why is there so many !diots..
 

http://www.tweakmonster.com/articles...pc65/index.htm

http://www.ocaddiction.com/reviews/w...reme/pg2.shtml

now correct me if i'm wrong.
why from PUMP > CPU BLOCK?
isn't it supposed to be PUMP > RAD > BLOCKS ?

since that the pump will dump heat into the water, we want to cool the block, not the pump..so the water from the rad will be the coolest..not pump..right... depending on the heat output of the pump, it might make a big diffrence..



:shrug:

Well actually thats still up for debate. I run mine: pump>rad>block>pump, but some people find more success doing it the other way.

Your logic is sound, it's just that the water from the pump is more turbulant (has more pressure/flow~speed as well???) which is just as important an aspect as H2o temperature change from the rad, so it's a toss~up, as long as the pump is close enough to the CPU~block to make the turbulance felt...

I Think!!...

dunno about the turbulance; but BillA once pointed me out in kind words that flow is a constant throughout a given loop. it makes sence once u think of it.

anyway; on topic; i don't know if it makes that much difference , where your pump is. i got my pump before the block because i wouldn't have to be using strange bends to be able to mount it after the WB.

if u have to start using ebows to be able to do it right, then don't, i'd say.

Yep!, scratch the flowspeed :) ...

Mount it the way you want. Usual water temp differentials are less than 1°C through the circuit. Getting less bends and less tubing is more important.

there is no "supposed" to on this matter. Since well, its not a set in stone subject.

I have mine pump -> rad, mostly because...thats what would work best in my situation.

Not that anyone gives a rats ass,but I have mine set up like so:

reservoir
radiator with twin 120 mm fans pulling
350 gph magdrive
waterblock
350 gph magdrive
back to reservoir

Not posting temps as I don't want credibility to be an issue. ;)

I was only this morning thinking of having two pumps at oposite ends of the system to keep pressure boosted throughout!, post your temps dude!, as long as you acknowledge the falibility of the mobo(or whatever) thermistdors it's OK. it's when you see a dodgy temp value used to 'sell' a product (as [h]ardware reviews do, no matter how much they deny it) it galls (IMO)...

i believe the flow doesn't really matter..since its a loop..
i don't think there's a diffrence if the pump is right before the block, or rad first then block..thats what i've been thinking..

anybody got solid facts to prove me wrong or right?

Ok...just fired up the rig. Its 57 f or 13.8c outside this morning in the shop here. The thermistor I have epoxied under the chip reports temps at 23.7c and Sandra reports the chip at 24 c.
Thats a TBird 1.4 at 1.75 volts on a K7T266 Pro2 running stock FSB.
I lied earlier when I said I have 2 pumps on this rig.... I actually have 3. Right after the WB I have a Shurflo 2.8gpm diagphram pump inline too... its not running right now as it causes my 5/16th hose to flatten out.
Temp update : 1 magdrive running= 25.5c
2 magdrives running= 24.3c

Remember my ambient is a bit nipply right now at probably 60f, so that helps alot. I'll post back when it heats up to mid day.
IMHO, 2 pumps is neccesary in a restricted system like mine... my radiator is 3 inches thick with 20 passes and all my pumps are external with the required plumbing, probably 12 feet of tubing in all.

The important thing it to report how temps were acquired and the range. If one has relative temps (I changed this part and noticed xx difference) then it can still be useful for them (or for others) to discuss their systems.

Sort of like if you use a normal Hg thermometer outside your window to decide whether you need a coat or not. Perfectly acceptable. If I used the same thermomter (with temps reported to 0.1C) to write a research paper on how climate change was affected by global warming then that is UNacceptable.

In theory, since the pump does add an amount of heat to the water, it would be best to go pump -> rad->block. With most hobby pumps it is a small amount of heat, and you may find that if you have to add a 90 or a lot of extra tubing to run it this way that the loss in flow rate affects CPU temps more than the extra 20W heat. Some are moving to much larger pumps nowadays, and they can produce far more heat (my little giant is 200W).

At any rate I can't imagine it making a huge difference (or perhaps even a detectable one) for most.

I figure that with a 30 watt pump and a 100gph flow, you should see less than a .05c increase in water temps from the pump.

pHaestus, if you don't mind me asking, what is you water temp over room temp using the little giant, and what are you using for a rad? Thanks in advance.

It couldn't matter less.

thats really helpful and informative.
i learned something.
thank you. :rolleyes:

Something I'm surprised nobody has touched on yet...
 

If I recall correctly, don't a lot of magdrive pumps have a tendency to reverse direction when turned on/off? If so, a pump->rad->wb system would become a pump->wb->rad system, depending on your luck.

Has anyone had temp differences in this type of setup with a magdrive, simply by restarting the pump?

Yes, your'e right, they do start at random direction, but the outlet is made somewhat centered, so it doesn't matter which way the rotor spins. It decreases some of the pump's abilities, but that's the only way (without the tweaks like Volenti did) to ensure the direction of flow regardless of direction of rotor spin.
Dunno how to explain it better, there was a thread concerning this...:shrug:

puzzdre says it all. maybe it is true the rotor spins in different directions. but my loop always flows in the same direction...

I recall that thread, but thought there were some pumps that didn't have the outlet centered. (I have been known to be wrong in the past, though :rolleyes: )

Dunno for the mag drive pump like that in cheap $$ range (I might be wrong too!:D ), it's possible that those have some additional electronics built in to force the spin always in one direction...
I had no luck in digging out that thread...Maybe it was on OCAU...dunno...:(

My lord, that case is quite 'bling bling.' I think he needs more stickas though as they add a good 10 horsepower each :evilaugh:

Aw c'mon, give the guy a break. He's only got 13 stickers on the front bezel, and I didn't see _any_ on the corner of the monitor that was in the picture. </sarcasm>

I suppose you would also make fun of the pewter dragon figurine he has on top of the PC, if you had noticed it. :p

Um, why do you come here as a n00b and start flaming everyone in topic subjects? Why do you assume that because you read something in an article that it is the truth? Why don't you traipse your browser over to [H] if that is the attitude you want to take.

You need to try it both ways in your machine before judging which is best. I've found the best results by having my block after the pump, and they've always been rather significant gains. Some people have found the opposite. This depends on the system, so if you don't have anything to back up what you're flaming everyone over besides articles that may or may not have the accuracy of a WB shootout!!! at [H], then back your truck up before you get shat upon.

Did anyone notice the waterblocks the guy in the TweakMonster article had on his memory? Interesting to note that they were not connected in any of the photos in the article.
I'm skeptical of any gains, but curious of his results, nonetheless.

http://www.bluebuzz.com/~rburgess/memory_wb0.jpg
http://www.bluebuzz.com/~rburgess/memory_wb2.jpg

Why the sarcasm? I'm simply stating that it makes no measurable difference if you've got good flow.