issue
using same monitoring technology for another.

Monitoring coolant flow in one location is an "enterprise-class"
item, while monitoring every part that is heat sensitive is a
low budget option?

Yes. You are proposing a "from scrach" solution that will
require more expensive parts, engineering to construct and
testing to validate. Show us a ready-to-use affordable,
small, suitable, etc, solution. Even a good theory on how
to do something must be weighted against the actual
implementation burden.

You don't seem to have just a whole lot of experience with this
kind of stuff.

That's funny.
SHOW US this grand concept you have- on water-cooled PC
class systems in actual commercial PC use. Show us sytems
that only use component thermal feedback control that failed
because of lack of your (implied necessary) water flow
sensor idea.

In short, provide any evidence that what you claim is
necessary, is, OR that it could even be implemented in a
timely or cost-effective manner. Plenty of water-cooled
systems out there are running using reasonable protection
that doesn't include your theoretically necessary control,
which I still contend is inferior to thermal feedback from
the cooled parts.

You still have not provided any reason/theory, let alone
evidence, that the thermal state of the cooling liquid is a
better control for the thermal state of the cooled
component, than consideration of the thermal state of that
component itself. On the contrary, we really DON"T CARE
what temp the water is nor how fast it's flowing so long as
the components stay cool. On the contrary, since the
system needs a mimal level of airflow either way, a system
with utmost noise reduction would turn off the water pump
completely if flow rate wasn't easily adjustable, IF the
thermals allowed it.

I've not suggested that option for one of the several
reasons your theory isn't practical- that requires
engineering of the solution and testing before being
considered reliable. Which year is it considered reliable?
It's being considered for use NOW, not after beta testing
and refinements at additional cost.

Different systems have different requirements. Some
distinctly different (non-PC, even non-computer) system's
use of water does not make that system-specific solution
applicable to water-cooling-at large.

 

[Floyd L. Davidson]

How many parts do you expect to be water-cooled? It's
usually not but a handful if more than 2. Plus, once one
has a single one actively monitored, it is not such an issue
using same monitoring technology for another.

Your statement was only regarding those items that a lack of
cooling will affect, which now you are attempting to limit to
only the items that are "water-cooled", but in fact it is far
more serious than that. *Everything* in the case is "in
jeopardy", not just those parts that are water cooled. The
entire system has to be monitored according to your concept!

Lets just consider one item... the difference between having a
coolant flow meter on the return entry to the surge tank, and
doing it your way. I'll limit it to that because this one item
is more significant than all of what you believe must be done.

Eventually one of the things that happens with liquid cooling
systems is that they leak. Hoses pop off, fittings break,
something slits, whatever... if we have a flow meter where it
should be, the *immediate* effect is to shut the system down.

Or... if the hose exiting the water block has broken and we
wait until the cpu gets hot... the entire surge tank will be
pumped dry, and all of that coolant will be sprayed into the
case before the cpu even begins to get warmer... Then, in about
2 seconds, the cpu is going to go from fully cooled to *no*
cooling, and may or may not actually shutdown before it fries
(depending on the cpu).

It isn't an "enterprise-class" item, it's simply good engineering
and cheap insurance compared to the price of everything else
inside the case. But, you are right that in an el cheapo,
disposable system, it isn't absolutely necessary. (But, is any
system that you've invest enough to put water cooling in still
what could be called "el cheapo, disposable"???)

Yes. You are proposing a "from scrach" solution that will

I'm suggesting that if a person wants to engineer their own
system, that is exactly what they need to do. Do you really
think pre-engineered systems were *not* monitored to prove the
design? Once proven, they sell it minus all the metering.

Not metering to prove the design is just simply *piss* *poor*
*engineering*.

require more expensive parts, engineering to construct and
testing to validate. Show us a ready-to-use affordable,
small, suitable, etc, solution. Even a good theory on how
to do something must be weighted against the actual
implementation burden.

The "ready-to-use" comes *after* the engineering model is
evaluated, not before. These home designed one of a kind
systems are *not* "ready-to-use". (Just *look* at the designs
that have been mentioned here as possible first attempts!)

That's funny.

Yes, I'll grant that some of your statements have been a bit
hilarious.

SHOW US this grand concept you have- on water-cooled PC
class systems in actual commercial PC use. Show us sytems
that only use component thermal feedback control that failed
because of lack of your (implied necessary) water flow
sensor idea.

