Yes, Another Car Radiator Thread....Major 56K Warning!

[chris.y2k.r1]

Thanks again for such a clear and well stated explanation Naja, what you said settled well and quickly, you should teach. You'd make an excellent teacher with the ability you have to express yourself so clearly and politely. Nice Job (All around that is). Hey, I don't want to de-rail your thread (Nor do I want to come off as a first-chair skin flutist), but do you have children? I was just thinking that you'd do well as "Daddy" again with your ability to communicate concepts; something young children would benefit greatly from.

So, anyway, I was thinking about something you "said" that didn't quite settle with me. It was an assumption you made about opening the third valve and expecting the temperature to rise as a result of reduced flow to the 2 CPUs/machines. I'm by no means trying to correct you, however, I don't think that CPU temperature is a very good indicator for flow. And this is why:



The above graph from Swiftech showing the flow rate versus temperate indicates that above 2 or 3 gallons per hour, flow has little affect on the temperature of your cooling target. Obviously this is getting into the realm of diminishing returns.

Just something I thought I might point out. You may perhaps already be aware of this. I didn't want to see you get a larger pump in an effort to increase the flow with lower temps being the motivator for such a decision.

I was disappointed when I found this out actually. I thought I had a great idea with the HF pump that I showed you. Man that thing can move some water. It's actually pretty big too.

As a final thought, my current project is making a huge reservoir out of Plexiglas. Huge meaning 8" Wide x 18" Long x 12" High constructed of 1/2" thick material. I have a huge case and I plan on installing the reservoir inside the case along with the evaporator from an office style water cooler into said reservoir. The compressor and condenser are both small enough to fit into the case as well. Of course the condenser will be fastened to the outside of the case, fastened to the bottom, between the casters it rolls on. So all the info you've shared is immediately useful in my case. I'll be glad to post up some pictures of my build in another thread of course.

Keep up the good work!

(Side Rant - It's people like yourself who take the time to help out fellow extremers that keep this place going. Not! the people who brag about the fact that they got lucky enough to get a great CPU and resultantly can out-clock everyone else. Then resfuse to give away any 'secrets' incase some else might be able to clock there's up as well - I mean WTH is that? That's not what these forums are supposed to be for... right...? Right!?)

Deep Breath

[/End Rant]

Seriously though this place does have a problem with that and I find it really annoying. Especially since it seems promoted here. Doh!

[Naja002]

Thanks again for such a clear and well stated explanation Naja, what you said settled well and quickly, you should teach. You'd make an excellent teacher with the ability you have to express yourself so clearly and politely. Nice Job (All around that is). Hey, I don't want to de-rail your thread (Nor do I want to come off as a first-chair skin flutist), but do you have children? I was just thinking that you'd do well as "Daddy" again with your ability to communicate concepts; something young children would benefit greatly from.

Honestly, my ability to teach depends directly on the student. I start getting stressed pretty easily when someone isn't "getting it" and, At Best I can only figure 5-6 ways max to try to explain something--if they aren't getting it by then--well, I'm a bit stressed and they are on their own with it!

So, anyway, I was thinking about something you "said" that didn't quite settle with me. It was an assumption you made about opening the third valve and expecting the temperature to rise as a result of reduced flow to the 2 CPUs/machines. I'm by no means trying to correct you, however, I don't think that CPU temperature is a very good indicator for flow. And this is why:



The above graph from Swiftech showing the flow rate versus temperate indicates that above 2 or 3 gallons per hour, flow has little affect on the temperature of your cooling target. Obviously this is getting into the realm of diminishing returns.

Just something I thought I might point out. You may perhaps already be aware of this. I didn't want to see you get a larger pump in an effort to increase the flow with lower temps being the motivator for such a decision.

I was disappointed when I found this out actually. I thought I had a great idea with the HF pump that I showed you. Man that thing can move some water. It's actually pretty big too.

