Pump heat

Sled_Dog · 08-11-2002, 12:42 AM

I turn my pump off to fill my resovoir pretty much everyday. a day or 2 ago I decided to leave it off while I slept. I had accidentally done it many times and figured why not. So off it went. I checked my idle temps a while later only to discover they ahd dropped! My CPU temp dropped about 5 degrees adn my mobo temp dropped about 2. I went to sleep and the next morning checked again. Only to discover they had fallen lower yet. It now is running wiht a few IE windows, Trillian, and MIRC at about 89F. At Idle it runs at about 87F that is 9 degrees cooler then with my pump on(pump on idles at 96) Now before you say that is still warm my room temp is normally 80 degrees and I accidentally forgot to set my voltage back when I clocked my chip back. So my voltage modded KT7A is running the chip at 2.3Vs for 100mhz oc! I have to restart someday and fix that but darn if I don't hate losing uptime. By the way my pump is a Danner model 5 500GPH. I know the probe isn't wrong cause my coolant lines right next to the block are cool.

kibbler · 08-11-2002, 03:16 AM

holy moly! you left your computer running overnight with the pump OFF?!?!

Intentionally?

Unless I misread something...

In a thread a good while back, I described how my CPU melted tygon, for crying out loud, tygon, when the pump was left off accidentally (thru no fault of mine, grr)

Sled_Dog · 08-11-2002, 08:02 AM

its been running for 2 days without it incident free. Blocks hot but its not blazing hot.

nexxo · 08-11-2002, 12:24 PM

Allrighty... that sounds sort of weird. However, it is a known physics phenomenon (try saying that quickly) that warming liquid expands and rises, and cooling liquid contracts and falls. In a closed circuit, this can cause circulation to occur spontaneously. Perhaps this is what is happening in your case?

Sled_Dog · 08-11-2002, 06:28 PM

you mean convection currents and hmm that is interesting theory

jaydee · 08-11-2002, 09:28 PM

Last time I shut my pump off to test what you are doing the CPU temp rose to 65C in a bout 2 mins. I then tunred the pump on needless to say. But the difference here is my system is not close/sealed as I have a open res. This sounds like an interesting experiment. I will seal off my system and give it shot sometime this week.

It really doesn't make since though....

koslov · 08-11-2002, 09:46 PM

Here's the only way I think that could possibly work.

You'd need a very high-flow radiator for this to work, and likely a large copper reservoir. In any case, the WB would have to be the lowest point in the system, or else the hot water would not go anywhere.

edit: Here's a more optimized plan that won't produce hot water traps.

Sled_Dog · 08-11-2002, 10:22 PM

my setup. the temp has slowly risen to almost 96F which is where it was idling with the pump, the room temp has also risen from like 78 to almost 85.

koslov · 08-12-2002, 12:19 AM

So I'm guessing this is how your system looks in a temperature snapshot. But I still can't make any sense out of it.

mfpmax · 08-12-2002, 04:18 AM

Quote:

Originally posted by nexxo
Allrighty... that sounds sort of weird. However, it is a known physics phenomenon (try saying that quickly) that warming liquid expands and rises, and cooling liquid contracts and falls. In a closed circuit, this can cause circulation to occur spontaneously. Perhaps this is what is happening in your case?

thats what I hoped would happen in case of pump failure...but it didn't when I accidently turned my pump off...twice...I now have a half dead K6-450 that does 600MHz but runs like a Pentium 100

Sled_Dog · 08-12-2002, 04:58 AM

I thought you guys had a good idea but were still wrong but I just touched my hoses and they are exactly as that M$Paint image is! The one from the chip to radiator is warm and half the rad is warm and everything else is cold.

redleader · 08-12-2002, 01:05 PM

Someone on Ars passively cooled a 300A in the day. They just got a big block, and a lot of copper pipe ziging over the outside of the case. Convection easily cooled it. Granted a 20w Celeron is a bit different then an Athlon, but at least it can be done.

nexxo · 08-12-2002, 01:11 PM

...nnnooope... convection still works. Water in the block and its connecting tubes expands. This displaces cooler, contracting water in radiator which is being pushed into the res. Water from the res, meanwhile, because of gravity, pushes down into the pump and upwards to the block again. This gives the aforementioned expanding warm water from the block a firm foothold to push off against, into the rad where it displaces the cooler water...

