I have a HUGE problem here (pump)

[Solarian League]

Because i'm no good with pumbering so i'm having my dad help with the watercooling system and stuff, anyway he's being very picky about pump.

Anyway correct me if i'm wrong but dosen't cather's cascide require at minum of 5 gallon/min to function at its best.

Anyway i searched ****ing high and low on the internet and i couldn't find any pump that fitted the spec.

it HAS to have 1/2" in and out.
It has to be rated for continious durty aka 24/7
It has to be able to pump between 5-7 gallon/min or rather 18-27 litter/min

looked at the ehams, the best one i found was about 370 gallon a min.

the iwaks are PERFECT, theyre adjustable so i can take their 8 gallon/min model and slow it down to about 5-6 gallon a min, surely an 1/2" hose can handle that.

anyway i cannot even do that option and i have searched HIGH AND LOW and can't find a ****ing 1/2" in/out pump within the range that i NEED

anyway i found quite a few pump that works but the problem is theyre all frigging 3/4" to 1" and my dad REFUSE to use em it HAS to be 1/2" port on the in and out.

[pHaestus]

not aware of any reference to a 5GPM flow rate with Cascade. Pretty sure Cathar said it does even better at low flow rates than the Whitewater... 5GPM is not really realistic; any pump that can generate that much flow is going to add so much heat to your loop that temps wont be great anyway.

[t00lb0x]

The only problem with pumps are that people recommend different pumps. Deciding on a pump has so many factors. I would wait until you get a few more recommendations and cross compare.

[Cathar]

Don't worry about it too much.

Anything like an Eheim 1048 (1/2" in and out) or better will give you very good results out of the Cascade block.

The thing with the Cascade block is that its performance scales quite well with stronger pumps.

This leads to some confusion. You could use an Eheim 1046, or heck, even the baby pump that I used, being a 50cm head 200lph pump, and still get good performance with the Cascade.

The trick here though is defining "good" in reference to "what?".
The block will perform well no matter what pump you throw at it. It's possible to eke out another couple of degrees by using a very strong pump and a large radiator setup, but by now we're well into the point of diminishing returns.

Given a radiator that will fit in your case, some moderate speed fans, and the Cascade, we're probably talking about less than a 1C difference in final CPU temperature from an Eheim 1048 to an Iwaki MD-30RZ. Given a much larger radiator setup the difference will be greater, but even there, maybe a 2C difference at best.

[Solarian League]

allright cool thanks for that clartfication

i have found an 5.5 gal/min (at 9 foot head) pump with in and out thread of 1/2" thread at the surplus center pump

i think that one will work good for my needs, its a mag drive pump which means no direct shaft so less wear i would think

cathar: a question, will you still make the friday date for the "building or manfuacting" of the waterblock?

[Cathar]

Why not just buy a Danner Mag Drive Model 3?

Around $45, with 1/2" NPT thread fitting. All you need to do is buy a pair of 1/2" tail to 1/2" NPT barbs, screw them on, and you're done.

The pump you're looking at is WAY over-powered for your needs. It'd be putting more heat into your water than your entire computer.

[Ewan]

Just a quick question for the people with practical experience:
I've read in a few places that oversizing the pump and getting mondo flow just raises the temperature of the water to no good end.

I'm wondering if there is any practical evidence of this. I would have thought that if the added flow would more than make up for the slightly higher base water temperature. At the higher flowrate the radiator becomes more efficient as well.

Just wondering

[airspirit]

Er, if you use a Danner 500 model (the PondMaster and MagDrives are very nice and silent if mounted right) that gives you around 8.5 GPM at 0 head. You should be at around 5-6 GPM after considering system resistance. I've used one of these for continuous use, and they can be run inline. I believe they use 1/2->5/8" connectors, but you can easily adapt that to 1/2->1/2 if you need.

Problem solved! And for only $50-60, it is a cheap solution as well!

edit: a WAY overpowered pump can cause problems due to heat. My pump injects around 100W into the water, but my system is unique ... depending on the system, an inline pump such as the danners above will make up more of a difference in flow rate than will be lost in the marginal increase in energy input to the coolant. Going from a 3 to a 5 would usually be a good thing, though I suppose in certain systems the opposite would be true. In a system that uses a high flow block, however, you REALLY want to get the proper amount of flow.

[redleader]

Rads are so cheap and so overpowered I doubt that a hot pump is a big concern if you plan ahead. However I agree that supergiant pumps are overkill.

