Which pump suited to pump through 75 feet of tubing ?

[Pudi]

Ok basically which pump would be better for this job? so I can have decent water flow through 75feet of 3/8 cooper coil tubing. Normal DCC will do, or should I invest towards some Iwaki MD30?

[NaeKuh]

your watercooling the condensor?

OR, your dropping a drum underground?

:D

You need an industrial pump my friend. I would grab something bigger then the MD30

[aspire.comptech]

I don't know if either will do it because of the amount of head your going to need...

EDIT: God, that sounds pretty dirty...

[Pudi]

your dropping a drum underground?

:D


Yeap new house new garden and a lots of new idea's has been born last week :D


You need an industrial pump my friend. I would grab something bigger then the MD30

Damn really :( so it looks like I'm going back to 45 gallon water drum idea, no cooper tubes for me :down:

[Serpentarius]

what's this? watercool your summer house?

[Anemone]

Three Iwaki's would do it :p

[aspire.comptech]

As would a triple cascade...

Cost about the same too...

[BlueAqua]

Using the ground as a heatsink would work but I think it's probably more troublesome than it's worth. I love projects and it would be fun, but I don't think it would be something that I would want day after day.

[Pudi]

I think it's been done before by Marci with 150 feet of cooper tubing and some Central Heating pump.

http://www.over-clock.com/ivb/index.php?showtopic=4135

i love the idea of silent watercooling, no more noisy fans :D

[NaeKuh]

As would a triple cascade...

Cost about the same too...

:rofl:
:ROTF:
:yepp:

the drum is more green friendly tho. And his coolant will probably be no higher then 5-10C depending on how deep and where he digs.

Pudi you may want to double up on the pumps. That could help. But you still need a strong pump.

[eXa]

Why only 3\8 tubing? for that lenght i would go way up. 5\8 atleast.
Edit: Meh... copper coil.. Do u need to use copper coil?

[Pudi]

Edit: Meh... copper coil.. Do u need to use copper coil?

I don't understand what you mean?

I thought cooper coil is the cheapest and the best option for heat transfering.

[silverphoenix]

you need this

http://www.pumpwarehouse.com/image/pwr.gif

[eXa]

I don't understand what you mean?

I thought cooper coil is the cheapest and the best option for heat transfering.

So the 75feet of copper coil is doing all the cooling? ur not using anything else?

[HotGore]

Maybe this?

http://www.canadiantire.ca/browse/product_detail.jsp?PRODUCT%3C%3Eprd_id=84552444215 4421&FOLDER%3C%3Efolder_id=1408474396672116&bmUID=1215706433320&deptid=1408474396672082&ctgrid=1408474396672086&subctgrid=1408474396672116

[Pudi]

That is sick! Hayward Power Pump 1.5 HP 26lbs and something crazy like 4000GPH :eek: no thank you :up: and yes I was thinking just about 75 feet of cooper tube buried 4-5 feet deep and spread over my garden, but i'm having a really hard time with finding any geothermal watercooling projects :(

[NaeKuh]

So the 75feet of copper coil is doing all the cooling? ur not using anything else?

you miss my first post? :rofl:

ive seen it all with water, so i knew exactly what he was gonna do. :up:

you need to go deeper.

like 10 - 15 feet.


mmm how was it pulled off? Well the drum had a line that you can fill and bleed. The drum was burried underground.

Inside the drum was coils of copper that would sit inside. The coils then came out.

So the drum had 4 openings.

2 for the drum itself, and 2 for the copper.

The drum coolant doesnt need to be replaced very often. So people would throw ethene + water. The copper coil is your water line.

So the coil takes water inside the drum, The drum cools it and the coil takes it out.

Every 2 yrs or so, you need to flush the drum out by replacing its coolant.


This is how i saw it pulled off. You need to becareful about the drum. I believe some soil is acidic so it can eat though over time. But were talking many many years.


Of course, people think im a noob, so you dont need to listen to me. :rofl:

[Chruschef]

wow thats an amazing idea. :shocked:
Copper coil underground in soil? thats a really great idea =O.

[eXa]

Id say, bury a drum and go with normal tubing instead.
With 5\8(internal) and the drum i think a MD30RZ would be plenty. Depends on how much ur hooking up in the hot hend tho.

[orclev]

Id say, bury a drum and go with normal tubing instead.
Long copper coil would give better heat dissipation with any sort of flow. A buried drum would work for a reservoir so long as it was deep enough and the fluid could sit in it for long enough, but the low surface area to volume ratio would make it a poor choice otherwise. Of course, in terms of raw thermal capacity, as opposed to thermal transfer, it can't be beat.

[NaeKuh]

Long copper coil would give better heat dissipation with any sort of flow. A buried drum would work for a reservoir so long as it was deep enough and the fluid could sit in it for long enough, but the low surface area to volume ratio would make it a poor choice otherwise. Of course, in terms of raw thermal capacity, as opposed to thermal transfer, it can't be beat.

the drum acts as a large res, and cooling area.

coils of copper inside the drum provide the surface area.

And i told him he needs to go deeper like 10-15feet, to keep below the frost line.

[eXa]

If ur using a very small container it would heat up after a while. But if its large enough it wont. I dont think coils of copper inside the drum would do much, since the coling area to the surroundings would still be the same. And water going into and out of the drum would circulate the water inside the drum anyways.