I've been working with liquid cooled electronics equipment for
40 years. You name the problem, I've seen it for real. Most of
the water cooled systems available for PC's are designed (with
light controllers, for example) to appeal to people who have an
ego and are impressed with eye candy.

I'll repeat what I said in another post, which is that anyone
who puts together a water cooling system had *better* have a
backup available or be able to tolerate the time it takes to
replace the entire system. Putting water cooling into something
that you cannot replace within the time range that you can
tolerate being without, is asking for a disaster.

The reason is because people are simply not willing to realize
that they *can't* properly engineer such a system without
significant expense just for monitoring. Look at what *you*
have proposed... spending several hundred dollar on piss poor
engineering that will destroy the whole system with a single
catastrophic failure of any one of many minor components!

....

which I still contend is inferior to thermal feedback from
the cooled parts.

What you "still contend" is of no significance. You are
the guy who wants to monitor every heat sensitive component,
who claims the coolant will boil, and says the surge tank
does not store heat in one article and claims it has to in
another.

The rest of this article isn't worth responding to, because
the questions you ask have all been answered previously.

 

[kony]

Your statement was only regarding those items that a lack of
cooling will affect,

No I have argued all along that there is need for a certain
minimum airflow regardless of water-cooling.

which now you are attempting to limit to
only the items that are "water-cooled", but in fact it is far
more serious than that. *Everything* in the case is "in
jeopardy", not just those parts that are water cooled. The
entire system has to be monitored according to your concept!

Actually, your idea to monitor water flow then goes one
worse in that water temp or flow rate is about the only
parameter that doesn't matter.

Lets just consider one item... the difference between having a
coolant flow meter on the return entry to the surge tank, and
doing it your way. I'll limit it to that because this one item
is more significant than all of what you believe must be done.

Eventually one of the things that happens with liquid cooling
systems is that they leak. Hoses pop off, fittings break,
something slits, whatever... if we have a flow meter where it
should be, the *immediate* effect is to shut the system down.

Unless it's not calibrated properly, or breaks, or the logic
was wrong, etc. Essentially, it does not displace or better
the monitoring of other parts, it only serves as a suitable
alarm against coolant loss, but not even that because
significant coolant will be lost from the (typical)
reservior with no drop in flow rate.

Or... if the hose exiting the water block has broken and we
wait until the cpu gets hot... the entire surge tank will be
pumped dry, and all of that coolant will be sprayed into the
case before the cpu even begins to get warmer...

For it to go dry, all the coolant was sprayed into the case
either way.

Then, in about
2 seconds, the cpu is going to go from fully cooled to *no*
cooling, and may or may not actually shutdown before it fries
(depending on the cpu).

Nope, the CPU will heat up far more gradually than a 2
second interval, in fact anyone and everyone can right now
stop their heatsink fan and expect dozens of seconds- quite
sufficient for thermal shutdown mechanisms already in place.

Regardless, I do agree that a detection method for coolant
loss can be a good idea, but only in the context of a simple
shutdown mechanism, with the thermal control of the parts
still being part of shutdown or fan RPM variable control.

Even then, the reliability of the system hinges on the less
relable or more costly (or both) flow sensor.

It isn't an "enterprise-class" item, it's simply good engineering
and cheap insurance compared to the price of everything else
inside the case. But, you are right that in an el cheapo,
disposable system, it isn't absolutely necessary. (But, is any
system that you've invest enough to put water cooling in still
what could be called "el cheapo, disposable"???)

Actually, the cheap systems is where it would be necessary
to have the leakage sensor as a properly set up good system
should not have as much a chance of cooling leakage as
multiple other mishaps. If you are simply suggesting we
cover all bases, then of course we need to enumerate all the
potential mishaps a system might suffer. Let's start out
with a burglar alarm, storm detection, a motion senor. What
else?

Good engineering requires quality parts and testing. You
don't just buy a cheap flow sensor and slap in in there. If
the system isn't as relable as an air-cooled system you have
a system hardly worth the bother- in this day and age there
is no reason for an unreliable system.

Even so, it's your toy and IF you spend the time it can be
done, but it is a bit presumptious to suggest things based
on the assumption that someone else wants to spend that
level or time or extra expense based on the ideal for your
toy.

