Which pump to buy?

[Lord Smack]

I did my research and decided to go with the swiftech MCW 462 mostly because I hadn't been able to find any complaints on any review sites.( although I will most likely get it for several weeks.) also I found a site that sold the Haden 1676 for $ 29.00 so all you OC sites out there selling it for $40-50 kiss my a** ! I got A shiney new tower and now I'm stumped! I was about to buy a mag drive 250 for $33.00, when someone tole me its loud as hell! They told mo to buy the Eheim 1046. that ist less noisy and that the magnetic field extends only a few inches compared to the Mag Drives 2 feet of field. Now I have also been told that the field will not affect anything either way. I have heard that the extra flow of the Mag cools better but I have also heard that the Ehiem's slower flow cools much better. price doesn't matter I had a budget of $50 foe my pump. Someone please help me choose! P.S. I do plan on adding a vid cooler in the near future if that makes a difference.

[hielko]

I don't believe a lower flowrate can be better. Compare it to aircooling, do you think less CFM can be better? I don't think so...

I'm not sure, but I think I have a possibly explanation why people are thinging this. Most people are using submergible pumps in a resevoir. A slow pomp will output just 10-25 Watt of heat where a realy powerful pomp can produce as much as 100-150 Watt! So the extra heat of the pump does increase the temp of the coolant, not the higher flowrate. This shouldn't be a problem by inline pumps.

I heard that the eheim 1250 is a very good pump. I don't know exactly how much $$$ it costs, but I think aroud the 50 dollar.

btw: don't worry about the magnetic field thing, it's really not strong enough to make a difference.

[Fig]

The reason a slower pump is said to be more efficient in some applications is becuase the water spends more time in the radiator/fan, where the greatest measure of cooling is done(for the water)

[hielko]

That's stupid imho

There will always be water in de radiator. And when the water is going faster the temp difference between te water entering the rad. and leaving is smaller. But so is the water what's heated up in the waterblock.

[Fig]

No need to get all fired up! Im just stating what I have read! Sheesh.

[UnaClocker]

I have the Mag Drive model 2, the 250gph pump. It's not silent, but I only hear it humming when the computer is turned off. The only time it does make noise is where there is air in the system. It's the best bang for the buck. Also, it's better to have too big of a pump than too small, you can always put in a flow valve and throttle back to your "sweet" spot..

[p[ara]d0x]

The Mag Drives are good if you like to do more DIY stuff. Another alternative is the OCWC Reesepumps, a bit more expensive, but nice.

[terabyte]

Well consider that in a car the themostat cuts the water flow until it reaches a certian temp allowing the water in the rad to thrughly cool it might make sence I had a old car that would over heat and I found out it had no thermostat I put one in and no more probs, I'm no thermall enginner but it might make sence that a slow rate where the water is in the rad longer might make a diffrence in efficiency.

[gmat]

I'm not exactly thermal engineer but engineer, yes. I think the flow rate has little relationship with temperature. Consider that a higher flow rate will give you a more uniform temperature all over the water circuit.
What counts is the thermal power of elements (negative or positive...). Say, if your WB pumps in 70W you'd better have a radiator which pumps out around 70W, or more.
One thing about low flow rate is, you gould quickly get a hot point in and after the WB. And you would not want water to boil inside it, would you ?
The higher the flow rate, the higher contact surface area you'll get. So you'll get more uniform temperatures around the circuit, and less 'hot spots'.
To make a long story short, consider that your WB/CPU interface temp has an inverse exponential relationship with flow rate. At low rates, a subtle variation in the flow will make big difference. At higher rates, it will make little difference.
With a single circuit (pump/wb/rad) and 10mm tubing you wont get much difference above 500l per hour... given you've got *real* 500 lph, which you'll never have. Mazes, turns and elevation all bring your flow rate down.
Thats it I hope it was clear enough...

[DeathFlame]

Ok guys, this guy did some tests. Check out his page.

Flow Rate Tests

Quick summary for you lazy ones: Put in a valve and "tweak" out your flow rate. Various different factors contribute to which is the sweet spot for you, so test it out and find out for yourself.

[Lord Smack]

thanks for the info guys it helped a lot

[LiquidCool]

Well, I have been using a 500gph submersible 12 volt pump for over a year now with my Celeron2/533 at 880 with 1.7 volts. At full load it keeps it 4-5 degrees celcius above ambient. This is with a 10"x5"x1/2" radiator and 1 4-7/8" fan running at 5 volts instead of 12. So there is maybe 25-30 CFM going through the radiator. I can run it all day at 100% with seti or prime and the temps never go any higher.