Honestly, it seems that maybe you were headed in one direction and the reversed on yourself. My checking temps in reference to flow was toward low flow. As Martin described above:

Flow rate itself will still cool plenty good down to 1GPM and even lower (only talking a couple of degrees between something like .8GPM and 2GPM), so as long as you have enough, the rest will be ambient driven.

If this system was pushing a low flow rate with both PCs hooked up--then opening the 3rd valve should have sufficiently decreased flow to valves #1 and #2 to cause an increase in temps. But since that didn't happen--I can assume that my flow rate is equal to or greater than 1gpm per PC. Keep in mind that the 3rd valve is just looped right back to the drain side. There are no blocks or anything. So, it is the shortest route to drain and offers the least resistance. Gas, water, electricity always take the path of least resistance. So, by opening that valve and not seeing an increase in temps--I'm allowed at this point to assume A) that both current setups are getting 1gpm or more, and B) that a third setup will still allow 1gpm or more per setup. There's no reason to think otherwise..... The only thing that I could see to cause an issue would be the resistance added by the 3rd set of blocks---but I really don't foresee that happening.

I am aware of the diminishing returns of increased flow rates, but Thank You for pointing it out! I'm not really interested in increasing my flow--unless it becomes necessary for some reason. Right now, the only reason I would buy another pump is because of the possibility of pump failure as I mentioned above. There really isn't much chance of 2 pumps going out at the same time--unless by lightening or something similar. If this pump goes out--my PCs will shutdown from overheating--not good. But since this pump can apparently handle the setup, another inline as a built in backup couldn't hurt.

As a final thought, my current project is making a huge reservoir out of Plexiglas. Huge meaning 8" Wide x 18" Long x 12" High constructed of 1/2" thick material. I have a huge case and I plan on installing the reservoir inside the case along with the evaporator from an office style water cooler into said reservoir. The compressor and condenser are both small enough to fit into the case as well. Of course the condenser will be fastened to the outside of the case, fastened to the bottom, between the casters it rolls on. So all the info you've shared is immediately useful in my case. I'll be glad to post up some pictures of my build in another thread of course.

Sounds like an ambitious project! You going to post it in the "Chilled Liquid Cooling" forum?

(Side Rant - It's people like yourself who take the time to help out fellow extremers that keep this place going. Not! the people who brag about the fact that they got lucky enough to get a great CPU and resultantly can out-clock everyone else. Then resfuse to give away any 'secrets' incase some else might be able to clock there's up as well - I mean WTH is that? That's not what these forums are supposed to be for... right...? Right!?)

Deep Breath

[/End Rant]

Seriously though this place does have a problem with that and I find it really annoying. Especially since it seems promoted here. Doh!

I understand your point and I really don't doubt there's an element of truth to it. I don't like trying to keep things secret. If I do Good then You do better--we can Both learn somethings! But ego is a fragile little egg and a major setback for humans. Sad, but true.

I've learned a lot here at XS though. Biggest reason that I don't post more is simply because there is just too much to keep up with and it changes every week! Sheesh! I don't mind being "corrected", but there's not much point in posting when I'm continuously stumbling along try to keep up with the right answers for this week!


Anyway, I hit Lowes earlier today. Got everything I needed to change the plumbing from the barrel to the pump, and run permanent electric to the out-building. I have a bunch of unrelated stuff to do before my work week starts. So, not sure how much or what I'll get done between now and Monday. We'll see....

[Naja002]

Ok, just a quick update......apologies for the poor quality pix. As I posted earlier, I went to Lowes yesterday and got the stuff to change the plumbing from the barrel to the pump. That's done, but I had an issue that may keep me shut down for a week or so..... I'll know in a few hrs, I guess....

Anyway, First:


WoWZa! Now that's a Union!





