Can someone with some real knowlege answer this question??? - Page 2

BillA · 10-11-2002, 02:56 PM

you know Ben,
that poor guy is going to be walking down the street and Peter is going to run up and bust him right in the face screaming "YOU ARROGANT BASTARD"

and it will be all your fault
now don't you feel guilty ?

utabintarbo · 10-11-2002, 03:21 PM

Quote:

Originally posted by unregistered
you know Ben,
that poor guy is going to be walking down the street and Peter is going to run up and bust him right in the face screaming "YOU ARROGANT BASTARD"

and it will be all your fault
now don't you feel guilty ?

WOW! 2 funny posts from BillA! (Check here) The apocalypse is upon us!

Joe · 10-11-2002, 03:52 PM

Quote:

Originally posted by nuclear
Might I had to joe post that most big iron server are in server rooms, where the climatisation system is more noisy than the server itself....
(dell from 1650 to 6550, compaq san, storatek wolfcreek, ibm p390, ibm as400 of many size, sun enterpise 220r to enterprise 6500, they were all less noisy than the climastisation system, even the 6500 that was in another room since the climatisation system was from the floor like most server room)
So even if they cool there server with redundant noisy fan, it doesn't matter at all since no one is in the server room to ear it.
I had a Dell 2450 and 2550 rack mount 2u server with 1 p3 on the first and 2 p3 tualatin on the second. They were noisy, really noisy, but still we were using them in the server room because noise isn't really a factor there.

Indeed! the through the floor leibert units in most data centers are nothing more than a BIG heat exchanger, and a big blower. There is chilled water run through them and thats how they cool. Byt that blower makes a good deal of noise.

I had a Dell 1650 at my desk to prep it, and it was FREAKING loud! I was amazed how loud that lil box is ( the Dell 4600's are quieter), but in the data center you cant even hear it.

BUT... the servers are made to function at room temp, they DO NOT need to run at sub 70F temps. The 2600 with the heat pipes, was run next to my chair in my cube, and at room temp ambient air, the HSF was still just slightly warm.

Just cause a server room is normally well conditioned, doesnt mean they made the servers any less redundant or less cooling just because of that. They expect room temp or better, and can still function well up to 90F for most servers and mainframes.

jaydee · 10-11-2002, 08:11 PM

Excellent input guys! Thanks!

godsdice · 10-11-2002, 11:44 PM

Actually, the reason why noone ever goes into the "server room" where I work is because it's always at 50F. Used to have a centrifuge in there because there wasn't room anywhere else...directly below the vent. Completely messed up results....argh.

Les · 10-12-2002, 03:03 AM

Quote:

Originally posted by unregistered

those not having followed Cathar's quest (20+ pg thread on OCAU) are encouraged to read it here
as most of the variables are discussed (or perhaps cryptically alluded to)...................................

maybe,also,here:
http://forums.overclockers.com.au/sh...5&pagenumber=2
here:
http://forums.overclockers.com.au/sh...threadid=98583
and here:
http://forums.overclockers.com.au/sh...threadid=93479

MadDogMe · 10-19-2002, 07:42 AM

What's the average temp over ambient at the mo'?, 10DegC?. it's a long way or a little way depending on how you look at it. that's how much H20 as a system has left to go.
The Block?, the way to work it out I think is to have a supply of ambient temp water and feed it through the block, how high over ambient is the CPU~temp now?, that's how far H20~blocks have left...

I'd like to know how close to ambient Cathars block holds temps, I'd guess it's pretty damn close...

jaydee · 10-19-2002, 10:05 AM

Quote:

Originally posted by MadDogMe

The Block?, the way to work it out I think is to have a supply of ambient temp water and feed it through the block, how high over ambient is the CPU~temp now?, that's how far H20~blocks have left...