If you go to my <A HREF="http://liquidcool.org">site</A> you can see some temp results. I also have these pumps for sale for $19.95 (normaly $29.95) as well as reservoirs and a very nice line of fittings. My fittings don't need any hose clamps and give a perfect leakfree seal but still cost less than a brass fitting with a hose clamp.

[DeathFlame]

You COULD not use hoseclamps. But why not spend a few cents and add that extra insurance. $1000 isn't worth nothing wet with sparks flying, just casue you saved a buck or two without hoseclamps.

They're made for a reason, don't tempt fate.

[LiquidCool]

Thats the whole thing, there is no way you could even use hose clamps with my fittings. The tubing goes inside the fitting and seals with an O-Ring. Then there are these small stainless steel teeth that hold the tube in place. No worries! they take up to 150psi so I doubt pressure is a problem. You cant pull them out by hand because of the way it's made. There is a collet at the head of the fitting that you push to release the hose. Buy a few from me and you'll never go back Here is a picture of the innards of them...

<CENTER> </CENTER>


And if you go to this page on my site you can see some instructions on how they work. It really is much simpler and safer than normal barbed fittings.

[DeathFlame]

It's nice to be wrong once in a while, especially about something as cool as that.

Thanks for shutting me up and showing me something maybe I need for my sistem.

[imagex]

I'm running two pumps one 1000 gph for my water block and one 500 gph for my rad, found that bildge pump work great and are cheap.1000 ghp =$26 500 gph =$9. i don't have to worry about the flow rate between my water block and my rad there two differant systems,and both pumps for $35 thats 1500 gph. later

[LiquidCool]

Wasnt saying you were wrong, I could have typed it better in the first place. But, go buy some of the fittings, they're pretty cheap for what you get and they don't hurt the tubing, so you can connect and disconnect them over and over again! I'm trying to get alot of people interested in them because the more I can buy at once, the cheaper I can get them, which means I can sell them cheaper... tell ya what I'll hook up anybody that orders a decent amount($30+) with a 10% discount just to get these things out... any takers, well? After you finish the order send me an email telling me you read this at procooling and you want 10% off, and I'll paypal the money back to you...

LiquidCool.Org

[gmat]

Wow.. I want some, definitely !
How can i order from your site (it's unclear)... Do you ship overseas ? Can i pay with my VISA ?

[LiquidCool]

Yeah, you can use your visa through PayPal.com. Here is my email and ICQ# and we will work somthing out...

Michael@liquidcool.org
ICQ 95735580

Let me know what you want to do...

[imagex]

hey here's what looks like a good deal. 500 gph for 14.99. http://cgi.ebay.com/aw-cgi/eBayISAPI...tem=1227741479

later

[LiquidCool]

ACTUALLY TO BE TECHNICAL, the thermostat in a car has nothing to do with cooling the water. The thermostat is there to keep the engine from getting to cold! Beleive it or not. But if the engine gets to cold it runs less efficient... so THERE, nah nah The reason a vehicle will overheat from a bad thermostat is if it is stuck closed, EG: not letting any water through at all...

[ondaedg]

you can buy that bilge pump at Walmart

[imagex]

also some engines have spaces that are hard for the water to reach ,flowing water flows the easys path, in this case they stop or slow the water so it goes to all areas.

[LiquidCool]

I was refering to the guy that said thermostats were in cars to let the water have time in the radiator. I don't know how he came to that conclusion but oh well...

I had a Dodge Dakota that I dropped a 1978 360 into. I got these 2 huge ass Aluminum radiators out of some 98? chevy Z71 and had the local radiator shop join them together in front of each other(EG: made it twice as thick. Son now I have a 4" thick 3' wide and 1-1/2' tall aluminum radiator. It was a monster, had to section the front of my frame to let it sit properly. Anyway, I had to get rid of it because it kept the engine TOO COOL! I could watch the temp gauge and when the thermostat would open up. my water temp would drop down around 110 120. Which of course would close the thermostat back up. Then the temp would slowly rise up around 190. Thermostat would open then BOOM back down to 110... If I was going down I-4 ( a large long highway near me) doing a "respectable :O" speed I could actually feel the difference in performance when that cold water hit. I also didn't want to crack my block. ANYWAY enough about my other toy