Edit: A 55gallon plastic drum has the surface area of about 23500 square centimeters, and when u get all the moist dirt surrounding the entire drum i think it would cool quite well.

[Eldonko]

I would go with a Eheim 1264. I have a 1250 and it has plenty of power for 15 feet tubing, 2 heatercores and 4 blocks. The 1264 has 4x the power @ 4500 I/h.

[Pudi]

Yeah but still only 3.5 m head

[Eldonko]

3.5 is a lot, is it not? Just trying to suggest something that wont cost a fortune in power.

[orclev]

The problem is more with how long the fluid sits inside the drum before it's recirculated inside the cooling lines. If it flows too fast you wouldn't get much cooling from the drum. Then again, assuming you're intake is at the top of the drum, and your outtake is at the bottom, the laws of thermodynamics would ensure that the coldest water was being pumped through the line. You would also be better off with a metal drum as opposed to plastic, the plastic would tend to act as an insulator, although it does have the nice property of holding up well over time when exposed to the elements, which cannot be said for the metal drums. One would also tend to wonder if you couldn't get a similar effect with a metal 75 gallon drum in a air conditioned room. Assuming the fluid in the drum was at ambient when you switched the system on, I wonder if the surface area of the drum, and the thermal capacity of the fluid would be enough to keep the fluid only a few degrees above ambient, even without any sort of radiator.

[Big Lar]

I have an Iwaki WMD20 RLT thats pushing 25' each way with 1/2" internal sized tube, going thru 3 blocks and back to my Rez. I get a full tube on the return, and would venture a guess at about 2 to 3 gpm. These things have more power than some realise.

Larry

[NaeKuh]

3.5 is a lot, is it not? Just trying to suggest something that wont cost a fortune in power.

not when your looking at 75feet

which is more then 25m.

Wait... 25m...

an RD-30 could do that.

Orclev you brought up some good points however.

1. if he gets below the frost line, his temps would be somewhere between 5-10C. What room do you have that gets that cold all year all time long?
2. If he has enough coolant as a battery in the drum, it would take a while for it to warm up. Not to mention he has a constant 5-10C cooling source.
3. The only problem is getting the drum under ground, and then having the tubing come out. I cant think of a nice and clean solution to that unless your a pro plumber.

[silverphoenix]

https://www.haywardnet.com/images/performance_data/110a.gif

This is also through a 2.5" ID tube lol

[cdawall]

sounds like a job fro a pond pump :p

http://www.homedepot.com/webapp/wcs/stores/servlet/ProductDisplay?storeId=10051&productId=100020661&langId=-1&catalogId=10053&ci_src=14110944&ci_sku=100020661&cm_mmc=1hd.com2froogle-_-product_feed-_-D28X-_-100020661

http://www.petsmart.com/product/index.jsp?productId=2754531&utm_medium=googleproduct&mr:referralID=NA&mr:trackingCode=941DA465-BF4D-DD11-98CA-001422107090&utm_source=cse

[majestik]

not when your looking at 75feet

which is more then 25m.
Unless I misread something, he's not planning on pumping water 25m VERTICALLY

[silverphoenix]

sounds like a job fro a pond pump :p

http://www.homedepot.com/webapp/wcs/stores/servlet/ProductDisplay?storeId=10051&productId=100020661&langId=-1&catalogId=10053&ci_src=14110944&ci_sku=100020661&cm_mmc=1hd.com2froogle-_-product_feed-_-D28X-_-100020661

http://www.petsmart.com/product/index.jsp?productId=2754531&utm_medium=googleproduct&mr:referralID=NA&mr:trackingCode=941DA465-BF4D-DD11-98CA-001422107090&utm_source=cse

Lol those have similar #'s to the Hydors. For 75' of tubing those are gonna barely trickle if at all lol. But srsly you may look into some heavy duty pond pumps(not these lol) or a small pool motor. The cool thing about most pool pumps is the motor is separate from the impeller housing drastically reducing heat dump by a 1/3HP motor lol. Martin's excel sheet only lets me use 25' of tubing lol but an RD30 at 24V with 25' (1/3 your tubing) gets 2.95gpm so with 75' your gonna get less than 1gpm with an RD30, this is before taking into account the coiling the blocks etc. Maybe Dual RD30's would do. But I'd go for a small but tough pool pump parked outside ontop of the ground where you're gonna bury this thing lol.

[Pudi]

So now, how am I suppose to dig a hole which is 15 feet deep, 15 freakin feet :rolleyes:

[NaeKuh]

So now, how am I suppose to dig a hole which is 15 feet deep, 15 freakin feet :rolleyes:

long arms.

or small midgets. :rofl:


Unless I misread something, he's not planning on pumping water 25m VERTICALLY

each section of tubing adds resistance. You need to account for that as well.

Have you tried blowing into 75feet of tubing.

ive done 50feet when i buy masterkleer, trust me its not easy.