Seems that life would not be complete if we could just get a well thought out list together, head to the hardware store and get everything we need. But I guess that would take a lot of the fun out of everything, huh? So, I got my list together, headed to Lowes got everything for the electrical, the plumbing and a couple of other unrelated projects--and actually did pretty good! I got one pvc part wrong--got a female thread instead of the male. Its the piece that screws into the pump intake. So, after scrounging around in my extra parts bins--I found this:





Eureeka! Got one! The show is back on the road!



But then ya know life is never that easy.....never. So, I head out there today to go a head and get this plumbing done, and get the show back on the road. I like to build pvc from both ends and then connect in the middle somewhere--just seems like the easiest route to me. So, one of the first things I do is screw that brass fitting into the pump intake. 3 or 4 turns and: C-R-A-C-K-!!!!!. It cracked the Frickin' threaded housing right down the side. Jeezus! Shat! Definitely not Happy about it. Started scratchin' my head wonderin' what the Heck I was gunna do...... I don't have any more tubing clamps that size. So, I thought: Ok, silicone. See if there's any tubing that will fit snuggly around that part and tie wraps. So, ultimately, that is what I did:






No, I'm not expecting it to work long term. I can pick up some clamps on my way home this weekend. But I am hoping it will hold for a few days. The tubing is the leftover 1 1/4" from the rad inlet tubing. It fits nice and snug. If I had some actual screw clamps--I'd have a bit more faith, but I'm doubting at the moment!

I guess the part that makes this so ridiculous is that I actually have the piece that I need. No, not the exact part, but the fix the screw-up part: a double threaded 3/4" pvc nipple. Shat! I didn't see it when I was scrounging around, but it jumped right out at me after I broke the frickin' housing. Sheesh. Ho-hum. I guess things were going too well initially. I'm going to give it a few hours for the silicone to set and then see about firing it up.


So, on with the upgrade:

Here's a pic of the new 2" pipe--not inserted into the barrel:

The part that goes into the barrel is 1 1/4", the rest is 2".



I added a couple of air bleeder holes so that only X-amount of air can get trapped in the barrel. Might make priming the pump a bit of a pain at times, but I know how to deal with that, so life is Good!



Added a different pump priming port:


Now that's an Opening!:

Yep, still got the toes on that foot......


So, this is how the system looks now:



Going to see how much a new pump cover will cost...... EDIT: $12.98 shipped. I guess I can live with that.

2nd EDIT: I fired it up about an hr ago. Yes, there's a leak, of course. I have the pump in a dish tub. Not sure whether it will shut things down or not. A little leak can add up to a lot of water over time, so we'll see.....going to check it again in another hr or so and see what the situation is........

3rd EDIT: Ok, I checked it a few minutes ago. Really hard to tell what the deal is going to be. I'll be gone 14-15 hrs tomorrow, so I guess I'll have to check it before I leave, get the built up water out and make a judgement call. I don't mind the tub filling up--I just don't want it overflowing onto the wood floor. I think it will be ok, but we'll see.....The PCs are up and running--have been since I turned the system back on.....fingers-crossed.....

[qdemn7]

This is so fabulous, but......everything is so huge, and that pump seems so tiny in comparison. Seems like it ought to be physically much bigger.

[Naja002]

This is so fabulous, but......everything is so huge, and that pump seems so tiny in comparison. Seems like it ought to be physically much bigger.

You need a Jeremy Clarkson Pump

http://www.xtremesystems.org/forums/...d.php?t=185915


.

2nd video ~34seconds in.......





Also, I just checked the link below that I posted earlier:

I just bought this:

80-85 SKYLARK, CITATION w/ HDC NEW RADIATOR--3 rows, copper/brass--$88.50 shipped

Lots of Copper/Brass radiators at Great prices--some damaged, some not--check the description:

ltiparts

They have zero radiators atm. They had like 2-3 dozen copper rads....now there's zip. Wondering if they got out of rad sales.......?


Or did you guys buy 'em all up.....!?!?