I don't see that being accurate. I am talking about AL and Cu. You would have to have a perfect joint between the block and CPU and a block material that conducts 100% efficent for this to be true. It is impossible to get to ambient or even close (close meaning 3Cish or better with modern CPU's). But there has to be a barrier that maxes out the Cu/AL capabilites and it stops accepting more heat.

Or maybe I am wrong here?

Since87 · 10-19-2002, 02:59 PM

Quote:

Originally posted by myv65
How about a high-efficiency peltier with a thermally-regulated supply voltage? Hey, if someone could do that, sign me up! Someday someone probably will, but that day hasn't come, yet.

By high efficiency, do you mean better efficiency than the pelts in common use today on CPU's?

If you mean the pelts in use today, there are already temperature controllers that will work. What are you willing to pay? The devices I know of that you can buy "off the shelf" are $500+.

I am going to build one for myself with a hygrometer built in to keep the minimum temperature above the dewpoint. I was thinking about developing it into a product, but I've pretty well concluded the market is not there. I couldn't afford to sell them for less than $180 and OC'ers tend to be inherently cheap.

Also pelts aren't keeping up with the increasing heat generation of CPU's. Unless something like this Coolchips stuff becomes a reality there doesn't appear to be a future for pelts. (I'm somewhat suspicious that the coolchips stuff is a scam.)

jaydee · 10-19-2002, 05:42 PM

Quote:

Originally posted by Since87
By high efficiency, do you mean better efficiency than the pelts in common use today on CPU's?

If you mean the pelts in use today, there are already temperature controllers that will work. What are you willing to pay? The devices I know of that you can buy "off the shelf" are $500+.

I am going to build one for myself with a hygrometer built in to keep the minimum temperature above the dewpoint. I was thinking about developing it into a product, but I've pretty well concluded the market is not there. I couldn't afford to sell them for less than $180 and OC'ers tend to be inherently cheap.

Also pelts aren't keeping up with the increasing heat generation of CPU's. Unless something like this Coolchips stuff becomes a reality there doesn't appear to be a future for pelts. (I'm somewhat suspicious that the coolchips stuff is a scam.)

Dude, check this out!!!

Quote:

Swiftech, the Overclockers best friend will be shipping their new MCX462+T by the end of the month.

Designed for Extreme Overclockers who shun liquid cooling, are not concerned about noise level, but still want everything they can squeeze out of their CPU's, Swiftech has heard your prayers!

The MCX462+T is a Thermoelectric Heatsink that employs a 226W Peltier unit that is pre-assembled and insulated to eliminate the possibility of condensation.

The MCX462+T as it's name implies, is for extreme air-cooling of your screaming AMD CPU and requires the availability of the four holes around the socket of your motherboard.

Intel P4 users don't fret; I have it on good authority that a MCX4000 T is in the works!

I just couldn't close this report without showing you the bottom of the MCX462+T, pure Swiftech beauty!

Here is where I heard about it and a couple pics.http://www.systemcooling.com/modules...rticle&sid=220
Scroll down the page a bit. Don't give up! There may not be much of a market now but if things keep heating up with CPU's then there will be!! Would be good to get a solid foot in the door before the market gets good like Danger Den did before water cooling started picking up. Don't let Swiftech take all the market!

Since87 · 10-19-2002, 08:09 PM

I've seen the specs on the Swiftech TEC/Heatsink. It appears to me that they are claiming that it can get a 100 Watt CPU to ambient with 80 CFM airflow. I am very skeptical. I've got a spreadsheet for calculating TEC cooling under various conditions, and that kind of performance is at the extreme end of what I can believe that heatsink is capable of. (Depends on the C/W of the heatsink when cooling a 50mm X 50mm pelt.)

