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

[Naja002]

As a comparison, here's the state of 2 PCs when I woke up today at ~9am:

E8400 @ 4.0(dual core)
P5E-VM HDMI
Swifty 120.2 w/ 74cfm push/pull fans
Apogee GT and MCW30
MCP350
Single bay res

Ambient: A/C on 23.3C/74F, Thermometer reading 25.5C/78F
43, 46C






Q9450 @ 3.4(quad core)
GA-G31M-S2L
Car Rad setup

Ambient: PC weather says ~20C/68F
39, 37, 38, 38C





Again, I'm really curious to see how things go when I hook up the next PC. Winter should be fun!

Also, these chips/mbs will be swapped out. The Q9450 and GA-G31M-S2L are both new to me, so I figured that I would check them out together before putting the PCs in their long-term setups.

[MomijiTMO]

Holy cow!

Great job mate.

[chris.y2k.r1]

I've always thought about doing a house wide cooling system that you simply tap into as needed. Having the drum (reservoir) in the basement is a great idea as cold air sinks.

You should try and put a wall unit AC in front of the radiator to see if you can inject some active cooling into the system. That huge amount of water would be hard to cool down but once cooled it would hold the low temps for a long time.

None the less, very ambitious project and nice execution.

FYI - I just bought a pump that is probably perfectly suited for your needs. It's a 1600 GPH (1585 to be exact), 100% duty cycle, submersible, epoxy-sealed pump. It's $49.99 at Harbor Freight (Internet order only). Despite Harbor Freight's reputation of sporting cheap Chinese merchandise I can vouch that this thing is incredibly well constructed. I cannot believe how much water this thing actually moves. It has a 13" lift capability. I ran it in my tub as a test and let's just say that was a bad idea as it blasted some tiles off the wall. Doh!

Here's the link to HF (The pump is much bigger than it looks in the picture - check out the dimensions. It's frikkin' beefy )

[Naja002]

Hi Chris,

HarborFreight has come up quite a bit since the chinese have started to understand the concept of "Quality". I gladly buy from HF these days, but I still pay attention to cheap v quality!

What really matters most in these applications is the pumps head. That's the pressure that the pump puts out and gives an indication of how it will perform against resistance. The pump you bought has 13' of head. The one I have has 14' of head. In theory, mine will handle this type of application slightly better. Why? Because it can handle a little bit more resistance. Of course, pumps vary from manufacturer to manufacturer. And manufacturers make claims that are barely true, etc, etc, etc. But all things being equal--head and reliability is what matters.

An aquarium "powerhead" may move 300gph in an aquarium with basically no resistance, but it won't do squat against head/resistance. Yet a 300gph pump will.......

A point on my particular setup is: I need an inline pump--not a submersible only pump. In order for me to use a submersible only type pump--I would have to cut open the top of the barrel in order to get it inside. Not really an issue, but then I would have to make sure that the barrel top is the highest point in the system. Otherwise it will flood whenever the power is off. Inline is just much easier in my setup.

That HF looks like a good little pump. From the looks of it, its probably made by Rio or Via Aqua.

[p2501]

Naja002, I think you've got the title of your thread all wrong! This is anything but "another" car rad thread!

The names I would call your project go along the lines of "massive", "industrial", "awesome" and so on... I like it!

[iboomalot]

13 foot or 14 foot head or lift wouldn't that depend on the tubing size????

Laing has alot of lift but 3/8" to 1/2" dia of water vs these large pumps with 3/4-1" outlets.

so would a 1" pump with 13 ft of lift be more likely to be 20-30ft of lift using 1/2" tubing vs 1" ????

water has weight and the larger dia of water would weight more????


Just wondering how apples to apples specs are

[NaeKuh]

Naja,

Your the first to pull off a car radiator where you didnt have anything for me to rant about.

Very nicely done Bro

[Naja002]

iboomalot--I'll respond to your post probably tomorrow. A bit tired right now and just wanting to post the below......

2nd PC is hooked up to the system. There's plenty of flow for both. I eventually opened both gate valves 100% and there was no change in temps=Good Flow! I'll post some pix of my little setup, but I'd like to get some organization going first!

The PCs are currently setup the same cpu/mb-wise. I'll swap them around later. I wanted to keep them the same while I checked out adding the second pc to the system.