[iboomalot]

I've seen people run 5 gallon buckets for reservoirs, but you've taken the lead with the 55 gallon barrell. It will probably take a couple of hours for the system to get up to full load temps with that much heat storage ability...that's cool!

**light came on**


I was going to a car rad setup but I might use something I have at my disposal.


30 ft from my study is my pool

I could run a skull + SLI and never heat that reservior up

its a salt pool so no clorine or algea to worry about.

Opinions????

[ripken204]

lol, it could work. as long as the salt does not corrode the blocks somehow.
you can get a pump just for that type of water so no problem there.
just make sure to put a filter on the inlet, you dont want any bugs getting pulled in and getting stuck in your blocks, lol.

[Naja002]

Just a quick update:

I hit the auto parts store on the way home and picked up 4 screw clamps. I just knew I was going to find a flood of some type when I got here, but---I didn't! So, life is Good! It leaked about a gallon of water into the dish tub, and, I guess, the negative pressure finally built up enough to stop the leak (luckily this crack is on the intake side! ) I cut off the tie-wraps and put 3 screw clamps on it. Tightened them down real well. I did not turn the system off--I left it running, so I expect the negative pressure to hold. Between the negative pressure and the clamps I think I'm good to go until the new pump cover gets here. Hopefully it will be here by Mon or Tues.


35C/95F when I got home today. Hottest core on the quad was 55C the rest were a couple degrees below. Same variance as the other temps above. The Duo runs a few degrees cooler. So far my max temp is 20C delta. I can live with that....

Headed to bed before long, so I can get up and do it again tomorrow....

[Naja002]

Went to Passive cooling about 36hrs ago. Saw ~4C increase in temps during the hottest part of the day--highest is on the Quad @ 59C. I would offer a screenshot, but I fell asleep in my chair. The Q9450/GA-G31M-S2L was locked up when I woke up about an hr ago. That's the 4th time in the last week. Not sure what's going on with it. I lowered the FSB 5mhz to 420. Since then it locked up this last time. I'll get a screenshot to post.

Took off the filter on the rad and turned off the fans and opened the shroud door. That seems to have given me a ambient/cpu delta of 24-25C when its hot outside. Its pretty stable at that delta, but right now, its ~19.4C/67F and the ambient/cpu delta is at ~32C (ie, 51, 47, 50, 48).

One thing that I've noticed is that things don't cool down as fast when ambient temps start dropping. The fans don't seem to do a whole lot on cpu temps, but seem to cool the system down overall a lot faster than just passive. When the fans are going I can "watch" the ambient temps drop by watching the cpu temps drop. With just passive--the ambient temps will drop noticeably outside, but its not really reflected in the cpu temp for a while. Either way, I can do passive 12 months of the yr, and I can definitely do it ~7 months of the yr!

The leak is still contained. Hopefully the pump cover will be here today or tomorrow. I picked up everything to run the permanent electric to the out-building and now to run 2x 20amp circuits into this room. The meter box is right there--so I don't know what I was thinking. The PCs are going on their own 20amp circuit. The AC is already on its own. Total, I'll have 4x 20amp circuits in this room and 2x 20amp in the out-building. Given a little bit of time, this whole area will start to hum and "glow".

[Sam__]

wow, great job on the whole thing here dude. amazing.

[Naja002]

wow, great job on the whole thing here dude. amazing.

Thanx--its been a lot of fun so far---for the most part!

Just a quick update:

The 2 electrical circuits to the out-building are installed and operational. Turned out to be a heck of a lot more time consuming then I realized. The Electrical update is in this thread:

Needing some Residential Electrical Help.....Major 56K Warning!

Lotsa pix.....

I had the PCs shut down for a few days while I was doing the electrical install and waiting on the new pump cover. Its here (Finally!), but I haven't installed it yet. Running the system as is has turned out fine, but I will be installing the new pump cover soon. Ambient temps have dropped outside the last few days (mid-80's), so my PC temps have been nice (low 40's).