I'm suggesting that if a person wants to engineer their own
system, that is exactly what they need to do. Do you really
think pre-engineered systems were *not* monitored to prove the
design? Once proven, they sell it minus all the metering.

Then it shouldn't leak.

Not metering to prove the design is just simply *piss* *poor*
*engineering*.

You seem to have misunderstood. I mean you have to test the
METERING before ever relying on it. Use the system but you
can't put an unknown design or quality meter in a dodgy
water cooling system and think you've done anything positive
UNTIL that metering system has been fully tested- not just
to work but to REMAIN working, including remaining working
long term with your choice of water and water additives.

The "ready-to-use" comes *after* the engineering model is
evaluated, not before. These home designed one of a kind
systems are *not* "ready-to-use". (Just *look* at the designs
that have been mentioned here as possible first attempts!)

This is not particularly newsworthy. It is obvious.
That's exactly why adding further parts without those parts
having been proven, is a further liability.

Yes, I'll grant that some of your statements have been a bit
hilarious.


I've been working with liquid cooled electronics equipment for
40 years. You name the problem, I've seen it for real. Most of
the water cooled systems available for PC's are designed (with
light controllers, for example) to appeal to people who have an
ego and are impressed with eye candy.

Never said otherwise. What's the point? If anything it
points to the questionable quality of these flow meters too.

I'll repeat what I said in another post, which is that anyone
who puts together a water cooling system had *better* have a
backup available or be able to tolerate the time it takes to
replace the entire system. Putting water cooling into something
that you cannot replace within the time range that you can
tolerate being without, is asking for a disaster.

That's an interesting concept and the idea of redundancy I
agree with. On the other hand, if you feel failure is that
imminent then it doesn't seem you yet have a handle on how
to make a PC cooling system reliable.

The reason is because people are simply not willing to realize
that they *can't* properly engineer such a system without
significant expense just for monitoring.

Exactly, which is why it's so questionable to focus on flow
meters that aren't qualified and demonstrated to be a good
contol instead of thermal parts monitoring. Perhaps
suupplimentary control but again only after extensive
testing, it is not a finished solution as you described
while whole cooling systems are.

Look at what *you*
have proposed... spending several hundred dollar on piss poor
engineering that will destroy the whole system with a single
catastrophic failure of any one of many minor components!

Actually that's your suggestion, you quite clearly described
piss poor engineering but not a pre-tested solution ready
for deployment. Failure revolves around weak links. Proper
engineering would fix those weak links not make vague
innuendos about someone else's realization that we don't
care about things like water temp.

...

What you "still contend" is of no significance. You are
the guy who wants to monitor every heat sensitive component,

No, only those water-cooled and with potential to exceed
their thermal threshold. That is trivial to do and very
reliable. In fact off the shelf kits with all the leads and
alarm exist.

who claims the coolant will boil,

You might want to set a shrink, your imagination is getting
in the way of reality here. I never claimed coolant will
boil.

and says the surge tank
does not store heat in one article and claims it has to in
another.

No I claimed it is not a significant pseudo-radiator
relative to the actual radiator.

The rest of this article isn't worth responding to, because
the questions you ask have all been answered previously.

I'm not asking questions I"m pointing how how far off on a
tangent you've gone instead of focusing on failure points.
If you have experienced damage, leakage, etc., from coolant
loss, then stop jerking around with flow meters will after
you find better parts.

 

[kony]

It isn't an "enterprise-class" item, it's simply good engineering
and cheap insurance compared to the price of everything else
inside the case. But, you are right that in an el cheapo,
disposable system, it isn't absolutely necessary. (But, is any
system that you've invest enough to put water cooling in still
what could be called "el cheapo, disposable"???)

In all fairness, I was more opposed to a generic suggestion
than a specific proven and tested implementation. Iif you
had a specific flow meter you'd tested long enough that you
could personally recommend it that would be more desirable.

On the other hand, that does not displace my opinon about
the need for thermal shutdown based on component temps.
There are more factors than merely whether there's water and
what the water temp is. For example, whether the block(s)
are making good contact with the components, not only right
after installation, during initial testing, but also later
in the system's life.