Here's a couple of screenshots--ambient temps outside are ~26C/79F, fans on high:

[Martinm210]

Awesome work, very creative and I think you've got plenty of radiator capacity there. The only thing I would watch is flow rate after running through all of that and then getting split into 3rds. With as much heat dissipation capabilities as you have, could could probably run several of those pumps without any effects of pump heat dump.

That's cool!

[Naja002]

13 foot or 14 foot head or lift wouldn't that depend on the tubing size????

Laing has alot of lift but 3/8" to 1/2" dia of water vs these large pumps with 3/4-1" outlets.

so would a 1" pump with 13 ft of lift be more likely to be 20-30ft of lift using 1/2" tubing vs 1" ????

water has weight and the larger dia of water would weight more????


Just wondering how apples to apples specs are

Ok, I'm going to go ahead and try to respond to your questions. I'm no fluid dynamics engineer or anything, so anybody feel free to let me know what I get wrong.

Basically, head or lift is an indication of a pump's ability to overcome resistance (or maintain pressure). A lift of 13' max is just that--the max. smaller tubing will not increase the amount of lift that a pump can provide, and may decrease the lift by adding resistance to the setup.

The 13' lift is basically @ zero resistance. If you look at a pump curve for lift and GPH: As GPH increase--lift decreases, and vice versa. Also, on a pump curve: a pump with a 13' lift will pump ~0GPH at 13' (max resistance). If it is rated @ 1,000GPH then it will pump ~1,000GPH @ zero feet (zero resistance).

so would a 1" pump with 13 ft of lift be more likely to be 20-30ft of lift using 1/2" tubing vs 1" ????

The answer is: No, it would not. A) the max lift you will get is ~13', and B) the smaller tubing would only offer greater resistance, so you may very well not get the full 13' of lift.

water has weight and the larger dia of water would weight more????

Think of it like this: What would happen if you put that pump into the middle of a 15' diameter pool? Would it move/lift all of the water or would it just push through the center of the water? It would just push through the center. The same kind of thing applies to pipes: A 2" diameter pipe is going to offer less resistance by allowing water to flow more easily through the center of the pipe than 1", 1/2", etc. The smaller pipes/tubes offer greater resistance at a given psi. The real question is: What is a pumps ability to maintain X-psi. That is kinda/sorta what head/lift tells us.


That's the layman's version...... I've had a lot of oddball projects over the yrs, learned a few things in the process, but I still may have some of the above wrong. Anybody care to confirm?....correct?

[Pedalmonkey]

I really like the setup. If the rad is in a different room you could find a squirel blower from and HVAC system. 1000+ cfm. could work better on the car rad with the high density fins.

[Naja002]

Awesome work, very creative and I think you've got plenty of radiator capacity there. The only thing I would watch is flow rate after running through all of that and then getting split into 3rds. With as much heat dissipation capabilities as you have, could could probably run several of those pumps without any effects of pump heat dump.

That's cool!

Hi Martin,

Care to offer any input about my last post--pump head/lift?


I don't think I would bother to buy another one of these pumps to run in series. I think I would check out other options. But one thing that has been kinda of toying with me is: pump failure. Power goes out--everything goes off. Pump failure=PC shutdown on overheat. So, it won't be long before I really start looking at that issue!

Anyway, I think you're right about the heat dump. Seems that I have a 20C delta (ambient/cpu). I've been thinking that maybe I'm actually not injecting enough heat into the system to get good efficiency. The opposite of diminishing returns. These rads are designed to dump some heat, so I'm willing to bet that this rad can handle one heck of a heat load. Any thoughts?

[Digz]

You need a Jeremy Clarkson Pump

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


.

[Martinm210]

Hi Martin,

Care to offer any input about my last post--pump head/lift?


I don't think I would bother to buy another one of these pumps to run in series. I think I would check out other options. But one thing that has been kinda of toying with me is: pump failure. Power goes out--everything goes off. Pump failure=PC shutdown on overheat. So, it won't be long before I really start looking at that issue!

Anyway, I think you're right about the heat dump. Seems that I have a 20C delta (ambient/cpu). I've been thinking that maybe I'm actually not injecting enough heat into the system to get good efficiency. The opposite of diminishing returns. These rads are designed to dump some heat, so I'm willing to bet that this rad can handle one heck of a heat load. Any thoughts?