As far as my pelt controller is concerned, I wish I'd been looking into this cooling stuff a couple years ago, because I probably could have sold enough of these things to make some money, but the window of opportunity is closing fast now. Besides I'm a design engineer. I like designing things. I don't want to deal with all the issues of running a business.

jaydee · 10-19-2002, 08:22 PM

Quote:

Originally posted by Since87
I've seen the specs on the Swiftech TEC/Heatsink. It appears to me that they are claiming that it can get a 100 Watt CPU to ambient with 80 CFM airflow. I am very skeptical. I've got a spreadsheet for calculating TEC cooling under various conditions, and that kind of performance is at the extreme end of what I can believe that heatsink is capable of. (Depends on the C/W of the heatsink when cooling a 50mm X 50mm pelt.)

As far as my pelt controller is concerned, I wish I'd been looking into this cooling stuff a couple years ago, because I probably could have sold enough of these things to make some money, but the window of opportunity is closing fast now. Besides I'm a design engineer. I like designing things. I don't want to deal with all the issues of running a business.

What I havn't figured out about that Swiftech is how that pelt is powered? Surely it isn't of the computer power supply? That would seem not possible unless the power supply was 600watt and 20+amps? 226 watt pelt should take some serious power to operate. I havn't seen the power usages of that pelt yet. My buddy is a design Engineer also. He is coming up with some interesting stuff not related to computer cooling. His real job requires him to design heating and cooling systems. I wish I knew half of what he knows. Worst part is we graduated out of the same high school and class in 1995! I am still making more han he is though, although I am sure not for long!

jaydee · 10-19-2002, 08:25 PM

Also that pelt controller can have far more uses than just computer cooling. There are a lot of other things it can be used for. But I hear ya about running a business and what not.

MadDogMe · 10-20-2002, 05:43 AM

Quote:

I don't see that being accurate. I am talking about AL and Cu. You would have to have a perfect joint between the block and CPU and a block material that conducts 100% efficent for this to be true. It is impossible to get to ambient or even close (close meaning 3Cish or better with modern CPU's). But there has to be a barrier that maxes out the Cu/AL capabilites and it stops accepting more heat.

It all comes down to thickness, a 1mm Cu base will conduct heat to the H2o more efficiently than a 1cm thick base yes?, it all depends on the design of the block, reaching ambient is the only real benchmark we have, it's not so much copper stops accepting heat, it's water stops accepting heat from the copper(not enough surface area, design again). Cu will carry on passing heat as long as it has somewhere to go, or until melting point is reached.

Reaching ambient(within a fraction of a Degree) should'nt be impossible, removing the heat from the water and passing it back through the CPU at ambient is another matter

...

bigben2k · 10-20-2002, 09:15 AM

Ya'll are getting off-topic, with this talk about TECs.

A normalized chiller, which could keep the water temp at 20 deg C (aka ambient), and a high flow direct die cooling option would give us the lowest temperature that a waterblock could achieve, in theory. Even then, as long as there's a baseplate, the CPU temps will be ever so slightly higher.

Either way, the point is that the lowest CPU temp that can be achieved with normal cooling, is several degrees above ambient. We just don't know how many degrees that is.

What we do know is that without any cooling whatsoever, the max temp is in excess of 350 deg C, and that's an important piece of information!

BillA · 10-20-2002, 10:21 AM

the numbers being 'quoted' here are make believe
(suggest you all DO some testing or, failing that, read about some testing)

understanding that any 'minimum' must also specify the flow rate and the applied power;
some illustrative graphs are in this article (on the last page)
- though not definitive relative to any absolute, as indeed nothing can be

you're trying to pick nits about wbs when the TIM joint is FAR more critical -> and variable

hello ? hello ? operator this line was disconnected . . . .

shaft01 · 10-20-2002, 10:46 AM

Quote:

Originally posted by jaydee116
What I havn't figured out about that Swiftech is how that pelt is powered? Surely it isn't of the computer power supply? That would seem not possible unless the power supply was 600watt and 20+amps? 226 watt pelt should take some serious power to operate. I havn't seen the power usages of that pelt yet. My buddy is a design Engineer also. He is coming up with some interesting stuff not related to computer cooling. His real job requires him to design heating and cooling systems. I wish I knew half of what he knows. Worst part is we graduated out of the same high school and class in 1995! I am still making more han he is though, although I am sure not for long!