I am going to install 2x 20amp circuits into this room directly from the meter box. The PCs will go on 1 of those circuits. Probably going to run a 1/2" water supply to the out-building real quick, so I don't have to deal with a bucket or hose--the hose bib is Waaaaaay-over-there, makes it a pita..... Its a no-brainer and too easy to do.

Otherwise, everything is still pretty much the way it was, I believe.....I am currently back to using the fans. The passive worked very well as far as I'm concerned. I expect to be hooking up timers to the fans. I haven't sorted out how I am going to use them yet--but that may require a little tinkering to see what's best. Probably going to work out a bit different during the hotter months v. the cooler months......

[Naja002]

Just some pix of the start of my mini-crunching farm--Only 2 PCs at the moment.

Still need the power buttons mounted, the HDs stacked and a fan solution--I found these 4" fans available at walmart ($6). I'm going to pick up 2--one on the top shelf blowing across the MBs and 1 on the bottom shelf blowing on the HDs and PSUs:











I've definitely got room for 1 more PC. Regrouping on the cash right now though. Most likely going to be either a Q9450 or Q6600 (probably the Q6600) on another GA-G31M-S2L. Seems like a Good little Crunching MB so far....

Finished installing the 2x 20amp circuits into this room. So, now the PCs are on their own circuit. That will definitely help me sleep better at night!

[Marco André]

OMG! Huge build

[[XC] Hicks121]

HEHE, no more extension cord brother?? Got it lookn sweet.

Whats this crap of you falling asleep in the chair a few posts up?? You must be up there with Movieman

[Naja002]

Thanx for the comment, Marco...

HEHE, no more extension cord brother?? Got it lookn sweet.

Whats this crap of you falling asleep in the chair a few posts up?? You must be up there with Movieman

Ya workin' me too hard with all this electrical stuff! Us old-folks can't handle the heat, ya know!

[Naja002]

Also, I just checked the link below that I posted earlier:

I just bought this:

80-85 SKYLARK, CITATION w/ HDC NEW RADIATOR--3 rows, copper/brass--$88.50 shipped

Lots of Copper/Brass radiators at Great prices--some damaged, some not--check the description:

ltiparts

They have zero radiators atm. They had like 2-3 dozen copper rads....now there's zip. Wondering if they got out of rad sales.......?


Or did you guys buy 'em all up.....!?!?

Looks like they are back with 3 dozen rads. Here's a Good one:

74-78 TORONADO NEW 3-CORE RADIATOR--$130.50 shipped

34" x 17" x 2" Triple core...... Inlet needs rounded.....

Another:

70's GALAXIE, LTD, MONTEREY NEW 3-CORE RADIATOR---$154 shipped



Sweeeeeeet............

[Naja002]

Just a quick video to show the actual water flow via trapped air through the tubing on this setup. This setup still has 2 PCs hooked up and has been running fine. Temps are still maintaining as previously posted. Q9450 @ 3.45 and an E8400 @ 4.1.

The setup does NOT run like it is shown in the video. The air is just to give a visual of the water flow. For some reason when I open gate valve #3--air starts traveling through the system. Gate valve #3 is just looped right back to the #3 output on the manifold--so I don't really understand why it does this, but it does.... Anyway, the flow in the video is actually reduced--because gate valve #3 is open--allowing a path of least resistance and all that. When I close gate valve #3--the flow to both PCs increases noticeably, so look at the video as less or reduced flow from what is normally going through the tubing:

Its a 1 min video that shows flow through both PCs. Audio has NO valve:


Click here to see Video


Again, the system does not run with air--that's just to show the flow for the video....

[[XC] Synthetickiller]

Slosh slosh slosh! wow, that's a crazy amount of air.