That's right head or lift on a pump is the maximum lift hight regardless of tubing size and even regardless of the fluid density, that's why the term "Head" is used. It's one of those fun little things that's hard to understand and most people will assume the size of the tubing matters, but it does not.

Like you said the only thing the size of the tubing affects is the friction and restriction as the fluid flows, and bigger is better in that respect, you're 1" sized lines is a great idea there to help minimize restriction from the longer tubes and lines you have between the rad system and PC.

Anyhow the added restriction from the tubing and pipe may not amount to much, I think splitting it off into 3 loops is what may be the biggest culprit on flow rate. I'd just run a quick bucket flow rate test and see if you're getting alright flow rate. 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.

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!

[Naja002]

Hi Martin,

Kudos to You for the clarification!

Well, temps are the same now as they were in the last screenshots (within a degree): Q9450=46, 44, 45, 44C E8400=39, 40C Ambient temps dropped overnight though, so we'll see what happens this afternoon!

There's Good flow! I drained the system some to add the second PC--apparently not quite enough though. However, when I turned the system back on the air that was trapped started racing through the tubing. I intially got one PC going by slightly opening the gate valve (about 1 full turn, maybe 1 1/2) and then got the 2nd PC going the same way. After watching the temps, I opened one gate valve more, then more, then 100%--no change in temps. Then did the 2nd PC--again, no change in temps. I have the 3rd loop--looped back to itself. I opened the gate vale on it 100%--and there was still no change in temps. The 3rd loop being looped right back to the outflow offers the path of least resistance--so, if there was not sufficient flow for all 3--it should have shown itself by increasing temps. It didn't.

As far as doing an actual flow rate test--I have one minor issue: I only used 3 gate valves, so in order to disconnect anything--I have to shutdown the system, drain it some and then figure out how to close off the 1 barb that has no shut-off. Here's the pic from above:



Short-sightedness on my part. I was looking at the extra $30 in cost and the lack of need long-term. I could just adjust the flow via the 1 gate valve per loop, right? Right. But that doesn't help with connecting/disconnecting to/from the system. Not sure when, but I would say that I am going to add 3 more gate valves to the manifold.

Another change that I may make is increasing the pipe size from the blue barrel down to the pump. Its 1" now, but I'm thinking of increasing it to 2" to act as somewhat of a pump reservior. The pump is getting plenty of flow, but it just seems like maybe it would benefit from more available water. However, in a closed loop, I cannot really sort out in my head whether or not there would actually be a benefit. Here's another pic from above:



Add 2" (or even 3") pvc after the ball valve/elbow down to the pump. Seems like more water would be available for the pump, but that water has to be replaced and it would be drawn through the 1" from the barrel. So, I am wondering if I can do as I'm thinking? Or, If I should replace everything with larger pipe completely--from the barrel to the pump? Hoping that makes sense!


Suppose to get up to 34C/93F today, so the real test will begin.......

[DIE-TEMP]

You should drop a few bags of ice cubes in that blue barrel mate, just for kicks.....see how low your temps go

[systemviper]

that 55 gal drum and all those pumps and switches kinda look like the
set up for my old garden...... ahhh the memories!

[TRG]

I really like these uberwaterloops, nice plumbing and great idea!

I hope to get myself a loop like that when things are more settled at home for the long term.

[chris.y2k.r1]

Hi Chris,

HarborFreight has come up quite a bit since the chinese have started to understand the concept of "Quality". I gladly buy from HF these days, but I still pay attention to cheap v quality!

What really matters most in these applications is the pumps head. That's the pressure that the pump puts out and gives an indication of how it will perform against resistance. The pump you bought has 13' of head. The one I have has 14' of head. In theory, mine will handle this type of application slightly better. Why? Because it can handle a little bit more resistance. Of course, pumps vary from manufacturer to manufacturer. And manufacturers make claims that are barely true, etc, etc, etc. But all things being equal--head and reliability is what matters.

An aquarium "powerhead" may move 300gph in an aquarium with basically no resistance, but it won't do squat against head/resistance. Yet a 300gph pump will.......

A point on my particular setup is: I need an inline pump--not a submersible only pump. In order for me to use a submersible only type pump--I would have to cut open the top of the barrel in order to get it inside. Not really an issue, but then I would have to make sure that the barrel top is the highest point in the system. Otherwise it will flood whenever the power is off. Inline is just much easier in my setup.