[cdawall]

Lol those have similar #'s to the Hydors. For 75' of tubing those are gonna barely trickle if at all lol. But srsly you may look into some heavy duty pond pumps(not these lol) or a small pool motor. The cool thing about most pool pumps is the motor is separate from the impeller housing drastically reducing heat dump by a 1/3HP motor lol. Martin's excel sheet only lets me use 25' of tubing lol but an RD30 at 24V with 25' (1/3 your tubing) gets 2.95gpm so with 75' your gonna get less than 1gpm with an RD30, this is before taking into account the coiling the blocks etc. Maybe Dual RD30's would do. But I'd go for a small but tough pool pump parked outside ontop of the ground where you're gonna bury this thing lol.

so something like htis

http://www.skycraftsurplus.com/index.asp?PageAction=VIEWPROD&ProdID=1615

[orclev]

Heh, Skycraft is around the corner from me. You can find some great stuff there, but there's also a ton of junk so you've got to be a bit careful.

[Pudi]

I've noticed one thing about powerful swimming pool pumps, it is advised that these pumps are NOT run continuously.

[eXa]

*Sigh* do we really need to get through head and flow again? Eheim... come on, go back to [H] :mad:

1 or 2 RD30, but u need a meanwell powersupply too. MD30RZ gives u 11m of head and runs of 120v. U can use 2 of them if u think 1 aint enough.

I know a plastic drum wont transfer the heat as well as an metall drum. But if id want to bury a drum i wouldnt want anything that could rust. If i figured out a 55gallon plastic drum wouldnt cut it i would just get a larger drum instead.
But i wouldnt be afraid to use an 55gallon plastic drum, even if it was made of copper, the surroundings would still hold the heat around the drum so i dont think it matters that much. I mean, its 200l of water ur trying to heat up. Even in an isolated enviroment that would take hours. When ur enviroment is 5-10c all around without much isolation... good luck.

[Pedalmonkey]

just get some dynamite, that should dig your hole. but serioulsy one of those giant drills they use for geothermal cooling and heating of houses, which is esentialy what your doing, would be perfect. wouldn't be cheap but easy and fast. we all like easy and fast dont we??? hell you could even set ur house up geothermal while ur at it. good for the earth and ur PC. :D

[eXa]

The problem is more with how long the fluid sits inside the drum before it's recirculated inside the cooling lines.
Oh yeah? so u would want a really slow pump in a normal WC system to keep the water as long as possible in the rad?

Seriously, it doesnt work like that.

[orclev]

Oh yeah? so u would want a really slow pump in a normal WC system to keep the water as long as possible in the rad?

Seriously, it doesnt work like that.
Actually yes you would, so long as it was able to flow through the waterblocks sufficiently. The longer it's in the radiator the closer to ambient it will be, likewise the longer it stays in the waterblock the more heat it can absorb, but of course with the waterblock you want to transfer the heat away from the chip as quickly as possible so a balance needs to be maintained between the fluid traveling too fast through the waterblock to sufficiently transfer the heat, and too slow to transport it away from the block.

[Chruschef]

i know a little about thermodynamics in watercooling, so correct me if i'm wrong.. But this is a closed watercooling loop so all the water temperatures should be the same, the faster the pump the better because more water moves across the heatblock absorbing more heat and thus lowering temperatures.

[eXa]

@Orclev
Chirst. Stop writing and come back when u figured out why u are wrong. Stop spreading misinformation. :mad:

Hint. Time spent in rad vs time spent in block is the same no matter how much flow u got.

[twwen2]

@Orclev
Chirst. Stop writing and come back when u figured out why u are wrong. Stop spreading misinformation. :mad:

Hint. Time spent in rad vs time spent in block is the same no matter how much flow u got.

QFT!:eek:

[NaeKuh]

i know a little about thermodynamics in watercooling, so correct me if i'm wrong.. But this is a closed watercooling loop so all the water temperatures should be the same, the faster the pump the better because more water moves across the heatblock absorbing more heat and thus lowering temperatures.

under normal conditions somewhat yes.

its recorded to be less then 1c at the inlet and exit of a rad.

however when talking about 75FT! :rofl:

There will be hot cold and warm sections thoughout the loop.

For you noobs to visual this:
THe average loop is at most 10 feet. Think of all that tubing in your computer and multiply it by 7. :D

[twwen2]

There will be hot cold and warm sections thoughout the loop.

Actually you've got a point there.

[tw33ter]

but if the water moves too quickly through the rad, does it get cooled as well as it would, moving slower so that it spends more time in the rad?

[NaeKuh]

[QUOTE=tw33ter;3131133]but if the water moves too quickly through the rad, does it get cooled as well as it would, moving slower so that it spends more time in the rad?

if your asking this question.

read andy's post above, mcoffey, and ask yourself this question again.

[road-runner]

So now, how am I suppose to dig a hole which is 15 feet deep, 15 freakin feet :rolleyes:

With one of these... :up:

http://equipmentbychesapeake.com/images/database/DD_daewoo_doosan_crawler_excavator_DX300.jpg

[NaeKuh]

With one of these... :up:


:rofl:

RR you need to sig your nice new processor. :up:

i think its very sig worthy. :yepp:

[road-runner]

:rofl:

RR you need to sig your nice new processor. :up:

i think its very sig worthy. :yepp:

Hows that? Where is yours?

[NaeKuh]

Hows that? Where is yours?

