BEST Iwaki pump?

[Cutless009]

Whats the best Iwaki pump out there, cost aside? Is it the RD-30 or is there a better one?

Is there any serious modifications you have to make in order to run it in a WC rig, or does it come with 1/2" barbs?

Does anyone make a custom top for it, or not a big enough market for it?

[Septim]

you don't need a custom top for it... impeller & volute design is already perfect...

inlet is around 5/8 thin barb...
http://farm4.static.flickr.com/3059/2950285053_28a463da99.jpg

here you see i've added some 5/8id 7/8od tubing over it...
http://farm4.static.flickr.com/3195/2950285299_fe312a7d72.jpg

then used a 3/4id tubing as feed from an ek400 res... with the bottom modded for bigger outlet to feed this hungry RD30...

btw with just 6 screws, the top comes off, and you can rotate the outlet barb to point where you want it to...

[ballz0r]

RD30 is the only one i tried...you can drill and tap the outlet to G1/4 too

http://lh4.ggpht.com/_5JXSu5Gfgo4/SQsd9vQqSGI/AAAAAAAAAyU/Gh6V3BdU13o/s800/31102008073.jpg

[Sadasius]

How is the noise? I am thinking of using two in my next build.

[Septim]

well quiet enough for me... mounting it on some rubber washers/gromets probably helps...

and of course bled fully...

[RedRaider]

RD-30 is a MANS pump, not for the squimish....

BTW, PM me if you want one, I have a pakage deal coming up with pump/meanwell PSU/modded EK RES/a-bracket.

Only available for RedRaider Technology Members.:clap:

[Snyxxx]

RD30 is the only one i tried...you can drill and tap the outlet to G1/4 too

http://lh4.ggpht.com/_5JXSu5Gfgo4/SQsd9vQqSGI/AAAAAAAAAyU/Gh6V3BdU13o/s800/31102008073.jpg

ballz0r,

What a nice and clean layout. Love the way you mounted that pump/res. Very efficient.

[Cutless009]

RD-30 is a MANS pump, not for the squimish....

BTW, PM me if you want one, I have a pakage deal coming up with pump/meanwell PSU/modded EK RES/a-bracket.

Only available for RedRaider Technology Members.:clap:

Shit man I wont be ready till late January when I start me new build :(

[Fitseries3]

if anyone can find me another RD30 or even a RD20 cheap i'm looking for another one.


oh and... they are AWESOME.

[link1896]

Whats the best Iwaki pump out there, cost aside? Is it the RD-30 or is there a better one?

Is there any serious modifications you have to make in order to run it in a WC rig, or does it come with 1/2" barbs?

Does anyone make a custom top for it, or not a big enough market for it?

Custom tops I wouldn't worry about, there is no way an armature computer water cooling pump "guy" is going to deign a better top then Iwaki.

there are better, bigger centrifugal pumps in the Iwaki range, i/e MD100R, or even better, diaphragm pumps. I've used the Iwaki LKN57, (not in a pc application, but I am using another diaphragm pump in a new build).

The LKN57 does 6 litres per minute (just under 2 GPM) at 45psi. 45psi is about 100 foot of head. But a LKN57 is HUGE. we're talking 12" long, 12" tall, 35 pounds.


A diaphragm pump, or any positive displacement pump, works a whole lot better in a multi block, multi rad loop with a HIGH amount of restriction then a centrifugal pump. A centrifugal pump ends up with too much sheer. This is where the pump is well outside its peak efficiency area, where a large amount of the pumps power gets used to sheer the liquid inside the pump chamber and not used to move the liquid through the loop.