It requires a 25 amp single output PS........the rated draw of the pelt is 24 amps.......So yes it requires a dedicated PS........

bigben2k · 10-20-2002, 11:25 AM

Quote:

Originally posted by unregistered
you're trying to pick nits about wbs when the TIM joint is FAR more critical -> and variable

Well, that's implied, isn't it?

Why don't you give us a sample (example) set of numbers, of the gradient that you've mentioned before? In other words, what's your take on the lowest temp that can be achieved? (without refering us to page 1 of this thread, or to Google).

Myv65 came closest, when he pointed out that 0 deltaT would be a requirement to cool the CPU completely and entirely. Of course this can't be achieved, so what can be achieved?

BillA · 10-20-2002, 12:59 PM

sorry Ben

Dave will do that, pHaestus will do that - but I will not

I taught for one semester at a Uni and learned that I was not temperamentally suited to the endless regurgitation of what I knew, when presenting it to those with little motivation

I'll tell you where it is
I will not summarize it, I will not paraphrase it, I will not distill it down to 'good' or 'bad'
that is for YOU to do

too much to read ? well try this:
if you can't be bothered to read, why the hell should I type it over again for you ?

"Well, that's implied, isn't it?"
no
I saw only one number (due to my ignore list), and it was tripe

I have posted tons of graphs, and you now need me to tell you what they mean ?
have you figured out what the offset due to the TIM joint is ?
why not ? its been posted and discussed many times

what I said before Ben,
you have to start actually learning some of this stuff instead of asking the same questions over and over
with your post count you can't play the newbie game

bigben2k · 10-20-2002, 01:20 PM

Fair enough.

I downloaded a few documents from Intel, and I'll print them tmo, and review them, and see what I can come up with.

Your point is still well taken: the TIM joint is critical, as it is single-handedly the biggest variable, based on what the users have posted.

I was hoping that you could give us a rough idea of what it is that we're looking at, but barring that, I'll give it a shot.

No promises.

BillA · 10-20-2002, 01:36 PM

relative to the 'minimum' question:

there is a facinating (to me natch) chart on the last pg which was initially posted with an explanation in a thread on this forum;
then the thread starter spiked the entire thread 'cause he did not like the heat deriving from his BS claims
(and put himself on my ignore list)

go to OCAU where I have extensively discussed T/M, (search "T/M" ?)
because the remainder is the TIM joint C/W
you WILL be interested

EDIT:
always remember Intel's "cascade of gradients"; then focus on the element of interest

jaydee · 10-20-2002, 07:01 PM

Quote:

Originally posted by unregistered

there is a facinating (to me natch) chart on the last pg which was initially posted with an explanation in a thread on this forum;
then the thread starter spiked the entire thread 'cause he did not like the heat deriving from his BS claims
(and put himself on my ignore list)

Well thats total BS, but I will not get into that. What I will say is your are a lesser person for continuing to bring it up without knowing the true circumstances and intentionally or not making this thread another possibility for a flame war which no one will win. Your intelligence may be better than most here but that does not warrant your ****ed up attitude! Of course you get told this time and time again and still do not learn so who are you to tell Ben he is not learning after being told over and over? Anyway don't bother replying to that question as it will clearly do no good.

Also the only reason I havn't put you on Ignore is some of the stuff you say is actually usefull and good to know. But that is starting to be outwayed by the attitude.

bigben2k · 10-20-2002, 07:12 PM

BillA · 10-20-2002, 07:30 PM

I 'think' this thread may be the best for an overview
rereading it now makes my head spin, hopefully it becones clearer as it progresses

I'll work through your list too

EDIT: damn, that BillA guy is a wordy bastard

MadDogMe · 10-21-2002, 04:21 AM

BillA!, if AS is such a big variable for testing , why use it?. The block and CPU would get along fine without it would'nt they?(albeit with a DegC or three rise

), but a Cu to silicon(sp?) join would'nt vary at all would it?, apart from the pressure...