Just curious, do you think this could passively cool a computer w/ a 5gallon container, same rad, but no air moving over it (rad positions under the desk)?
This just seems like the perfect way to cool something quietly. I guess you can't tell the computers are on noise wise, save for the fans?

Amazing work. Keep it up. Can't wait for winter!

[Naja002]

Slosh slosh slosh! wow, that's a crazy amount of air.

Just curious, do you think this could passively cool a computer w/ a 5gallon container, same rad, but no air moving over it (rad positions under the desk)?
This just seems like the perfect way to cool something quietly. I guess you can't tell the computers are on noise wise, save for the fans?

Amazing work. Keep it up. Can't wait for winter!

Hi SyntheticKiller,

The short answer: As far as I know, the answer is: Yes. AFAIK, its already been done more than once and works well.

The Long Answer: One of the things that I really like about this setup is that the rad is in another building, so all the heat is pumped out of this building/room. I currently run 3 rigs (2 OCed crunchers) and they produce plenty of heat. I pay to run these rigs, of course, but then I was paying to cool this room too. The little A/C in this room draws ~500w. Both fans combined in the out-building draw ~72w on medium. The pump (I haven't actually checked it via killawattt) is rated at 93w. So, I think its safe to say that over the course of the summer--I'll save a few $$ overall.

Back in Post #59, I went Passive for a short time with both OCed rigs going. My temps increased ~4C overall. The biggest thing I noticed was that the system didn't cool down very quickly as ambient temps dropped outside. With the fans going, I can pretty much tell if the outside ambient is going up or down just by watching the cpu temps. Passive--its not so easy. The setup retains the heat longer, but does cool down over time. Keep in mind though that the highest temps never went over ~60-62C (which was on the quad) even during the hottest part of the day ~95F/35C.

I'm not sure, but from watching the behavior of this setup and using a little automotive common-sense: I think the reason that the temps are as high for me as they are is because these auto rads are designed for a different application. In other words, they are designed to dump a lot of heat at a higher temp range. So, my guess is that the more heat that's pumped into the rad--the better it will perform. I said it somewhere before: I think that this setup could handle 10 OCed rigs and not miss a beat. I don't think temps would actually rise by much--if at all. I think the rad would simply become more efficient. 123Bob cools his entire crunching farm with a car rad....

I guess my point is: With 1 PC, I think it would do what you want it to do without a problem. Under the desk though there will be little air movement and you'll be pumping the heat into the room. If its "at your feet", so-to-speak, then you may notice the temp rise in your general area. But if you're just surfing, email and that kind of stuff--then you should see very sweet temps. But if that's the case--then I think you can accomplish the same thing with application-specific (PC) type hardware and eliminate all the adapter hassles.

Another thing worth mentioning is: You mention "a 5gallon container". If you mean a 5g bucket--then good luck! Those things can be extremely difficult to get sealed in this type of application. If its some other type of container--like a car-boy or something, then I can't really comment on sealing it--just think it through before hand. Just keep in mind that you can put positive pressure into the container by pumping into it, and negative pressure by pumping out of it. Negative pressure will work toward your sealing benefit! If you are looking to just passively cool 1 PC with a car rad--I think its safe to say that you can skip the 5g container all together and make your life much easier.

HTH

[[XC] Synthetickiller]

Hi SyntheticKiller,

The short answer: As far as I know, the answer is: Yes. AFAIK, its already been done more than once and works well.

The Long Answer: One of the things that I really like about this setup is that the rad is in another building, so all the heat is pumped out of this building/room. I currently run 3 rigs (2 OCed crunchers) and they produce plenty of heat. I pay to run these rigs, of course, but then I was paying to cool this room too. The little A/C in this room draws ~500w. Both fans combined in the out-building draw ~72w on medium. The pump (I haven't actually checked it via killawattt) is rated at 93w. So, I think its safe to say that over the course of the summer--I'll save a few $$ overall.