 

[rhys]

if anyone is curious I documented all the steps and components that I used
to build a silent air cooled pc for games:
http://silentgamingreview.com/components.htm

Silent Gaming Review
http://silentgamingreview.com

Great post. I will be building a dual Opteron or Xeon workstation
before the end of the year and I am putting a premium on quiet. Most
of your ideas are applicable to me and I look forward to an end (or
severe reduction) to the ear buzzing I get from fans/hard drive
noises.

R.

 

[Floyd L. Davidson]

Actually, your idea to monitor water flow then goes one
worse in that water temp or flow rate is about the only
parameter that doesn't matter.

A lack of coolant flow on the return line is the *best*
indication that a disaster is happening.

Unless it's not calibrated properly, or breaks, or the logic
was wrong, etc. Essentially, it does not displace or better
the monitoring of other parts, it only serves as a suitable
alarm against coolant loss, but not even that because
significant coolant will be lost from the (typical)
reservior with no drop in flow rate.

It's inexpensive insurance for expensive equipment.

For it to go dry, all the coolant was sprayed into the case
either way.

That *doesn't* happen with a careful design.

Nope, the CPU will heat up far more gradually than a 2
second interval, in fact anyone and everyone can right now
stop their heatsink fan and expect dozens of seconds- quite
sufficient for thermal shutdown mechanisms already in place.

On some cpu's. Not on others. Note that the water block
does *not* provide the kind of minimal protection that air
flow heat sinks do.

Regardless, I do agree that a detection method for coolant
loss can be a good idea, but only in the context of a simple
shutdown mechanism,

So why is it like pulling teeth to get you to admit that?

with the thermal control of the parts
still being part of shutdown or fan RPM variable control.

Nobody, except *you* has claimed anything other about the
shutdown mechanism. As to the thermal control, you still have
yet to understand how the system works:

1) heat exchange system #1 (waterblocks on cpu etc)
transfers heat *to* the coolant.

2) heat exchange system #2 (radiator, fans, etc)
transfers heat *from* the coolant.

*You* want to control the second part with metering in the first
part. A *better* way is separate metering and control for
both parts; and since the mechanism to implement it is required
anyway, proper design is a trivial addition.

Good engineering requires quality parts and testing. You
don't just buy a cheap flow sensor and slap in in there. If
the system isn't as relable as an air-cooled system you have
a system hardly worth the bother- in this day and age there
is no reason for an unreliable system.

Water cooled systems are *not* as reliable as air cooled
systems, and nobody should misunderstand that fact. Water
cooling can result in lower temperatures (which overclockers
like) or in lower noise (which has been the main focus in these
threads), but if for /higher/ /reliability/ go with air cooling
and the higher temps and more noise.

You seem to have misunderstood. I mean you have to test the
METERING before ever relying on it. Use the system but you
can't put an unknown design or quality meter in a dodgy
water cooling system and think you've done anything positive
UNTIL that metering system has been fully tested- not just
to work but to REMAIN working, including remaining working
long term with your choice of water and water additives.

More mumbling... most of which I've snipped out, but since
this one was particularly ridiculous and off topic, I thought
I'd point out out how unfocused and non-topical most of your
comments are. *You* are the one suggesting that metering of
component parts of the system is unnecessary. That *is* the
only way to test. My point all along has been to build in
the testing (particularly where it is trivial to add it), while
*you* keep saying it is unnecessary.

Your comments are running in circles on this and several other
subtopics. (Which is why I've suggested that you just don't
understand water cooling well enough to be making critical
comments.)

No, only those water-cooled and with potential to exceed
their thermal threshold. That is trivial to do and very
reliable. In fact off the shelf kits with all the leads and
alarm exist.

So tell me, if it is trivial to monitor *all* water-cooled
parts, why is it not trivial to monitor the functional parts of
the system? It's the same temperature probes...

You might want to set a shrink, your imagination is getting
in the way of reality here. I never claimed coolant will
boil.

"and that fan is also being controlled, then all
passively cooled components much also be weighed in
the determination of proper fan speed control, but
still, NEVER the water temp. Water is absolutely
the very last thing anyone should care about until
it starts boiling- and if your water is boiling you
have far larger problems than which fans are
spinning."

No I claimed it is not a significant pseudo-radiator
relative to the actual radiator.

You did that too, but that has nothing to do with going in
circles over the heat storage issue. One of your circular
comments does not cancel another of them.