[Fitseries3]

holy shit... http://cgi.ebay.com/IWAKI-America-Single-Phase-Magnetic-Pump-Motor-MD-100R_W0QQitemZ270319282302QQcmdZViewItemQQptZBI_Pu mps?hash=item270319282302&_trksid=p3286.c0.m14&_trkparms=72%3A1205|66%3A2|65%3A12|39%3A1|240%3A13 18|301%3A0|293%3A1|294%3A50

[Fitseries3]

is this one decent? http://cgi.ebay.com/Pump-Diaphragm-Shurflo-1-8GPM-50PSI-12VDC_W0QQitemZ110326924265QQcmdZViewItemQQptZBI_P umps?hash=item110326924265&_trksid=p3286.c0.m14&_trkparms=72%3A1205|66%3A2|65%3A12|39%3A1|240%3A13 18|301%3A0|293%3A1|294%3A50

[Sadasius]

is this one decent? http://cgi.ebay.com/Pump-Diaphragm-Shurflo-1-8GPM-50PSI-12VDC_W0QQitemZ110326924265QQcmdZViewItemQQptZBI_P umps?hash=item110326924265&_trksid=p3286.c0.m14&_trkparms=72%3A1205|66%3A2|65%3A12|39%3A1|240%3A13 18|301%3A0|293%3A1|294%3A50

At 50 psi....Good pump but not for watercooling. Will pop your rad if a hose don't come of first. Too much pressure. An RD-30 running at full speed at 24v is 15psi.

[Fitseries3]

what would be acceptable?

[Cutless009]

At 50 psi....Good pump but not for watercooling. Will pop your rad if a hose don't come of first. Too much pressure. An RD-30 running at full speed at 24v is 15psi.

I bet you a thermochill could hold 50psi without popping, their stuff is high quality, and if you knew how to properly seal fittings, you could probably pull off 50psi in your loop, but hell that would be extreme, I bet the pumps would shake the whole system on start-up.

[Sadasius]

Well there was a guy that popped his TC rad on regular tap pressure which is about 60psi. Made all the tubes round....lol

[Fitseries3]

at 7gpm even if you have a lot of water in your loop the entire loop would cycle in a matter of seconds. no time to remove any heat from the components at all.

[Cutless009]

hey I didnt say it was practical, I said it was EXTREME lol :) Run 1/4" tubing :P That oughta restrict everything a bit :P

[jtf2]

http://www.overclock.net/sale/407206-ultimate-wc-setup-now-parting-out.html

[link1896]

At 50 psi....Good pump but not for watercooling. Will pop your rad if a hose don't come of first. Too much pressure. An RD-30 running at full speed at 24v is 15psi.

do you realize that the pressure, i.e 50psi, is what the pump CAN do, not what it WILL be doing in a given loop? The pressure in the loop, measured at the pumps outlet, is dictated by the loops restriction! (i.e what blocks, rads, etc, you have plumbed in)


If we take a positive displacement pump that CAN do 50psi before it's over pressure switch turns it off, and configure it with no restriction on the outlet, so effectively ZERO head, it will be producing practically zero pressure (lets ignore the restriction due to the outlet barb)

A 1.8gpm pump that CAN do 50psi before its maximum pressure safety switch cuts in (this is to protect it's plastic diaphragm from tearing) with a typical 120x3 rad, cpu block, gpu block and maybe a nb block, will produce nowhere near 50psi (or 100 feet) of restriction at 1.8gpm.

[link1896]

at 7gpm even if you have a lot of water in your loop the entire loop would cycle in a matter of seconds. no time to remove any heat from the components at all.

the higher the flow, the higher the turbulence is inside a water block, which gives us GREATER heat transfer.

Look at martin's testing. Do you see scenarios where where the coolant temp is even close to the die temps?

What you were probably wanting to say was a higher flow rate will reduce the time the coolant stays in the radiator, but again this is a common misconception.

Look at the below graph, compare the trends of the different flow rates. This data is for an industrial cooler from Lytron

http://www.lytron.com/uploadedImages/Heat_Exchangers/Standard_Heat_Exchangers/Standard_Product_Categories/Standard_Products/Pics,_Select,_Graphs,_DWG,_Specs/heat_exchanger_6340_graph.gif

[Fitseries3]

then whats going on with my rig? i run the pump at a low speed and temps are lower and the system has more stability but the water gets HOT as hell over time.

if i run my pump at high speed then my system is not as stable and i cant get as high of OCs. the water does stay cool though.

i've tried to find an equilibrium but have yet to find anything better than the slow pump setup.