Hehe, looks like my sig on anandtech. :P

dont sig my comps. :yepp:

i like to keep that newbie look. :rofl:

[road-runner]

Hehe, looks like my sig on anandtech. :P

dont sig my comps. :yepp:

i like to keep that newbie look. :rofl:

I usually dont, there is not enough room to put all my rigs in there... :up:

[YugenM]

I don't think Blade Runner from Zero Fan Zone did 15ft, did he? More like 6ft if IIRC

[silverphoenix]

Doesn't work like that man. One molecule of fluid can only absorb so much heat. At any one place in time, there only so many molecules present as it equates to volume. So the quicker you move that volume thru the block, the more heat it can absorb, and the quicker it goes thru the rad, the more heat it can get rid of. Now, after a point, there's only so much heat to absorb and get rid of, so higher flow stops adding greater benifit.

That's an over simplification, and I've tried to explain it in common terms, but that''s the way it basically works.

QFT also the big thing is you never reach the thermal load limit of the water, the higher the delta the increase in rate of energy being transferred (colder air cools faster than warmer air), Ideally you want the water to move fast through the cpu and stay in the rad as long as possible. The only way to do this is to have large Radiators, some of it can be made up with higher airflow through the radiator.

[orclev]

Hint. Time spent in rad vs time spent in block is the same no matter how much flow u got.
Flow is really the wrong term here. Flow is a constant for the system, what is however variable is velocity. If you can arrange it so that the fluid is travelling fast through the waterblock (say by having a smaller diameter tubing), versus what's going through the radiator (either by making the radiator larger, or increasing the diameter of the radiator tubing which is essentially the same thing), the fluid will spend more time in the radiator and therefor radiate more heat.


Doesn't work like that man. One molecule of fluid can only absorb so much heat. At any one place in time, there only so many molecules present as it equates to volume. So the quicker you move that volume thru the block, the more heat it can absorb, and the quicker it goes thru the rad, the more heat it can get rid of. Now, after a point, there's only so much heat to absorb and get rid of, so higher flow stops adding greater benifit.

That's an over simplification, and I've tried to explain it in common terms, but that''s the way it basically works.

See the following.


QFT also the big thing is you never reach the thermal load limit of the water, the higher the delta the increase in rate of energy being transferred (colder air cools faster than warmer air), Ideally you want the water to move fast through the cpu and stay in the rad as long as possible. The only way to do this is to have large Radiators, some of it can be made up with higher airflow through the radiator.

Yes, the longer it's in the radiator, the more time it has to dissipate the heat, likewise the longer it's in the waterblock the more time it has to absorb heat, but in the case of the waterblock you don't want it to reach equilibrium (unlike the radiator) so you don't actually want it to stay in the waterblock for very long. The ideal with the waterblock is high velocity, high volume, but barring that, high velocity low volume is the second best. With the radiator low velocity high volume is the ideal. Thankfully the Bernoulli principle is working in our favor in this case because we can increase the volume and it decreases the velocity at the same time. This same cannot be said for the waterblock, because we want to maximize both volume and velocity, so ultimately the waterblock is the limiting factor. Now, in a truly closed loop, the entire system is under the same pressure, so you can't vary the velocity except if you also vary the volume, but if you have a sufficiently large reservoir the pressure need not be uniform across the entire system, particularly if you have the reservoir feeding multiple loops.

[Stigma]

The big issue here isn't the 75feet of tubing, but rather how much head (ie. the difference between the highest and lowest point in your loop) you are going to need. if the 75feet of tubing was laid out flat on the ground, then even the most basic cheap pump would be able to push it (albeit probably very slowly). if the tubing was going straight up and down on the other hand, the same pump would stall after just a few feet.

So, from what I've read here you are going to need like 5 feet of head, and then a generous amount of extra power beyond that to account for the pressure loss that the 75feet will incur.

I would recommend you try an Iwaki MD-30RLZ (the high pressure version, RLZ, instead or RLT, very important). It has a max head of 26feet, so it should be able to handle this. It costs around 200USD new, but I recently saw one going (used) for a measly 30 bucks on ebay. Since the Iwakis are magpumps and also very high-quality they are nearly indestructible, so its usually not a problem that they have been used. Ebay is also a good source for even stronger Iwaki pumps really cheap, but ou have to keep in mind that the MD-30 allready draws like 120watts, and the even stronger ones draw even more...

-Stigma

[twwen2]

the fluid will spend more time in the radiator and therefor radiate more heat.

But you could argue that a higher velocity (and thus more turbulence) through the radiator will mean greater heat transfer from the water to the chambers and fins. Isn't that why the HWLabs GTX use thin tubes in their rads? Increase restriction, increase velocity/turbulence, higher heat transfer :shrug:

[CyberDruid]

An option might be to source a used dive tank. These are very heavy gauge aluminum. I would just run a 50% Xerex mix right through that...Use a pickup tube at the bottom where it's coolest and that's
that. No coil. Let the tank actually do the cooling. This makes the hole a lot smaller: you could rent a gas powered auger and get down pretty far.

