Cooling Pump issues...

Thanks for the input, guys!

The head pressure on this thing is rather large (don't have the figures in front of me, but as previously stated, hit the ceiling without slowing down). It's 3/4" fittings, which are nice but a little TOO big.

I'm going to make an attempt at this, but I may follow the advice of attaching it in a separate box for the time being. I doubt it's going to add SO much heat as to be a problem, and the flow coupled with the head pressure will allow me to add an extra rad with no problems if it does become an issue (maybe one before the reservoir and another directly after the pump). I can give it a try because, heh, it was free and if it doesn't work I'm out nothing and just have to get off my cheap a$$ and go get me an actual pump.

I have also thought of leaving it IN the computer, but putting it in its own soundproof padded little cell, where it can't get TOO noisy. The heat could be dissipated by a 120mm fan on the side of that little cage, blowing out the hot air (or 2 80s, one in, one out, low power setting for noise). Thoughts on this idea?

 

That sounds like a danner magdrive 950 you have there, is it? I got a magdrive 1800 sitting here thats going to be used for an aquarium in the future but is doing nothing right now. Its rated 114 liters/min and 27 liters/min at 16' of head... it sounds like it could shoot water over my house . If you are going to use that pump dont use a dimmer to try to slow it down. To adjust the speed of a motor like whats in that pump you need to adjust the line frequency, but the cost of a variable frequency drive is insane so forget about it. Either plan a very large water loop or get a smaller pump. I'd go crazy on a big loop though .

 

You are correct, biff!! I have located more info about it, and it is a danner magdrive 950.

If I may ask (I'm no electricity guru), why can't I dim it down a bit?

Heh...what kind of loop do you think I'll need here? I'm figuring on just a cpu, gfx, northbridge, maybe hdds. The beauty of the pump is for all of its size, it's REALLY quiet (particularly when going in a soundproof insulated little plexi box inside the case). I'm not really all that much a gamer (more a tweaker), so I don't have a board capable of SLI or anything.

 

How the motor works is like this. Your impeller is attached to a permanent magnet. Surrounding that magnet is the windings. The winding are arranged so that it creates a rotating magnetic field. Actually thats in 3phase units, in a single phase unit its analternating magnetic field with a another winding that is set at an angle and then the current is phase shifted from the fundamental motor current to crudely simulate a second phase and hence a rotating magnetic field, just to make sure the impeller actually rotates rather than flettering back and forth. Any way the permanent magnet of the impeller stays "locked" to this rotating magnetic field and rotates at the same rate. Now here's the meat... The speed of the magnetic field and hence the impeller is based on the frequency of the line voltage... a slower frequency means a slower rotating magnetic field. The higher or lower the line voltage increases or decreases the strength of the magnetic coupling between the rotating magnetic field and the impeller. So if you still had 60Hz coming in but reduced the voltage to say 60V your impeller would still turn at the same speed. BUT! if the impeller is loaded enough to break this smaller magnetic coupling then the impeller is no longer locked to the rotating current. In a real situation the impeller would most likely try to stop and reverse to meet up with the next approaching opposite polarity magnetic field. So to make a long story short(er?), if the coupling breaks there is a significant if not complete loss of power from the pump and it will probably get quite noisy with a fluttering impeller to boot. And there you have it! Thats why you need to reduce the frequency not the voltage to slow down the pump.

You dont need to read this part but just to add to above. If you decrease the frequency you will actually have to reduce the voltage aswell since at the lower frequency there isn't as much inductive reactance in the windings and if the voltage stays the same the current will rise and so will the power dissipation.

 

Ok...I think I get the idea. So I need to regulate frequency as much as I need to regulate voltage...any idea on a ratio for this? I have some buddies who are Elec Engineers and could build whatever I need them to (I think)...it'd be great to build a control mechanism for this.

 

That I can't tell you because it will be based on the specifics of the motor, but your friends should be able to figure it out pretty easily.

 

Earlier he said the pump will hit the cieling with a 1/2 inch stream. That is a good 4 to 6 feet of lift. with that kind of lift it might need a tank on the output as a pressure stabilizer? That powerful a pump might give a serious surge on start up- ever hear of water hammer?

 

Man, if you have that sort of pump, place the pump and the radiator outside your house! Best would be to make a res from a big compressed air gas bottle (empty one), set the inlet at the top and the oulet from the bottom, connect the pump accordingly and then there will be only two tubes comming in and out from your room
Best temps you can get from mother nature 24/7/365!