If this is how it works, you should find that things will go ass-up if you lower the position of the res (so perhaps you don't want to do that...

)

bigben2k · 08-12-2002, 01:20 PM

If the block is hot, then your mobo temp probe might not be working.

In other words, your CPU is running hot, you just don't know it yet.

Which processor are you using, and what clock speed is it at?

What temp probe are you using (internal diode or CPU socket probe)?

The self circulating theory doesn't work: the hot water would have to go down in the rad, to come back out.

airspirit · 08-12-2002, 02:19 PM

Ben is right. That makes no sense. You might have flash burned your thermistor. The only heat xfer would be laterally through the fluid, and heat wouldn't pump away fast enough to prevent your system from frying. You're probably air cooling it via the water block. Odds are, your processor is running just south of nuclear meltdown, and the thermistor is too damaged from the heat to understand what is happening. Put the pump on and see what happens to the temps. If it suddenly says you're running sub-ambient, you have a severe problem.

At any rate, TURN YOUR PUMP ON BEFORE YOU MAKE A KEYCHAIN!

nikhsub1 · 08-12-2002, 03:18 PM

Something is totally wrong here. I just did an experiment, I unplugged my Eheim 1048 while folding (24/7) and my temp went from 39C (normal, with pump) to 50C in less than 1 minute. This is with an on die reading from my 8K3A mobo. Something is clearly wrong with your temp sensor and I don't recommend you run the thing without the pump.

bigben2k · 08-12-2002, 03:36 PM

What some people don't realize is that the CPU will not die instantly with the pump. The heat spreads in the block, then through the water EXACTLY THE WAY IT IS NOW!

God man, do you want to put a webcam on it so that we can watch a system fry live?

airspirit · 08-12-2002, 05:26 PM

What amazes me is that he says his system has been up for a couple of days this way. If my reckoning is correct, he's running an Athlon XP chip which is known to be able to run right up to the point where it bakes into silicon toast. I don't know how his system hasn't shut down yet. I'd like for him to shut it off and take a look (and a pic) of his processor in all of its toasty goodness just as a lesson for those people that think there is a natural cooling mechanism in their WC setup. Either that or the webcam idea: that way we can all holler "WE TOLD YOU SO!" in unison when his comp does the reboot of death and refuses to post. A sad end to the experiment, but my would it be satisfying. His WC system is serving as a really bad heat sink ... it would be interesting to see how long it would go before it shut down for good.

airspirit · 08-12-2002, 05:30 PM

One thing is for certain: if he was running either a TBird or a TBred, it would have been cooked long before now.

Sled Dog: can you humor us and post some water temps? I'd almost be willing to start a $1.00 pool for the guys ... I'm guessing 77C. Any takers?

(This is assuming that this whole experience isn't total and utter rubbish).

kibbler · 08-12-2002, 05:37 PM

Quote:

God man, do you want to put a webcam on it so that we can watch a system fry live?

That'd be good for a few lols. There's a video a while back from someplace that showed what happened to Intel and AMD procs when the hsf was taken off. Smokin!

Wish I could find it just now

koslov · 08-12-2002, 06:49 PM

Here it is: http://www17.tomshardware.com/cpu/01...917/index.html

For entertainment purposes only. Not to be considered a credible source of information.