I'm not asking questions

....

And you should be!

I"m pointing how how far off on a
tangent you've gone instead of focusing on failure points.

Another hilariously inaccurate statement.

 

[Floyd L. Davidson]

On the other hand, that does not displace my opinon about
the need for thermal shutdown based on component temps.
There are more factors than merely whether there's water and
what the water temp is.

This is a prime example of flawed logic. The issues of
/thermal/ /control/ devices and /safety/ /shutdown/ devices are
separate. Just because a safety shutdown device that is totally
separate from the thermal control mechanism is discussed does
*not* mean that no thermal control mechanism is necessary, yet
you reject all such devices that are not directly tied to the
thermal control system as if that is a valid dichotomy.

You keep trying to tie *everything* to a thermal probe on the
CPU, and that is simply *not* the best way to design such a
system.

There are *two* heat exchanging systems. It is possible, as you
wish to do, to operate both of them based on metering of only
one; but that does not provide optimal functionality. Indeed,
the typical system does that to a great degree, and in the
process uses either gross granularity or even no control at all
on one of the two systems. Commonly that means excessive noise
is generated because there is little or no linkage to the actual
need for air flow.

If the object is to reduce fan noise, it is necessary to use
separate logical control systems for the two separate heat
exchange systems. And since the major physical components
necessarily are already there, it is relatively trivial in an
original design.

(Perhaps I should point out that there are other ways to reduce
fan noise... one of which is to eliminate the fan entirely;
which can be done with an external radiator of appropriate size
or location. One trick is to bury a large enough tank in the
ground...)

 

[stormrider]

For my next trick I will build a silent dual-core CPU, top of the line GPU
gaming machine - so stay tuned!
The more I think about it the less I like the idea of leaks and the
potential maintanence headaches of water cooling, so it will most likely be
air cooled.
I am playing with the idea of custom air ducts and double sound padding.
And thanks to Floyd the fans will be software controlled this time.
http://silentgamingreview.com

 

[kony]

This is a prime example of flawed logic. The issues of
/thermal/ /control/ devices and /safety/ /shutdown/ devices are
separate. Just because a safety shutdown device that is totally
separate from the thermal control mechanism is discussed does
*not* mean that no thermal control mechanism is necessary, yet
you reject all such devices that are not directly tied to the
thermal control system as if that is a valid dichotomy.

You keep trying to tie *everything* to a thermal probe on the
CPU, and that is simply *not* the best way to design such a
system.

Yes a cooling system necessarily has to be tied to the temp
of the cooled part. Any avoidance of this is simply a
defective design. I"m not apathetic about your ramblings so
I'm done with this topic.

 

[Floyd L. Davidson]

Yes a cooling system necessarily has to be tied to the temp
of the cooled part.

You *still* haven't understood that the there are *two* cooling
systems, and the "cooled part" in one of them is the liquid.

Any avoidance of this is simply a
defective design. I"m not apathetic about your ramblings so
I'm done with this topic.

Well I *am* apathetic about your ramblings, so I agree the
topic is finish.

 

[Fisher]

if anyone is curious I documented all the steps and components that I used
to build a silent air cooled pc for games:
http://silentgamingreview.com/components.htm

Did you ever try the Scythe SCNJ-1000 "Ninja" heatpipe fanless cpu
cooler? It looks less bulky than the one you chose and the review I
read was favorable.

 

[David Maynard]

You *still* haven't understood that the there are *two* cooling
systems, and the "cooled part" in one of them is the liquid.

In any design there are cost vs benefit considerations and the point here
is that if the cooling system is operational (as opposed to an experimental
design) then CPU temp provides the most benefit (for a cost that's already
included in virtually all modern motherboards) with little, if anything,
gained by adding the expense to also monitor water temp. If the water is
'over temp' then the CPU will be too so CPU monitoring catches the problem.
If the water is within spec, however, you still don't know if the CPU temp
is. It 'should be' but monitoring the CPU ensures it while water temp does not.

Using your logic of monitoring 'all the cooled parts' one could also demand
a radiator sensor as it's a 'cooled part' too. And while that, too, might
be a useful diagnostic tool for debugging the design it's an added cost to
an operational system that, again, adds little to the task of detecting a
general cooling system failure.