[link1896]

then whats going on with my rig? i run the pump at a low speed and temps are lower and the system has more stability but the water gets HOT as hell over time.

if i run my pump at high speed then my system is not as stable and i cant get as high of OCs. the water does stay cool though.

i've tried to find an equilibrium but have yet to find anything better than the slow pump setup.

overclocked power draw + full speed pump power draw is pushing the PSU to the limits is my guess. Tried a separate power supply for the pump(s)?

[Fitseries3]

um... my water rig is external and it's completely independent from the computer.

it has its own PSU's for both the pump and the fans. pump is powered by a 240watt/24v meanwell and fans are powered by an 80watt 12v meanwell.

the computer runs off of 2 psu's itself. one 1000watt pcpower and cooling for gpu's and one 860watt pc power and cooling for the rest of the machine.

[Snyxxx]

then whats going on with my rig? i run the pump at a low speed and temps are lower and the system has more stability but the water gets HOT as hell over time.

if i run my pump at high speed then my system is not as stable and i cant get as high of OCs. the water does stay cool though.

i've tried to find an equilibrium but have yet to find anything better than the slow pump setup.

At low speed, maybe it just takes longer for the heat dump from the pump to heat the water.

At high speed, then the heat dump from the pump is going into the water faster for which you see sooner.

I added a second DDC pump in series and saw no benefit even though the flow rate was higher. The extra heat dump canceled the flow rate benefit.

[link1896]

At low speed, maybe it just takes longer for the heat dump from the pump to heat the water.

At high speed, then the heat dump from the pump is going into the water faster for which you see sooner.

I added a second DDC pump in series and saw no benefit even though the flow rate was higher. The extra heat dump canceled the flow rate benefit.

sounds quite plausible re Fitseries3's problems.

Snyxxx, did you try adding more airflow at the same time you added another pump?

I'll soon have my diaphragm pump working with my novel approach to cooling the motor frame.

[Snyxxx]

sounds quite plausible re Fitseries3's problems.

Snyxxx, did you try adding more airflow at the same time you added another pump?

I'll soon have my diaphragm pump working with my novel approach to cooling the motor frame.

No, I kept all conditions identical just to see the affect of adding a second pump. As others had reported, the heat did go into the water.

The net difference was about a wash in CPU and GPU temperatures. Therefore, not worth the extra hassle of running two pumps. However, some like redundancy with DDC pumps due to the bad reputation from the early days it still cannot quite shake.:rolleyes:

[Cutless009]

the higher the flow, the higher the turbulence is inside a water block, which gives us GREATER heat transfer.

Look at martin's testing. Do you see scenarios where where the coolant temp is even close to the die temps?

What you were probably wanting to say was a higher flow rate will reduce the time the coolant stays in the radiator, but again this is a common misconception.

Look at the below graph, compare the trends of the different flow rates. This data is for an industrial cooler from Lytron

http://www.lytron.com/uploadedImages/Heat_Exchangers/Standard_Heat_Exchangers/Standard_Product_Categories/Standard_Products/Pics,_Select,_Graphs,_DWG,_Specs/heat_exchanger_6340_graph.gif


So according to the results in this graph the higher the flow rate the better?

[link1896]

So according to the results in this graph the higher the flow rate the better?

yep. but heat dump from the pump can be an issue.
if you move away from standard pc water cooling pumps and into industrial pumps heat dump into the liquid being cooled is less of a problem. Just scale the radiator up accordingly otherwise

[Snyxxx]

So according to the results in this graph the higher the flow rate the better?

Higher flow rate is always better, but there are diminishing returns after a certain point.

[bigslappy]

Well there was a guy that popped his TC rad on regular tap pressure which is about 60psi. Made all the tubes round....lol

That BE ME ! & it was a GTX480 expensive lesson that was .....

[link1896]

That BE ME ! & it was a GTX480 expensive lesson that was .....

I can understand a join in the rad failing, but I can't come at the "made all the tubes round". Tell me its like a fishing story and their rads don't use 0.0001mm walled tubes

[Cutless009]

Mmmm, Im think RD-30's all around in my next build... maybe... three loops total? CPU, GPU's, and motherboard? Sounds awesome, YEP awesome. Only $230 per pump! YAY!