The bloke who had done this before is of course this guy:
http://zfz.com/



Now in the real world , a really powerfull pump might try to push all the tubes off the fittings so be carefull and tight them up really well with warm driven clips or something!

B

 

Alright since everyone is all bent up on the head pressure and flowrate of his pump here's a link... he has the 9.5 model, and here's a link for the specifications of his pump. Since someone mentioned it, at 6' of lift that pump will put out 720 GPH.

Nothing says you need a big res for this pump, the only point to a res is to help work the air bubbles out of the lines. The only thing you have to worry about your res design is that the water flow wont cause a whirlpool effect or churn up the water so much that air will be pulled down into the lines. Although this pump seems big to us water coolers is really kinda small when it comes to aquariums and I haven't heard of anybody blowing apart their plumbing on their aquarium. This pump has a maximum pressure ouput of about 6 P.S.I, which any reasonable hose clamp will have no problems with.

Alfter giving it some more thought, it might be worth it to give a dimmer a try for two reasons. First, like I said you only need a magnetic field strong enough to keep the impeller locked with the rotating field so you might be able to lessen that field and keep the pump working fine and save yourself some watts. Second, and this is a long shot but, you may find a point that the impeller starts to "slip" on the rotating field but keeps turning at a reduced rpm. For a couple bucks for a cheapie knob type dimmer you might get a pleasant surprise. Just listen to the pump really well and stop if you hear odd noises. Also if it the flow rate you're worried about and not the wattage of the pump, it's perfectly fine to restrict the OUTPUT with a ball or gate valve.

Hope all this helps!

 

Im afrade how high the water comes out of the unrestricted pump has very little to do with head pressure.

slater..

edit: 13' for 93 Watts!, Thats like adding another CPU to your loop.

(D4 = 24 W,10 ft)

 

biff, your replies keep getting more and more useful. thanks for the input, and keep it coming!

slater, I think you might be right about the heat output, but I wonder if there's a way to tone that down. Not to mention, this pump has 3/4" inner diameter inlet and outlet, so truth be told I could probably run this thing with a LOT of water to absorb that heat. I wonder what the difference in efficiency would be that would make the d4 so close in pressure for so much less energy transferrence? Aren't they basically the same idea of setup?

My case is getting built today and over the next week (pictures are already taken for the initial portion), so the pump will be finding its home soon. I really appreciate everyone's help in figuring out the best way to run this! When I start posting, I'll drop a link on here for the project log thread and everyone can see what's happening with it.

Quick question, though: Will insulating the pump in its own box for sound protection increase the heat it transfers to the water or have any other unseen effect? I can't think that it would, but... Barring that, would it be better to put it in the res? it can be a submerged or in-line pump...what are the benefits to either?

 

The difference btween the d4 and yours is the d4 shuts off at 10' yours will do 400 GPH at 10'. Also the 93W is max not continuous operating power. It will take the same amount of watts to move the same amount of water to the same height at a particular flow rate. Past that efficiency comes into play, and I doubt the D4 is 4x more efficient than your magdrive.

Using alot of water will not lessen the heat put into the water by the pump it will just slow down how fast the water heats up which won't help you any.

Read my previous post about using a dimmer to tone down your wattage.... it MAY be possible.

Insulating the pump in its own box will cause the pump to run much hotter and some of that heat will be transfered to the water. The pump will consume some amount of watts depending on the load you give it and that heat energy has to go somewhere. If you had it in a perfectly thermally sealed box the temp would rise untill it all melted down. That whole damn conservation of energy thing!

Submerged will be more quiet but will add more heat to the water and also possibly increase the life span of the pump, the opposite is true for inline.

 

Well, I wasn't thinking perfectly insulated, actually just a box that had venting to it but had egg-crate styrofoam in there to dampen the sound. How much heat would running the pump submerged add to the system? Are we talking about 1 or 2 degrees or are we talking 10?

Will examine the dimmer switch idea further.

 

I would say closer to 1 or 2

You allredy have the pump and therefore theres not rely anypoint in me telling you that its not the 'ideal' pump for the job. Im sure it will do a pretty good job expecialy if you have a large radiator to ofsett the heat input into the water.

slater..

 

I'd give it a go inline first. Although I haven't fired up my magdrive 18 yet, the whole reason I got this pump is quality and its supposed to be extremely quiet.