Has anyone tested just the block on CPU with constant temp water before?(no rad), I'd be interested to know how far above ambient it holds the CPU(on one of the 'great5' blocks

), it'd give an idea of the % of the frailtys of the rest of the system would'nt it?...

What kinda temps do people get with H2o nowadays?, above their ambient I mean?, I average 10DegC with a poorly tuned setup. What is a good No?, 7c?, 5c?... does'nt leave a lot left in the block when you take the rad's inefficiency into count...

10-19-2002, 08:25 PM	#38
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	Also that pelt controller can have far more uses than just computer cooling. There are a lot of other things it can be used for. But I hear ya about running a business and what not. Last edited by jaydee116; 10-19-2002 at 08:35 PM.

10-20-2002, 09:15 AM	#40
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Ya'll are getting off-topic, with this talk about TECs. A normalized chiller, which could keep the water temp at 20 deg C (aka ambient), and a high flow direct die cooling option would give us the lowest temperature that a waterblock could achieve, in theory. Even then, as long as there's a baseplate, the CPU temps will be ever so slightly higher. Either way, the point is that the lowest CPU temp that can be achieved with normal cooling, is several degrees above ambient. We just don't know how many degrees that is. What we do know is that without any cooling whatsoever, the max temp is in excess of 350 deg C, and that's an important piece of information! __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

10-20-2002, 01:20 PM	#45
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Fair enough. I downloaded a few documents from Intel, and I'll print them tmo, and review them, and see what I can come up with. Your point is still well taken: the TIM joint is critical, as it is single-handedly the biggest variable, based on what the users have posted. I was hoping that you could give us a rough idea of what it is that we're looking at, but barring that, I'll give it a shot. No promises. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

10-20-2002, 07:12 PM	#48
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Private discussions aside... I ran the search, and have been (trying to) read through it all. I'm obviously not done!!! In the mean time, I'd like to share links to OCAU forums, of the relevant threads: http://forums.overclockers.com.au/sh...m&pagenumber=1 http://forums.overclockers.com.au/sh...m&pagenumber=1 http://forums.overclockers.com.au/sh...m&pagenumber=1 http://forums.overclockers.com.au/sh...&highlight=tim http://forums.overclockers.com.au/sh...&highlight=tim http://forums.overclockers.com.au/sh...&highlight=tim http://forums.overclockers.com.au/sh...&highlight=tim http://forums.overclockers.com.au/sh...&highlight=tim http://forums.overclockers.com.au/sh...&highlight=tim I'm about halfway through it all... __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

10-20-2002, 07:30 PM	#49
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	I 'think' this thread may be the best for an overview rereading it now makes my head spin, hopefully it becones clearer as it progresses I'll work through your list too EDIT: damn, that BillA guy is a wordy bastard Last edited by BillA; 10-20-2002 at 10:37 PM.

10-11-2002, 02:56 PM	#26
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	you know Ben, that poor guy is going to be walking down the street and Peter is going to run up and bust him right in the face screaming "YOU ARROGANT BASTARD" and it will be all your fault now don't you feel guilty ?

10-11-2002, 08:11 PM	#29
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	Excellent input guys! Thanks!

10-11-2002, 11:44 PM	#30
godsdice Cooling Neophyte Join Date: Jan 2002 Posts: 29	Actually, the reason why noone ever goes into the "server room" where I work is because it's always at 50F. Used to have a centrifuge in there because there wasn't room anywhere else...directly below the vent. Completely messed up results....argh.