[Sadasius]

I can understand a join in the rad failing, but I can't come at the "made all the tubes round". Tell me its like a fishing story and their rads don't use 0.0001mm walled tubes


He has a picture of it. I remember seeing it. He said something that he heard pinging noises from the kitchen and discovered that the tubes were rounding out and the fins disconnecting. I guess I got the TC part wrong though. It was posted on XS some time ago.

[bigslappy]

He has a picture of it. I remember seeing it. He said something that he heard pinging noises from the kitchen and discovered that the tubes were rounding out and the fins disconnecting. I guess I got the TC part wrong though. It was posted on XS some time ago.

still got that rad round tubes all over & it leaks now

[bigslappy]

Mmmm, Im think RD-30's all around in my next build... maybe... three loops total? CPU, GPU's, and motherboard? Sounds awesome, YEP awesome. Only $230 per pump! YAY!
can be had for way less talk to RR

[Eternalightwith]

I'm interested. I want the EK400 res though. aka theBIGone

RD-30 is a MANS pump, not for the squimish....

BTW, PM me if you want one, I have a pakage deal coming up with pump/meanwell PSU/modded EK RES/a-bracket.

Only available for RedRaider Technology Members.:clap:

[bigslappy]

he's a mod for the EK400 into an Iwaki
http://www.realredraider.com/vbulletin/showthread.php?t=432

[NAekuh]

to be honest... since ive owned other iwaki's...

the other guys dont make any sense... there way too big... so unless you have a freezing tower like slappy does, i dont recomend anything besides the RD-30.

And if you dont know why your getting the RD-30, and your just getting it because its the RD-30 then... umm please PM me... I need to do some teaching...

[Cutless009]

to be honest... since ive owned other iwaki's...

the other guys dont make any sense... there way too big... so unless you have a freezing tower like slappy does, i dont recomend anything besides the RD-30.

And if you dont know why your getting the RD-30, and your just getting it because its the RD-30 then... umm please PM me... I need to do some teaching...

Uh, getting it because it has the best pressure, meaning I can use more rads, more restrictive blocks, and put more into a loop, and still maintain > 1gpm flow, with higher flow rates for less restrictive loops o.o.

But feel free to pm me, I do need to learn more about pumps, and Im always up for some learnin :up:

[NAekuh]

Uh, getting it because it has the best pressure, meaning I can use more rads, more restrictive blocks, and put more into a loop, and still maintain > 1gpm flow, with higher flow rates for less restrictive loops o.o.

But feel free to pm me, I do need to learn more about pumps, and Im always up for some learnin :up:

no thats why i said if your getting it because its a "RD-30" to pm me.. lol...

Seems like you know the reason for getting it.

[Cutless009]

no thats why i said if your getting it because its a "RD-30" to pm me.. lol...

Seems like you know the reason for getting it.

Its all good, Im still very nab, in fact, Ive never booted a WC computer, tonight will be the first time I do that! Im a researcher for sure, I think I spent at LEAST a 1,000 hours (reasonably and in all seriousness) researching parts before I began Formal Black, and it is indeed my very first WC build EVER. Cant wait to boot it up!

But yah if you ever see me say anything noobish please feel free to pm me with why and what the correct answer is :up: always eager to learn more ^_^

[Ricey]

The need for pressure is only a function of the restrictiveness of blocks.

Nobody can convince me, beyond a certain extent, that super restrictive blocks are just plain silly and IMO totally sub-standard.

I probably pissed off 90 percent of this forum. Guess what? I don't give a shit.

I always separate my cpu block from everything else. I have no need for that pressure because my blocks flow well enough. Pressure by itself does all of jack shit for cooling.

Its time to reverse the shit wisdom that has permeated XS for all these years, and everyone believes that shit espoused at XS as gospel.

No matter what the diminish returns are, more flow is always better. Heat dump is a function of pump design. High heat dump can always be negated with radiator capacity.

Piss poor flow rates like 1 or 1.5 GPM cannot be negated.

I can do much much better with flow with multiple Laings, and with lower heat dump too, than whatever Iwaki that is heralded. Flow is not important, right?