CPU temperature is sufficient to determine if it's doing the job, or not,
as that's the purpose of it, and all those other intermediate 'cooled
parts': to keep the CPU cool.

Now, another purpose past diagnostic that monitoring other parts might
serve is in detecting an impending failure 'earlier' so there's more time
to shut down but I don't know that's of much benefit to a typical PC.

I am assuming you're talking about a constant flow design and not trying
some kind of closed loop flow rate control.

 

[Floyd L. Davidson]

Given the length of this thread, and the lack of focus, I can
see why it is confusing...

In any design there are cost vs benefit considerations and the
point here is that if the cooling system is operational (as
opposed to an experimental design)

Right. Of course what we are talking about *is* an experimental
design. One of a kind, first ever... This is *not* a drop in
pre-engineered and beta tested system.

Hence the point here is not what you apparently have addressed.
That makes most of your comments invalid for what we have been
looking at, even though they are indeed correct for what *you*
are talking about.

then CPU temp provides the
most benefit (for a cost that's already included in virtually
all modern motherboards) with little, if anything, gained by

You missed the fact that we are discussing a system where the
*intention* is to reduce noise. That's the target: minimum
noise.

Which is to say, the goal posts aren't the same as it would
be for an overclocker or someone who simply wants to see a
water cooled cpu system functioning.

adding the expense to also monitor water temp. If the water is

Note that the "expense" to also monitor water temperature,
assuming that the system monitors *any* temperature, is minimal.
Temperature probes cost a couple dollars. It might actually
cost 3-4 more bucks to mount a probe (for example in a water
jacket). Overall, it isn't a significant expense.

'over temp' then the CPU will be too so CPU monitoring catches
the problem. If the water is within spec, however, you still
don't know if the CPU temp is. It 'should be' but monitoring the
CPU ensures it while water temp does not.

But monitoring the CPU won't help decide when and how to reduce
the fan noise associated with cooling the water! Since that was
the entire point of this design, we it makes no sense to leave
out that part of the design.

Using your logic of monitoring 'all the cooled parts' one could

NO, NO, *DO NOT* pin that crap on me. That was kony's statement,
which he then confuses because he doesn't realize how many
"cooled parts" there are! As I noted, the cpu is *not* the
only "cooled part".

also demand a radiator sensor as it's a 'cooled part' too. And

Smart idea! (There is hope for you!)

Given that it cost only a couple bucks and some wire, it really
is a good idea to measure the water temperature going into and
coming out of any heat exchanger intended to change the water
temperature. That difference is what can be used to control the
fan noise with fine enough granularity to make a "quiet" system.

(Of course, since I'm a geeky kinda guy, I'll spend another $5
and measure the air temperature on each side of the heat
exchanger too, just because it makes a really interesting graph.)

while that, too, might be a useful diagnostic tool for debugging
the design it's an added cost to an operational system that,

Nobody has discussed adding that to an operational system which
is *not* trying to be super quiet. (Which is what I assume you
mean, because otherwise your statement doesn't make good sense.)

again, adds little to the task of detecting a general cooling
system failure.

Nobody has claimed that it has *anything* to do with "system
failure". Other than kuny's strawman claim that failure
detection systems were somehow intended to replace control
systems. In addition to not being able to separate the *two*
cooling systems (one for the cpu and one for the water), he also
could not separate failure monitoring from control monitoring.

CPU temperature is sufficient to determine if it's doing the
job, or not, as that's the purpose of it, and all those other
intermediate 'cooled parts': to keep the CPU cool.

That simply isn't true. That isn't true even for systems that
overclockers want! It takes a significant amount of design
effort to balance the cooling such that each specific part that
is cooled gets exactly what it needs, and no more or no less.
Have you looked at some of the more complex systems? Water
jackets on disk drives and in power supplies, as well as on hot
chips!

There is no way to simply calculate that, throw a system
together, and end up with a optimal configuration. The
variation is in *how* it is monitored, not if.

For example, if software design is not part of the project (and
in fact not even all programmers are able to write the software
required, much less hardware hackers who don't do serious
programming), some of the manual probes might be used, or just
simple software, and only a few parameters measured at a time
over many tests to eventually compile a full set of data. It
will work just as well. But by the same token if complex
software design is part of the project, it doesn't cost
significantly more to monitor one or a couple *dozen*
temperature probes all at once!