[orclev]

But you could argue that a higher velocity (and thus more turbulence) through the radiator will mean greater heat transfer from the water to the chambers and fins. Isn't that why the HWLabs GTX use thin tubes in their rads? Increase restriction, increase velocity/turbulence, higher heat transfer :shrug:

Actually they use thin tubes because it gives greater surface to volume ratio. Turbulence wouldn't add any significant amount of heat transfer other than by more uniformly distributing the heat within the fluid, but that would be such a minor variance when talking about volumes of this magnitude that its effect is negligible. A sphere will of course have the lowest surface to volume ratio which is why fluids naturally form spheres in microgravity, and a flat plane is the greatest surface to volume ratio, so the closer to a flat plane (flattened tube in this case), the more surface area the fluid is exposed to, and the greater the heat transfer. Given a large enough delta between the two time spent in the radiator won't really matter much, but to achieve the closest to equilibrium (which in the radiator will be ambient) requires a longer time spent in the radiator.

Edit: For a graphic example of this, see this image from the Guide To WaterCooling and Leak Testing sticky on this forum (posted by maxxxracer):
http://www.xtremesystems.org/forums/attachment.php?attachmentid=45743&stc=1&d=1144713834

[eXa]

Flow is really the wrong term here. Flow is a constant for the system, what is however variable is velocity. If you can arrange it so that the fluid is travelling fast through the waterblock (say by having a smaller diameter tubing), versus what's going through the radiator (either by making the radiator larger, or increasing the diameter of the radiator tubing which is essentially the same thing), the fluid will spend more time in the radiator and therefor radiate more heat.



See the following.



Yes, the longer it's in the radiator, the more time it has to dissipate the heat, likewise the longer it's in the waterblock the more time it has to absorb heat, but in the case of the waterblock you don't want it to reach equilibrium (unlike the radiator) so you don't actually want it to stay in the waterblock for very long. The ideal with the waterblock is high velocity, high volume, but barring that, high velocity low volume is the second best. With the radiator low velocity high volume is the ideal. Thankfully the Bernoulli principle is working in our favor in this case because we can increase the volume and it decreases the velocity at the same time. This same cannot be said for the waterblock, because we want to maximize both volume and velocity, so ultimately the waterblock is the limiting factor. Now, in a truly closed loop, the entire system is under the same pressure, so you can't vary the velocity except if you also vary the volume, but if you have a sufficiently large reservoir the pressure need not be uniform across the entire system, particularly if you have the reservoir feeding multiple loops.

Wow... so much writing for nothing. This still doesnt mean u want a slow pump because then the water would stay longer in the rad. It doesnt work like that. I told u to come back when u got that, time isnt there yet i see.

[eXa]

The big issue here isn't the 75feet of tubing, but rather how much head (ie. the difference between the highest and lowest point in your loop) you are going to need. if the 75feet of tubing was laid out flat on the ground, then even the most basic cheap pump would be able to push it (albeit probably very slowly). if the tubing was going straight up and down on the other hand, the same pump would stall after just a few feet.

So, from what I've read here you are going to need like 5 feet of head, and then a generous amount of extra power beyond that to account for the pressure loss that the 75feet will incur.

I would recommend you try an Iwaki MD-30RLZ (the high pressure version, RLZ, instead or RLT, very important). It has a max head of 26feet, so it should be able to handle this. It costs around 200USD new, but I recently saw one going (used) for a measly 30 bucks on ebay. Since the Iwakis are magpumps and also very high-quality they are nearly indestructible, so its usually not a problem that they have been used. Ebay is also a good source for even stronger Iwaki pumps really cheap, but ou have to keep in mind that the MD-30 allready draws like 120watts, and the even stronger ones draw even more...

-Stigma

It will be a closed loop so elevation doesnt mean anything. But because of so much tubing and the restriction that tubing itself introduce, you would want a powerfull pump. I agree on the MD30 pump...

[orclev]

Wow... so much writing for nothing. This still doesnt mean u want a slow pump because then the water would stay longer in the rad. It doesnt work like that. I told u to come back when u got that, time isnt there yet i see.

Just because you ignore what I say doesn't mean you're right. You're also making a straw man. I never said you'd want a slow pump. What I said was that a "slow" pump would be best assuming that you still had proper flow through the waterblocks. With only one pump that's not possible, but if you had multiple pumps and reservoirs acting to break them into separate sections you could have a high flow pump feeding your waterblocks, and a low flow pump feeding your radiators.

[eXa]

I dont ignore what u said. u are plain wrong, now ur even trying to add ress and different pumps into it too.
The higher the flow, the better the rad perfoms. Just because the water moves faster through the rad doesnt make it cool worse. the water flows faster, but it aslo makes more runs through the rad. What im saying is that no matter how much flow u got, the water will still spend the same time inside the rad. More flow gives less surface tension, it transfer the heat from the water to the copper better.

Edit: For ur statement about the rad to be true, would mean that u are using "fresh" water all the time and u needed all the time in the rad as u could get because the water aint coming back to the rad anymore. That is not the case. It is a closed loop, u are using the same water over and over again. And making 2 loops(only connected by the ress), 1 for rad and 1 for block is just stupid. U would have water coming from the cpu beeing sucked right back into the cpu loop again without beeing cooled by the rad. So it wouldnt make any difference. In fact it would be worse since the rad is performing worse and because u have a second uneeded pump dumping heat into the water.

