Did I overkill it?

[Talik]

So, I just ordered the Rio 3100, 900 gph pump because it seems like it should do the job and have enough power left over to water my garden. The problem, however, came when I realized that this pump might create heat of itself. Does anyone know if this pump is going to add undue heat to my water?

I don't mind spending the extra money, unless this pump is actually going to do a worse job than a cheaper counterpart.

On a side note, this is my first watercooling system and I was wondering if I should mount my resevoir above the radiator so that air doesn't get trapped in the rad?
And speaking of radiators, I was planning on paying a visit to my local junkyard and pulling something out of a car or motorcycle. Is there anything special I should look for when choosing one? I have plenty of room in my way oversized case for a fairly big radiator.

[bigben2k]

Heatercores out of SUVs are always good!

The Rio is designed to be run submerged, but some run it in-line. I've seen a few people report problems with some of the larger units, in-line.

Note that that pump has a max head of 10 feet (~3m). For $50, I think you can do better.

[Talik]

What do you suggest to do a better job? I want a lot of pressure as I intend to start off cooling my P4 2.0 GHz and eventually upgrade to cool my geforce4 ti 4200, 480 gb 5 drive raid 5, 120 gb 3 drive raid 0, and a 120 gb single drive. (storage, photoshop swap, and boot in case your wondering why the hell I have so many drives in my system)

so, as you can see, I'll need something a little bit beefy.

[Axle]

Ney, I don't think you do. Even with 3(that would be raid 5 & 3 drives..*sigh* I'm so dyslexic) HD coolers (which, I may add, is a might mad). Personally I'd stick with a good old Maxi Jet 1200. $20, small, quiet, 290GPH. It still has amazing flow in my system, which leaves me, well, amazed. I run one with: 18ft custom 1/2" rad + 24ft 1/2" sprial rad + 10ft or so 3/8"Tygon + TC4 + 3/8"od GF4 (something like 1/4?) and it runs like a champ. And for $20 you can't go wrong.

When I was first going to set up WCing I, too, was going for something around 500GPH, and wow, you just don't need it. 900GPH, espically in a classically HOT Via, would be heating your water as well as way overkill. JMHO..

[Talik]

I got a Rio, not a Via.. unless they are the same thing? same company, maybe?
To get the drive to run in-line, is there any special modding I have to do? I would assume that inline, the heat would dissipate into the air and not so much into the water possibly eliminating my problem. You know any more specifics about those reported problems with running these inline, bigben2k?

Damn... after all the research I though I did, it still wasn't enough.

[Blackeagle]

First off don't go to a junk yard for a rad. You have no idea what sorts of deposits inside it or leaks you are getting. And you can buy a new one for about $20-$30 at Napa auto parts.

A single pass heater core with the inlets/outlets at opposite corners is what I'd suggest. Check on early 70's Chevy pick-ups heater cores to get one of these that will take 2 fans per side if you wish.

[bigben2k]

Originally posted by Talik
I got a Rio, not a Via.. unless they are the same thing? same company, maybe?
To get the drive to run in-line, is there any special modding I have to do? I would assume that inline, the heat would dissipate into the air and not so much into the water possibly eliminating my problem. You know any more specifics about those reported problems with running these inline, bigben2k?

Damn... after all the research I though I did, it still wasn't enough.

Oh I spend a lot of time going over many, many specs, and I'm still not close to a perfect solution. I found this great pump once, but it was $400.00 .

I can tell you that the pump must be a mag-drive: anything else is too noisy, or nowhere near powerful enough.

Running pumps in series is a good way to increase the pressure, which is what all rigs need. Keeping cost in mind, if you get a good deal on a small pump, start lining them up! Keep in mind though, that you may have to reinforce some parts of it, because the increased pressure will push the pump beyond what it was designed to do by itself: you'll have leaks where you didn't even think leaks were possible!

The reports I've heard mostly come from the OC forums. Users report noises appearing after a few months, high heat conditions, and the odd failure. You can run a search, it shouldn't be too hard to find.

The Rio 180 was very popular, some time back, and reportedly runs great. Unfortunately, it's grossly underpowered, by today's standard. The nice thing about the Rios is that they're cheap, and you can save even more by buying them in bulk. ProCooling even had a discount with GreystoneCreations, some time back (it might still apply!). Excellent prices.

You can also get lucky, if you know where to shop: I picked up a Little Giant 2-MDQ-SC, a large pump that borders on the (good) side of "insane", for $40 at a local flea market (brand spankin' new, in the box!)

My research wasn't fruitless; I would have never known that that pump was any good, if I hadn't researched it, and I have to thank all the members right here at ProCooling for it.

[airspirit]

If you don't need that much power, you are going overkill. I hope you're running a few parallel channels, and have at least a 5/8" ID width on the heatercore. Look at the heatercore dbase sticky at the top of this forum to find what heatercore you're looking for. Do NOT buy one used unless you want to risk leaks and assorted nastiness in your system.

That pump will need to be submerged or its life span will be cut short significantly. It uses too much power and doesn't have adequate external cooling. I've done a lot of research on this class of pump, so trust me on that.