Now, another purpose past diagnostic that monitoring other parts
might serve is in detecting an impending failure 'earlier' so
there's more time to shut down but I don't know that's of much
benefit to a typical PC.

It serves the purpose of controlling the noise making apparatus,
the fans, with much finer granularity.

I am assuming you're talking about a constant flow design and
not trying some kind of closed loop flow rate control.

You did assume something right!

I suspect from what you have said that if you adjust to the
actual target here, you'll see some differences.

Of course, there are also other ways to reduce fan noise, and
one of those is to *not* use fans to cool the water. For
example, where I live the outside temperature is always low
enough to allow convection cooling if the heat exchanger is
large enough. Hence I could mount a large enough outdoor
radiator and probably not even use a fan to cool the water. A
relatively low speed (quiet) fan could then be the only fan
needed to keep the case temperature down.

Others have done things like bury a surge tank in the ground to
accomplish the same thing. It works well. (Heh, I could bury a
tank in the permafrost, and end up having to *heat* the incoming
coolant rather than cool it!)

In those cases the initial stages require metering, but once a
good data set is obtained there is little need to continue
monitoring (other than for curiosity). But the same is not
really true for any system using a fan to provide cooling for
the liquid as long as the objective is to control the fan for
minimum noise.

 

[David Maynard]

Given the length of this thread, and the lack of focus, I can
see why it is confusing...

Near as I can tell the only 'confusion' is you changing the goal posts.

Right. Of course what we are talking about *is* an experimental
design. One of a kind, first ever... This is *not* a drop in
pre-engineered and beta tested system.

It appears to be what *you* are talking about, at least at this stage of
the game, but it wasn't what the original poster said with the rather
benign comment that he was "tossing around the idea of using a water
cooling system."

I suppose it's possible he meant to begin an extensive R&D project
inventing, as you now put it, a "One of a kind, first ever..." never been
done before experimental dynamic flow control water cooling system but it's
usually the case that when someone mentions "water cooling," with no
further clarification, they're speaking of something a bit more conventional.

Hence the point here is not what you apparently have addressed.
That makes most of your comments invalid for what we have been
looking at,

Define "we."

even though they are indeed correct for what *you*
are talking about.




You missed the fact that we are discussing a system where the
*intention* is to reduce noise. That's the target: minimum
noise.

You have an annoying tendency to accuse people of 'missing' things simply
because they have a different opinion.

I didn't 'miss' it at all and, again, when people talk of minimizing noise
with a water cooling system they are generally speaking of the
conventional, well established, means of doing so.

This case might be different but, near as I can tell, no one has talked
about embarking on a nobel prize effort other than you.

Which is to say, the goal posts aren't the same as it would
be for an overclocker or someone who simply wants to see a
water cooled cpu system functioning.

I rather suspect the original poster was contemplating a process that is
already known to work.

The point of contention began when, in response to the question of what
happens in the even of pump failure, you suggested that flow monitoring
would detect it and Kony suggested it was better to monitor the device
being cooled.

I imagine he presumed, since I would, that a poster in this news group is
not likely an 'enterprise' operation.

Note that the "expense" to also monitor water temperature,
assuming that the system monitors *any* temperature, is minimal.
Temperature probes cost a couple dollars. It might actually
cost 3-4 more bucks to mount a probe (for example in a water
jacket). Overall, it isn't a significant expense.

That assumes the motherboard has 'spare' temperature monitoring inputs and
that is certainly not universally true.

But monitoring the CPU won't help decide when and how to reduce
the fan noise associated with cooling the water! Since that was
the entire point of this design, we it makes no sense to leave
out that part of the design.

Reinventing the wheel does, indeed, seem to be a part of your design but,
as I said, I'm not so sure the original poster had in mind becoming the
Thomas Edison of PC water cooling.

NO, NO, *DO NOT* pin that crap on me. That was kony's statement,

Kony never made any such statement. It came from your 'admonition' that the
water must be monitored because it's a "cooled part."

which he then confuses because he doesn't realize how many
"cooled parts" there are! As I noted, the cpu is *not* the
only "cooled part".

Your assumption of what 'confuses' Kony is self serving nonsense.