That HF looks like a good little pump. From the looks of it, its probably made by Rio or Via Aqua.

Heya Naja,

Thanks for the informative and quick reply. I appreciate you taking that time. I saw the 'head' ratings but I have no idea what that actually translates to. What is a single 'head' actually measuring? Is that 1 foot of height in 1" tube or something like that?

Since you definitely know your pumps, let me ask you another quick question in relation to your 1" to 2" pipe upgrade. Wouldn't that extra volume of water coming down the tube (once pulled up out of the barrel) create positive pressure like a siphoning effect? What I'm wondering is, if you had 2" of pipe and then 1/2 way down to the pump reduced to 1", wouldn't that increase the pressure? You need more pressure right?

The other question I had is what role (if any) does siphoning play in a closed loop system with a pump and how does diameter effect that system (assuming that it's large enough not to cause any restrictions/resistance to begin with)?

Thanks!

[mr2sw21]

subscribed!

I can't believe what I am seeing, great job.

[chris.y2k.r1]

Hi Chris,

HarborFreight has come up quite a bit since the chinese have started to understand the concept of "Quality". I gladly buy from HF these days, but I still pay attention to cheap v quality!

What really matters most in these applications is the pumps head. That's the pressure that the pump puts out and gives an indication of how it will perform against resistance. The pump you bought has 13' of head. The one I have has 14' of head. In theory, mine will handle this type of application slightly better. Why? Because it can handle a little bit more resistance. Of course, pumps vary from manufacturer to manufacturer. And manufacturers make claims that are barely true, etc, etc, etc. But all things being equal--head and reliability is what matters.

An aquarium "powerhead" may move 300gph in an aquarium with basically no resistance, but it won't do squat against head/resistance. Yet a 300gph pump will.......

A point on my particular setup is: I need an inline pump--not a submersible only pump. In order for me to use a submersible only type pump--I would have to cut open the top of the barrel in order to get it inside. Not really an issue, but then I would have to make sure that the barrel top is the highest point in the system. Otherwise it will flood whenever the power is off. Inline is just much easier in my setup.

That HF looks like a good little pump. From the looks of it, its probably made by Rio or Via Aqua.

Heya Naja,

Thanks for the informative and quick reply. I appreciate you taking that time. I saw the 'head' ratings but I have no idea what that actually translates to. What is a single 'head' actually measuring? Is that 1 foot of height in 1" tube or something like that?

Since you definitely know your pumps, let me ask you another quick question in relation to your 1" to 2" pipe upgrade. Wouldn't that extra volume of water coming down the tube (once pulled up out of the barrel) create positive pressure like a siphoning effect? What I'm wondering is, if you had 2" of pipe and then 1/2 way down to the pump reduced to 1", wouldn't that increase the pressure? You need more pressure right?

The other question I had is what role (if any) does siphoning play in a closed loop system with a pump and how does diameter effect that system (assuming that it's large enough not to cause any restrictions/resistance to begin with)?

Ya know, that's probably a perfect setup for active cooling since you don't have space limitations you could put a huge coil in there and keep the temps low but reasonable so that you would get minimum condensation.


Thanks!

[l33t p1mp]

Dang man, very interesting results. About how much did all this cost you?

[Naja002]

Ok, I need a bit of Electrical help here........ I made a post in the " Wampeteers, Foma, & Granfaloons" forum:

http://www.xtremesystems.org/forums/...93#post3036793

Just unsure about connecting to the meter box---Any Electricians in the house...?






Hi Chris,

Heya Naja,

Thanks for the informative and quick reply. I appreciate you taking that time. I saw the 'head' ratings but I have no idea what that actually translates to. What is a single 'head' actually measuring? Is that 1 foot of height in 1" tube or something like that?

Check out Posts #35 and #39 in this thread, if you still have questions--lets us know....

Since you definitely know your pumps, let me ask you another quick question in relation to your 1" to 2" pipe upgrade. Wouldn't that extra volume of water coming down the tube (once pulled up out of the barrel) create positive pressure like a siphoning effect? What I'm wondering is, if you had 2" of pipe and then 1/2 way down to the pump reduced to 1", wouldn't that increase the pressure? You need more pressure right?