Sled_Dog · 08-12-2002, 07:39 PM

well when I said my block is hot I didn't mean liek really hot. With my old watercooling setup on this same setup I accidentally turned the pump off and the chip hit 212F! This chip is an old AXIA Thunderbird 1000. I don't give a #$%* if it burns out, if it does its XP time. But I have been running for probably 4 days now like this. The block is higher then room temp prolly in the 100 something range. You can touch it and continue touching it for as long as you want it is not really hot at all. When I hit 212 the one time I couldn't touch the block for more then say 1 second. I will try and hunt down a thermometer and get a water temp. Don't have inline probes for that. I am using the On Board probe in my old ABIT KT7A. I am really starting to buy the Convection current theory. My radiator is a direct straight up line from the CPU outlet which is at a higher point on the block then the inlet. EXPERIMENT IDEA(since this seems to have turned from a funny thing into an experiment): I will turn my radiator fan off and see if the temp rises.

bigben2k · 08-12-2002, 08:56 PM

Webcam (oops, I mean drumroll) please!

What's the wattage of that CPU, at that voltage? I can't find Radiate anywhere.

Sled_Dog · 08-12-2002, 09:02 PM

I have no idea I will try and hunt down Radiate. I turned the fan off and in the past 5 minutes since I did so it has risen 1 degree on the cpu(makes sense if it were convection currents it would move slowly, and 2 degrees on the motherboard(probably cause of the less airflow now). just found radiate and it said 27.4Ws something about that sounds wrong.

jaydee · 08-12-2002, 09:15 PM

Quote:

Originally posted by Sled_Dog
just found radiate and it said 27.4Ws something about that sounds wrong.

That is very wrong. Even the Duron600 uses more watts. That T-Boird should be at 1.75Vcore stock.

08-11-2002, 12:42 AM	#1
Sled_Dog Cooling Savant Join Date: Jul 2001 Location: Snowyland(not Alaska) Posts: 238	Pump heat I turn my pump off to fill my resovoir pretty much everyday. a day or 2 ago I decided to leave it off while I slept. I had accidentally done it many times and figured why not. So off it went. I checked my idle temps a while later only to discover they ahd dropped! My CPU temp dropped about 5 degrees adn my mobo temp dropped about 2. I went to sleep and the next morning checked again. Only to discover they had fallen lower yet. It now is running wiht a few IE windows, Trillian, and MIRC at about 89F. At Idle it runs at about 87F that is 9 degrees cooler then with my pump on(pump on idles at 96) Now before you say that is still warm my room temp is normally 80 degrees and I accidentally forgot to set my voltage back when I clocked my chip back. So my voltage modded KT7A is running the chip at 2.3Vs for 100mhz oc! I have to restart someday and fix that but darn if I don't hate losing uptime. By the way my pump is a Danner model 5 500GPH. I know the probe isn't wrong cause my coolant lines right next to the block are cool. __________________ pro/chat has a website now? wow

08-11-2002, 08:02 AM	#3
Sled_Dog Cooling Savant Join Date: Jul 2001 Location: Snowyland(not Alaska) Posts: 238	its been running for 2 days without it incident free. Blocks hot but its not blazing hot. __________________ pro/chat has a website now? wow

08-11-2002, 12:24 PM	#4
nexxo Cooling Savant Join Date: Jul 2002 Location: Brimingham, UK Posts: 385	Allrighty... that sounds sort of weird. However, it is a known physics phenomenon (try saying that quickly) that warming liquid expands and rises, and cooling liquid contracts and falls. In a closed circuit, this can cause circulation to occur spontaneously. Perhaps this is what is happening in your case? __________________ "There is a thin line between magic and madness"

08-11-2002, 06:28 PM	#5
Sled_Dog Cooling Savant Join Date: Jul 2001 Location: Snowyland(not Alaska) Posts: 238	you mean convection currents and hmm that is interesting theory __________________ pro/chat has a website now? wow

08-11-2002, 10:22 PM	#8
Sled_Dog Cooling Savant Join Date: Jul 2001 Location: Snowyland(not Alaska) Posts: 238	my setup. the temp has slowly risen to almost 96F which is where it was idling with the pump, the room temp has also risen from like 78 to almost 85. __________________ pro/chat has a website now? wow