I am sure that Kony made the reasonable assumption that when a posters says
he is "tossing around the idea of using a water cooling system" that they
mean the same kind of conventional water cooling system everyone else means
when they say it. And, under that criteria, his comment about monitoring
CPU temp is perfectly valid.

Smart idea! (There is hope for you!)

Given that it cost only a couple bucks and some wire, it really
is a good idea to measure the water temperature going into and
coming out of any heat exchanger intended to change the water
temperature.

Beyond the fact that not all motherboards have a spare input for your
'couple of bucks and some wire', anyone who's done an 'experimental' system
should know that nothing is 'a couple of bucks' by the time you finish
figuring out why it doesn't work as originally envisioned.

That difference is what can be used to control the
fan noise with fine enough granularity to make a "quiet" system.

That's a possibility. It's also possible that straight CPU temp fan control
would suffice. And then it's possible that simply using a large, fixed
speed, low RPM fan like everyone else does would be sufficient too.

(Of course, since I'm a geeky kinda guy, I'll spend another $5
and measure the air temperature on each side of the heat
exchanger too, just because it makes a really interesting graph.)

Well, let's see, we're now up to at least 4 (water, radiator, exchanger air
inlet and outlet) 'spare' temperature inputs without even trying.

Nobody has discussed adding that to an operational system which
is *not* trying to be super quiet. (Which is what I assume you
mean, because otherwise your statement doesn't make good sense.)

I didn't say anyone had discussed it. It's simply an example of how one can
monitor themselves to death as long as cost/benefit is of no consideration.
You did me two better with the inlet/oulet air temps.

Nobody has claimed that it has *anything* to do with "system
failure".

In your eagerness to explain how others 'miss' things and get so 'confused'
you've lost track of the discussion. The discussion between you and Kony of
what to monitor began in response to "What do you think about the pump
stopping, though. First it will fry the components."

Sounds like a failure to me.

Other than kuny's strawman claim that failure
detection systems were somehow intended to replace control
systems.

The strawman is yours as Kony never made such a claim.

In addition to not being able to separate the *two*
cooling systems (one for the cpu and one for the water), he also
could not separate failure monitoring from control monitoring.

It's you who haven't been able to separate what someone says from what you
misinterpret.

That simply isn't true. That isn't true even for systems that
overclockers want!

Of course it's true.

It takes a significant amount of design
effort to balance the cooling such that each specific part that
is cooled gets exactly what it needs, and no more or no less.

Nonsense. There is no problem with 'overcooling' one of the components.

Have you looked at some of the more complex systems? Water
jackets on disk drives and in power supplies, as well as on hot
chips!

Yes. And I assure you they didn't worry about any of them getting 'more
cooling' that what it 'needs', just as you won't find air blocks in your
typical ventilated case because, "Oh my god, the northbridge is only 4C
above ambient without one."

There is no way to simply calculate that,

Sure there is. Watts in, watts out.

throw a system
together, and end up with a optimal configuration. The
variation is in *how* it is monitored, not if.

For your 'one of a kind never been done before' experimental R&D project,
yes, assuming anyone wants to embark on it.

For example, if software design is not part of the project (and
in fact not even all programmers are able to write the software
required, much less hardware hackers who don't do serious
programming),

That should be a pretty dern good clue that your average everyday home PC
user didn't have this in mind when they said they were "tossing around the
idea of using a water cooling system." Or, if they did, you should be
warning them about the magnitude of the effort.

some of the manual probes might be used, or just
simple software, and only a few parameters measured at a time
over many tests to eventually compile a full set of data. It
will work just as well. But by the same token if complex
software design is part of the project, it doesn't cost
significantly more to monitor one or a couple *dozen*
temperature probes all at once!

Yeah. I've heard that one before. "Now that we're hip deep in [complex
software design] costs we'd never planned for then piling more on won't hurt."

It serves the purpose of controlling the noise making apparatus,
the fans, with much finer granularity.

You have yet to establish any required degree of granularity.

You did assume something right!

I suspect from what you have said that if you adjust to the
actual target here, you'll see some differences.

It should be obvious I know both ends of the matter but I suspect you've
way over estimated what a normal person means when they're "tossing around
the idea of using a water cooling system."

Btw, not that I want to get into designing one but, considering the
potential thermal lag in the water cycle it might be better to use
electronic component temps to control the loop anyway. Especially since
it's 'free'.

<snip of more design theories>