Back in Post #59, I went Passive for a short time with both OCed rigs going. My temps increased ~4C overall. The biggest thing I noticed was that the system didn't cool down very quickly as ambient temps dropped outside. With the fans going, I can pretty much tell if the outside ambient is going up or down just by watching the cpu temps. Passive--its not so easy. The setup retains the heat longer, but does cool down over time. Keep in mind though that the highest temps never went over ~60-62C (which was on the quad) even during the hottest part of the day ~95F/35C.

I'm not sure, but from watching the behavior of this setup and using a little automotive common-sense: I think the reason that the temps are as high for me as they are is because these auto rads are designed for a different application. In other words, they are designed to dump a lot of heat at a higher temp range. So, my guess is that the more heat that's pumped into the rad--the better it will perform. I said it somewhere before: I think that this setup could handle 10 OCed rigs and not miss a beat. I don't think temps would actually rise by much--if at all. I think the rad would simply become more efficient. 123Bob cools his entire crunching farm with a car rad....

I guess my point is: With 1 PC, I think it would do what you want it to do without a problem. Under the desk though there will be little air movement and you'll be pumping the heat into the room. If its "at your feet", so-to-speak, then you may notice the temp rise in your general area. But if you're just surfing, email and that kind of stuff--then you should see very sweet temps. But if that's the case--then I think you can accomplish the same thing with application-specific (PC) type hardware and eliminate all the adapter hassles.

Another thing worth mentioning is: You mention "a 5gallon container". If you mean a 5g bucket--then good luck! Those things can be extremely difficult to get sealed in this type of application. If its some other type of container--like a car-boy or something, then I can't really comment on sealing it--just think it through before hand. Just keep in mind that you can put positive pressure into the container by pumping into it, and negative pressure by pumping out of it. Negative pressure will work toward your sealing benefit! If you are looking to just passively cool 1 PC with a car rad--I think its safe to say that you can skip the 5g container all together and make your life much easier.

HTH

Sounds good, or rather silent. I am thinking that probably buying the smallest car radiator is the ticket. CPU, NB, & GPU would not even be an issue w/ so much surface area.

Thanks.

[orclev]

Since a number of good questions were asked about fluid dynamics I thought I'd toss out a couple bits of info (if I get any of this wrong let me know, I've made mistakes with regard to open versus closed loops before ).

First, flow. Flow is a bit hard for some people to wrap their heads around because it's a measurement of volume over time. Most people when they think of flow picture velocity, which is a measure of distance over time. These two are not interchangeable which is where a lot of the confusion comes from. Think of it like this, it's the difference between a foot of copper wire, and a cubic foot of copper wire, one tells us the length of one dimension, and the other tells us how much there actually is, but not any of it's measurements in any dimensions.

Now, Flow has a relationship with Velocity, which is determined by the Bernoulli Principle which says that when the speed of a fluid increases, the pressure of the fluid decreases. In a closed loop such as a PCs cooling loop, pressure is constant (more or less), and flow is constant, which means that the speed is the variable and is determined by the volume the fluid is traveling through (in practice this is usually a tube, so the diameter of the tube). The smaller the volume (cross section) the faster the fluid will travel.

Now, the point of this. Earlier it was asked if decreasing the size of the tubing would increase the pressure, and the answer is no (actually due to the Bernoulli Principle it would actually decrease slightly). The velocity of the fluid traveling through that point of the loop would increase, but as soon as it got to a wider part of the loop it would decrease again. Remember flow in the loop is constant (it's actually the inequality in force between the pump and the resistance of the loop), the only way to increase it is to get a better pump, or decrease the resistance of the loop.

Second, head pressure. The explanations were pretty good, but I think maybe another way of explaining it might help with understanding why head pressure is used when evaluating pumps. It's important to remember that the flow is generally speaking the result of subtracting one force (resistance) from another one (the pump pressure). One force, the pumps pressure, is generally fixed (variable pumps can change it, but it still has a minimum and maximum defined for the pump), the opposing force(s) are the variable, and the difference between the two determines the flow.