08-11-2002, 03:16 AM	#2
kibbler Cooling Savant Join Date: Jan 2002 Posts: 107	holy moly! you left your computer running overnight with the pump OFF?!?! Intentionally? Unless I misread something... In a thread a good while back, I described how my CPU melted tygon, for crying out loud, tygon, when the pump was left off accidentally (thru no fault of mine, grr)

08-11-2002, 09:28 PM	#6
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	Last time I shut my pump off to test what you are doing the CPU temp rose to 65C in a bout 2 mins. I then tunred the pump on needless to say. But the difference here is my system is not close/sealed as I have a open res. This sounds like an interesting experiment. I will seal off my system and give it shot sometime this week. It really doesn't make since though....

08-12-2002, 04:58 AM	#11
Sled_Dog Cooling Savant Join Date: Jul 2001 Location: Snowyland(not Alaska) Posts: 238	I thought you guys had a good idea but were still wrong but I just touched my hoses and they are exactly as that M$Paint image is! The one from the chip to radiator is warm and half the rad is warm and everything else is cold. __________________ pro/chat has a website now? wow

08-12-2002, 01:05 PM	#12
redleader Thermophile Join Date: Jun 2001 Location: The deserts of Tucson, Az Posts: 1,264	Someone on Ars passively cooled a 300A in the day. They just got a big block, and a lot of copper pipe ziging over the outside of the case. Convection easily cooled it. Granted a 20w Celeron is a bit different then an Athlon, but at least it can be done.

08-12-2002, 01:11 PM	#13
nexxo Cooling Savant Join Date: Jul 2002 Location: Brimingham, UK Posts: 385	...nnnooope... convection still works. Water in the block and its connecting tubes expands. This displaces cooler, contracting water in radiator which is being pushed into the res. Water from the res, meanwhile, because of gravity, pushes down into the pump and upwards to the block again. This gives the aforementioned expanding warm water from the block a firm foothold to push off against, into the rad where it displaces the cooler water... If this is how it works, you should find that things will go ass-up if you lower the position of the res (so perhaps you don't want to do that... ) __________________ "There is a thin line between magic and madness"

08-12-2002, 01:20 PM	#14
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	If the block is hot, then your mobo temp probe might not be working. In other words, your CPU is running hot, you just don't know it yet. Which processor are you using, and what clock speed is it at? What temp probe are you using (internal diode or CPU socket probe)? The self circulating theory doesn't work: the hot water would have to go down in the rad, to come back out. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

08-12-2002, 02:19 PM	#15
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	Ben is right. That makes no sense. You might have flash burned your thermistor. The only heat xfer would be laterally through the fluid, and heat wouldn't pump away fast enough to prevent your system from frying. You're probably air cooling it via the water block. Odds are, your processor is running just south of nuclear meltdown, and the thermistor is too damaged from the heat to understand what is happening. Put the pump on and see what happens to the temps. If it suddenly says you're running sub-ambient, you have a severe problem. At any rate, TURN YOUR PUMP ON BEFORE YOU MAKE A KEYCHAIN!

08-12-2002, 03:18 PM	#16
nikhsub1 c00ling p00n Join Date: Jun 2002 Location: L.A. Posts: 758	Something is totally wrong here. I just did an experiment, I unplugged my Eheim 1048 while folding (24/7) and my temp went from 39C (normal, with pump) to 50C in less than 1 minute. This is with an on die reading from my 8K3A mobo. Something is clearly wrong with your temp sensor and I don't recommend you run the thing without the pump. __________________ :-.,_,.-:'``':-.,_,.-:'``':-.,_,.-:'``':-.,_,.-:'``':-.,_,.-: E6700 @ 3.65Ghz / P5W DH Deluxe / 2GB 667 TeamGroup / 1900XTX PC Power & Cooling Turbo 510 Deluxe Mountain Mods U2-UFO Cube Storm G5 --> MP-01 --> PA 120.3 --> 2x DDC Ultras in Series --> Custom Clear Res "Artificial intelligence is no match for natural stupidity." 1,223,460+ Ghz Folding@Home aNonForums :-.,_,.-:'``':-.,_,.-:'``':-.,_,.-:'``':-.,_,.-:'``':-.,_,.-:

08-12-2002, 03:36 PM	#17
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	What some people don't realize is that the CPU will not die instantly with the pump. The heat spreads in the block, then through the water EXACTLY THE WAY IT IS NOW! God man, do you want to put a webcam on it so that we can watch a system fry live? __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

08-12-2002, 05:26 PM	#18
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	What amazes me is that he says his system has been up for a couple of days this way. If my reckoning is correct, he's running an Athlon XP chip which is known to be able to run right up to the point where it bakes into silicon toast. I don't know how his system hasn't shut down yet. I'd like for him to shut it off and take a look (and a pic) of his processor in all of its toasty goodness just as a lesson for those people that think there is a natural cooling mechanism in their WC setup. Either that or the webcam idea: that way we can all holler "WE TOLD YOU SO!" in unison when his comp does the reboot of death and refuses to post. A sad end to the experiment, but my would it be satisfying. His WC system is serving as a really bad heat sink ... it would be interesting to see how long it would go before it shut down for good.

08-12-2002, 05:30 PM	#19
airspirit Been /.'d... have you? Join Date: Jul 2002 Location: Moscow, ID Posts: 1,986	One thing is for certain: if he was running either a TBird or a TBred, it would have been cooked long before now. Sled Dog: can you humor us and post some water temps? I'd almost be willing to start a $1.00 pool for the guys ... I'm guessing 77C. Any takers? (This is assuming that this whole experience isn't total and utter rubbish).

08-12-2002, 06:49 PM	#21
koslov Cooling Savant Join Date: Jul 2002 Location: Posts: 264	Here it is: http://www17.tomshardware.com/cpu/01...917/index.html For entertainment purposes only. Not to be considered a credible source of information.

08-12-2002, 07:39 PM	#22
Sled_Dog Cooling Savant Join Date: Jul 2001 Location: Snowyland(not Alaska) Posts: 238	well when I said my block is hot I didn't mean liek really hot. With my old watercooling setup on this same setup I accidentally turned the pump off and the chip hit 212F! This chip is an old AXIA Thunderbird 1000. I don't give a #$%* if it burns out, if it does its XP time. But I have been running for probably 4 days now like this. The block is higher then room temp prolly in the 100 something range. You can touch it and continue touching it for as long as you want it is not really hot at all. When I hit 212 the one time I couldn't touch the block for more then say 1 second. I will try and hunt down a thermometer and get a water temp. Don't have inline probes for that. I am using the On Board probe in my old ABIT KT7A. I am really starting to buy the Convection current theory. My radiator is a direct straight up line from the CPU outlet which is at a higher point on the block then the inlet. EXPERIMENT IDEA(since this seems to have turned from a funny thing into an experiment): I will turn my radiator fan off and see if the temp rises. __________________ pro/chat has a website now? wow

08-12-2002, 08:56 PM	#23
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Webcam (oops, I mean drumroll) please! What's the wattage of that CPU, at that voltage? I can't find Radiate anywhere. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

08-12-2002, 09:02 PM	#24
Sled_Dog Cooling Savant Join Date: Jul 2001 Location: Snowyland(not Alaska) Posts: 238	I have no idea I will try and hunt down Radiate. I turned the fan off and in the past 5 minutes since I did so it has risen 1 degree on the cpu(makes sense if it were convection currents it would move slowly, and 2 degrees on the motherboard(probably cause of the less airflow now). just found radiate and it said 27.4Ws something about that sounds wrong. __________________ pro/chat has a website now? wow

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