[silverphoenix]

Just because you ignore what I say doesn't mean you're right. You're also making a straw man. I never said you'd want a slow pump. What I said was that a "slow" pump would be best assuming that you still had proper flow through the waterblocks. With only one pump that's not possible, but if you had multiple pumps and reservoirs acting to break them into separate sections you could have a high flow pump feeding your waterblocks, and a low flow pump feeding your radiators.

technically what you are saying is correct but unless In a really long loop like this one with significant temperature differences in the water it won't make a difference. Due to the fact that water's specific heat is so High. But in this case yes you may not want too much GPM, but I think that is the least of his problems, he needs to set GPM goal through his blocks and try to reach that but not go over to maximize the cooling potential of the long coil esp since it's technically passive.

There is an optimal GPM for wc that depends on 2 non linear trends one depending on the block's efficiency and the other the radiators. Now usually the radiators aren't large enough to make that big of a dent because the 2 trends cross where the graph would normally be flatter, now by keeping one side the same(aka the blocks) but massively increasing the surface area of the "radiator" (coil) you drastically change how GPM affects not only both the blocks, but the radiators dissipation, the 4th factor is the air flow or in this case surrounding permafrost that will take away the heat, the permafrost can't be made colder by you or turned up like a fan. But considering the overall surface area compared to a radiator(with all its rows and fins). I'd say the difference in GPM is nill and you should aim for your target flow, whatever it may be >1.5GPm prolly, after that the bigger thing is how you can maximize surface area of the coils like stated before round isn't best but it's free flowing.

[eXa]

@Mcoffey

His theroy has some single points that are correct but its flawed. The overall is just plain wrong.

[orclev]

I dont ignore what u said. u are plain wrong, now ur even trying to add ress and different pumps into it too.
The higher the flow, the better the rad perfoms. Just because the water moves faster through the rad doesnt make it cool worse. the water flows faster, but it aslo makes more runs through the rad. What im saying is that no matter how much flow u got, the water will still spend the same time inside the rad. More flow gives less surface tension, it transfer the heat from the water to the copper better.
Ah, now there's a good argument. I hadn't considered surface tension.

I think the problem is one of arguing two different problems. I can see how a smaller amount of heat dissipation per pass, but more passes would be equivalent to fewer passes with more heat dissipation, although you would dissipate heat faster the longer the water was in the radiator, it's just with higher velocity and a closed loop you achieve the same the same effect. The problem was I was arguing the point from the standpoint of a single pass and not taking the entire system into account, and you are arguing for the entire system and ignoring the single pass.

I'm correct with regards to a single pass through the system, but wrong for the system as a whole. You're correct with regards to the entire system, but wrong about a single pass.

[eXa]

Yes i havent been talking about a single pass at all. It really aint that relevant in a closed loop anyways.

[eXa]

Yeah. Simply:
More flow gives more of the red area(moving hot water) and less of the blue area(still standing cold water). And that aplies to both waterblocks and radiators.
http://www.xtremesystems.org/forums/attachment.php?attachmentid=45743&stc=1&d=1144713834

[orclev]

Rght, which is exactly what I was saying. Slowing down the flow so it stays in the rad longer doesn't work as it applies to WC'ing. And to make such a statement is wrong, misleading, and basically a myth.

andyc

Actually it does work, it just doesn't make a difference, the system performs identically either way, it's just more of a hassle to achieve that effect, and wasted energy in the form of a more complex loop and more pumps.

[eXa]

No... do i need to explain why that 2 loop, 2 pump, 1 ress idea dont work? did u read my edit?

[orclev]

No... do i need to explain why that 2 loop, 2 pump, 1 ress idea dont work? did u read my edit?

I just read the edit, but what I should have said was 2 ress, 2 pumps... I had said that earlier but it doesn't matter at this point, it's really a moot point. Even with 1 ress and 2 pumps it would work, although the efficiency would be worse. Assuming you could set it up so that the water flowed through the radiator slower than it did going through the blocks it still performs as well as if it flowed the same speed through the entire system. The added complexity of changing the flow through the various parts though would hurt the system as a whole.

[NaeKuh]

orclev

im sorry to say this, but your wrong, and andy is right.
And he knows more then you do, as do we.

And it was shown by martin that increase in flow results in better temps.
This is why we try to keep flow at least 1gpm. Your trying to resurrect that arguement that went on in SLI Zone.
Your digging yourself a very big and deep hole.

[orclev]

Saying "it just doesn't work that way" isn't really helpful. Saying, "it doesn't matter in this case because" however is very helpful. As we've established already, it does in fact work that way, but, it doesn't matter in this case because the rules change depending on whether it's a closed loop or open loop. I take part in these discussions because it improves my understanding of the theory, and sometimes, as is the case now, teaches me areas where I've overlooked something.

Fun things I've learned today:
Surface Tension decreases with velocity.
Diamonds conduct heat better than Copper (a lot better).
In a closed loop efficiency of a single pass must be multiplied by the flow to determine the efficiency of the entire system.

And just for the heck of it, I calculated the surface area to volume ratios of a 1 ft. piece of round 1/2 ID tubing and 0.1x1.97 rectangular tubing (which have ignoring rounding errors, identical volume), and they are ~7.99 and ~20.12 respectively. This also assumes that the walls of both pieces of tubing have 0 thickness, in practice that would need to be accounted for as well.