[Talik]

Does anyone done any tests to see how much heat any pumps put out? This seems like an important number to know when choosing one.
so, what do you guys think of the Mag-Drive #7(700 gph)?
I could probably get by with the #3, (350 gph) but I don't mind spending the extra $15 bucks difference if the heat generation is not gonna be a problem.
or would the Eheim 1250 generate less heat?
or maybe two inline mag-drive #3's?.. AAH, Too many choices! I can't decide!

ps. I'm not really worrying at all about noise generation, I've already got 2-3 dozen fans in my case allready, so the only factors I'm worrying about is flow and temp increase.

pps. I've decided to go with the Balkamp Heater core as it should fit nicely into my case. Is this a good heater core? http://www.napaonline.com/cgi-bin/nc...grpid=45685334

[Graystar]

If flow is all you're really concerned about then there are other pumps that are better suited. Iwaki, Custom Sealife (Laing), March, and others are a better fit for high flow.

However, the graph of flow rate vs. performance is not linear. more flow does not equal better performance. Every combination of pump, block, and radiator has its unique performance curve, and the shape is parabolic. The top of the curve is the "best" flow rate where your components will work most efficiently. More or less flow will be detrimental. Lots of experimentation is usually required to find out what that rate is.

This article list the flow rates of various common pumps.
http://www.overclockers.com/articles723/
As you see, the actual rates are far lower than the rated flow. This is considered very bad for the pump. The pump works most efficiently at 80%-85% of max flow. But our pumps are working between 9% and 27% of their rated flow. This stresses the pump and creates heat. What you want is high pressure and low flow. The Iwaki MD-6Z is pretty good. Also the March 893-07 is a good match. Both these pumps are around 150 bucks but they are very well made and suitable for 24x7 operation.

[Talik]

Ack, money is not a huge object, but it is still an object. Lets assume that I don't want to spend any more than $60 on the pump(I'd prefer less). I can justify building my own cooling system only as long as it costs me less than a vapochill.

If I were to go with my current pump(the rio; which would be easier as I wouldn't have to sell it or buy a new one), and run it before the radiator, would the radiator sufficiently cool the extra heat from the pump? Would the heat off of the processor wear out the pump?

[bigben2k]

All right, I've got to step in now. I knew this was headed this way!

There are more than one heat source, as far as a pump is concerned, but I'll talk about the only two relevant ones:

1-Heat from the pumping action. Yes, that's right, all that sloshing of the water actually creates heat. It can be calculated, based on flow and pressure, and can be measured in Watts. Many people have run the test, where they circulate water in a closed loop, and it does heat up, very exactly with the power applied to moving the water.

2-Heat from the pump motor coil. Those electrical coils do actually heat up, and that's the heat that we perceive as localised to the pump. If you know #1, you can then compare your pump's actual electrical power draw, to figure out just how efficient it is. But I digress... This heat is usually very well isolated from the water, because plastic is a poor transmitter for heat. Let me make this even clearer: even if the plastic reaches a temp of say, 65 deg C, there is only a little bit of energy that can travel through plastic. This is not a significant source of heat (unless the pump housing is metallic).

Bad or cheap pumps will run hot. As a result, some end up having to be cooled, by running submerged. Usually, this is not a problem, if you are running a pond, or an aquarium.

A good pump will have a very precisely wound coil, minimizing energy losses.

I ended up with a Little Giant 2-MDQ-SC which, by any standard, is a power hog: 182 Watts, with max flow of 550gph, and a max head of 14+feet. Luckily, there is a magnetic drive that turns the impeller, so the motor is completely isolated from the water, which circulates through a "glass-filled polypropylene" housing.

[redleader]

Does anyone done any tests to see how much heat any pumps put out? This seems like an important number to know when choosing one.

Yeah I tested my Eheim 1250 (320GPH). It was 9w IIRC. The rating is just startup power, not actual use. Your Rio is probably in that range.

IOW pump heat is virtually nothing unless you have a truely huge pump.

[Talik]

Thanks, bigben2k, for that breakdown of pump heat creation. The rio finally arrived yesterday, so I threw it in a bucket of water to see how fast it would heat up. The water was at 10 above ambient after a couple of hours when I turned it off. Now, in a computer application I would be running it inline, and before a radiator, which will have to be my next test. but I havn't purchased the radiator or the tubes, yet, so it'll have to wait.

The pump claims to draw 73 watts. (I'm not quite sure what to do with that number as I'm an artist, engineering is only a hobby.)

Hopefully, run inline, this pump won't create nearly as much heat, or is most of the heat transfered through the impeller?

I looked into your little giant, and I couldn't find any lower than $81, any cheaper places I should know about? How big is that pump? the pictures make it look fairly large, but they can be decieving. I trust that you did the proper research and found the best pump for the job, so I might go with the little giant if I can sell the rio. (I have a friend who's building a pond)

[bigben2k]

Well, in all honesty, I can't recomend the 2-mdq-sc, because it's very large: ~10 inches long, ~4 1/2" diameter. The main issue is the vents for the motor, which spew out hot air, so I'm having to make a special housing to route this hot air out of the case.

As for your Rio, I think you'll find that the heat comes from the pumping action, so you should get a similar temp. To test it, sit the pump on a chair, and run tubes in and out of a bucket: you'll see what I mean.

The difference in heat that you'll notice, will come from the pump's innefficiency at converting electrical power into a movement of water. That's mainly composed of the motor coils heating up, but like I said, there's not a lot of power transmitted through plastic, so you might not be able to measure a difference.


To clarify a bit: this heat that comes from the pumping action, will appear at the most restrictive points within a cooling loop. In most of our cases, that's the waterblock, followed by the rad. Not the pump.