10-19-2002, 07:42 AM	#32
MadDogMe Thermophile Join Date: Aug 2002 Location: Just shut up ;) ... Posts: 1,068	What's the average temp over ambient at the mo'?, 10DegC?. it's a long way or a little way depending on how you look at it. that's how much H20 as a system has left to go. The Block?, the way to work it out I think is to have a supply of ambient temp water and feed it through the block, how high over ambient is the CPU~temp now?, that's how far H20~blocks have left... I'd like to know how close to ambient Cathars block holds temps, I'd guess it's pretty damn close...

10-19-2002, 08:09 PM	#36
Since87 Pro/Guru - Uber Mod Join Date: Sep 2002 Location: Indiana Posts: 834	I've seen the specs on the Swiftech TEC/Heatsink. It appears to me that they are claiming that it can get a 100 Watt CPU to ambient with 80 CFM airflow. I am very skeptical. I've got a spreadsheet for calculating TEC cooling under various conditions, and that kind of performance is at the extreme end of what I can believe that heatsink is capable of. (Depends on the C/W of the heatsink when cooling a 50mm X 50mm pelt.) As far as my pelt controller is concerned, I wish I'd been looking into this cooling stuff a couple years ago, because I probably could have sold enough of these things to make some money, but the window of opportunity is closing fast now. Besides I'm a design engineer. I like designing things. I don't want to deal with all the issues of running a business.

10-20-2002, 10:21 AM	#41
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	the numbers being 'quoted' here are make believe (suggest you all DO some testing or, failing that, read about some testing) understanding that any 'minimum' must also specify the flow rate and the applied power; some illustrative graphs are in this article (on the last page) - though not definitive relative to any absolute, as indeed nothing can be you're trying to pick nits about wbs when the TIM joint is FAR more critical -> and variable hello ? hello ? operator this line was disconnected . . . .

10-20-2002, 12:59 PM	#44
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	sorry Ben Dave will do that, pHaestus will do that - but I will not I taught for one semester at a Uni and learned that I was not temperamentally suited to the endless regurgitation of what I knew, when presenting it to those with little motivation I'll tell you where it is I will not summarize it, I will not paraphrase it, I will not distill it down to 'good' or 'bad' that is for YOU to do too much to read ? well try this: if you can't be bothered to read, why the hell should I type it over again for you ? "Well, that's implied, isn't it?" no I saw only one number (due to my ignore list), and it was tripe I have posted tons of graphs, and you now need me to tell you what they mean ? have you figured out what the offset due to the TIM joint is ? why not ? its been posted and discussed many times what I said before Ben, you have to start actually learning some of this stuff instead of asking the same questions over and over with your post count you can't play the newbie game

10-20-2002, 01:36 PM	#46
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	relative to the 'minimum' question: there is a facinating (to me natch) chart on the last pg which was initially posted with an explanation in a thread on this forum; then the thread starter spiked the entire thread 'cause he did not like the heat deriving from his BS claims (and put himself on my ignore list) go to OCAU where I have extensively discussed T/M, (search "T/M" ?) because the remainder is the TIM joint C/W you WILL be interested EDIT: always remember Intel's "cascade of gradients"; then focus on the element of interest

10-21-2002, 04:21 AM	#50
MadDogMe Thermophile Join Date: Aug 2002 Location: Just shut up ;) ... Posts: 1,068	BillA!, if AS is such a big variable for testing , why use it?. The block and CPU would get along fine without it would'nt they?(albeit with a DegC or three rise ), but a Cu to silicon(sp?) join would'nt vary at all would it?, apart from the pressure... Has anyone tested just the block on CPU with constant temp water before?(no rad), I'd be interested to know how far above ambient it holds the CPU(on one of the 'great5' blocks ), it'd give an idea of the % of the frailtys of the rest of the system would'nt it?... What kinda temps do people get with H2o nowadays?, above their ambient I mean?, I average 10DegC with a poorly tuned setup. What is a good No?, 7c?, 5c?... does'nt leave a lot left in the block when you take the rad's inefficiency into count...

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)