[twwen2]

Saying "it just doesn't work that way" isn't really helpful. Saying, "it doesn't matter in this case because" however is very helpful.

The problem is that he already explained why it doesn't work, but you were'nt willing to accept his explanation. That's why he had to resort to a statement like that. I wish more people would respect the well-established opinions (let alone the well-established theory) of the experts round here.

[silverphoenix]

The problem is that he already explained why it doesn't work, but you were'nt willing to accept his explanation. That's why he had to resort to a statement like that. I wish more people would respect the well-established opinions (let alone the well-established theory) of the experts round here.

True but you shouldn't alway just accept it. I like how he remained civil about the whole thing, that way you get a positive learning experience and sometimes the "established" facts do change. Always question everything, but don't do it in an obnoxious stupid way lol.

[eXa]

The problem with 2loop,2pump,1ress is that it is still a closed loop and ur still always using the same water. So in the end, the time spent in the rad vs the time spent in the block is still equal. But because the water flows more slowly through the rad makes it perform worse(surface tension and stuff) and ur also dumping more heat into the water because u got another usless pump. Therefor that system doesnt work.
I dont mean the cpu would overheat when i say "doesnt work". The entire point with that system is to make the rad perform better than in a normal 1 loop 1 pump 1 ress(or whatever) system, wich it doesnt. It does the opposite, so the system dont work.;)

[twwen2]

True but you shouldn't alway just accept it. I like how he remained civil about the whole thing, that way you get a positive learning experience and sometimes the "established" facts do change. Always question everything, but don't do it in an obnoxious stupid way lol.

Also true. I'll give you credit orclev, you behaved well and remained polite during the dabate, more than can be said for some :shakes:. I do agree that it's good to have constructive discussions, we all learn more that way.

[orclev]

What he said was:

@Orclev
Chirst. Stop writing and come back when u figured out why u are wrong. Stop spreading misinformation. :mad:

Hint. Time spent in rad vs time spent in block is the same no matter how much flow u got.
Which really didn't explain anything, and was on a different point entirely (that is, time in radiator versus time in blocks). Had he said that within the loop total time spent in the radiator is constant no matter what the flow is so the amount of heat dissipated over time doesn't change, that would have showed exactly where my theory went wrong. Instead I thought he was trying to make a point about flow being constant so a lot of time was wasted while I tried to figure out exactly what he was talking about. But none of that matters at this point as we've already worked all this out.

I respect anyone that is also respectful. Politely pointing out mistakes, or even better providing links to other threads or sites that explain exactly why a theory is wrong or right is the way to earn my respect. If you have practical experience with something I also respect that, although that would lead me to question various theories and try to figure out exactly why some behavior was being observed. Just saying come back when you figure it out is roughly equivalent to saying STFU noob, GTFO, and is not the way to get my respect. The fact that eXa took the time to debate this with me means that I do respect his opinion, even if he was a bit short with me in the beginning, and likewise mcoffey. I hope you (twwen2) also learned something from all this because you obviously didn't understand the reason for the flattened tubes in radiators as opposed to the rounded ones.

[orclev]

The problem with 2loop,2pump,1ress is that it is still a closed loop and ur still always using the same water. So in the end, the time spent in the rad vs the time spent in the block is still equal. But because the water flows more slowly through the rad makes it perform worse(surface tension and stuff) and ur also dumping more heat into the water because u got another usless pump. Therefor that system doesnt work.
I dont mean the cpu would overheat when i say "doesnt work". The entire point with that system is to make the rad perform better than in a normal 1 loop 1 pump 1 ress(or whatever) system, wich it doesnt. It does the opposite, so the system dont work.;)
Correct. Once again, thank you for spending the time to work this out with me, I'm definitely going to have to spend some more time on trying to work out optimum configurations with respect to the entire system instead of simply trying to improve the efficiency of a single pass. I hadn't realized previously that improving a single pass wouldn't necessarily lead to an improvement of the entire system.

[twwen2]

Just saying come back when you figure it out is roughly equivalent to saying STFU noob, GTFO, and is not the way to get my respect.

Fair enough. I'll just let you know we've had a few people come through here recently with guns blazing (and on other forums too), which can stir the pot a bit.



I hope you (twwen2) also learned something from all this because you obviously didn't understand the reason for the flattened tubes in radiators as opposed to the rounded ones.

Absolutely. <i just got my post above in before you posted ;) I probably learnt as much as you during the last two pages!:up:

[eXa]

Im sorry for being short with u in the start. But seeing alot of new people showing up giving wrong advises has gotten me a little irritated lately. Escpecially on matters wich have been discussed plenty before and have been laid dead several times. unfortainly u where the one who got the short stick... or whatever they say.

[twwen2]

But seeing alot of new people showing up giving wrong advises has gotten me a little irritated lately. Escpecially on matters wich have been discussed plenty before and have been laid dead several times. unfortainly u where the one who got the short stick... or whatever they say.

My point exactly. Nothing personal, but you can see why some people might be on a shorter fuse in recent times.