The Koolaid probably tastes great! Thank God I've never drunk that urine.

[bigslappy]

buy one , use it,

u'll never go back


;p

[CrazyJoe]

After seeing your mod ... I may need to buy two !

[Metal]

Buy 3 bigger E-Peen lol

[Ricey]

Let me fill you in on a tip. Iwakis die. Two people with dead RD30s already. Phcpp had two die. So, I guess you guys don't need redundancy huh?

So don't give me flack about Iwakis being bulletproof. They are just Japanese pumps. I'll take German equipment any day of the week.

Now.. Anyone care to hear my opinions about Sanyo Denki? :) :)

[Kayin]

The need for pressure is only a function of the restrictiveness of blocks.

Nobody can convince me, beyond a certain extent, that super restrictive blocks are just plain silly and IMO totally sub-standard.

I probably pissed off 90 percent of this forum. Guess what? I don't give a shit.

I always separate my cpu block from everything else. I have no need for that pressure because my blocks flow well enough. Pressure by itself does all of jack shit for cooling.

Its time to reverse the shit wisdom that has permeated XS for all these years, and everyone believes that shit espoused at XS as gospel.

No matter what the diminish returns are, more flow is always better. Heat dump is a function of pump design. High heat dump can always be negated with radiator capacity.

Piss poor flow rates like 1 or 1.5 GPM cannot be negated.

I can do much much better with flow with multiple Laings, and with lower heat dump too, than whatever Iwaki that is heralded. Flow is not important, right?

The Koolaid probably tastes great! Thank God I've never drunk that urine.

Quoted for truth. Ricey, I think we are the only ones here who understand why flow triumphs. Those cross-drilled blocks certainly didn't give much surface area, did they?

[RedRaider]

Kayin, What you talking about Willis ??? :obama:

Bro, I sang the virtues of HIGH FLOW from the mountain tops on the "other" forum...:fact:

[Kayin]

I'm saying just what Ricey did. Flow cannot be replaced (though certain designs do not care as much for it) but pressure? Not really necessary, except with a few designs on the market today. My personal jury is still out on whether it's worth it or not (I think a lapped Supreme is the way to go, or the AC block) because I don't have the items in hand, but as far as pressure? Theres a reason that the Laing DDC and the Laing D5 (the two most common pumps) as well as the Eheim series, and all the way back through our roots to the Danner Mag-Drive (still have one in use) and the Hydor L30 all do not have high pressure-it's not needed.

Red, you learn quickly, you're not dumb, and you never take it "because X said so." Therefore, I'm here and not there. However, Ricey and I learned this lesson back when WC was a newer hobby, it was mostly DIY and we all experimented to figure shit out. I do not share every opinion of his, but we learned this one separately by doing the same damn thing. Flow moves water past, if the water is moving, and it's a closed loop, under normal circumstances pressure means jack-all. We started having pressure show up with the impingement designs (Little River WhiteWater, to whit, and its clone the RBX) and those had problems all their own. However, I ran an RBX on an OCed 3000+ A64 at 2.9 from 1.8, with a Danner Mag3 and it tore ass even though it's a pump that's not meant to encounter any resistance.

I'll look up the principles involved, but suffice to say if you wanted an RD-30, you sought reliability. Pressure should be a non-issue.

I have more experimenting to do, but we all know we will both be trying, no?

[Ricey]

I have never been a DIY kind of guy because I'ma klutz. I took a hole saw to an old steel case to try, and I made a mess of it.

What I do know is that most of this water cooling thing is just common sense. A lot of people hide behind a lot of mumbo jumbo science, and all the fancy charts and graphs, and those math formulae do not replace basic common sense.

The whole hobby has evolved since its inception. However, certain elements, including Crediki, Nikhsub.. not to mention Cathar.. have come up with their own version of what is good and proper. I respect them for that, but don't expect me to drink the Koolaid like another of those fucking XS sheep. Those people can be led to the abattoir and slaughtered into lamb chops, or led right off the edge of the cliff. Not me.

I'm not blindly subscribing to Iwaki and San Ace and all the other prescribed XS nonsense without some intense scrutiny.