[pHaestus]

I have been using a Little giant industrial pump 3E-12NYS in my test loop, and it is speced at 500GPH and 18' max head. I measure about 3.5GPM MAX out of that system. Not sure you are going to get 5GPM out of anything other than a positive displacement pump that is speced for 5GPM operation honestly. Even 3GPM is a stretch, especially using 1/2" hosebarbs (their ID is ~3/8). I have everything pretty much plumbed to 5/8" and need to finish up drilling out all those barbs even...

[Blackeagle]

The Iwaki MD-15 would be a good choice to go with the Cascade.

Long service life of the Iwaki.
11.1 ft. of max head
5gpm free flow
1/2" inlet and outlet
30 watts power draw, so not going to add to much heat.
quiet

Only problem with Iwaki pumps is cost, but still would be a good match for a Cascade block.

[Zhentar]

I'm using the Little Giant PCL-020 Cost me about $55 at my local Lowe's; Its got 5 GPM free flow and 12.2' head rating. 47 watt draw though, but its pretty quiet. The input/output are both 3/8" though.

[Solarian League]

i like the iwaki pumps, but i never did find an md-15 pump all i saw at a few site was an md-10 pump which was too small, and a md-20 pump which had 3/4" attachment.

anyway if i can find an MD-15 pump i probably will end up using that thing.


talked with my dad and he's pretty set on geting that pump, that i posted above, the little giant one. anyway...

what sort of problem should i expect to get? etc....

another thing, how long do those little giant pump last?

[Quarter]

Little Giant makes good products, but the pump your looking at is going to put out some heat.
@ 1.5 amp, your looking at around 172 watts.

Iwaki makes some very fine gear and for comparison, the MD 15 is only 33 watts.

[Solarian League]

um i wonder if my radator core will be able to handle 175 watt + at minum 80 + 40 some watt.... ouch

anyway that radator people said should be able to handle atleast one 226 watt pelter + cpu + GPU/pelter + northbridge so i think i'm safe on my radator but my water will be little warm that is for sure.

the problem with the iwaks is that i can't find an md-15 anywhere, to buy, i can find its spec yes but... anyway the md-20 would be perfect but my dad hates the idea of redcludition, thinks it puts too much strain on the pump, and i ditto about that but i think its sort of stilly but he's the one who is building the water loop, i'm no good with that sort of stuff.

[Since87]

Originally posted by Quarter
Little Giant makes good products, but the pump your looking at is going to put out some heat.
@ 1.5 amp, your looking at around 172 watts.

Iwaki makes some very fine gear and for comparison, the MD 15 is only 33 watts.

You can't necessarily determine the power consumption based on the voltage and current ratings for an AC pump. The power consumption may be much less than 172 Watts depending on the phase angle of the pump's current draw.

[Blackeagle]

www.underwatercreations.com

This is the site where I purchased my MD-15. They have a pretty good selection of Iwaki pumps.

If part of your father's concern is the quality of the pump, then you can do no better than a Iwaki pump.

[BrianW]

I got mine there as well. They had to special order it, as the japanese version is behind production in orders. Probably why you do not see alot in retail. But these guys must have something worked out w/ Iwaki-Welchum.

BrianW

[Solarian League]

i think i'm pretty much stuck with that pump in that link. so... we gotta to get a solution that includes that pump. anyway do you guys think my radator will be able to handle it.

this is my radator link

anyway i got this small wind tunnel in the top part of my computer, the 550 watt antec PSU only ate up 1/2 of the PSU sliot and i don't need the top 4x 5.25 inch bays so i and my dad converted it into an mini wind tunnel with two 120mm 114 CFM parsonic pafinfio fans at about 32 dBa so it should work if that's not enough air flow to keep the water cooled to say 30-40 average then i probably will stick in an stronger 190 cfm delta.

anyway everything in my system is using 1/2" hose, anyway will my radator core be able to handle the load? they say it should be able to handle 2 or so high powered pelts so...

[Long Haired Git]

1. In my opinion, 5 LPM is a good target flow rate. That's "litres" people, as in the 20th century base ten measurement. You poor people stuck back in the 19th century, using base {anything} for measurements will keep getting confused. There are even multiple definitions of how much a gallon is. US = 3.785, English = 4.545 and then I think there are multiple versions of gallons anyway. Assuming standard US definition, 18 LPM (5 GPM) is HUGE, way too much.