Yes, but there is already "positive pressure". What I am looking for is greater quantity and reduced resistance. Running the 2" pvc down to the pump and then immediately reducing it to the 3/4" pump inlet will allow the pump access to a larger quantity of water. But I'm not sure how it would play outon a closed-loop--IF--I leave the pipe coming out of the barrel @ 1". If I change everything to 2" from the barrel to the pump--then it should offer the pump more access (water) then it can use--so the pump will be able to draw as much water as it can, as fast as it can, and as easily as it can. Make sense? Right now, the pump has to try to draw X-amount of water through the 1" pipe. It does that via increased flow, but the increased flow creates increased resistance. 2" pipe would reduce/eliminate that resistance and increase available quantity making it easier for the pump to pump as much as it can. To "be all it can be"......

Same basic principle of having the res in the pc feed the pump......

Pumps of this nature really don't like having inflow restricted. They can handle the output being restricted much better than the input. Just from listening to the pump--I think the bigger pipe would help!

The other question I had is what role (if any) does siphoning play in a closed loop system with a pump and how does diameter effect that system (assuming that it's large enough not to cause any restrictions/resistance to begin with)?

In a closed loop--gravity takes care of the static pressure. Martin answered that question for me in Alex's thread. Basically, for every Foot (or inch, cm, mm, etc) of vertical head (resistance) there is a foot of gravity pulling down in some other part of the system. Otherwise--it cannot be a closed-loop. Think of a Circle--0--there is an equal up to down. Its just gravity at work. And in a closed-loop--the ends have to meet back up--just like in a circle. Otherwise there's no way to close the loop. Diameter of tubing/pipe doesn't matter--except for it effect on the pump's ability to do its job.

Ya know, that's probably a perfect setup for active cooling since you don't have space limitations you could put a huge coil in there and keep the temps low but reasonable so that you would get minimum condensation.

Right now, I'm experiencing an ambient/cpu delta of ~18-20C. Lower would be nice, but I'm going to have a few "issues" this winter. If the temp outside is 0C/32F, then the delta should offer cpu temps ~20/68F. If the room temp in here is ~24C/75F--then I think I have to watch for condensation. I may have to go to Passive cooling during the colder months by just turning off the fans! Not sure yet.

I tossed around the idea yesterday of adding a chiller, but I really don't want to bear that purchase expense or added electrical cost. Plus it would have to go after the radiator and I'm not really sure how it would play out.

Right now, then fans are on medium speed, the ambient is 32C/90F and the quad is running 48-52C, so I'm still running a 16-20C delta.

Going out to turn the fans on high to see if it even makes a difference.....

[[XC] Hicks121]

Naja!! This thing rocks man!! Completely awsome!!! Keep up the good work brotha & get that electrical done now!!! Break out that shovel man!!!

[Naja002]

Naja!! This thing rocks man!! Completely awsome!!! Keep up the good work brotha & get that electrical done now!!! Break out that shovel man!!!

Thanx to the Kindness and Generosity of Mr. Hicks--the electrical situation has been sorted out and will be completed by Tuesday! Yeah, Baby!

Yes, we discussed my love for anything that involves a shovel......

Many Thanx, Bro!

Dang man, very interesting results. About how much did all this cost you?

Hard to say really, but here is somewhat of a breakdown:

Car Rad--$88.50 shipped
Pump--$60 shipped
Plumbing--Some pieces I had, some I didn't, some I may replace (2" barrel-pump). The gate valves are ~$8 each after tax, so there's $24 right there. Unions are $4-5each--I think I used 4=~$20. All told--probably--~$80 in plumbing parts. Tubing, clamps, etc. that I didn't already have....
Blue Barrel Free. These are also available for free or very low cost (sometimes ~$15) from you local coca-cola/pepsi facility.
Electrical installation-- I've been wanting to run electrical to that out-building for a long time--just never really had a valid reason to....well, now I do! I already have about $40 in the 2 boxes and 2 breakers. Going to pick up another ~35amp breaker, wire and conduit, so I figure another $100 or less. So, about $140 or less total.


So, with the electrical: ~$368.50
Without the electrical: ~$228.50

Not sure whether to count the electrical or not--it will provide other benefits and I've been wanting to do it anyway for a long time.....


Here's some warm weather up dates:

Ambient temp outside is: ~33C/92F


Q9450: 53, 49, 51, 50C





E8400: 48, 49C




Still maintaining the 16-20C ambient/cpu delta......