[cdawall]

what about this the max head on it is 44ft :shrug: and its made to run 24/7 :shrug:

http://www.drsfostersmith.com/product/prod_display.cfm?pcatid=4614

[Waterlogged]

I think Kapt hit on a very good point and one that's been overlooked so far. Why not dig your hole to just below the frost line (which varies from location to location, here in WNY, ~4' IIRC) pour a bit of concrete and while it's drying, set your copper tubing coil into it and then completely cover the rest of the coil with concrete. Totally maintenance free geothermal cooling.

I also think NaeKuh is right about the RD-30 handling this loop.

[NaeKuh]

what about this the max head on it is 44ft :shrug: and its made to run 24/7 :shrug:

http://www.drsfostersmith.com/product/prod_display.cfm?pcatid=4614


that works, but its loud!

[cdawall]

that works, but its loud!

so leave it outside i thought the whole point was find something that could do it

[orclev]

Why not dig your hole to just below the frost line (which varies from location to location, here in WNY, ~4' IIRC) pour a bit of concrete and while it's drying, set your copper tubing coil into it and then completely cover the rest of the coil with concrete. Totally maintenance free geothermal cooling.

That would protect the copper but what sort of impact would the concrete have on the heat dissipation? According to this (http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html) concrete isn't a particularly good thermal conductor, but then again it's not particularly bad (interestingly dirt seems to be on par with porcelain). I guess when you're talking about a heatsink that size the conductivity doesn't really matter that much, so I guess the question is does the reduced maintenance outweigh the hit to performance (I'm inclined to think yes)?

[NaeKuh]

I think Kapt hit on a very good point and one that's been overlooked so far. Why not dig your hole to just below the frost line (which varies from location to location, here in WNY, ~4' IIRC) pour a bit of concrete and while it's drying, set your copper tubing coil into it and then completely cover the rest of the coil with concrete. Totally maintenance free geothermal cooling.

I also think NaeKuh is right about the RD-30 handling this loop.

:rofl:

WL what happens if he has an earthquake by some fluke and the block cracks.
its gonna take the tubing with it.

And he's not gonna be happy that he's dumping a ton of ethyene glycol into his lawn.

[WhiteFireDragon]

this underground cooling is a great idea and i never thought of it. i've seen another creative WC idea somewhere where some guy ran tubes from his pool into the comp. pool is also a huge res + rad. not sure how safe that is because of the ions in the pool...

[Waterlogged]

Well, if it's a 1 acre block, then yeah, that'd be a problem. I'm thinking along the lines of 6'l x 2'h x 2'w.

[orclev]

this underground cooling is a great idea and i never thought of it. i've seen another creative WC idea somewhere where some guy ran tubes from his pool into the comp. pool is also a huge res + rad. not sure how safe that is because of the ions in the pool...
I'd be more concerned with gunk from the pool gumming up the waterblocks and pump then I would ions. Have you seen the sludge that builds up in pools if you're not on top of the maintenance (and even sometimes when you are)? You'd need to invest in a pretty good filter and you'd be constantly cleaning it.

[NaeKuh]

this underground cooling is a great idea and i never thought of it. i've seen another creative WC idea somewhere where some guy ran tubes from his pool into the comp. pool is also a huge res + rad. not sure how safe that is because of the ions in the pool...

all of this has been done.

The pool caused problems with corrosion cuz the guy used alu and copper. i remember a documentation about this a long time ago.

The drum underground was also done, and it was done with an air tank. It was pulled off very nicely, but it requires a lot of mods.

What he's trying to do is recreate something that people saw as being too difficult to do. It works really well if you get it setup, and your performance will be top, however the cost and labor associated with it....

You better be cooling more then just 1 computer. More like Road Runner.

Waterlog, when did you become mayor of our town?!?!?!

[majestik]

what about this the max head on it is 44ft :shrug: and its made to run 24/7 :shrug:

http://www.drsfostersmith.com/product/prod_display.cfm?pcatid=4614

445W pump, :ROTF:
If your electricity costs 10c/kWh, that pump will cost close to $400 a year in electricity.

[orclev]

...

Just thought I'd mention the idea for any WC'rs out their building a new house.

andyc

:rofl: Now there's a selling point. "This house comes pre-wired with Cat6 for gigabit networking, and in the basement is a underground heatsink for watercooling your computer(s)".

... Actually, come to think of it that would be pretty cool :D.

[orclev]

445W pump, :ROTF:
If your electricity costs 10c/kWh, that pump will cost close to $400 a year in electricity.

Yeah, and the real kick in the pants is the pump doesn't even come with a power cord. It does make me wonder why every other model of pump listed on there comes with the cord except the 445W version. Maybe they figure it's industrial use or something and it's cheaper to leave the cord out?

[cdawall]

here are the figures on the pumps

http://a1272.g.akamai.net/7/1272/1121/20031026073736/www.drsfostersmith.com/images/LittleGiantQuarium.GIF

http://www.f3images.com/IMD/250/LG2353/LG2353_99.jpg

samn thing still is pushing 1680GPH with a 20ft head height

[qdemn7]

If you're going to go this route, you might want to think about a ground loop, instead of a big reservoir. Companies that do geothermal cooling systems install them all the time. http://www.groundloop.com/geothermal.htm

[drbhatti]

dont know whether you have noticed, but someone is selling a MD70RZT on the bay - might be just what you are looking for!