My rule of thumb is:

Flow <= 1.0 LPM = too small, try again. Throw the pump out, or at least check for restrictions and kinks.
1.0 LPM < Flow < 3 LPM = acceptable, but borderline. Should have gone for a better pump
3.0 LPM < Flow < 8 LPM = Sweet spot
9.0 LPM < Flow < 10 LPM = acceptable, but borderline. What gains has the expensive pump got?
10 LPM < Flow = What are you trying to acheive here?

I will gladly accept donations of an entire model-range of a manufacturers pumps to test it out....

[Althornin]

Its just that for the higher flow rates, you need a bigger radiator to get any gains, or the extra heat the pump adds to the system negates the additional cooling higher flowrates gets you.

[#Rotor]

think of it this way. Cathar's designs will fly high, with the puniest little pump you can imagine. any high resistance, turbulence inducing block, can fly under severe flow-deprived conditions. It is the free flowing open chamber blocks, that absolutely has to have mega flow, in order to keep up with the restrictive ones..

[Solarian League]

Originally posted by #Rotor
think of it this way. Cathar's designs will fly high, with the puniest little pump you can imagine. any high resistance, turbulence inducing block, can fly under severe flow-deprived conditions. It is the free flowing open chamber blocks, that absolutely has to have mega flow, in order to keep up with the restrictive ones..

ah that's cool i do have an danger den maze 4 for my GPU so i probably will need an decent flow for that area but atleast i will know that cathar's block won't be the problem anyway i hope it get here soon.

[airspirit]

Er, to the above "metric evangelist", many of us are from the states and like our old school measurements. Further, many of the pumps sold here are sold using GPM. You will notice, however, that we use degC since, again, it has become a relative standard in our circles. You may want to think about what you are saying before you charge in and call many established experts in this area (and senior members of this forward) a bunch of backward ninnies because YOU don't prefer our units of measurements. Prove and establish yourself before making a useless comment such as that, because from where I stand, it just looks like you want to be written off as a troll.

Some blocks like to have a good 3-4 GPM running through them. Some like more. 1 GPM (around 4 LPM) is absolutely NOT a good idea for many of the blocks out there currently and would cause instant meltdown in some of the systems that we are trying to cool.

If I cut my bypass off in my cooling chest it lowers my flow rate to around 1.3 GPM per block (over 5 LPM, IIRC). Doing this causes my temps to go up around 6-7C. I like to keep my flow rates at least over 2.5 GPM per block to ensure that I'm cooling adequately. Are these results a figment of my imagination?

To further understand this, go read some of BillA's reviews on overclockers.com. The graphs might open your eyes to the fact that your ideas, however grounded in fact you may think they are, are incorrect.

[Cathar]

Originally posted by airspirit
Some blocks like to have a good 3-4 GPM running through them. Some like more. 1 GPM (around 4 LPM) is absolutely NOT a good idea for many of the blocks out there currently and would cause instant meltdown in some of the systems that we are trying to cool.

1GPM (~3.9lpm) causing instant melt-down? Surely you jest?

For a real 100W heat load, water flowing at 1 US GPM will heat up by around 0.36C as it absorbs the heat from inlet to outlet of the heat source.

Not exactly melt-down. The majority of water-coolers in Germany are getting by quite happily with down to even 0.5lpm (1/8th GPM) and cooling "okay". Many of these guys are using traditional Maze-style blocks too.

Originally posted by airspirit

If I cut my bypass off in my cooling chest it lowers my flow rate to around 1.3 GPM per block (over 5 LPM, IIRC). Doing this causes my temps to go up around 6-7C. I like to keep my flow rates at least over 2.5 GPM per block to ensure that I'm cooling adequately. Are these results a figment of my imagination?

Over 10lpm is a fairly rushing flow rate for a computer cooling setup. I think you'd find that in a quick poll of members here at this forum that many are at or well below the 2gpm mark and aren't suffering for it.

What waterblock(s) do you have where you are witnessing a ~6C difference between 1.3GPM and 2.5GPM?

Originally posted by airspirit

To further understand this, go read some of BillA's reviews on overclockers.com. The graphs might open your eyes to the fact that your ideas, however grounded in fact you may think they are, are incorrect.

Which graphs were you looking at? Pretty much all graphs of BillA's I saw had a definite tendency to start tailing off between 3-6LPM, and past 8LPM almost all tested blocks are picking up very minimal gains as the flow rate is increased.

4-6lpm is ~1-1.5gpm, and 8lpm is ~2gpm.

I don't mean to be rude, but as you're knocking LHG for stating that the conversions from Imperial to Metric can be confusing and you're claiming they're not, is it perhaps possible that you have in fact confused BillA's graphs and what you think you're reading as GPM, really is LPM?