For the most part, in a water cooling loop, you're dealing with two forces, one of which naturally cancels itself (because it's a closed loop). The two forces are, resistance, and gravity. Because the loop is closed, the resistance due to gravity is naturally canceled due to the downward force also applied by gravity (what goes up must come down etc.). What this means is that for practical purposes we only need to worry about resistance of the loop (gravity does play a role in starting the flow, but once you have any flow at all gravity cancels itself).

Now, I kind of got off on a tangent there, but some of that is important for understanding this next point. So, if the resistance of the loop is variable, but the force of the pump is fixed, and flow is the difference between the two, how do we measure and compare pumps? We could of course pick some arbitrary amount of resistance and use that to compare all pumps, but that's not particularly useful. What we do have is a force that scales linearly (more or less) and therefore can be used as the opposing force in calculating pump force. In order to compare pumps, we just need to find the point at which our opposing force exactly equals the force of the pump, and then the value of that opposing force for different pumps can be compared. The force that's used on pumps is gravity, specifically the gravity of the earth (which is 9.8m/s^2 or 32.2ft/s^2). The actual force at a particular distance from the surface of the Earth can be calculated but for our purposes doesn't actually matter. What does matter is that for a particular height that force is constant. This gives us a convenient way to compare our pumps, based on the force at a particular height, which is exactly what head pressure is. Head pressure is the height at which the force of gravity equals the force exerted by the pump.

For the curious, this site http://www.engineeringtoolbox.com/pu...ure-d_663.html has a formula you can use to convert head pressure into various other measurements such as PSI. Of particular note see the formula for converting head in meter to pressure in bar. Notice that 0.0981, look familiar?

Now that I'm done hijacking the thread, I'm going to go back to lurking.

[Asgard_thor]

how does your wife let you Spend this money?!!?! lol haha

very nice build

I think you should be able to add 4-5 more crunchers if you put a AC in the window, it would take some time to cool just like the other person said, but I would say its worth it

[orclev]

how does your wife let you Spend this money?!!?! lol haha

very nice build

I think you should be able to add 4-5 more crunchers if you put a AC in the window, it would take some time to cool just like the other person said, but I would say its worth it

Mmmmm phase change goodness. The trade off with that taking some time to cool is that it would also take some time to heat up, so it all balances. This also assumes you'd be running it constantly. If you're planning to be shutting it on and off all the time you'd see a lot less benefit. Out of curiosity do you have any idea what the actual water temperature is?

[orclev]

Sounds good, or rather silent. I am thinking that probably buying the smallest car radiator is the ticket. CPU, NB, & GPU would not even be an issue w/ so much surface area.

Thanks.

Hmm, maybe a motorcycle radiator? They're pretty small as they usually mount them on the front of the bike underneath the front forks. Hard part would be to find one made out of copper as the increased weight versus aluminum would tend to make that an unpopular choice for motorcycles.

[Naja002]

how does your wife let you Spend this money?!!?! lol haha

very nice build

I think you should be able to add 4-5 more crunchers if you put a AC in the window, it would take some time to cool just like the other person said, but I would say its worth it

I've considered putting an A/C in that window. I put up this post about a month ago seeking some guidance, but not much response to it:

Contemplating adding AC/Chiller to my setup......

Honestly, I think that even putting the little 5K btu ac in there that I have--it would just cycle on/off too much, plus the resevior would still be collecting heat from the out-building ambient. If I made an enclosure around the rad and barrel then there may be enough heat and air-space for it to work much better. I may do that, but it won't be any time soon. Busy with other stuff and I think I would rather put a 3rd cruncher on this setup first. I've just been spending too much overall and need to regroup for a while! Plus winter is right around the corner! In ~2 months night time temps should start to drop noticeably...... I'm lookin' forward to see how this works more or less as it is....